Authors: Ouail Kitouni, Niklas Nolte, Diane Bouchacourt, Adina Williams, Mike Rabbat, Mark Ibrahim
Abstract: Today's best language models still struggle with hallucinations: factually incorrect generations, which impede their ability to reliably retrieve information seen during training. The reversal curse, where models cannot recall information when probed in a different order than was encountered during training, exemplifies this in information retrieval. We reframe the reversal curse as a factorization curse - a failure of models to learn the same joint distribution under different factorizations. Through a series of controlled experiments with increasing levels of realism including WikiReversal, a setting we introduce to closely simulate a knowledge intensive finetuning task, we find that the factorization curse is an inherent failure of the next-token prediction objective used in popular large language models. Moreover, we demonstrate reliable information retrieval cannot be solved with scale, reversed tokens, or even naive bidirectional-attention training. Consequently, various approaches to finetuning on specialized data would necessarily provide mixed results on downstream tasks, unless the model has already seen the right sequence of tokens. Across five tasks of varying levels of complexity, our results uncover a promising path forward: factorization-agnostic objectives can significantly mitigate the reversal curse and hint at improved knowledge storage and planning capabilities.
Authors: Aashish Arora, Elsbeth Turcan
Abstract: Data augmentation has the potential to improve the performance of machine learning models by increasing the amount of training data available. In this study, we evaluated the effectiveness of different data augmentation techniques for a multi-label emotion classification task using a low-resource dataset. Our results showed that Back Translation outperformed autoencoder-based approaches and that generating multiple examples per training instance led to further performance improvement. In addition, we found that Back Translation generated the most diverse set of unigrams and trigrams. These findings demonstrate the utility of Back Translation in enhancing the performance of emotion classification models in resource-limited situations.
Authors: Valentin Thomas, Junwei Ma, Rasa Hosseinzadeh, Keyvan Golestan, Guangwei Yu, Maksims Volkovs, Anthony Caterini
Abstract: Tabular data is a pervasive modality spanning a wide range of domains, and the inherent diversity poses a considerable challenge for deep learning. Recent advancements using transformer-based in-context learning have shown promise on smaller and less complex datasets, but have struggled to scale to larger and more complex ones. To address this limitation, we propose a combination of retrieval and fine-tuning: we can adapt the transformer to a local subset of the data by collecting nearest neighbours, and then perform task-specific fine-tuning with this retrieved set of neighbours in context. Using TabPFN as the base model -- currently the best tabular in-context learner -- and applying our retrieval and fine-tuning scheme on top results in what we call a locally-calibrated PFN, or LoCalPFN. We conduct extensive evaluation on 95 datasets curated by TabZilla from OpenML, upon which we establish a new state-of-the-art with LoCalPFN -- even with respect to tuned tree-based models. Notably, we show a significant boost in performance compared to the base in-context model, demonstrating the efficacy of our approach and advancing the frontier of deep learning in tabular data.
Authors: Junwei Ma, Apoorv Dankar, George Stein, Guangwei Yu, Anthony Caterini
Abstract: Advances in deep generative modelling have not translated well to tabular data. We argue that this is caused by a mismatch in structure between popular generative models and discriminative models of tabular data. We thus devise a technique to turn TabPFN -- a highly performant transformer initially designed for in-context discriminative tabular tasks -- into an energy-based generative model, which we dub TabPFGen. This novel framework leverages the pre-trained TabPFN as part of the energy function and does not require any additional training or hyperparameter tuning, thus inheriting TabPFN's in-context learning capability. We can sample from TabPFGen analogously to other energy-based models. We demonstrate strong results on standard generative modelling tasks, including data augmentation, class-balancing, and imputation, unlocking a new frontier of tabular data generation.
Authors: Yibo Yang, Xiaojie Li, Motasem Alfarra, Hasan Hammoud, Adel Bibi, Philip Torr, Bernard Ghanem
Abstract: Relieving the reliance of neural network training on a global back-propagation (BP) has emerged as a notable research topic due to the biological implausibility and huge memory consumption caused by BP. Among the existing solutions, local learning optimizes gradient-isolated modules of a neural network with local errors and has been proved to be effective even on large-scale datasets. However, the reconciliation among local errors has never been investigated. In this paper, we first theoretically study non-greedy layer-wise training and show that the convergence cannot be assured when the local gradient in a module w.r.t. its input is not reconciled with the local gradient in the previous module w.r.t. its output. Inspired by the theoretical result, we further propose a local training strategy that successively regularizes the gradient reconciliation between neighboring modules without breaking gradient isolation or introducing any learnable parameters. Our method can be integrated into both local-BP and BP-free settings. In experiments, we achieve significant performance improvements compared to previous methods. Particularly, our method for CNN and Transformer architectures on ImageNet is able to attain a competitive performance with global BP, saving more than 40% memory consumption.
Authors: Yibo Yang, Xiaojie Li, Zhongzhu Zhou, Shuaiwen Leon Song, Jianlong Wu, Liqiang Nie, Bernard Ghanem
Abstract: Current parameter-efficient fine-tuning (PEFT) methods build adapters without considering the context of downstream task to learn, or the context of important knowledge to maintain. As a result, there is often a performance gap compared to full-parameter finetuning, and meanwhile the finetuned model suffers from catastrophic forgetting of the pre-trained world knowledge. In this paper, we propose CorDA, a Context-oriented Decomposition Adaptation method that builds learnable adapters from weight decomposition oriented by the context of downstream task or world knowledge. Concretely, we collect a few data samples, and perform singular value decomposition for each linear layer of a pre-trained LLM multiplied by the covariance matrix of the input activation using these samples. By doing so, the context of the representative samples is captured through deciding the factorizing orientation. Our method enables two options, the knowledge-preserved adaptation and the instruction-previewed adaptation. For the former, we use question-answering samples to obtain the covariance matrices, and use the decomposed components with the smallest $r$ singular values to initialize a learnable adapter, with the others frozen such that the world knowledge is better preserved. For the latter, we use the instruction data from the finetuning task, such as math or coding, to orientate the decomposition and train the largest $r$ components that capture the main characteristics of the task to learn. We conduct extensive experiments on Math, Code, and Instruction Following tasks. Our knowledge-preserved adaptation not only achieves better performance than LoRA on finetuning tasks, but also mitigates the forgetting of world knowledge. Our instruction-previewed adaptation is able to further enhance the finetuning performance, surpassing full-parameter finetuning and the state-of-the-art PEFT methods.
Authors: Zhiyang Wang, Juan Cervino, Alejandro Ribeiro
Abstract: Convolutional neural networks have been successfully extended to operate on graphs, giving rise to Graph Neural Networks (GNNs). GNNs combine information from adjacent nodes by successive applications of graph convolutions. GNNs have been implemented successfully in various learning tasks while the theoretical understanding of their generalization capability is still in progress. In this paper, we leverage manifold theory to analyze the statistical generalization gap of GNNs operating on graphs constructed on sampled points from manifolds. We study the generalization gaps of GNNs on both node-level and graph-level tasks. We show that the generalization gaps decrease with the number of nodes in the training graphs, which guarantees the generalization of GNNs to unseen points over manifolds. We validate our theoretical results in multiple real-world datasets.
Authors: Philippe Chlenski, Quentin Chu, Itsik Pe'er
Abstract: We extend decision tree and random forest algorithms to mixed-curvature product spaces. Such spaces, defined as Cartesian products of Euclidean, hyperspherical, and hyperbolic manifolds, can often embed points from pairwise distances with much lower distortion than in single manifolds. To date, all classifiers for product spaces fit a single linear decision boundary, and no regressor has been described. Our method overcomes these limitations by enabling simple, expressive classification and regression in product manifolds. We demonstrate the superior accuracy of our tool compared to Euclidean methods operating in the ambient space for component manifolds covering a wide range of curvatures, as well as on a selection of product manifolds.
Authors: Kevin Kuo, Arian Raje, Kousik Rajesh, Virginia Smith
Abstract: Low-rank adaptation (LoRA) is a natural method for finetuning in communication-constrained machine learning settings such as cross-device federated learning. Prior work that has studied LoRA in the context of federated learning has focused on improving LoRA's robustness to heterogeneity and privacy. In this work, we instead consider techniques for further improving communication-efficiency in federated LoRA. Unfortunately, we show that centralized ML methods that improve the efficiency of LoRA through unstructured pruning do not transfer well to federated settings. We instead study a simple approach, \textbf{FLASC}, that applies sparsity to LoRA during communication while allowing clients to locally fine-tune the entire LoRA module. Across four common federated learning tasks, we demonstrate that this method matches the performance of dense LoRA with up to $10\times$ less communication. Additionally, despite being designed primarily to target communication, we find that this approach has benefits in terms of heterogeneity and privacy relative to existing approaches tailored to these specific concerns. Overall, our work highlights the importance of considering system-specific constraints when developing communication-efficient finetuning approaches, and serves as a simple and competitive baseline for future work in federated finetuning.
Authors: Steven Cho, Seaton Cousins-Baxter, Stefano Ruberto, Valerio Terragni
Abstract: Machine Learning (ML) has become an integral part of our society, commonly used in critical domains such as finance, healthcare, and transportation. Therefore, it is crucial to evaluate not only whether ML models make correct predictions but also whether they do so for the correct reasons, ensuring our trust that will perform well on unseen data. This concept is known as trustworthiness in ML. Recently, explainable techniques (e.g., LIME, SHAP) have been developed to interpret the decision-making processes of ML models, providing explanations for their predictions (e.g., words in the input that influenced the prediction the most). Assessing the plausibility of these explanations can enhance our confidence in the models' trustworthiness. However, current approaches typically rely on human judgment to determine the plausibility of these explanations. This paper proposes TOWER, the first technique to automatically create trustworthiness oracles that determine whether text classifier predictions are trustworthy. It leverages word embeddings to automatically evaluate the trustworthiness of a model-agnostic text classifiers based on the outputs of explanatory techniques. Our hypothesis is that a prediction is trustworthy if the words in its explanation are semantically related to the predicted class. We perform unsupervised learning with untrustworthy models obtained from noisy data to find the optimal configuration of TOWER. We then evaluated TOWER on a human-labeled trustworthiness dataset that we created. The results show that TOWER can detect a decrease in trustworthiness as noise increases, but is not effective when evaluated against the human-labeled dataset. Our initial experiments suggest that our hypothesis is valid and promising, but further research is needed to better understand the relationship between explanations and trustworthiness issues.
Authors: Jing Liu, Andrew Lowy, Toshiaki Koike-Akino, Kieran Parsons, Ye Wang
Abstract: The recent developments of Diffusion Models (DMs) enable generation of astonishingly high-quality synthetic samples. Recent work showed that the synthetic samples generated by the diffusion model, which is pre-trained on public data and fully fine-tuned with differential privacy on private data, can train a downstream classifier, while achieving a good privacy-utility tradeoff. However, fully fine-tuning such large diffusion models with DP-SGD can be very resource-demanding in terms of memory usage and computation. In this work, we investigate Parameter-Efficient Fine-Tuning (PEFT) of diffusion models using Low-Dimensional Adaptation (LoDA) with Differential Privacy. We evaluate the proposed method with the MNIST and CIFAR-10 datasets and demonstrate that such efficient fine-tuning can also generate useful synthetic samples for training downstream classifiers, with guaranteed privacy protection of fine-tuning data. Our source code will be made available on GitHub.
Authors: Oshin Dutta, Ritvik Gupta, Sumeet Agarwal
Abstract: In recent years, there has been a growing emphasis on compressing large pre-trained transformer models for resource-constrained devices. However, traditional pruning methods often leave the embedding layer untouched, leading to model over-parameterization. Additionally, they require extensive compression time with large datasets to maintain performance in pruned models. To address these challenges, we propose VTrans, an iterative pruning framework guided by the Variational Information Bottleneck (VIB) principle. Our method compresses all structural components, including embeddings, attention heads, and layers using VIB-trained masks. This approach retains only essential weights in each layer, ensuring compliance with specified model size or computational constraints. Notably, our method achieves upto 70% more compression than prior state-of-the-art approaches, both task-agnostic and task-specific. We further propose faster variants of our method: Fast-VTrans utilizing only 3% of the data and Faster-VTrans, a time efficient alternative that involves exclusive finetuning of VIB masks, accelerating compression by upto 25 times with minimal performance loss compared to previous methods. Extensive experiments on BERT, ROBERTa, and GPT-2 models substantiate the efficacy of our method. Moreover, our method demonstrates scalability in compressing large models such as LLaMA-2-7B, achieving superior performance compared to previous pruning methods. Additionally, we use attention-based probing to qualitatively assess model redundancy and interpret the efficiency of our approach. Notably, our method considers heads with high attention to special and current tokens in un-pruned model as foremost candidates for pruning while retained heads are observed to attend more to task-critical keywords.
Authors: Samuel Deng, Daniel Hsu, Jingwen Liu
Abstract: We study the problem of online multi-group learning, a learning model in which an online learner must simultaneously achieve small prediction regret on a large collection of (possibly overlapping) subsequences corresponding to a family of groups. Groups are subsets of the context space, and in fairness applications, they may correspond to subpopulations defined by expressive functions of demographic attributes. In contrast to previous work on this learning model, we consider scenarios in which the family of groups is too large to explicitly enumerate, and hence we seek algorithms that only access groups via an optimization oracle. In this paper, we design such oracle-efficient algorithms with sublinear regret under a variety of settings, including: (i) the i.i.d. setting, (ii) the adversarial setting with smoothed context distributions, and (iii) the adversarial transductive setting.
Authors: Abhiram Anand Thiruthummal, Sergiy Shelyag, Eun-jin Kim
Abstract: We propose a novel method for fast and accurate training of physics-informed neural networks (PINNs) to find solutions to boundary value problems (BVPs) and initial boundary value problems (IBVPs). By combining the methods of training deep neural networks (DNNs) and Extreme Learning Machines (ELMs), we develop a model which has the expressivity of DNNs with the fine-tuning ability of ELMs. We showcase the superiority of our proposed method by solving several BVPs and IBVPs which include linear and non-linear ordinary differential equations (ODEs), partial differential equations (PDEs) and coupled PDEs. The examples we consider include a stiff coupled ODE system where traditional numerical methods fail, a 3+1D non-linear PDE, Kovasznay flow and Taylor-Green vortex solutions to incompressible Navier-Stokes equations and pure advection solution of 1+1 D compressible Euler equation. The Theory of Functional Connections (TFC) is used to exactly impose initial and boundary conditions (IBCs) of (I)BVPs on PINNs. We propose a modification to the TFC framework named Reduced TFC and show a significant improvement in the training and inference time of PINNs compared to IBCs imposed using TFC. Furthermore, Reduced TFC is shown to be able to generalize to more complex boundary geometries which is not possible with TFC. We also introduce a method of applying boundary conditions at infinity for BVPs and numerically solve the pure advection in 1+1 D Euler equations using these boundary conditions.
Authors: Abhiram Anand Thiruthummal, Eun-jin Kim, Sergiy Shelyag
Abstract: Artificial neural networks (ANNs) are powerful tools capable of approximating any arbitrary mathematical function, but their interpretability remains limited, rendering them as black box models. To address this issue, numerous methods have been proposed to enhance the explainability and interpretability of ANNs. In this study, we introduce the application of information geometric framework to investigate phase transition-like behavior during the training of ANNs and relate these transitions to overfitting in certain models. The evolution of ANNs during training is studied by looking at the probability distribution of its parameters. Information geometry utilizing the principles of differential geometry, offers a unique perspective on probability and statistics by considering probability density functions as points on a Riemannian manifold. We create this manifold using a metric based on Fisher information to define a distance and a velocity. By parameterizing this distance and velocity with training steps, we study how the ANN evolves as training progresses. Utilizing standard datasets like MNIST, FMNIST and CIFAR-10, we observe a transition in the motion on the manifold while training the ANN and this transition is identified with over-fitting in the ANN models considered. The information geometric transitions observed is shown to be mathematically similar to the phase transitions in physics. Preliminary results showing finite-size scaling behavior is also provided. This work contributes to the development of robust tools for improving the explainability and interpretability of ANNs, aiding in our understanding of the variability of the parameters these complex models exhibit during training.
Authors: Carole-Jean Wu, Bilge Acun, Ramya Raghavendra, Kim Hazelwood
Abstract: Barroso's seminal contributions in energy-proportional warehouse-scale computing launched an era where modern datacenters have become more energy efficient and cost effective than ever before. At the same time, modern AI applications have driven ever-increasing demands in computing, highlighting the importance of optimizing efficiency across the entire deep learning model development cycle. This paper characterizes the carbon impact of AI, including both operational carbon emissions from training and inference as well as embodied carbon emissions from datacenter construction and hardware manufacturing. We highlight key efficiency optimization opportunities for cutting-edge AI technologies, from deep learning recommendation models to multi-modal generative AI tasks. To scale AI sustainably, we must also go beyond efficiency and optimize across the life cycle of computing infrastructures, from hardware manufacturing to datacenter operations and end-of-life processing for the hardware.
Authors: Chaolv Zeng, Zhanyu Liu, Guanjie Zheng, Linghe Kong
Abstract: In recent years, significant progress has been made in multivariate time series forecasting using Linear-based, Transformer-based, and Convolution-based models. However, these approaches face notable limitations: linear forecasters struggle with representation capacities, attention mechanisms suffer from quadratic complexity, and convolutional models have a restricted receptive field. These constraints impede their effectiveness in modeling complex time series, particularly those with numerous variables. Additionally, many models adopt the Channel-Independent (CI) strategy, treating multivariate time series as uncorrelated univariate series while ignoring their correlations. For models considering inter-channel relationships, whether through the self-attention mechanism, linear combination, or convolution, they all incur high computational costs and focus solely on weighted summation relationships, neglecting potential proportional relationships between channels. In this work, we address these issues by leveraging the newly introduced state space model and propose \textbf{C-Mamba}, a novel approach that captures cross-channel dependencies while maintaining linear complexity without losing the global receptive field. Our model consists of two key components: (i) channel mixup, where two channels are mixed to enhance the training sets; (ii) channel attention enhanced patch-wise Mamba encoder that leverages the ability of the state space models to capture cross-time dependencies and models correlations between channels by mining their weight relationships. Our model achieves state-of-the-art performance on seven real-world time series datasets. Moreover, the proposed mixup and attention strategy exhibits strong generalizability across other frameworks.
Authors: Ruisi Cai, Yuandong Tian, Zhangyang Wang, Beidi Chen
Abstract: This paper tackles the memory hurdle of processing long context sequences in Large Language Models (LLMs), by presenting a novel approach, Dropping In Convolutions for Long Context Compression (LoCoCo). LoCoCo employs only a fixed-size Key-Value (KV) cache, and can enhance efficiency in both inference and fine-tuning stages. Diverging from prior methods that selectively drop KV pairs based on heuristics, LoCoCo leverages a data-driven adaptive fusion technique, blending previous KV pairs with incoming tokens to minimize the loss of contextual information and ensure accurate attention modeling. This token integration is achieved through injecting one-dimensional convolutional kernels that dynamically calculate mixing weights for each KV cache slot. Designed for broad compatibility with existing LLM frameworks, LoCoCo allows for straightforward "drop-in" integration without needing architectural modifications, while incurring minimal tuning overhead. Experiments demonstrate that LoCoCo maintains consistently outstanding performance across various context lengths and can achieve a high context compression rate during both inference and fine-tuning phases. During inference, we successfully compressed up to 3482 tokens into a 128-size KV cache, while retaining comparable performance to the full sequence - an accuracy improvement of up to 0.2791 compared to baselines at the same cache size. During post-training tuning, we also effectively extended the context length from 4K to 32K using a KV cache of fixed size 512, achieving performance similar to fine-tuning with entire sequences.
Authors: Junghwan Lee, Chen Xu, Yao Xie
Abstract: We present a conformal prediction method for time series using the Transformer architecture to capture long-memory and long-range dependencies. Specifically, we use the Transformer decoder as a conditional quantile estimator to predict the quantiles of prediction residuals, which are used to estimate the prediction interval. We hypothesize that the Transformer decoder benefits the estimation of the prediction interval by learning temporal dependencies across past prediction residuals. Our comprehensive experiments using simulated and real data empirically demonstrate the superiority of the proposed method compared to the existing state-of-the-art conformal prediction methods.
Authors: Chenyi Zi, Haihong Zhao, Xiangguo Sun, Yiqing Lin, Hong Cheng, Jia Li
Abstract: Artificial general intelligence on graphs has shown significant advancements across various applications, yet the traditional 'Pre-train & Fine-tune' paradigm faces inefficiencies and negative transfer issues, particularly in complex and few-shot settings. Graph prompt learning emerges as a promising alternative, leveraging lightweight prompts to manipulate data and fill the task gap by reformulating downstream tasks to the pretext. However, several critical challenges still remain: how to unify diverse graph prompt models, how to evaluate the quality of graph prompts, and to improve their usability for practical comparisons and selection. In response to these challenges, we introduce the first comprehensive benchmark for graph prompt learning. Our benchmark integrates SIX pre-training methods and FIVE state-of-the-art graph prompt techniques, evaluated across FIFTEEN diverse datasets to assess performance, flexibility, and efficiency. We also present 'ProG', an easy-to-use open-source library that streamlines the execution of various graph prompt models, facilitating objective evaluations. Additionally, we propose a unified framework that categorizes existing graph prompt methods into two main approaches: prompts as graphs and prompts as tokens. This framework enhances the applicability and comparison of graph prompt techniques. The code is available at: https://github.com/sheldonresearch/ProG.
Authors: Huiling Meng, Ningyuan Chen, Xuefeng Gao
Abstract: Intensity control is a type of continuous-time dynamic optimization problems with many important applications in Operations Research including queueing and revenue management. In this study, we adapt the reinforcement learning framework to intensity control using choice-based network revenue management as a case study, which is a classical problem in revenue management that features a large state space, a large action space and a continuous time horizon. We show that by utilizing the inherent discretization of the sample paths created by the jump points, a unique and defining feature of intensity control, one does not need to discretize the time horizon in advance, which was believed to be necessary because most reinforcement learning algorithms are designed for discrete-time problems. As a result, the computation can be facilitated and the discretization error is significantly reduced. We lay the theoretical foundation for the Monte Carlo and temporal difference learning algorithms for policy evaluation and develop policy gradient based actor critic algorithms for intensity control. Via a comprehensive numerical study, we demonstrate the benefit of our approach versus other state-of-the-art benchmarks.
Authors: Wenhao Xu, Xuefeng Gao, Xuedong He
Abstract: Risk-sensitive linear quadratic regulator is one of the most fundamental problems in risk-sensitive optimal control. In this paper, we study online adaptive control of risk-sensitive linear quadratic regulator in the finite horizon episodic setting. We propose a simple least-squares greedy algorithm and show that it achieves $\widetilde{\mathcal{O}}(\log N)$ regret under a specific identifiability assumption, where $N$ is the total number of episodes. If the identifiability assumption is not satisfied, we propose incorporating exploration noise into the least-squares-based algorithm, resulting in an algorithm with $\widetilde{\mathcal{O}}(\sqrt{N})$ regret. To our best knowledge, this is the first set of regret bounds for episodic risk-sensitive linear quadratic regulator. Our proof relies on perturbation analysis of less-standard Riccati equations for risk-sensitive linear quadratic control, and a delicate analysis of the loss in the risk-sensitive performance criterion due to applying the suboptimal controller in the online learning process.
Authors: Lukas Weigand, Tim Roith, Martin Burger
Abstract: A popular method to perform adversarial attacks on neuronal networks is the so-called fast gradient sign method and its iterative variant. In this paper, we interpret this method as an explicit Euler discretization of a differential inclusion, where we also show convergence of the discretization to the associated gradient flow. To do so, we consider the concept of p-curves of maximal slope in the case $p=\infty$. We prove existence of $\infty$-curves of maximum slope and derive an alternative characterization via differential inclusions. Furthermore, we also consider Wasserstein gradient flows for potential energies, where we show that curves in the Wasserstein space can be characterized by a representing measure on the space of curves in the underlying Banach space, which fulfill the differential inclusion. The application of our theory to the finite-dimensional setting is twofold: On the one hand, we show that a whole class of normalized gradient descent methods (in particular signed gradient descent) converge, up to subsequences, to the flow, when sending the step size to zero. On the other hand, in the distributional setting, we show that the inner optimization task of adversarial training objective can be characterized via $\infty$-curves of maximum slope on an appropriate optimal transport space.
Authors: Zhaoru Ke, Hang Yu, Jianguo Li, Haipeng Zhang
Abstract: Current directed graph embedding methods build upon undirected techniques but often inadequately capture directed edge information, leading to challenges such as: (1) Suboptimal representations for nodes with low in/out-degrees, due to the insufficient neighbor interactions; (2) Limited inductive ability for representing new nodes post-training; (3) Narrow generalizability, as training is overly coupled with specific tasks. In response, we propose DUPLEX, an inductive framework for complex embeddings of directed graphs. It (1) leverages Hermitian adjacency matrix decomposition for comprehensive neighbor integration, (2) employs a dual GAT encoder for directional neighbor modeling, and (3) features two parameter-free decoders to decouple training from particular tasks. DUPLEX outperforms state-of-the-art models, especially for nodes with sparse connectivity, and demonstrates robust inductive capability and adaptability across various tasks. The code is available at https://github.com/alipay/DUPLEX.
Authors: Vahid MohammadZadeh Eivaghi, Mahdi Aliyari Shoorehdeli
Abstract: The Fisher Information Matrix (FIM) provides a way for quantifying the information content of an observable random variable concerning unknown parameters within a model that characterizes the variable. When parameters in a model are directly linked to individual features, the diagonal elements of the FIM can signify the relative importance of each feature. However, in scenarios where feature interactions may exist, a comprehensive exploration of the full FIM is necessary rather than focusing solely on its diagonal elements. This paper presents an end-to-end black box system identification approach that integrates the FIM into the training process to gain insights into dynamic importance and overall model structure. A decision module is added to the first layer of the network to determine the relevance scores using the entire FIM as input. The forward propagation is then performed on element-wise multiplication of inputs and relevance scores. Simulation results demonstrate that the proposed methodology effectively captures various types of interactions between dynamics, outperforming existing methods limited to polynomial interactions. Moreover, the effectiveness of this novel approach is confirmed through its application in identifying a real-world industrial system, specifically the PH neutralization process.
Authors: Sungwoo Park, Dongjun Kim, Ahmed Alaa
Abstract: In this paper, we introduce a new class of score-based generative models (SGMs) designed to handle high-cardinality data distributions by leveraging concepts from mean-field theory. We present mean-field chaos diffusion models (MF-CDMs), which address the curse of dimensionality inherent in high-cardinality data by utilizing the propagation of chaos property of interacting particles. By treating high-cardinality data as a large stochastic system of interacting particles, we develop a novel score-matching method for infinite-dimensional chaotic particle systems and propose an approximation scheme that employs a subdivision strategy for efficient training. Our theoretical and empirical results demonstrate the scalability and effectiveness of MF-CDMs for managing large high-cardinality data structures, such as 3D point clouds.
Authors: Shai Feldman, Yaniv Romano
Abstract: We develop a method to generate prediction sets with a guaranteed coverage rate that is robust to corruptions in the training data, such as missing or noisy variables. Our approach builds on conformal prediction, a powerful framework to construct prediction sets that are valid under the i.i.d assumption. Importantly, naively applying conformal prediction does not provide reliable predictions in this setting, due to the distribution shift induced by the corruptions. To account for the distribution shift, we assume access to privileged information (PI). The PI is formulated as additional features that explain the distribution shift, however, they are only available during training and absent at test time. We approach this problem by introducing a novel generalization of weighted conformal prediction and support our method with theoretical coverage guarantees. Empirical experiments on both real and synthetic datasets indicate that our approach achieves a valid coverage rate and constructs more informative predictions compared to existing methods, which are not supported by theoretical guarantees.
Authors: Qinting Jiang, Chuyang Ye, Dongyan Wei, Yuan Xue, Jingyan Jiang, Zhi Wang
Abstract: Despite progress, deep neural networks still suffer performance declines under distribution shifts between training and test domains, leading to a substantial decrease in Quality of Experience (QoE) for multimedia applications. Existing test-time adaptation (TTA) methods are challenged by dynamic, multiple test distributions within batches. This work provides a new perspective on analyzing batch normalization techniques through class-related and class-irrelevant features, our observations reveal combining source and test batch normalization statistics robustly characterizes target distributions. However, test statistics must have high similarity. We thus propose Discover Your Neighbours (DYN), the first backward-free approach specialized for dynamic TTA. The core innovation is identifying similar samples via instance normalization statistics and clustering into groups which provides consistent class-irrelevant representations. Specifically, Our DYN consists of layer-wise instance statistics clustering (LISC) and cluster-aware batch normalization (CABN). In LISC, we perform layer-wise clustering of approximate feature samples at each BN layer by calculating the cosine similarity of instance normalization statistics across the batch. CABN then aggregates SBN and TCN statistics to collaboratively characterize the target distribution, enabling more robust representations. Experimental results validate DYN's robustness and effectiveness, demonstrating maintained performance under dynamic data stream patterns.
Authors: George Ma, Emmanuel Bengio, Yoshua Bengio, Dinghuai Zhang
Abstract: GFlowNets have exhibited promising performance in generating diverse candidates with high rewards. These networks generate objects incrementally and aim to learn a policy that assigns probability of sampling objects in proportion to rewards. However, the current training pipelines of GFlowNets do not consider the presence of isomorphic actions, which are actions resulting in symmetric or isomorphic states. This lack of symmetry increases the amount of samples required for training GFlowNets and can result in inefficient and potentially incorrect flow functions. As a consequence, the reward and diversity of the generated objects decrease. In this study, our objective is to integrate symmetries into GFlowNets by identifying equivalent actions during the generation process. Experimental results using synthetic data demonstrate the promising performance of our proposed approaches.
Authors: Qi Lv, Xiang Deng, Gongwei Chen, Michael Yu Wang, Liqiang Nie
Abstract: While the conditional sequence modeling with the transformer architecture has demonstrated its effectiveness in dealing with offline reinforcement learning (RL) tasks, it is struggle to handle out-of-distribution states and actions. Existing work attempts to address this issue by data augmentation with the learned policy or adding extra constraints with the value-based RL algorithm. However, these studies still fail to overcome the following challenges: (1) insufficiently utilizing the historical temporal information among inter-steps, (2) overlooking the local intrastep relationships among states, actions and return-to-gos (RTGs), (3) overfitting suboptimal trajectories with noisy labels. To address these challenges, we propose Decision Mamba (DM), a novel multi-grained state space model (SSM) with a self-evolving policy learning strategy. DM explicitly models the historical hidden state to extract the temporal information by using the mamba architecture. To capture the relationship among state-action-RTG triplets, a fine-grained SSM module is designed and integrated into the original coarse-grained SSM in mamba, resulting in a novel mamba architecture tailored for offline RL. Finally, to mitigate the overfitting issue on noisy trajectories, a self-evolving policy is proposed by using progressive regularization. The policy evolves by using its own past knowledge to refine the suboptimal actions, thus enhancing its robustness on noisy demonstrations. Extensive experiments on various tasks show that DM outperforms other baselines substantially.
Authors: Nick Hauptvogel, Christian Igel
Abstract: Bayesian neural networks address epistemic uncertainty by learning a posterior distribution over model parameters. Sampling and weighting networks according to this posterior yields an ensemble model referred to as Bayes ensemble. Ensembles of neural networks (deep ensembles) can profit from the cancellation of errors effect: Errors by ensemble members may average out and the deep ensemble achieves better predictive performance than each individual network. We argue that neither the sampling nor the weighting in a Bayes ensemble are particularly well-suited for increasing generalization performance, as they do not support the cancellation of errors effect, which is evident in the limit from the Bernstein-von~Mises theorem for misspecified models. In contrast, a weighted average of models where the weights are optimized by minimizing a PAC-Bayesian generalization bound can improve generalization performance. This requires that the optimization takes correlations between models into account, which can be achieved by minimizing the tandem loss at the cost that hold-out data for estimating error correlations need to be available. The PAC-Bayesian weighting increases the robustness against correlated models and models with lower performance in an ensemble. This allows us to safely add several models from the same learning process to an ensemble, instead of using early-stopping for selecting a single weight configuration. Our study presents empirical results supporting these conceptual considerations on four different classification datasets. We show that state-of-the-art Bayes ensembles from the literature, despite being computationally demanding, do not improve over simple uniformly weighted deep ensembles and cannot match the performance of deep ensembles weighted by optimizing the tandem loss, which additionally come with non-vacuous generalization guarantees.
Authors: Gianluca Fabiani, Ioannis G. Kevrekidis, Constantinos Siettos, Athanasios N. Yannacopoulos
Abstract: Deep Operator Networks (DeepOnets) have revolutionized the domain of scientific machine learning for the solution of the inverse problem for dynamical systems. However, their implementation necessitates optimizing a high-dimensional space of parameters and hyperparameters. This fact, along with the requirement of substantial computational resources, poses a barrier to achieving high numerical accuracy. Here, inpsired by DeepONets and to address the above challenges, we present Random Projection-based Operator Networks (RandONets): shallow networks with random projections that learn linear and nonlinear operators. The implementation of RandONets involves: (a) incorporating random bases, thus enabling the use of shallow neural networks with a single hidden layer, where the only unknowns are the output weights of the network's weighted inner product; this reduces dramatically the dimensionality of the parameter space; and, based on this, (b) using established least-squares solvers (e.g., Tikhonov regularization and preconditioned QR decomposition) that offer superior numerical approximation properties compared to other optimization techniques used in deep-learning. In this work, we prove the universal approximation accuracy of RandONets for approximating nonlinear operators and demonstrate their efficiency in approximating linear nonlinear evolution operators (right-hand-sides (RHS)) with a focus on PDEs. We show, that for this particular task, RandONets outperform, both in terms of numerical approximation accuracy and computational cost, the ``vanilla" DeepOnets.
Authors: Yurui Lai, Xiaoyang Lin, Renchi Yang, Hongtao Wang
Abstract: In recent years, graph neural networks (GNNs) have emerged as a potent tool for learning on graph-structured data and won fruitful successes in varied fields. The majority of GNNs follow the message-passing paradigm, where representations of each node are learned by recursively aggregating features of its neighbors. However, this mechanism brings severe over-smoothing and efficiency issues over high-degree graphs (HDGs), wherein most nodes have dozens (or even hundreds) of neighbors, such as social networks, transaction graphs, power grids, etc. Additionally, such graphs usually encompass rich and complex structure semantics, which are hard to capture merely by feature aggregations in GNNs. Motivated by the above limitations, we propose TADA, an efficient and effective front-mounted data augmentation framework for GNNs on HDGs. Under the hood, TADA includes two key modules: (i) feature expansion with structure embeddings, and (ii) topology- and attribute-aware graph sparsification. The former obtains augmented node features and enhanced model capacity by encoding the graph structure into high-quality structure embeddings with our highly-efficient sketching method. Further, by exploiting task-relevant features extracted from graph structures and attributes, the second module enables the accurate identification and reduction of numerous redundant/noisy edges from the input graph, thereby alleviating over-smoothing and facilitating faster feature aggregations over HDGs. Empirically, TADA considerably improves the predictive performance of mainstream GNN models on 8 real homophilic/heterophilic HDGs in terms of node classification, while achieving efficient training and inference processes.
Authors: Xinqiang Yu, Chuanguang Yang, Chengqing Yu, Libo Huang, Zhulin An, Yongjun Xu
Abstract: Policy Distillation (PD) has become an effective method to improve deep reinforcement learning tasks. The core idea of PD is to distill policy knowledge from a teacher agent to a student agent. However, the teacher-student framework requires a well-trained teacher model which is computationally expensive.In the light of online knowledge distillation, we study the knowledge transfer between different policies that can learn diverse knowledge from the same environment.In this work, we propose Online Policy Distillation (OPD) with Decision-Attention (DA), an online learning framework in which different policies operate in the same environment to learn different perspectives of the environment and transfer knowledge to each other to obtain better performance together. With the absence of a well-performance teacher policy, the group-derived targets play a key role in transferring group knowledge to each student policy. However, naive aggregation functions tend to cause student policies quickly homogenize. To address the challenge, we introduce the Decision-Attention module to the online policies distillation framework. The Decision-Attention module can generate a distinct set of weights for each policy to measure the importance of group members. We use the Atari platform for experiments with various reinforcement learning algorithms, including PPO and DQN. In different tasks, our method can perform better than an independent training policy on both PPO and DQN algorithms. This suggests that our OPD-DA can transfer knowledge between different policies well and help agents obtain more rewards.
Authors: Hong Xiong, Feng Wu, Leon Deng, Megan Su, Li-wei H Lehman
Abstract: In the context of medical decision making, counterfactual prediction enables clinicians to predict treatment outcomes of interest under alternative courses of therapeutic actions given observed patient history. Prior machine learning approaches for counterfactual predictions under time-varying treatments focus on static time-varying treatment regimes where treatments do not depend on previous covariate history. In this work, we present G-Transformer, a Transformer-based framework supporting g-computation for counterfactual prediction under dynamic and time-varying treatment strategies. G-Transfomer captures complex, long-range dependencies in time-varying covariates using a Transformer architecture. G-Transformer estimates the conditional distribution of relevant covariates given covariate and treatment history at each time point using an encoder architecture, then produces Monte Carlo estimates of counterfactual outcomes by simulating forward patient trajectories under treatment strategies of interest. We evaluate G-Transformer extensively using two simulated longitudinal datasets from mechanistic models, and a real-world sepsis ICU dataset from MIMIC-IV. G-Transformer outperforms both classical and state-of-the-art counterfactual prediction models in these settings. To the best of our knowledge, this is the first Transformer-based architecture for counterfactual outcome prediction under dynamic and time-varying treatment strategies. Code will be released upon publication of the paper.
Authors: Dou Hu, Lingwei Wei, Wei Zhou, Songlin Hu
Abstract: This paper proposes an information-theoretic representation learning framework, named conditional information flow maximization, to extract noise-invariant sufficient representations for the input data and target task. It promotes the learned representations have good feature uniformity and sufficient predictive ability, which can enhance the generalization of pre-trained language models (PLMs) for the target task. Firstly, an information flow maximization principle is proposed to learn more sufficient representations by simultaneously maximizing both input-representation and representation-label mutual information. In contrast to information bottleneck, we handle the input-representation information in an opposite way to avoid the over-compression issue of latent representations. Besides, to mitigate the negative effect of potential redundant features, a conditional information minimization principle is designed to eliminate negative redundant features while preserve noise-invariant features from the input. Experiments on 13 language understanding benchmarks demonstrate that our method effectively improves the performance of PLMs for classification and regression. Extensive experiments show that the learned representations are more sufficient, robust and transferable.
Authors: Hengguan Huang, Xing Shen, Songtao Wang, Dianbo Liu, Hao Wang
Abstract: Faced with complex problems, the human brain demonstrates a remarkable capacity to transcend sensory input and form latent understandings of perceived world patterns. However, this cognitive capacity is not explicitly considered or encoded in current large language models (LLMs). As a result, LLMs often struggle to capture latent structures and model uncertainty in complex compositional reasoning tasks. This work introduces a novel Bayesian prompting approach that facilitates training-free Bayesian inference with LLMs by using a verbalized Probabilistic Graphical Model (PGM). While traditional Bayesian approaches typically depend on extensive data and predetermined mathematical structures for learning latent factors and dependencies, our approach efficiently reasons latent variables and their probabilistic dependencies by prompting LLMs to adhere to Bayesian principles. We evaluated our model on several compositional reasoning tasks, both close-ended and open-ended. Our results indicate that the model effectively enhances confidence elicitation and text generation quality, demonstrating its potential to improve AI language understanding systems, especially in modeling uncertainty.
Authors: Biqing Qi, Junqi Gao, Jianxing Liu, Ligang Wu, Bowen Zhou
Abstract: From the perspective of information bottleneck (IB) theory, we propose a novel framework for performing black-box transferable adversarial attacks named IBTA, which leverages advancements in invariant features. Intuitively, diminishing the reliance of adversarial perturbations on the original data, under equivalent attack performance constraints, encourages a greater reliance on invariant features that contributes most to classification, thereby enhancing the transferability of adversarial attacks. Building on this motivation, we redefine the optimization of transferable attacks using a novel theoretical framework that centers around IB. Specifically, to overcome the challenge of unoptimizable mutual information, we propose a simple and efficient mutual information lower bound (MILB) for approximating computation. Moreover, to quantitatively evaluate mutual information, we utilize the Mutual Information Neural Estimator (MINE) to perform a thorough analysis. Our experiments on the ImageNet dataset well demonstrate the efficiency and scalability of IBTA and derived MILB. Our code is available at https://github.com/Biqing-Qi/Enhancing-Adversarial-Transferability-via-Information-Bottleneck-Constraints.
Authors: Biqing Qi, Yang Luo, Junqi Gao, Pengfei Li, Kai Tian, Zhiyuan Ma, Bowen Zhou
Abstract: Deep State Space Models (SSMs) have proven effective in numerous task scenarios but face significant security challenges due to Adversarial Perturbations (APs) in real-world deployments. Adversarial Training (AT) is a mainstream approach to enhancing Adversarial Robustness (AR) and has been validated on various traditional DNN architectures. However, its effectiveness in improving the AR of SSMs remains unclear. While many enhancements in SSM components, such as integrating Attention mechanisms and expanding to data-dependent SSM parameterizations, have brought significant gains in Standard Training (ST) settings, their potential benefits in AT remain unexplored. To investigate this, we evaluate existing structural variants of SSMs with AT to assess their AR performance. We observe that pure SSM structures struggle to benefit from AT, whereas incorporating Attention yields a markedly better trade-off between robustness and generalization for SSMs in AT compared to other components. Nonetheless, the integration of Attention also leads to Robust Overfitting (RO) issues. To understand these phenomena, we empirically and theoretically analyze the output error of SSMs under AP. We find that fixed-parameterized SSMs have output error bounds strictly related to their parameters, limiting their AT benefits, while input-dependent SSMs may face the problem of error explosion. Furthermore, we show that the Attention component effectively scales the output error of SSMs during training, enabling them to benefit more from AT, but at the cost of introducing RO due to its high model complexity. Inspired by this, we propose a simple and effective Adaptive Scaling (AdS) mechanism that brings AT performance close to Attention-integrated SSMs without introducing the issue of RO.
Authors: Junqi Gao, Biqing Qi, Yao Li, Zhichang Guo, Dong Li, Yuming Xing, Dazhi Zhang
Abstract: The transferability of adversarial perturbations provides an effective shortcut for black-box attacks. Targeted perturbations have greater practicality but are more difficult to transfer between models. In this paper, we experimentally and theoretically demonstrated that neural networks trained on the same dataset have more consistent performance in High-Sample-Density-Regions (HSDR) of each class instead of low sample density regions. Therefore, in the target setting, adding perturbations towards HSDR of the target class is more effective in improving transferability. However, density estimation is challenging in high-dimensional scenarios. Further theoretical and experimental verification demonstrates that easy samples with low loss are more likely to be located in HSDR. Perturbations towards such easy samples in the target class can avoid density estimation for HSDR location. Based on the above facts, we verified that adding perturbations to easy samples in the target class improves targeted adversarial transferability of existing attack methods. A generative targeted attack strategy named Easy Sample Matching Attack (ESMA) is proposed, which has a higher success rate for targeted attacks and outperforms the SOTA generative method. Moreover, ESMA requires only 5% of the storage space and much less computation time comparing to the current SOTA, as ESMA attacks all classes with only one model instead of seperate models for each class. Our code is available at https://github.com/gjq100/ESMA.
Authors: Maryam Ghasemian, Erman Ayday
Abstract: This study investigates the optimal selection of parameters for collaborative clustering while ensuring data privacy. We focus on key clustering algorithms within a collaborative framework, where multiple data owners combine their data. A semi-trusted server assists in recommending the most suitable clustering algorithm and its parameters. Our findings indicate that the privacy parameter ($\epsilon$) minimally impacts the server's recommendations, but an increase in $\epsilon$ raises the risk of membership inference attacks, where sensitive information might be inferred. To mitigate these risks, we implement differential privacy techniques, particularly the Randomized Response mechanism, to add noise and protect data privacy. Our approach demonstrates that high-quality clustering can be achieved while maintaining data confidentiality, as evidenced by metrics such as the Adjusted Rand Index and Silhouette Score. This study contributes to privacy-aware data sharing, optimal algorithm and parameter selection, and effective communication between data owners and the server.
Authors: Yihao Zhang, Zeming Wei, Meng Sun
Abstract: In modern machine (ML) learning systems, Transformer-based architectures have achieved milestone success across a broad spectrum of tasks, yet understanding their operational mechanisms remains an open problem. To improve the transparency of ML systems, automata extraction methods, which interpret stateful ML models as automata typically through formal languages, have proven effective for explaining the mechanism of recurrent neural networks (RNNs). However, few works have been applied to this paradigm to Transformer models. In particular, understanding their processing of formal languages and identifying their limitations in this area remains unexplored. In this paper, we propose an automata extraction algorithm specifically designed for Transformer models. Treating the Transformer model as a black-box system, we track the model through the transformation process of their internal latent representations during their operations, and then use classical pedagogical approaches like L* algorithm to interpret them as deterministic finite-state automata (DFA). Overall, our study reveals how the Transformer model comprehends the structure of formal languages, which not only enhances the interpretability of the Transformer-based ML systems but also marks a crucial step toward a deeper understanding of how ML systems process formal languages. Code and data are available at https://github.com/Zhang-Yihao/Transfomer2DFA.
Authors: Zongbin Wang, Bin Pan, Zhenwei Shi
Abstract: Domain generalization aims to develop a model that can perform well on unseen target domains by learning from multiple source domains. However, recent-proposed domain generalization models usually rely on domain labels, which may not be available in many real-world scenarios. To address this challenge, we propose a Discriminant Risk Minimization (DRM) theory and the corresponding algorithm to capture the invariant features without domain labels. In DRM theory, we prove that reducing the discrepancy of prediction distribution between overall source domain and any subset of it can contribute to obtaining invariant features. To apply the DRM theory, we develop an algorithm which is composed of Bayesian inference and a new penalty termed as Categorical Discriminant Risk (CDR). In Bayesian inference, we transform the output of the model into a probability distribution to align with our theoretical assumptions. We adopt sliding update approach to approximate the overall prediction distribution of the model, which enables us to obtain CDR penalty. We also indicate the effectiveness of these components in finding invariant features. We evaluate our algorithm against various domain generalization methods on multiple real-world datasets, providing empirical support for our theory.
Authors: Zongbin Wang, Bin Pan, Shiyu Shen, Tianyang Shi, Zhenwei Shi
Abstract: Domain generalization (DG) aims to train a model from limited source domains, allowing it to generalize to unknown target domains. Typically, DG models only employ large-scale pre-trained models during the initialization of fine-tuning. However, large-scale pre-trained models already possess the ability to resist domain shift. If we reference pre-trained models continuously during fine-tuning to maintain this ability, it could further enhance the generalization ability of the DG model. For this purpose, we introduce a new method called Fine-Tune with Large-scale pre-trained Priors (FT-LP), which incorporates the pre-trained model as a prior into the DG fine-tuning process, ensuring that the model refers to its pre-trained model at each optimization step. FT-LP comprises a theoretical framework and a simple implementation strategy. In theory, we verify the rationality of FT-LP by introducing a generalization error bound with the pre-trained priors for DG. In implementation, we utilize an encoder to simulate the model distribution, enabling the use of FT-LP when only pre-trained weights are available. In summary, we offer a new fine-tuning method for DG algorithms to utilize pre-trained models throughout the fine-tuning process. Through experiments on various datasets and DG models, our proposed method exhibits significant improvements, indicating its effectiveness.
Authors: Sana Ayromlou, Teresa Tsang, Purang Abolmaesumi, Xiaoxiao Li
Abstract: In real-world clinical settings, traditional deep learning-based classification methods struggle with diagnosing newly introduced disease types because they require samples from all disease classes for offline training. Class incremental learning offers a promising solution by adapting a deep network trained on specific disease classes to handle new diseases. However, catastrophic forgetting occurs, decreasing the performance of earlier classes when adapting the model to new data. Prior proposed methodologies to overcome this require perpetual storage of previous samples, posing potential practical concerns regarding privacy and storage regulations in healthcare. To this end, we propose a novel data-free class incremental learning framework that utilizes data synthesis on learned classes instead of data storage from previous classes. Our key contributions include acquiring synthetic data known as Continual Class-Specific Impression (CCSI) for previously inaccessible trained classes and presenting a methodology to effectively utilize this data for updating networks when introducing new classes. We obtain CCSI by employing data inversion over gradients of the trained classification model on previous classes starting from the mean image of each class inspired by common landmarks shared among medical images and utilizing continual normalization layers statistics as a regularizer in this pixel-wise optimization process. Subsequently, we update the network by combining the synthesized data with new class data and incorporate several losses, including an intra-domain contrastive loss to generalize the deep network trained on the synthesized data to real data, a margin loss to increase separation among previous classes and new ones, and a cosine-normalized cross-entropy loss to alleviate the adverse effects of imbalanced distributions in training data.
Authors: Kartik Choudhary, Dhawal Gupta, Philip S. Thomas
Abstract: We present ICU-Sepsis, an environment that can be used in benchmarks for evaluating reinforcement learning (RL) algorithms. Sepsis management is a complex task that has been an important topic in applied RL research in recent years. Therefore, MDPs that model sepsis management can serve as part of a benchmark to evaluate RL algorithms on a challenging real-world problem. However, creating usable MDPs that simulate sepsis care in the ICU remains a challenge due to the complexities involved in acquiring and processing patient data. ICU-Sepsis is a lightweight environment that models personalized care of sepsis patients in the ICU. The environment is a tabular MDP that is widely compatible and is challenging even for state-of-the-art RL algorithms, making it a valuable tool for benchmarking their performance. However, we emphasize that while ICU-Sepsis provides a standardized environment for evaluating RL algorithms, it should not be used to draw conclusions that guide medical practice.
Authors: Alkis Kalavasis, Amin Karbasi, Argyris Oikonomou, Katerina Sotiraki, Grigoris Velegkas, Manolis Zampetakis
Abstract: As ML models become increasingly complex and integral to high-stakes domains such as finance and healthcare, they also become more susceptible to sophisticated adversarial attacks. We investigate the threat posed by undetectable backdoors in models developed by insidious external expert firms. When such backdoors exist, they allow the designer of the model to sell information to the users on how to carefully perturb the least significant bits of their input to change the classification outcome to a favorable one. We develop a general strategy to plant a backdoor to neural networks while ensuring that even if the model's weights and architecture are accessible, the existence of the backdoor is still undetectable. To achieve this, we utilize techniques from cryptography such as cryptographic signatures and indistinguishability obfuscation. We further introduce the notion of undetectable backdoors to language models and extend our neural network backdoor attacks to such models based on the existence of steganographic functions.
Authors: Binchuan Qi, Li Li, Wei Gong
Abstract: There remain numerous unanswered research questions on deep learning (DL) within the classical learning theory framework. These include the remarkable generalization capabilities of overparametrized neural networks (NNs), the efficient optimization performance despite non-convexity of objectives, the mechanism of flat minima in generalization, and the exceptional performance of deep architectures, among others. This paper introduces a novel theoretical learning framework known as General Distribution Learning (GD Learning), which is designed to address a comprehensive range of machine learning and statistical tasks, including classification, regression and parameter estimation. Departing from statistical machine learning, GD Learning focuses on the true underlying distribution. In GD Learning, learning error, corresponding to the expected error in classical statistical learning framework, is divided into fitting errors caused by models and fitting algorithms, as well as sampling errors introduced by limited sampling data. The framework significantly incorporates prior knowledge, especially in scenarios characterized by data scarcity. This integration of external knowledge helps to minimize learning errors across the entire dataset, thereby enhancing performance. Within the GD Learning framework, we demonstrate that the global optimal solution to non-convex optimization problems, such as minimizing fitting error, can be approached by minimizing the gradient norm and the non-uniformity of the eigenvalues of the model's Jacobian matrix. This insight has led to the development of the gradient structure control algorithm. GD Learning also offers a fresh perspective on the questions on deep learning, including overparameterization and non-convex optimizations, bias-variance trade-off, and the mechanism of flat minima.
Authors: Philip Sosnin, Mark N. M\"uller, Maximilian Baader, Calvin Tsay, Matthew Wicker
Abstract: Modern machine learning pipelines leverage large amounts of public data, making it infeasible to guarantee data quality and leaving models open to poisoning and backdoor attacks. However, provably bounding model behavior under such attacks remains an open problem. In this work, we address this challenge and develop the first framework providing provable guarantees on the behavior of models trained with potentially manipulated data. In particular, our framework certifies robustness against untargeted and targeted poisoning as well as backdoor attacks for both input and label manipulations. Our method leverages convex relaxations to over-approximate the set of all possible parameter updates for a given poisoning threat model, allowing us to bound the set of all reachable parameters for any gradient-based learning algorithm. Given this set of parameters, we provide bounds on worst-case behavior, including model performance and backdoor success rate. We demonstrate our approach on multiple real-world datasets from applications including energy consumption, medical imaging, and autonomous driving.
Authors: Ran Xu, Yiwen Lu, Chang Liu, Yong Chen, Yan Sun, Xiao Hu, Joyce C Ho, Carl Yang
Abstract: Electronic Health Records (EHRs) contain rich patient information and are crucial for clinical research and practice. In recent years, deep learning models have been applied to EHRs, but they often rely on massive features, which may not be readily available for all patients. We propose HTP-Star, which leverages hypergraph structures with a pretrain-then-finetune framework for modeling EHR data, enabling seamless integration of additional features. Additionally, we design two techniques, namely (1) Smoothness-inducing Regularization and (2) Group-balanced Reweighting, to enhance the model's robustness during fine-tuning. Through experiments conducted on two real EHR datasets, we demonstrate that HTP-Star consistently outperforms various baselines while striking a balance between patients with basic and extra features.
Authors: Qi Qi, Quanqi Hu, Qihang Lin, Tianbao Yang
Abstract: This paper studies learning fair encoders in a self-supervised learning (SSL) setting, in which all data are unlabeled and only a small portion of them are annotated with sensitive attribute. Adversarial fair representation learning is well suited for this scenario by minimizing a contrastive loss over unlabeled data while maximizing an adversarial loss of predicting the sensitive attribute over the data with sensitive attribute. Nevertheless, optimizing adversarial fair representation learning presents significant challenges due to solving a non-convex non-concave minimax game. The complexity deepens when incorporating a global contrastive loss that contrasts each anchor data point against all other examples. A central question is ``{\it can we design a provable yet efficient algorithm for solving adversarial fair self-supervised contrastive learning}?'' Building on advanced optimization techniques, we propose a stochastic algorithm dubbed SoFCLR with a convergence analysis under reasonable conditions without requring a large batch size. We conduct extensive experiments to demonstrate the effectiveness of the proposed approach for downstream classification with eight fairness notions.
Authors: Ambrus Tam\'as, Szabolcs Szentp\'eteri, Bal\'azs Csan\'ad Cs\'aji
Abstract: Stochastic multi-armed bandits (MABs) provide a fundamental reinforcement learning model to study sequential decision making in uncertain environments. The upper confidence bounds (UCB) algorithm gave birth to the renaissance of bandit algorithms, as it achieves near-optimal regret rates under various moment assumptions. Up until recently most UCB methods relied on concentration inequalities leading to confidence bounds which depend on moment parameters, such as the variance proxy, that are usually unknown in practice. In this paper, we propose a new distribution-free, data-driven UCB algorithm for symmetric reward distributions, which needs no moment information. The key idea is to combine a refined, one-sided version of the recently developed resampled median-of-means (RMM) method with UCB. We prove a near-optimal regret bound for the proposed anytime, parameter-free RMM-UCB method, even for heavy-tailed distributions.
Authors: David R Wessels, David M Knigge, Samuele Papa, Riccardo Valperga, Sharvaree Vadgama, Efstratios Gavves, Erik J Bekkers
Abstract: Recently, Neural Fields have emerged as a powerful modelling paradigm to represent continuous signals. In a conditional neural field, a field is represented by a latent variable that conditions the NeF, whose parametrisation is otherwise shared over an entire dataset. We propose Equivariant Neural Fields based on cross attention transformers, in which NeFs are conditioned on a geometric conditioning variable, a latent point cloud, that enables an equivariant decoding from latent to field. Our equivariant approach induces a steerability property by which both field and latent are grounded in geometry and amenable to transformation laws if the field transforms, the latent represents transforms accordingly and vice versa. Crucially, the equivariance relation ensures that the latent is capable of (1) representing geometric patterns faitfhully, allowing for geometric reasoning in latent space, (2) weightsharing over spatially similar patterns, allowing for efficient learning of datasets of fields. These main properties are validated using classification experiments and a verification of the capability of fitting entire datasets, in comparison to other non-equivariant NeF approaches. We further validate the potential of ENFs by demonstrate unique local field editing properties.
Authors: Zijia Song, Zelin Zang, Yelin Wang, Guozheng Yang, Jiangbin Zheng, Kaicheng yu, Wanyu Chen, Stan Z. Li
Abstract: Multimodal fusion breaks through the barriers between diverse modalities and has already yielded numerous impressive performances. However, in various specialized fields, it is struggling to obtain sufficient alignment data for the training process, which seriously limits the use of previously elegant models. Thus, semi-supervised learning attempts to achieve multimodal alignment with fewer matched pairs but traditional methods like pseudo-labeling are difficult to apply in domains with no label information. To address these problems, we transform semi-supervised multimodal alignment into a manifold matching problem and propose a new method based on CLIP, named Gentle-CLIP. Specifically, we design a novel semantic density distribution loss to explore implicit semantic alignment information from unpaired multimodal data by constraining the latent representation distribution with fine granularity, thus eliminating the need for numerous strictly matched pairs. Meanwhile, we introduce multi-kernel maximum mean discrepancy as well as self-supervised contrastive loss to pull separate modality distributions closer and enhance the stability of the representation distribution. In addition, the contrastive loss used in CLIP is employed on the supervised matched data to prevent negative optimization. Extensive experiments conducted on a range of tasks in various fields, including protein, remote sensing, and the general vision-language field, demonstrate the effectiveness of our proposed Gentle-CLIP.
Authors: Sravanti Addepalli, Priyam Dey, R. Venkatesh Babu
Abstract: The need for abundant labelled data in supervised Adversarial Training (AT) has prompted the use of Self-Supervised Learning (SSL) techniques with AT. However, the direct application of existing SSL methods to adversarial training has been sub-optimal due to the increased training complexity of combining SSL with AT. A recent approach, DeACL, mitigates this by utilizing supervision from a standard SSL teacher in a distillation setting, to mimic supervised AT. However, we find that there is still a large performance gap when compared to supervised adversarial training, specifically on larger models. In this work, investigate the key reason for this gap and propose Projected Feature Adversarial Training (ProFeAT) to bridge the same. We show that the sub-optimal distillation performance is a result of mismatch in training objectives of the teacher and student, and propose to use a projection head at the student, that allows it to leverage weak supervision from the teacher while also being able to learn adversarially robust representations that are distinct from the teacher. We further propose appropriate attack and defense losses at the feature and projector, alongside a combination of weak and strong augmentations for the teacher and student respectively, to improve the training data diversity without increasing the training complexity. Through extensive experiments on several benchmark datasets and models, we demonstrate significant improvements in both clean and robust accuracy when compared to existing SSL-AT methods, setting a new state-of-the-art. We further report on-par/ improved performance when compared to TRADES, a popular supervised-AT method.
Authors: Yinan Huang, Siqi Miao, Pan Li
Abstract: Machine learning on graphs has recently found extensive applications across domains. However, the commonly used Message Passing Neural Networks (MPNNs) suffer from limited expressive power and struggle to capture long-range dependencies. Graph transformers offer a strong alternative due to their global attention mechanism, but they come with great computational overheads, especially for large graphs. In recent years, State Space Models (SSMs) have emerged as a compelling approach to replace full attention in transformers to model sequential data. It blends the strengths of RNNs and CNNs, offering a) efficient computation, b) the ability to capture long-range dependencies, and c) good generalization across sequences of various lengths. However, extending SSMs to graph-structured data presents unique challenges due to the lack of canonical node ordering in graphs. In this work, we propose Graph State Space Convolution (GSSC) as a principled extension of SSMs to graph-structured data. By leveraging global permutation-equivariant set aggregation and factorizable graph kernels that rely on relative node distances as the convolution kernels, GSSC preserves all three advantages of SSMs. We demonstrate the provably stronger expressiveness of GSSC than MPNNs in counting graph substructures and show its effectiveness across 10 real-world, widely used benchmark datasets, where GSSC achieves best results on 7 out of 10 datasets with all significant improvements compared to the state-of-the-art baselines and second-best results on the other 3 datasets. Our findings highlight the potential of GSSC as a powerful and scalable model for graph machine learning. Our code is available at https://github.com/Graph-COM/GSSC.
Authors: Simon Schug, Seijin Kobayashi, Yassir Akram, Jo\~ao Sacramento, Razvan Pascanu
Abstract: Transformers can under some circumstances generalize to novel problem instances whose constituent parts might have been encountered during training but whose compositions have not. What mechanisms underlie this ability for compositional generalization? By reformulating multi-head attention as a hypernetwork, we reveal that a low-dimensional latent code specifies key-query specific operations. We find empirically that this latent code is highly structured, capturing information about the subtasks performed by the network. Using the framework of attention as a hypernetwork we further propose a simple modification of multi-head linear attention that strengthens the ability for compositional generalization on a range of abstract reasoning tasks. In particular, we introduce a symbolic version of the Raven Progressive Matrices human intelligence test on which we demonstrate how scaling model size and data enables compositional generalization and gives rise to a functionally structured latent code in the transformer.
Authors: Huikang Liu, Peng Wang, Longxiu Huang, Qing Qu, Laura Balzano
Abstract: We study the problem of symmetric positive semi-definite low-rank matrix completion (MC) with deterministic entry-dependent sampling. In particular, we consider rectified linear unit (ReLU) sampling, where only positive entries are observed, as well as a generalization to threshold-based sampling. We first empirically demonstrate that the landscape of this MC problem is not globally benign: Gradient descent (GD) with random initialization will generally converge to stationary points that are not globally optimal. Nevertheless, we prove that when the matrix factor with a small rank satisfies mild assumptions, the nonconvex objective function is geodesically strongly convex on the quotient manifold in a neighborhood of a planted low-rank matrix. Moreover, we show that our assumptions are satisfied by a matrix factor with i.i.d. Gaussian entries. Finally, we develop a tailor-designed initialization for GD to solve our studied formulation, which empirically always achieves convergence to the global minima. We also conduct extensive experiments and compare MC methods, investigating convergence and completion performance with respect to initialization, noise level, dimension, and rank.
Authors: Wei Li, Pin-Yu Chen, Sijia Liu, Ren Wang
Abstract: Deep neural networks are susceptible to backdoor attacks, where adversaries manipulate model predictions by inserting malicious samples into the training data. Currently, there is still a lack of direct filtering methods for identifying suspicious training data to unveil potential backdoor samples. In this paper, we propose a novel method, Prediction Shift Backdoor Detection (PSBD), leveraging an uncertainty-based approach requiring minimal unlabeled clean validation data. PSBD is motivated by an intriguing Prediction Shift (PS) phenomenon, where poisoned models' predictions on clean data often shift away from true labels towards certain other labels with dropout applied during inference, while backdoor samples exhibit less PS. We hypothesize PS results from neuron bias effect, making neurons favor features of certain classes. PSBD identifies backdoor training samples by computing the Prediction Shift Uncertainty (PSU), the variance in probability values when dropout layers are toggled on and off during model inference. Extensive experiments have been conducted to verify the effectiveness and efficiency of PSBD, which achieves state-of-the-art results among mainstream detection methods.
Authors: Xinshu Li, Mingling Gong, Lina Yao
Abstract: The advancements in disentangled representation learning significantly enhance the accuracy of counterfactual predictions by granting precise control over instrumental variables, confounders, and adjustable variables. An appealing method for achieving the independent separation of these factors is mutual information minimization, a task that presents challenges in numerous machine learning scenarios, especially within high-dimensional spaces. To circumvent this challenge, we propose the Self-Distilled Disentanglement framework, referred to as $SD^2$. Grounded in information theory, it ensures theoretically sound independent disentangled representations without intricate mutual information estimator designs for high-dimensional representations. Our comprehensive experiments, conducted on both synthetic and real-world datasets, confirms the effectiveness of our approach in facilitating counterfactual inference in the presence of both observed and unobserved confounders.
Authors: Shreyas Basavatia, Keerthiram Murugesan, Shivam Ratnakar
Abstract: Interactive fiction games have emerged as an important application to improve the generalization capabilities of language-based reinforcement learning (RL) agents. Existing environments for interactive fiction games are domain-specific or time-consuming to generate and do not train the RL agents to master a specific set of skills. In this work, we introduce an interactive environment for self-supervised RL, STARLING, for text-based games that bootstraps the text-based RL agents with automatically generated games (based on the seed set of game ideas) to boost the performance and generalization capabilities to reach a goal of the target environment. These games let the agent hone their skills on a predefined set of tasks. We create and test an environment with 100 games, generated using this automated framework that uses large language models (GPT-3) and an interactive fiction game engine (based on Inform7) to provide the user with the ability to generate more games under minimal human supervision. Experimental results based on both the human participants and baseline text-based RL agents reveal that current state-of-the-art text-based RL agents cannot use previously learned skills in new situations at the level humans can. These results enforce STARLING's potential to serve as a sandbox environment for further research in self-supervised text-based RL.
Authors: Utsav Singh, Pramit Bhattacharyya, Vinay P. Namboodiri
Abstract: Developing interactive systems that leverage natural language instructions to solve complex robotic control tasks has been a long-desired goal in the robotics community. Large Language Models (LLMs) have demonstrated exceptional abilities in handling complex tasks, including logical reasoning, in-context learning, and code generation. However, predicting low-level robotic actions using LLMs poses significant challenges. Additionally, the complexity of such tasks usually demands the acquisition of policies to execute diverse subtasks and combine them to attain the ultimate objective. Hierarchical Reinforcement Learning (HRL) is an elegant approach for solving such tasks, which provides the intuitive benefits of temporal abstraction and improved exploration. However, HRL faces the recurring issue of non-stationarity due to unstable lower primitive behaviour. In this work, we propose LGR2, a novel HRL framework that leverages language instructions to generate a stationary reward function for the higher-level policy. Since the language-guided reward is unaffected by the lower primitive behaviour, LGR2 mitigates non-stationarity and is thus an elegant method for leveraging language instructions to solve robotic control tasks. To analyze the efficacy of our approach, we perform empirical analysis and demonstrate that LGR2 effectively alleviates non-stationarity in HRL. Our approach attains success rates exceeding 70$\%$ in challenging, sparse-reward robotic navigation and manipulation environments where the baselines fail to achieve any significant progress. Additionally, we conduct real-world robotic manipulation experiments and demonstrate that CRISP shows impressive generalization in real-world scenarios.
Authors: Igor Melnyk, Youssef Mroueh, Brian Belgodere, Mattia Rigotti, Apoorva Nitsure, Mikhail Yurochkin, Kristjan Greenewald, Jiri Navratil, Jerret Ross
Abstract: Current LLM alignment techniques use pairwise human preferences at a sample level, and as such, they do not imply an alignment on the distributional level. We propose in this paper Alignment via Optimal Transport (AOT), a novel method for distributional preference alignment of LLMs. AOT aligns LLMs on unpaired preference data by making the reward distribution of the positive samples stochastically dominant in the first order on the distribution of negative samples. We introduce a convex relaxation of this first-order stochastic dominance and cast it as an optimal transport problem with a smooth and convex cost. Thanks to the one-dimensional nature of the resulting optimal transport problem and the convexity of the cost, it has a closed-form solution via sorting on empirical measures. We fine-tune LLMs with this AOT objective, which enables alignment by penalizing the violation of the stochastic dominance of the reward distribution of the positive samples on the reward distribution of the negative samples. We analyze the sample complexity of AOT by considering the dual of the OT problem and show that it converges at the parametric rate. Empirically, we show on a diverse set of alignment datasets and LLMs that AOT leads to state-of-the-art models in the 7B family of models when evaluated with Open LLM Benchmarks and AlpacaEval.
Authors: Youssef Mroueh
Abstract: Policy alignment of large language models refers to constrained policy optimization, where the policy is optimized to maximize a reward while staying close to a reference policy with respect to an $f$-divergence such as the $\mathsf{KL}$ divergence. The best of $n$ alignment policy selects a sample from the reference policy that has the maximum reward among $n$ independent samples. For both cases (policy alignment and best of $n$), recent works showed empirically that the reward improvement of the aligned policy on the reference one scales like $\sqrt{\mathsf{KL}}$, with an explicit bound in $n$ on the $\mathsf{KL}$ for the best of $n$ policy. We show in this paper that the $\sqrt{\mathsf{KL}}$ information theoretic upper bound holds if the reward under the reference policy has sub-gaussian tails. Moreover, we prove for the best of $n$ policy, that the $\mathsf{KL}$ upper bound can be obtained for any $f$-divergence via a reduction to exponential order statistics owing to the R\'enyi representation of order statistics, and a data processing inequality. If additional information is known on the tails of the aligned policy we show that tighter control on the reward improvement can be obtained via the R\'enyi divergence. Finally we demonstrate how these upper bounds transfer from proxy rewards to golden rewards which results in a decrease in the golden reward improvement due to overestimation and approximation errors of the proxy reward.
Authors: Georgios Tsoumplekas, Christos L. Athanasiadis, Dimitrios I. Doukas, Antonios Chrysopoulos, Pericles A. Mitkas
Abstract: Despite the rapid expansion of smart grids and large volumes of data at the individual consumer level, there are still various cases where adequate data collection to train accurate load forecasting models is challenging or even impossible. This paper proposes adapting an established model-agnostic meta-learning algorithm for short-term load forecasting in the context of few-shot learning. Specifically, the proposed method can rapidly adapt and generalize within any unknown load time series of arbitrary length using only minimal training samples. In this context, the meta-learning model learns an optimal set of initial parameters for a base-level learner recurrent neural network. The proposed model is evaluated using a dataset of historical load consumption data from real-world consumers. Despite the examined load series' short length, it produces accurate forecasts outperforming transfer learning and task-specific machine learning methods by $12.5\%$. To enhance robustness and fairness during model evaluation, a novel metric, mean average log percentage error, is proposed that alleviates the bias introduced by the commonly used MAPE metric. Finally, a series of studies to evaluate the model's robustness under different hyperparameters and time series lengths is also conducted, demonstrating that the proposed approach consistently outperforms all other models.
Authors: Harrison Lam, Yuanjie Chen, Noboru Kanazawa, Mohammad Chowdhury, Anna Battista, Stephan Waldert
Abstract: We explored the challenge of predicting and explaining the occurrence of events within sequences of data points. Our focus was particularly on scenarios in which unknown triggers causing the occurrence of events may consist of non-consecutive, masked, noisy data points. This scenario is akin to an agent tasked with learning to predict and explain the occurrence of events without understanding the underlying processes or having access to crucial information. Such scenarios are encountered across various fields, such as genomics, hardware and software verification, and financial time series prediction. We combined analytical, simulation, and machine learning (ML) approaches to investigate, quantify, and provide solutions to this challenge. We deduced and validated equations generally applicable to any variation of the underlying challenge. Using these equations, we (1) described how the level of complexity changes with various parameters (e.g., number of apparent and hidden states, trigger length, confidence, etc.) and (2) quantified the data needed to successfully train an ML model. We then (3) proved our ML solution learns and subsequently identifies unknown triggers and predicts the occurrence of events. If the complexity of the challenge is too high, our ML solution can identify trigger candidates to be used to interactively probe the system under investigation to determine the true trigger in a way considerably more efficient than brute force methods. By sharing our findings, we aim to assist others grappling with similar challenges, enabling estimates on the complexity of their problem, the data required and a solution to solve it.
Authors: Harish Haresamudram, Hrudhai Rajasekhar, Nikhil Murlidhar Shanbhogue, Thomas Ploetz
Abstract: The astonishing success of Large Language Models (LLMs) in Natural Language Processing (NLP) has spurred their use in many application domains beyond text analysis, including wearable sensor-based Human Activity Recognition (HAR). In such scenarios, often sensor data are directly fed into an LLM along with text instructions for the model to perform activity classification. Seemingly remarkable results have been reported for such LLM-based HAR systems when they are evaluated on standard benchmarks from the field. Yet, we argue, care has to be taken when evaluating LLM-based HAR systems in such a traditional way. Most contemporary LLMs are trained on virtually the entire (accessible) internet -- potentially including standard HAR datasets. With that, it is not unlikely that LLMs actually had access to the test data used in such benchmark experiments.The resulting contamination of training data would render these experimental evaluations meaningless. In this paper we investigate whether LLMs indeed have had access to standard HAR datasets during training. We apply memorization tests to LLMs, which involves instructing the models to extend given snippets of data. When comparing the LLM-generated output to the original data we found a non-negligible amount of matches which suggests that the LLM under investigation seems to indeed have seen wearable sensor data from the benchmark datasets during training. For the Daphnet dataset in particular, GPT-4 is able to reproduce blocks of sensor readings. We report on our investigations and discuss potential implications on HAR research, especially with regards to reporting results on experimental evaluation
Authors: Emilia Agis Lerner, Florian E. Dorner, Elliott Ash, Naman Goel
Abstract: There is a growing body of work on learning from human feedback to align various aspects of machine learning systems with human values and preferences. We consider the setting of fairness in content moderation, in which human feedback is used to determine how two comments -- referencing different sensitive attribute groups -- should be treated in comparison to one another. With a novel dataset collected from Prolific and MTurk, we find significant gaps in fairness preferences depending on the race, age, political stance, educational level, and LGBTQ+ identity of annotators. We also demonstrate that demographics mentioned in text have a strong influence on how users perceive individual fairness in moderation. Further, we find that differences also exist in downstream classifiers trained to predict human preferences. Finally, we observe that an ensemble, giving equal weight to classifiers trained on annotations from different demographics, performs better for different demographic intersections; compared to a single classifier that gives equal weight to each annotation.
Authors: Burak Var{\i}c{\i}, Emre Acart\"urk, Karthikeyan Shanmugam, Ali Tajer
Abstract: Despite the multifaceted recent advances in interventional causal representation learning (CRL), they primarily focus on the stylized assumption of single-node interventions. This assumption is not valid in a wide range of applications, and generally, the subset of nodes intervened in an interventional environment is fully unknown. This paper focuses on interventional CRL under unknown multi-node (UMN) interventional environments and establishes the first identifiability results for general latent causal models (parametric or nonparametric) under stochastic interventions (soft or hard) and linear transformation from the latent to observed space. Specifically, it is established that given sufficiently diverse interventional environments, (i) identifiability up to ancestors is possible using only soft interventions, and (ii) perfect identifiability is possible using hard interventions. Remarkably, these guarantees match the best-known results for more restrictive single-node interventions. Furthermore, CRL algorithms are also provided that achieve the identifiability guarantees. A central step in designing these algorithms is establishing the relationships between UMN interventional CRL and score functions associated with the statistical models of different interventional environments. Establishing these relationships also serves as constructive proof of the identifiability guarantees.
Authors: Ziang Chen, Xiaohan Chen, Jialin Liu, Xinshang Wang, Wotao Yin
Abstract: Quadratic programming (QP) is the most widely applied category of problems in nonlinear programming. Many applications require real-time/fast solutions, though not necessarily with high precision. Existing methods either involve matrix decomposition or use the preconditioned conjugate gradient method. For relatively large instances, these methods cannot achieve the real-time requirement unless there is an effective precondition. Recently, graph neural networks (GNNs) opened new possibilities for QP. Some promising empirical studies of applying GNNs for QP tasks show that GNNs can capture key characteristics of an optimization instance and provide adaptive guidance accordingly to crucial configurations during the solving process, or directly provide an approximate solution. Despite notable empirical observations, theoretical foundations are still lacking. In this work, we investigate the expressive or representative power of GNNs, a crucial aspect of neural network theory, specifically in the context of QP tasks, with both continuous and mixed-integer settings. We prove the existence of message-passing GNNs that can reliably represent key properties of quadratic programs, including feasibility, optimal objective value, and optimal solution. Our theory is validated by numerical results.
Authors: Sobhan Mohammadpour, Emma Frejinger, Pierre-Luc Bacon
Abstract: Regularized reinforcement learning (RL), particularly the entropy-regularized kind, has gained traction in optimal control and inverse RL. While standard unregularized RL methods remain unaffected by changes in the number of actions, we show that it can severely impact their regularized counterparts. This paper demonstrates the importance of decoupling the regularizer from the action space: that is, to maintain a consistent level of regularization regardless of how many actions are involved to avoid over-regularization. Whereas the problem can be avoided by introducing a task-specific temperature parameter, it is often undesirable and cannot solve the problem when action spaces are state-dependent. In the state-dependent action context, different states with varying action spaces are regularized inconsistently. We introduce two solutions: a static temperature selection approach and a dynamic counterpart, universally applicable where this problem arises. Implementing these changes improves performance on the DeepMind control suite in static and dynamic temperature regimes and a biological sequence design task.
Authors: Yixin Song, Haotong Xie, Zhengyan Zhang, Bo Wen, Li Ma, Zeyu Mi, Haibo Chen
Abstract: Exploiting activation sparsity is a promising approach to significantly accelerating the inference process of large language models (LLMs) without compromising performance. However, activation sparsity is determined by activation functions, and commonly used ones like SwiGLU and GeGLU exhibit limited sparsity. Simply replacing these functions with ReLU fails to achieve sufficient sparsity. Moreover, inadequate training data can further increase the risk of performance degradation. To address these challenges, we propose a novel dReLU function, which is designed to improve LLM activation sparsity, along with a high-quality training data mixture ratio to facilitate effective sparsification. Additionally, we leverage sparse activation patterns within the Feed-Forward Network (FFN) experts of Mixture-of-Experts (MoE) models to further boost efficiency. By applying our neuron sparsification method to the Mistral and Mixtral models, only 2.5 billion and 4.3 billion parameters are activated per inference iteration, respectively, while achieving even more powerful model performance. Evaluation results demonstrate that this sparsity achieves a 2-5x decoding speedup. Remarkably, on mobile phones, our TurboSparse-Mixtral-47B achieves an inference speed of 11 tokens per second. Our models are available at \url{https://huggingface.co/PowerInfer}
Authors: Alexandre Hayderi, Amin Saberi, Ellen Vitercik, Anders Wikum
Abstract: Online Bayesian bipartite matching is a central problem in digital marketplaces and exchanges, including advertising, crowdsourcing, ridesharing, and kidney exchange. We introduce a graph neural network (GNN) approach that emulates the problem's combinatorially-complex optimal online algorithm, which selects actions (e.g., which nodes to match) by computing each action's value-to-go (VTG) -- the expected weight of the final matching if the algorithm takes that action, then acts optimally in the future. We train a GNN to estimate VTG and show empirically that this GNN returns high-weight matchings across a variety of tasks. Moreover, we identify a common family of graph distributions in spatial crowdsourcing applications, such as rideshare, under which VTG can be efficiently approximated by aggregating information within local neighborhoods in the graphs. This structure matches the local behavior of GNNs, providing theoretical justification for our approach.
Authors: Haoran You, Yipin Guo, Yichao Fu, Wei Zhou, Huihong Shi, Xiaofan Zhang, Souvik Kundu, Amir Yazdanbakhsh, Yingyan Lin
Abstract: Large language models (LLMs) have shown impressive performance on language tasks but face challenges when deployed on resource-constrained devices due to their extensive parameters and reliance on dense multiplications, resulting in high memory demands and latency bottlenecks. Shift-and-add reparameterization offers a promising solution by replacing costly multiplications with hardware-friendly primitives in both the attention and multi-layer perceptron (MLP) layers of an LLM. However, current reparameterization techniques require training from scratch or full parameter fine-tuning to restore accuracy, which is resource-intensive for LLMs. To address this, we propose accelerating pretrained LLMs through post-training shift-and-add reparameterization, creating efficient multiplication-free models, dubbed ShiftAddLLM. Specifically, we quantize each weight matrix into binary matrices paired with group-wise scaling factors. The associated multiplications are reparameterized into (1) shifts between activations and scaling factors and (2) queries and adds according to the binary matrices. To reduce accuracy loss, we present a multi-objective optimization method to minimize both weight and output activation reparameterization errors. Additionally, based on varying sensitivity across layers to reparameterization, we develop an automated bit allocation strategy to further reduce memory usage and latency. Experiments on five LLM families and eight tasks consistently validate the effectiveness of ShiftAddLLM, achieving average perplexity improvements of 5.6 and 22.7 points at comparable or lower latency compared to the most competitive quantized LLMs at 3 and 2 bits, respectively, and more than 80% memory and energy reductions over the original LLMs. Codes and models are available at https://github.com/GATECH-EIC/ShiftAddLLM.
Authors: Yusif Ibrahimov, Tarique Anwar, Tommy Yuan
Abstract: Mental health constitutes a complex and pervasive global challenge, affecting millions of lives and often leading to severe consequences. In this paper, we conduct a thorough survey to explore the intersection of data science, artificial intelligence, and mental healthcare, focusing on the recent developments of mental disorder detection through online social media (OSM). A significant portion of the population actively engages in OSM platforms, creating a vast repository of personal data that holds immense potential for mental health analytics. The paper navigates through traditional diagnostic methods, state-of-the-art data- and AI-driven research studies, and the emergence of explainable AI (XAI) models for mental healthcare. We review state-of-the-art machine learning methods, particularly those based on modern deep learning, while emphasising the need for explainability in healthcare AI models. The experimental design section provides insights into prevalent practices, including available datasets and evaluation approaches. We also identify key issues and challenges in the field and propose promising future research directions. As mental health decisions demand transparency, interpretability, and ethical considerations, this paper contributes to the ongoing discourse on advancing XAI in mental healthcare through social media. The comprehensive overview presented here aims to guide researchers, practitioners, and policymakers in developing the area of mental disorder detection.
Authors: Takayuki Osa, Tatsuya Harada
Abstract: Recent studies on online reinforcement learning (RL) have demonstrated the advantages of learning multiple behaviors from a single task, as in the case of few-shot adaptation to a new environment. Although this approach is expected to yield similar benefits in offline RL, appropriate methods for learning multiple solutions have not been fully investigated in previous studies. In this study, we therefore addressed the problem of finding multiple solutions from a single task in offline RL. We propose algorithms that can learn multiple solutions in offline RL, and empirically investigate their performance. Our experimental results show that the proposed algorithm learns multiple qualitatively and quantitatively distinctive solutions in offline RL.
Authors: Zhen Qin, Zhihui Zhu
Abstract: Recently, a tensor-on-tensor (ToT) regression model has been proposed to generalize tensor recovery, encompassing scenarios like scalar-on-tensor regression and tensor-on-vector regression. However, the exponential growth in tensor complexity poses challenges for storage and computation in ToT regression. To overcome this hurdle, tensor decompositions have been introduced, with the tensor train (TT)-based ToT model proving efficient in practice due to reduced memory requirements, enhanced computational efficiency, and decreased sampling complexity. Despite these practical benefits, a disparity exists between theoretical analysis and real-world performance. In this paper, we delve into the theoretical and algorithmic aspects of the TT-based ToT regression model. Assuming the regression operator satisfies the restricted isometry property (RIP), we conduct an error analysis for the solution to a constrained least-squares optimization problem. This analysis includes upper error bound and minimax lower bound, revealing that such error bounds polynomially depend on the order $N+M$. To efficiently find solutions meeting such error bounds, we propose two optimization algorithms: the iterative hard thresholding (IHT) algorithm (employing gradient descent with TT-singular value decomposition (TT-SVD)) and the factorization approach using the Riemannian gradient descent (RGD) algorithm. When RIP is satisfied, spectral initialization facilitates proper initialization, and we establish the linear convergence rate of both IHT and RGD.
Authors: Peng Xia, Ze Chen, Juanxi Tian, Yangrui Gong, Ruibo Hou, Yue Xu, Zhenbang Wu, Zhiyuan Fan, Yiyang Zhou, Kangyu Zhu, Wenhao Zheng, Zhaoyang Wang, Xiao Wang, Xuchao Zhang, Chetan Bansal, Marc Niethammer, Junzhou Huang, Hongtu Zhu, Yun Li, Jimeng Sun, Zongyuan Ge, Gang Li, James Zou, Huaxiu Yao
Abstract: Artificial intelligence has significantly impacted medical applications, particularly with the advent of Medical Large Vision Language Models (Med-LVLMs), sparking optimism for the future of automated and personalized healthcare. However, the trustworthiness of Med-LVLMs remains unverified, posing significant risks for future model deployment. In this paper, we introduce CARES and aim to comprehensively evaluate the Trustworthiness of Med-LVLMs across the medical domain. We assess the trustworthiness of Med-LVLMs across five dimensions, including trustfulness, fairness, safety, privacy, and robustness. CARES comprises about 41K question-answer pairs in both closed and open-ended formats, covering 16 medical image modalities and 27 anatomical regions. Our analysis reveals that the models consistently exhibit concerns regarding trustworthiness, often displaying factual inaccuracies and failing to maintain fairness across different demographic groups. Furthermore, they are vulnerable to attacks and demonstrate a lack of privacy awareness. We publicly release our benchmark and code in https://github.com/richard-peng-xia/CARES.
Authors: Zewen Liu, Yunxiao Li, Mingyang Wei, Guancheng Wan, Max S. Y. Lau, Wei Jin
Abstract: EpiLearn is a Python toolkit developed for modeling, simulating, and analyzing epidemic data. Although there exist several packages that also deal with epidemic modeling, they are often restricted to mechanistic models or traditional statistical tools. As machine learning continues to shape the world, the gap between these packages and the latest models has become larger. To bridge the gap and inspire innovative research in epidemic modeling, EpiLearn not only provides support for evaluating epidemic models based on machine learning, but also incorporates comprehensive tools for analyzing epidemic data, such as simulation, visualization, transformations, etc. For the convenience of both epidemiologists and data scientists, we provide a unified framework for training and evaluation of epidemic models on two tasks: Forecasting and Source Detection. To facilitate the development of new models, EpiLearn follows a modular design, making it flexible and easy to use. In addition, an interactive web application is also developed to visualize the real-world or simulated epidemic data. Our package is available at https://github.com/Emory-Melody/EpiLearn.
Authors: Da Zheng, Xiang Song, Qi Zhu, Jian Zhang, Theodore Vasiloudis, Runjie Ma, Houyu Zhang, Zichen Wang, Soji Adeshina, Israt Nisa, Alejandro Mottini, Qingjun Cui, Huzefa Rangwala, Belinda Zeng, Christos Faloutsos, George Karypis
Abstract: Graph machine learning (GML) is effective in many business applications. However, making GML easy to use and applicable to industry applications with massive datasets remain challenging. We developed GraphStorm, which provides an end-to-end solution for scalable graph construction, graph model training and inference. GraphStorm has the following desirable properties: (a) Easy to use: it can perform graph construction and model training and inference with just a single command; (b) Expert-friendly: GraphStorm contains many advanced GML modeling techniques to handle complex graph data and improve model performance; (c) Scalable: every component in GraphStorm can operate on graphs with billions of nodes and can scale model training and inference to different hardware without changing any code. GraphStorm has been used and deployed for over a dozen billion-scale industry applications after its release in May 2023. It is open-sourced in Github: https://github.com/awslabs/graphstorm.
Authors: Donghu Kim, Hojoon Lee, Kyungmin Lee, Dongyoon Hwang, Jaegul Choo
Abstract: Recently, various pre-training methods have been introduced in vision-based Reinforcement Learning (RL). However, their generalization ability remains unclear due to evaluations being limited to in-distribution environments and non-unified experimental setups. To address this, we introduce the Atari Pre-training Benchmark (Atari-PB), which pre-trains a ResNet-50 model on 10 million transitions from 50 Atari games and evaluates it across diverse environment distributions. Our experiments show that pre-training objectives focused on learning task-agnostic features (e.g., identifying objects and understanding temporal dynamics) enhance generalization across different environments. In contrast, objectives focused on learning task-specific knowledge (e.g., identifying agents and fitting reward functions) improve performance in environments similar to the pre-training dataset but not in varied ones. We publicize our codes, datasets, and model checkpoints at https://github.com/dojeon-ai/Atari-PB.
Authors: Zeyi Xu, Yifei Li
Abstract: Machine learning methods for physical simulation have achieved significant success in recent years. We propose a new universal architecture based on Graph Neural Network, the Message Passing Transformer, which incorporates a Message Passing framework, employs an Encoder-Processor-Decoder structure, and applies Graph Fourier Loss as loss function for model optimization. To take advantage of the past message passing state information, we propose Hadamard-Product Attention to update the node attribute in the Processor, Hadamard-Product Attention is a variant of Dot-Product Attention that focuses on more fine-grained semantics and emphasizes on assigning attention weights over each feature dimension rather than each position in the sequence relative to others. We further introduce Graph Fourier Loss (GFL) to balance high-energy and low-energy components. To improve time performance, we precompute the graph's Laplacian eigenvectors before the training process. Our architecture achieves significant accuracy improvements in long-term rollouts for both Lagrangian and Eulerian dynamical systems over current methods.
Authors: Fudong Lin, Kaleb Guillot, Summer Crawford, Yihe Zhang, Xu Yuan, Nian-Feng Tzeng
Abstract: Precise crop yield predictions are of national importance for ensuring food security and sustainable agricultural practices. While AI-for-science approaches have exhibited promising achievements in solving many scientific problems such as drug discovery, precipitation nowcasting, etc., the development of deep learning models for predicting crop yields is constantly hindered by the lack of an open and large-scale deep learning-ready dataset with multiple modalities to accommodate sufficient information. To remedy this, we introduce the CropNet dataset, the first terabyte-sized, publicly available, and multi-modal dataset specifically targeting climate change-aware crop yield predictions for the contiguous United States (U.S.) continent at the county level. Our CropNet dataset is composed of three modalities of data, i.e., Sentinel-2 Imagery, WRF-HRRR Computed Dataset, and USDA Crop Dataset, for over 2200 U.S. counties spanning 6 years (2017-2022), expected to facilitate researchers in developing versatile deep learning models for timely and precisely predicting crop yields at the county-level, by accounting for the effects of both short-term growing season weather variations and long-term climate change on crop yields. Besides, we develop the CropNet package, offering three types of APIs, for facilitating researchers in downloading the CropNet data on the fly over the time and region of interest, and flexibly building their deep learning models for accurate crop yield predictions. Extensive experiments have been conducted on our CropNet dataset via employing various types of deep learning solutions, with the results validating the general applicability and the efficacy of the CropNet dataset in climate change-aware crop yield predictions.
Authors: Pierre-Fran\c{c}ois Massiani, Sebastian Trimpe, Friedrich Solowjow
Abstract: The consistency of a learning method is usually established under the assumption that the observations are a realization of an independent and identically distributed (i.i.d.) or mixing process. Yet, kernel methods such as support vector machines (SVMs), Gaussian processes, or conditional kernel mean embeddings (CKMEs) all give excellent performance under sampling schemes that are obviously non-i.i.d., such as when data comes from a dynamical system. We propose the new notion of empirical weak convergence (EWC) as a general assumption explaining such phenomena for kernel methods. It assumes the existence of a random asymptotic data distribution and is a strict weakening of previous assumptions in the field. Our main results then establish consistency of SVMs, kernel mean embeddings, and general Hilbert-space valued empirical expectations with EWC data. Our analysis holds for both finite- and infinite-dimensional outputs, as we extend classical results of statistical learning to the latter case. In particular, it is also applicable to CKMEs. Overall, our results open new classes of processes to statistical learning and can serve as a foundation for a theory of learning beyond i.i.d. and mixing.
Authors: Lucas Slot, Stefan Tiegel, Manuel Wiedmer
Abstract: Rubinfeld & Vasilyan recently introduced the framework of testable learning as an extension of the classical agnostic model. It relaxes distributional assumptions which are difficult to verify by conditions that can be checked efficiently by a tester. The tester has to accept whenever the data truly satisfies the original assumptions, and the learner has to succeed whenever the tester accepts. We focus on the setting where the tester has to accept standard Gaussian data. There, it is known that basic concept classes such as halfspaces can be learned testably with the same time complexity as in the (distribution-specific) agnostic model. In this work, we ask whether there is a price to pay for testably learning more complex concept classes. In particular, we consider polynomial threshold functions (PTFs), which naturally generalize halfspaces. We show that PTFs of arbitrary constant degree can be testably learned up to excess error $\varepsilon > 0$ in time $n^{\mathrm{poly}(1/\varepsilon)}$. This qualitatively matches the best known guarantees in the agnostic model. Our results build on a connection between testable learning and fooling. In particular, we show that distributions that approximately match at least $\mathrm{poly}(1/\varepsilon)$ moments of the standard Gaussian fool constant-degree PTFs (up to error $\varepsilon$). As a secondary result, we prove that a direct approach to show testable learning (without fooling), which was successfully used for halfspaces, cannot work for PTFs.
Authors: Xiang Li, Jiexi Liu, Xinrui Wang, Songcan Chen
Abstract: In reality, data often exhibit associations with multiple labels, making multi-label learning (MLL) become a prominent research topic. The last two decades have witnessed the success of MLL, which is indispensable from complete and accurate supervised information. However, obtaining such information in practice is always laborious and sometimes even impossible. To circumvent this dilemma, incomplete multi-label learning (InMLL) has emerged, aiming to learn from incomplete labeled data. To date, enormous InMLL works have been proposed to narrow the performance gap with complete MLL, whereas a systematic review for InMLL is still absent. In this paper, we not only attempt to fill the lacuna but also strive to pave the way for innovative research. Specifically, we retrospect the origin of InMLL, analyze the challenges of InMLL, and make a taxonomy of InMLL from the data-oriented and algorithm-oriented perspectives, respectively. Besides, we also present real applications of InMLL in various domains. More importantly, we highlight several potential future trends, including four open problems that are more in line with practice and three under-explored/unexplored techniques in addressing the challenges of InMLL, which may shed new light on developing novel research directions in the field of InMLL.
Authors: Yujee Song, Donghyun Lee, Rui Meng, Won Hwa Kim
Abstract: A Marked Temporal Point Process (MTPP) is a stochastic process whose realization is a set of event-time data. MTPP is often used to understand complex dynamics of asynchronous temporal events such as money transaction, social media, healthcare, etc. Recent studies have utilized deep neural networks to capture complex temporal dependencies of events and generate embedding that aptly represent the observed events. While most previous studies focus on the inter-event dependencies and their representations, how individual events influence the overall dynamics over time has been under-explored. In this regime, we propose a Decoupled MTPP framework that disentangles characterization of a stochastic process into a set of evolving influences from different events. Our approach employs Neural Ordinary Differential Equations (Neural ODEs) to learn flexible continuous dynamics of these influences while simultaneously addressing multiple inference problems, such as density estimation and survival rate computation. We emphasize the significance of disentangling the influences by comparing our framework with state-of-the-art methods on real-life datasets, and provide analysis on the model behavior for potential applications.
Authors: Kim A. Nicoli, Christopher J. Anders, Lena Funcke, Tobias Hartung, Karl Jansen, Stefan K\"uhn, Klaus-Robert M\"uller, Paolo Stornati, Pan Kessel, Shinichi Nakajima
Abstract: In this paper, we propose a novel and powerful method to harness Bayesian optimization for Variational Quantum Eigensolvers (VQEs) -- a hybrid quantum-classical protocol used to approximate the ground state of a quantum Hamiltonian. Specifically, we derive a VQE-kernel which incorporates important prior information about quantum circuits: the kernel feature map of the VQE-kernel exactly matches the known functional form of the VQE's objective function and thereby significantly reduces the posterior uncertainty. Moreover, we propose a novel acquisition function for Bayesian optimization called Expected Maximum Improvement over Confident Regions (EMICoRe) which can actively exploit the inductive bias of the VQE-kernel by treating regions with low predictive uncertainty as indirectly ``observed''. As a result, observations at as few as three points in the search domain are sufficient to determine the complete objective function along an entire one-dimensional subspace of the optimization landscape. Our numerical experiments demonstrate that our approach improves over state-of-the-art baselines.
Authors: Daniel Kunin, Allan Ravent\'os, Cl\'ementine Domin\'e, Feng Chen, David Klindt, Andrew Saxe, Surya Ganguli
Abstract: While the impressive performance of modern neural networks is often attributed to their capacity to efficiently extract task-relevant features from data, the mechanisms underlying this rich feature learning regime remain elusive, with much of our theoretical understanding stemming from the opposing lazy regime. In this work, we derive exact solutions to a minimal model that transitions between lazy and rich learning, precisely elucidating how unbalanced layer-specific initialization variances and learning rates determine the degree of feature learning. Our analysis reveals that they conspire to influence the learning regime through a set of conserved quantities that constrain and modify the geometry of learning trajectories in parameter and function space. We extend our analysis to more complex linear models with multiple neurons, outputs, and layers and to shallow nonlinear networks with piecewise linear activation functions. In linear networks, rapid feature learning only occurs with balanced initializations, where all layers learn at similar speeds. While in nonlinear networks, unbalanced initializations that promote faster learning in earlier layers can accelerate rich learning. Through a series of experiments, we provide evidence that this unbalanced rich regime drives feature learning in deep finite-width networks, promotes interpretability of early layers in CNNs, reduces the sample complexity of learning hierarchical data, and decreases the time to grokking in modular arithmetic. Our theory motivates further exploration of unbalanced initializations to enhance efficient feature learning.
Authors: Zuzanna Fendor, Bas H. M. van der Velden, Xinxin Wang, Andrea Jr. Carnoli, Osman Mutlu, Ali H\"urriyeto\u{g}lu
Abstract: Research in the food domain is at times limited due to data sharing obstacles, such as data ownership, privacy requirements, and regulations. While important, these obstacles can restrict data-driven methods such as machine learning. Federated learning, the approach of training models on locally kept data and only sharing the learned parameters, is a potential technique to alleviate data sharing obstacles. This systematic review investigates the use of federated learning within the food domain, structures included papers in a federated learning framework, highlights knowledge gaps, and discusses potential applications. A total of 41 papers were included in the review. The current applications include solutions to water and milk quality assessment, cybersecurity of water processing, pesticide residue risk analysis, weed detection, and fraud detection, focusing on centralized horizontal federated learning. One of the gaps found was the lack of vertical or transfer federated learning and decentralized architectures.
Authors: Xiaoting Lyu, Yufei Han, Wei Wang, Jingkai Liu, Yongsheng Zhu, Guangquan Xu, Jiqiang Liu, Xiangliang Zhang
Abstract: Federated Learning (FL) is a collaborative machine learning technique where multiple clients work together with a central server to train a global model without sharing their private data. However, the distribution shift across non-IID datasets of clients poses a challenge to this one-model-fits-all method hindering the ability of the global model to effectively adapt to each client's unique local data. To echo this challenge, personalized FL (PFL) is designed to allow each client to create personalized local models tailored to their private data. While extensive research has scrutinized backdoor risks in FL, it has remained underexplored in PFL applications. In this study, we delve deep into the vulnerabilities of PFL to backdoor attacks. Our analysis showcases a tale of two cities. On the one hand, the personalization process in PFL can dilute the backdoor poisoning effects injected into the personalized local models. Furthermore, PFL systems can also deploy both server-end and client-end defense mechanisms to strengthen the barrier against backdoor attacks. On the other hand, our study shows that PFL fortified with these defense methods may offer a false sense of security. We propose \textit{PFedBA}, a stealthy and effective backdoor attack strategy applicable to PFL systems. \textit{PFedBA} ingeniously aligns the backdoor learning task with the main learning task of PFL by optimizing the trigger generation process. Our comprehensive experiments demonstrate the effectiveness of \textit{PFedBA} in seamlessly embedding triggers into personalized local models. \textit{PFedBA} yields outstanding attack performance across 10 state-of-the-art PFL algorithms, defeating the existing 6 defense mechanisms. Our study sheds light on the subtle yet potent backdoor threats to PFL systems, urging the community to bolster defenses against emerging backdoor challenges.
Authors: Xuyang Zhao, Huiyuan Wang, Weiran Huang, Wei Lin
Abstract: We provide a statistical analysis of regularization-based continual learning on a sequence of linear regression tasks, with emphasis on how different regularization terms affect the model performance. We first derive the convergence rate for the oracle estimator obtained as if all data were available simultaneously. Next, we consider a family of generalized $\ell_2$-regularization algorithms indexed by matrix-valued hyperparameters, which includes the minimum norm estimator and continual ridge regression as special cases. As more tasks are introduced, we derive an iterative update formula for the estimation error of generalized $\ell_2$-regularized estimators, from which we determine the hyperparameters resulting in the optimal algorithm. Interestingly, the choice of hyperparameters can effectively balance the trade-off between forward and backward knowledge transfer and adjust for data heterogeneity. Moreover, the estimation error of the optimal algorithm is derived explicitly, which is of the same order as that of the oracle estimator. In contrast, our lower bounds for the minimum norm estimator and continual ridge regression show their suboptimality. A byproduct of our theoretical analysis is the equivalence between early stopping and generalized $\ell_2$-regularization in continual learning, which may be of independent interest. Finally, we conduct experiments to complement our theory.
Authors: Masahiro Fujisawa, Futoshi Futami
Abstract: Nonparametric estimation with binning is widely employed in the calibration error evaluation and the recalibration of machine learning models. Recently, theoretical analyses of the bias induced by this estimation approach have been actively pursued; however, the understanding of the generalization of the calibration error to unknown data remains limited. In addition, although many recalibration algorithms have been proposed, their generalization performance lacks theoretical guarantees. To address this problem, we conduct a generalization analysis of the calibration error under the probably approximately correct (PAC) Bayes framework. This approach enables us to derive a first optimizable upper bound for the generalization error in the calibration context. We then propose a generalization-aware recalibration algorithm based on our generalization theory. Numerical experiments show that our algorithm improves the Gaussian-process-based recalibration performance on various benchmark datasets and models.
Authors: C. Metz, O. Bichler, A. Dupret
Abstract: This paper explores the combination of neural network quantization and entropy coding for memory footprint minimization. Edge deployment of quantized models is hampered by the harsh Pareto frontier of the accuracy-to-bitwidth tradeoff, causing dramatic accuracy loss below a certain bitwidth. This accuracy loss can be alleviated thanks to mixed precision quantization, allowing for more flexible bitwidth allocation. However, standard mixed precision benefits remain limited due to the 1-bit frontier, that forces each parameter to be encoded on at least 1 bit of data. This paper introduces an approach that combines mixed precision, zero-point quantization and entropy coding to push the compression boundary of Resnets beyond the 1-bit frontier with an accuracy drop below 1% on the ImageNet benchmark. From an implementation standpoint, a compact decoder architecture features reduced latency, thus allowing for inference-compatible decoding.
Authors: Alaia Solko-Breslin, Seewon Choi, Ziyang Li, Neelay Velingker, Rajeev Alur, Mayur Naik, Eric Wong
Abstract: Many computational tasks can be naturally expressed as a composition of a DNN followed by a program written in a traditional programming language or an API call to an LLM. We call such composites "neural programs" and focus on the problem of learning the DNN parameters when the training data consist of end-to-end input-output labels for the composite. When the program is written in a differentiable logic programming language, techniques from neurosymbolic learning are applicable, but in general, the learning for neural programs requires estimating the gradients of black-box components. We present an algorithm for learning neural programs, called ISED, that only relies on input-output samples of black-box components. For evaluation, we introduce new benchmarks that involve calls to modern LLMs such as GPT-4 and also consider benchmarks from the neurosymolic learning literature. Our evaluation shows that for the latter benchmarks, ISED has comparable performance to state-of-the-art neurosymbolic frameworks. For the former, we use adaptations of prior work on gradient approximations of black-box components as a baseline, and show that ISED achieves comparable accuracy but in a more data- and sample-efficient manner.
Authors: Shikai Qiu, Andres Potapczynski, Marc Finzi, Micah Goldblum, Andrew Gordon Wilson
Abstract: Dense linear layers are the dominant computational bottleneck in foundation models. Identifying more efficient alternatives to dense matrices has enormous potential for building more compute-efficient models, as exemplified by the success of convolutional networks in the image domain. In this work, we systematically explore structured matrices as replacements for dense matrices. We show that different structures often require drastically different initialization scales and learning rates, which are crucial to performance, especially as models scale. Using insights from the Maximal Update Parameterization, we determine the optimal scaling for initialization and learning rates of these unconventional layers. Finally, we measure the scaling laws of different structures to compare how quickly their performance improves with compute. We propose a novel matrix family containing Monarch matrices, the Block Tensor-Train (BTT), which we show performs better than dense matrices for the same compute on multiple tasks. On CIFAR-10/100 with augmentation, BTT achieves exponentially lower training loss than dense when training MLPs and ViTs. BTT matches dense ViT-S/32 performance on ImageNet-1k with 3.8 times less compute and is more efficient than dense for training small GPT-2 language models.
Authors: Zhenliang Xue, Yixin Song, Zeyu Mi, Le Chen, Yubin Xia, Haibo Chen
Abstract: This paper introduces PowerInfer-2, a framework designed for high-speed inference of Large Language Models (LLMs) on smartphones, particularly effective for models whose sizes exceed the device's memory capacity. The key insight of PowerInfer-2 is to utilize the heterogeneous computation, memory, and I/O resources in smartphones by decomposing traditional matrix computations into fine-grained neuron cluster computations. Specifically, PowerInfer-2 features a polymorphic neuron engine that adapts computational strategies for various stages of LLM inference. Additionally, it introduces segmented neuron caching and fine-grained neuron-cluster-level pipelining, which effectively minimize and conceal the overhead caused by I/O operations. The implementation and evaluation of PowerInfer-2 demonstrate its capability to support a wide array of LLM models on two smartphones, achieving up to a 29.2x speed increase compared with state-of-the-art frameworks. Notably, PowerInfer-2 is the first system to serve the TurboSparse-Mixtral-47B model with a generation rate of 11.68 tokens per second on a smartphone. For models that fit entirely within the memory, PowerInfer-2 can achieve approximately a 40% reduction in memory usage while maintaining inference speeds comparable to llama.cpp and MLC-LLM. For more details, including a demonstration video, please visit the project site at www.powerinfer.ai/v2.
Authors: Wyatt Mackey, Ioannis Schizas, Jared Deighton, David L. Boothe, Jr., Vasileios Maroulas
Abstract: A common technique for ameliorating the computational costs of running large neural models is sparsification, or the removal of neural connections during training. Sparse models are capable of maintaining the high accuracy of state of the art models, while functioning at the cost of more parsimonious models. The structures which underlie sparse architectures are, however, poorly understood and not consistent between differently trained models and sparsification schemes. In this paper, we propose a new technique for sparsification of recurrent neural nets (RNNs), called moduli regularization, in combination with magnitude pruning. Moduli regularization leverages the dynamical system induced by the recurrent structure to induce a geometric relationship between neurons in the hidden state of the RNN. By making our regularizing term explicitly geometric, we provide the first, to our knowledge, a priori description of the desired sparse architecture of our neural net. We verify the effectiveness of our scheme for navigation and natural language processing RNNs. Navigation is a structurally geometric task, for which there are known moduli spaces, and we show that regularization can be used to reach 90% sparsity while maintaining model performance only when coefficients are chosen in accordance with a suitable moduli space. Natural language processing, however, has no known moduli space in which computations are performed. Nevertheless, we show that moduli regularization induces more stable recurrent neural nets with a variety of moduli regularizers, and achieves high fidelity models at 98% sparsity.
Authors: Denis Tarasov, Kirill Brilliantov, Dmitrii Kharlapenko
Abstract: In deep Reinforcement Learning (RL), value functions are typically approximated using deep neural networks and trained via mean squared error regression objectives to fit the true value functions. Recent research has proposed an alternative approach, utilizing the cross-entropy classification objective, which has demonstrated improved performance and scalability of RL algorithms. However, existing study have not extensively benchmarked the effects of this replacement across various domains, as the primary objective was to demonstrate the efficacy of the concept across a broad spectrum of tasks, without delving into in-depth analysis. Our work seeks to empirically investigate the impact of such a replacement in an offline RL setup and analyze the effects of different aspects on performance. Through large-scale experiments conducted across a diverse range of tasks using different algorithms, we aim to gain deeper insights into the implications of this approach. Our results reveal that incorporating this change can lead to superior performance over state-of-the-art solutions for some algorithms in certain tasks, while maintaining comparable performance levels in other tasks, however for other algorithms this modification might lead to the dramatic performance drop. This findings are crucial for further application of classification approach in research and practical tasks.
Authors: Seyedhamidreza Mousavi, Mohammad Hasan Ahmadilivani, Jaan Raik, Maksim Jenihhin, Masoud Daneshtalab
Abstract: Deep Neural Networks (DNNs) are extensively employed in safety-critical applications where ensuring hardware reliability is a primary concern. To enhance the reliability of DNNs against hardware faults, activation restriction techniques significantly mitigate the fault effects at the DNN structure level, irrespective of accelerator architectures. State-of-the-art methods offer either neuron-wise or layer-wise clipping activation functions. They attempt to determine optimal clipping thresholds using heuristic and learning-based approaches. Layer-wise clipped activation functions cannot preserve DNNs resilience at high bit error rates. On the other hand, neuron-wise clipping activation functions introduce considerable memory overhead due to the addition of parameters, which increases their vulnerability to faults. Moreover, the heuristic-based optimization approach demands numerous fault injections during the search process, resulting in time-consuming threshold identification. On the other hand, learning-based techniques that train thresholds for entire layers concurrently often yield sub-optimal results. In this work, first, we demonstrate that it is not essential to incorporate neuron-wise activation functions throughout all layers in DNNs. Then, we propose a hybrid clipped activation function that integrates neuron-wise and layer-wise methods that apply neuron-wise clipping only in the last layer of DNNs. Additionally, to attain optimal thresholds in the clipping activation function, we introduce ProAct, a progressive training methodology. This approach iteratively trains the thresholds on a layer-by-layer basis, aiming to obtain optimal threshold values in each layer separately.
Authors: Basem Suleiman, Muhammad Johan Alibasa, Rizka Widyarini Purwanto, Lewis Jeffries, Ali Anaissi, Jacky Song
Abstract: Federated Learning (FL) enables local devices to collaboratively learn a shared predictive model by only periodically sharing model parameters with a central aggregator. However, FL can be disadvantaged by statistical heterogeneity produced by the diversity in each local devices data distribution, which creates different levels of Independent and Identically Distributed (IID) data. Furthermore, this can be more complex when optimising different combinations of FL parameters and choosing optimal aggregation. In this paper, we present an empirical analysis of different FL training parameters and aggregators over various levels of statistical heterogeneity on three datasets. We propose a systematic data partition strategy to simulate different levels of statistical heterogeneity and a metric to measure the level of IID. Additionally, we empirically identify the best FL model and key parameters for datasets of different characteristics. On the basis of these, we present recommended guidelines for FL parameters and aggregators to optimise model performance under different levels of IID and with different datasets
Authors: Ashka Shah, Adela DePavia, Nathaniel Hudson, Ian Foster, Rick Stevens
Abstract: The aim in many sciences is to understand the mechanisms that underlie the observed distribution of variables, starting from a set of initial hypotheses. Causal discovery allows us to infer mechanisms as sets of cause and effect relationships in a generalized way -- without necessarily tailoring to a specific domain. Causal discovery algorithms search over a structured hypothesis space, defined by the set of directed acyclic graphs, to find the graph that best explains the data. For high-dimensional problems, however, this search becomes intractable and scalable algorithms for causal discovery are needed to bridge the gap. In this paper, we define a novel causal graph partition that allows for divide-and-conquer causal discovery with theoretical guarantees. We leverage the idea of a superstructure -- a set of learned or existing candidate hypotheses -- to partition the search space. We prove under certain assumptions that learning with a causal graph partition always yields the Markov Equivalence Class of the true causal graph. We show our algorithm achieves comparable accuracy and a faster time to solution for biologically-tuned synthetic networks and networks up to ${10^4}$ variables. This makes our method applicable to gene regulatory network inference and other domains with high-dimensional structured hypothesis spaces.
Authors: Denys Pushkin, Rapha\"el Berthier, Emmanuel Abbe
Abstract: We investigate the out-of-domain generalization of random feature (RF) models and Transformers. We first prove that in the `generalization on the unseen (GOTU)' setting, where training data is fully seen in some part of the domain but testing is made on another part, and for RF models in the small feature regime, the convergence takes place to interpolators of minimal degree as in the Boolean case (Abbe et al., 2023). We then consider the sparse target regime and explain how this regime relates to the small feature regime, but with a different regularization term that can alter the picture in the non-Boolean case. We show two different outcomes for the sparse regime with q-ary data tokens: (1) if the data is embedded with roots of unities, then a min-degree interpolator is learned like in the Boolean case for RF models, (2) if the data is not embedded as such, e.g., simply as integers, then RF models and Transformers may not learn minimal degree interpolators. This shows that the Boolean setting and its roots of unities generalization are special cases where the minimal degree interpolator offers a rare characterization of how learning takes place. For more general integer and real-valued settings, a more nuanced picture remains to be fully characterized.
Authors: Yelysei Bondarenko, Riccardo Del Chiaro, Markus Nagel
Abstract: Large language models (LLMs) are omnipresent, however their practical deployment is challenging due to their ever increasing computational and memory demands. Quantization is one of the most effective ways to make them more compute and memory efficient. Quantization-aware training (QAT) methods, generally produce the best quantized performance, however it comes at the cost of potentially long training time and excessive memory usage, making it impractical when applying for LLMs. Inspired by parameter-efficient fine-tuning (PEFT) and low-rank adaptation (LoRA) literature, we propose LR-QAT -- a lightweight and memory-efficient QAT algorithm for LLMs. LR-QAT employs several components to save memory without sacrificing predictive performance: (a) low-rank auxiliary weights that are aware of the quantization grid; (b) a downcasting operator using fixed-point or double-packed integers and (c) checkpointing. Unlike most related work, our method (i) is inference-efficient, leading to no additional overhead compared to traditional PTQ; (ii) can be seen as a general extended pretraining framework, meaning that the resulting model can still be utilized for any downstream task afterwards; (iii) can be applied across a wide range of quantization settings, such as different choices quantization granularity, activation quantization, and seamlessly combined with many PTQ techniques. We apply LR-QAT to the LLaMA-2/3 and Mistral model families and validate its effectiveness on several downstream tasks. Our method outperforms common post-training quantization (PTQ) approaches and reaches the same model performance as full-model QAT at the fraction of its memory usage. Specifically, we can train a 7B LLM on a single consumer grade GPU with 24GB of memory.
Authors: Junhao Zheng, Shengjie Qiu, Chengming Shi, Qianli Ma
Abstract: As the applications of large language models (LLMs) expand across diverse fields, the ability of these models to adapt to ongoing changes in data, tasks, and user preferences becomes crucial. Traditional training methods, relying on static datasets, are increasingly inadequate for coping with the dynamic nature of real-world information. Lifelong learning, also known as continual or incremental learning, addresses this challenge by enabling LLMs to learn continuously and adaptively over their operational lifetime, integrating new knowledge while retaining previously learned information and preventing catastrophic forgetting. This survey delves into the sophisticated landscape of lifelong learning, categorizing strategies into two primary groups: Internal Knowledge and External Knowledge. Internal Knowledge includes continual pretraining and continual finetuning, each enhancing the adaptability of LLMs in various scenarios. External Knowledge encompasses retrieval-based and tool-based lifelong learning, leveraging external data sources and computational tools to extend the model's capabilities without modifying core parameters. The key contributions of our survey are: (1) Introducing a novel taxonomy categorizing the extensive literature of lifelong learning into 12 scenarios; (2) Identifying common techniques across all lifelong learning scenarios and classifying existing literature into various technique groups within each scenario; (3) Highlighting emerging techniques such as model expansion and data selection, which were less explored in the pre-LLM era. Through a detailed examination of these groups and their respective categories, this survey aims to enhance the adaptability, reliability, and overall performance of LLMs in real-world applications.
Authors: Dora Zhao, Morgan Klaus Scheuerman, Pooja Chitre, Jerone T. A. Andrews, Georgia Panagiotidou, Shawn Walker, Kathleen H. Pine, Alice Xiang
Abstract: Despite extensive efforts to create fairer machine learning (ML) datasets, there remains a limited understanding of the practical aspects of dataset curation. Drawing from interviews with 30 ML dataset curators, we present a comprehensive taxonomy of the challenges and trade-offs encountered throughout the dataset curation lifecycle. Our findings underscore overarching issues within the broader fairness landscape that impact data curation. We conclude with recommendations aimed at fostering systemic changes to better facilitate fair dataset curation practices.
Authors: Zhong Li, Simon Geisler, Yuhang Wang, Stephan G\"unnemann, Matthijs van Leeuwen
Abstract: Predictions made by graph neural networks (GNNs) usually lack interpretability due to their complex computational behavior and the abstract nature of graphs. In an attempt to tackle this, many GNN explanation methods have emerged. Their goal is to explain a model's predictions and thereby obtain trust when GNN models are deployed in decision critical applications. Most GNN explanation methods work in a post-hoc manner and provide explanations in the form of a small subset of important edges and/or nodes. In this paper we demonstrate that these explanations can unfortunately not be trusted, as common GNN explanation methods turn out to be highly susceptible to adversarial perturbations. That is, even small perturbations of the original graph structure that preserve the model's predictions may yield drastically different explanations. This calls into question the trustworthiness and practical utility of post-hoc explanation methods for GNNs. To be able to attack GNN explanation models, we devise a novel attack method dubbed \textit{GXAttack}, the first \textit{optimization-based} adversarial attack method for post-hoc GNN explanations under such settings. Due to the devastating effectiveness of our attack, we call for an adversarial evaluation of future GNN explainers to demonstrate their robustness.
Authors: David Berghaus, Kostadin Cvejoski, Patrick Seifner, Cesar Ojeda, Ramses J. Sanchez
Abstract: Markov jump processes are continuous-time stochastic processes which describe dynamical systems evolving in discrete state spaces. These processes find wide application in the natural sciences and machine learning, but their inference is known to be far from trivial. In this work we introduce a methodology for zero-shot inference of Markov jump processes (MJPs), on bounded state spaces, from noisy and sparse observations, which consists of two components. First, a broad probability distribution over families of MJPs, as well as over possible observation times and noise mechanisms, with which we simulate a synthetic dataset of hidden MJPs and their noisy observation process. Second, a neural network model that processes subsets of the simulated observations, and that is trained to output the initial condition and rate matrix of the target MJP in a supervised way. We empirically demonstrate that one and the same (pretrained) model can infer, in a zero-shot fashion, hidden MJPs evolving in state spaces of different dimensionalities. Specifically, we infer MJPs which describe (i) discrete flashing ratchet systems, which are a type of Brownian motors, and the conformational dynamics in (ii) molecular simulations, (iii) experimental ion channel data and (iv) simple protein folding models. What is more, we show that our model performs on par with state-of-the-art models which are finetuned to the target datasets.
Authors: Xiaodong Wu, Wenyi Yu, Chao Zhang, Philip Woodland
Abstract: Approximate Natural Gradient Descent (NGD) methods are an important family of optimisers for deep learning models, which use approximate Fisher information matrices to pre-condition gradients during training. The empirical Fisher (EF) method approximates the Fisher information matrix empirically by reusing the per-sample gradients collected during back-propagation. Despite its ease of implementation, the EF approximation has its theoretical and practical limitations. This paper first investigates the inversely-scaled projection issue of EF, which is shown to be a major cause of the poor empirical approximation quality. An improved empirical Fisher (iEF) method, motivated as a generalised NGD method from a loss reduction perspective, is proposed to address this issue, meanwhile retaining the practical convenience of EF. The exact iEF and EF methods are experimentally evaluated using practical deep learning setups, including widely-used setups for parameter-efficient fine-tuning of pre-trained models (T5-base with LoRA and Prompt-Tuning on GLUE tasks, and ViT with LoRA for CIFAR100). Optimisation experiments show that applying exact iEF as an optimiser provides strong convergence and generalisation. It achieves the best test performance and the lowest training loss for majority of the tasks, even when compared with well-tuned AdamW/Adafactor baselines. Additionally, under a novel empirical evaluation framework, the proposed iEF method shows consistently better approximation quality to the exact Natural Gradient updates than both EF and the more expensive sampled Fisher (SF). Further investigation also shows that the superior approximation quality of iEF is robust to damping across tasks and training stages. Improving existing approximate NGD optimisers with iEF is expected to lead to better convergence ability and stronger robustness to choice of damping.
Authors: Jason Shuo Zhang, Benjamin Howson, Panayiota Savva, Eleanor Loh
Abstract: Personalised discount codes provide a powerful mechanism for managing customer relationships and operational spend in e-commerce. Bandits are well suited for this product area, given the partial information nature of the problem, as well as the need for adaptation to the changing business environment. Here, we introduce DISCO, an end-to-end contextual bandit framework for personalised discount code allocation at ASOS.com. DISCO adapts the traditional Thompson Sampling algorithm by integrating it within an integer program, thereby allowing for operational cost control. Because bandit learning is often worse with high dimensional actions, we focused on building low dimensional action and context representations that were nonetheless capable of good accuracy. Additionally, we sought to build a model that preserved the relationship between price and sales, in which customers increasing their purchasing in response to lower prices ("negative price elasticity"). These aims were achieved by using radial basis functions to represent the continuous (i.e. infinite armed) action space, in combination with context embeddings extracted from a neural network. These feature representations were used within a Thompson Sampling framework to facilitate exploration, and further integrated with an integer program to allocate discount codes across ASOS's customer base. These modelling decisions result in a reward model that (a) enables pooled learning across similar actions, (b) is highly accurate, including in extrapolation, and (c) preserves the expected negative price elasticity. Through offline analysis, we show that DISCO is able to effectively enact exploration and improves its performance over time, despite the global constraint. Finally, we subjected DISCO to a rigorous online A/B test, and find that it achieves a significant improvement of >1% in average basket value, relative to the legacy systems.
Authors: Pratyush Maini, Hengrui Jia, Nicolas Papernot, Adam Dziedzic
Abstract: The proliferation of large language models (LLMs) in the real world has come with a rise in copyright cases against companies for training their models on unlicensed data from the internet. Recent works have presented methods to identify if individual text sequences were members of the model's training data, known as membership inference attacks (MIAs). We demonstrate that the apparent success of these MIAs is confounded by selecting non-members (text sequences not used for training) belonging to a different distribution from the members (e.g., temporally shifted recent Wikipedia articles compared with ones used to train the model). This distribution shift makes membership inference appear successful. However, most MIA methods perform no better than random guessing when discriminating between members and non-members from the same distribution (e.g., in this case, the same period of time). Even when MIAs work, we find that different MIAs succeed at inferring membership of samples from different distributions. Instead, we propose a new dataset inference method to accurately identify the datasets used to train large language models. This paradigm sits realistically in the modern-day copyright landscape, where authors claim that an LLM is trained over multiple documents (such as a book) written by them, rather than one particular paragraph. While dataset inference shares many of the challenges of membership inference, we solve it by selectively combining the MIAs that provide positive signal for a given distribution, and aggregating them to perform a statistical test on a given dataset. Our approach successfully distinguishes the train and test sets of different subsets of the Pile with statistically significant p-values < 0.1, without any false positives.
Authors: Antoine Siraudin, Fragkiskos D. Malliaros, Christopher Morris
Abstract: Discrete-state denoising diffusion models led to state-of-the-art performance in graph generation, especially in the molecular domain. Recently, they have been transposed to continuous time, allowing more flexibility in the reverse process and a better trade-off between sampling efficiency and quality. Here, to leverage the benefits of both approaches, we propose Cometh, a continuous-time discrete-state graph diffusion model, integrating graph data into a continuous-time diffusion model framework. Empirically, we show that integrating continuous time leads to significant improvements across various metrics over state-of-the-art discrete-state diffusion models on a large set of molecular and non-molecular benchmark datasets.
Authors: Agustinus Kristiadi, Felix Strieth-Kalthoff, Sriram Ganapathi Subramanian, Vincent Fortuin, Pascal Poupart, Geoff Pleiss
Abstract: Bayesian optimization (BO) is an integral part of automated scientific discovery -- the so-called self-driving lab -- where human inputs are ideally minimal or at least non-blocking. However, scientists often have strong intuition, and thus human feedback is still useful. Nevertheless, prior works in enhancing BO with expert feedback, such as by incorporating it in an offline or online but blocking (arrives at each BO iteration) manner, are incompatible with the spirit of self-driving labs. In this work, we study whether a small amount of randomly arriving expert feedback that is being incorporated in a non-blocking manner can improve a BO campaign. To this end, we run an additional, independent computing thread on top of the BO loop to handle the feedback-gathering process. The gathered feedback is used to learn a Bayesian preference model that can readily be incorporated into the BO thread, to steer its exploration-exploitation process. Experiments on toy and chemistry datasets suggest that even just a few intermittent, asynchronous expert feedback can be useful for improving or constraining BO. This can especially be useful for its implication in improving self-driving labs, e.g. making them more data-efficient and less costly.
Authors: Emmanuel Abbe, Samy Bengio, Aryo Lotfi, Colin Sandon, Omid Saremi
Abstract: Can Transformers predict new syllogisms by composing established ones? More generally, what type of targets can be learned by such models from scratch? Recent works show that Transformers can be Turing-complete in terms of expressivity, but this does not address the learnability objective. This paper puts forward the notion of 'distribution locality' to capture when weak learning is efficiently achievable by regular Transformers, where the locality measures the least number of tokens required in addition to the tokens histogram to correlate nontrivially with the target. As shown experimentally and theoretically under additional assumptions, distributions with high locality cannot be learned efficiently. In particular, syllogisms cannot be composed on long chains. Furthermore, we show that (i) an agnostic scratchpad cannot help to break the locality barrier, (ii) an educated scratchpad can help if it breaks the locality at each step, (iii) a notion of 'inductive scratchpad' can both break the locality and improve the out-of-distribution generalization, e.g., generalizing to almost double input size for some arithmetic tasks.
Authors: Mehrdad Kiamari, Mohammad Kiamari, Bhaskar Krishnamachari
Abstract: We introduce Graph Kolmogorov-Arnold Networks (GKAN), an innovative neural network architecture that extends the principles of the recently proposed Kolmogorov-Arnold Networks (KAN) to graph-structured data. By adopting the unique characteristics of KANs, notably the use of learnable univariate functions instead of fixed linear weights, we develop a powerful model for graph-based learning tasks. Unlike traditional Graph Convolutional Networks (GCNs) that rely on a fixed convolutional architecture, GKANs implement learnable spline-based functions between layers, transforming the way information is processed across the graph structure. We present two different ways to incorporate KAN layers into GKAN: architecture 1 -- where the learnable functions are applied to input features after aggregation and architecture 2 -- where the learnable functions are applied to input features before aggregation. We evaluate GKAN empirically using a semi-supervised graph learning task on a real-world dataset (Cora). We find that architecture generally performs better. We find that GKANs achieve higher accuracy in semi-supervised learning tasks on graphs compared to the traditional GCN model. For example, when considering 100 features, GCN provides an accuracy of 53.5 while a GKAN with a comparable number of parameters gives an accuracy of 61.76; with 200 features, GCN provides an accuracy of 61.24 while a GKAN with a comparable number of parameters gives an accuracy of 67.66. We also present results on the impact of various parameters such as the number of hidden nodes, grid-size, and the polynomial-degree of the spline on the performance of GKAN.
Authors: Nicole Hayes, Ekaterina Merkurjev, Guo-Wei Wei
Abstract: Data sets with imbalanced class sizes, often where one class size is much smaller than that of others, occur extremely often in various applications, including those with biological foundations, such as drug discovery and disease diagnosis. Thus, it is extremely important to be able to identify data elements of classes of various sizes, as a failure to detect can result in heavy costs. However, many data classification algorithms do not perform well on imbalanced data sets as they often fail to detect elements belonging to underrepresented classes. In this paper, we propose the BTDT-MBO algorithm, incorporating Merriman-Bence-Osher (MBO) techniques and a bidirectional transformer, as well as distance correlation and decision threshold adjustments, for data classification problems on highly imbalanced molecular data sets, where the sizes of the classes vary greatly. The proposed method not only integrates adjustments in the classification threshold for the MBO algorithm in order to help deal with the class imbalance, but also uses a bidirectional transformer model based on an attention mechanism for self-supervised learning. Additionally, the method implements distance correlation as a weight function for the similarity graph-based framework on which the adjusted MBO algorithm operates. The proposed model is validated using six molecular data sets, and we also provide a thorough comparison to other competing algorithms. The computational experiments show that the proposed method performs better than competing techniques even when the class imbalance ratio is very high.
Authors: Songlin Yang, Bailin Wang, Yu Zhang, Yikang Shen, Yoon Kim
Abstract: Transformers with linear attention (i.e., linear transformers) and state-space models have recently been suggested as a viable linear-time alternative to transformers with softmax attention. However, these models still underperform transformers especially on tasks that require in-context retrieval. While more expressive variants of linear transformers which replace the additive outer-product update in linear transformers with the delta rule have been found to be more effective at associative recall, existing algorithms for training such models do not parallelize over sequence length and are thus inefficient to train on modern hardware. This work describes a hardware-efficient algorithm for training linear transformers with the delta rule, which exploits a memory-efficient representation for computing products of Householder matrices. This algorithm allows us to scale up DeltaNet to standard language modeling settings. We train a 1.3B model for 100B tokens and find that it outperforms recent linear-time baselines such as Mamba and GLA in terms of perplexity and zero-shot performance on downstream tasks (including on tasks that focus on recall). We also experiment with two hybrid models which combine DeltaNet layers with (1) sliding-window attention layers every other layer or (2) two global attention layers, and find that these hybrid models outperform strong transformer baselines.
Authors: Edoardo Calvello, Nikola B. Kovachki, Matthew E. Levine, Andrew M. Stuart
Abstract: Transformers, and the attention mechanism in particular, have become ubiquitous in machine learning. Their success in modeling nonlocal, long-range correlations has led to their widespread adoption in natural language processing, computer vision, and time-series problems. Neural operators, which map spaces of functions into spaces of functions, are necessarily both nonlinear and nonlocal if they are universal; it is thus natural to ask whether the attention mechanism can be used in the design of neural operators. Motivated by this, we study transformers in the function space setting. We formulate attention as a map between infinite dimensional function spaces and prove that the attention mechanism as implemented in practice is a Monte Carlo or finite difference approximation of this operator. The function space formulation allows for the design of transformer neural operators, a class of architectures designed to learn mappings between function spaces, for which we prove a universal approximation result. The prohibitive cost of applying the attention operator to functions defined on multi-dimensional domains leads to the need for more efficient attention-based architectures. For this reason we also introduce a function space generalization of the patching strategy from computer vision, and introduce a class of associated neural operators. Numerical results, on an array of operator learning problems, demonstrate the promise of our approaches to function space formulations of attention and their use in neural operators.
Authors: Dutch Hansen, Siddartha Devic, Preetum Nakkiran, Vatsal Sharan
Abstract: Calibration is a well-studied property of predictors which guarantees meaningful uncertainty estimates. Multicalibration is a related notion -- originating in algorithmic fairness -- which requires predictors to be simultaneously calibrated over a potentially complex and overlapping collection of protected subpopulations (such as groups defined by ethnicity, race, or income). We conduct the first comprehensive study evaluating the usefulness of multicalibration post-processing across a broad set of tabular, image, and language datasets for models spanning from simple decision trees to 90 million parameter fine-tuned LLMs. Our findings can be summarized as follows: (1) models which are calibrated out of the box tend to be relatively multicalibrated without any additional post-processing; (2) multicalibration post-processing can help inherently uncalibrated models; and (3) traditional calibration measures may sometimes provide multicalibration implicitly. More generally, we also distill many independent observations which may be useful for practical and effective applications of multicalibration post-processing in real-world contexts.
Authors: Gennaro Gala, Cassio de Campos, Antonio Vergari, Erik Quaeghebeur
Abstract: Probabilistic integral circuits (PICs) have been recently introduced as probabilistic models enjoying the key ingredient behind expressive generative models: continuous latent variables (LVs). PICs are symbolic computational graphs defining continuous LV models as hierarchies of functions that are summed and multiplied together, or integrated over some LVs. They are tractable if LVs can be analytically integrated out, otherwise they can be approximated by tractable probabilistic circuits (PC) encoding a hierarchical numerical quadrature process, called QPCs. So far, only tree-shaped PICs have been explored, and training them via numerical quadrature requires memory-intensive processing at scale. In this paper, we address these issues, and present: (i) a pipeline for building DAG-shaped PICs out of arbitrary variable decompositions, (ii) a procedure for training PICs using tensorized circuit architectures, and (iii) neural functional sharing techniques to allow scalable training. In extensive experiments, we showcase the effectiveness of functional sharing and the superiority of QPCs over traditional PCs.
Authors: Calarina Muslimani, Bram Grooten, Deepak Ranganatha Sastry Mamillapalli, Mykola Pechenizkiy, Decebal Constantin Mocanu, Matthew E. Taylor
Abstract: For autonomous agents to successfully integrate into human-centered environments, agents should be able to learn from and adapt to humans in their native settings. Preference-based reinforcement learning (PbRL) is a promising approach that learns reward functions from human preferences. This enables RL agents to adapt their behavior based on human desires. However, humans live in a world full of diverse information, most of which is not relevant to completing a particular task. It becomes essential that agents learn to focus on the subset of task-relevant environment features. Unfortunately, prior work has largely ignored this aspect; primarily focusing on improving PbRL algorithms in standard RL environments that are carefully constructed to contain only task-relevant features. This can result in algorithms that may not effectively transfer to a more noisy real-world setting. To that end, this work proposes R2N (Robust-to-Noise), the first PbRL algorithm that leverages principles of dynamic sparse training to learn robust reward models that can focus on task-relevant features. We study the effectiveness of R2N in the Extremely Noisy Environment setting, an RL problem setting where up to 95% of the state features are irrelevant distractions. In experiments with a simulated teacher, we demonstrate that R2N can adapt the sparse connectivity of its neural networks to focus on task-relevant features, enabling R2N to significantly outperform several state-of-the-art PbRL algorithms in multiple locomotion and control environments.
Authors: Oishi Banerjee, Hong-Yu Zhou, Subathra Adithan, Stephen Kwak, Kay Wu, Pranav Rajpurkar
Abstract: Recent advances in generative vision-language models (VLMs) have exciting potential implications for AI in radiology, yet VLMs are also known to produce hallucinations, nonsensical text, and other unwanted behaviors that can waste clinicians' time and cause patient harm. Drawing on recent work on direct preference optimization (DPO), we propose a simple method for modifying the behavior of pretrained VLMs performing radiology report generation by suppressing unwanted types of generations. We apply our method to the prevention of hallucinations of prior exams, addressing a long-established problem behavior in models performing chest X-ray report generation. Across our experiments, we find that DPO fine-tuning achieves a 3.2-4.8x reduction in lines hallucinating prior exams while maintaining model performance on clinical accuracy metrics. Our work is, to the best of our knowledge, the first work to apply DPO to medical VLMs, providing a data- and compute- efficient way to suppress problem behaviors while maintaining overall clinical accuracy.
Authors: Philipp Misof, Pan Kessel, Jan E. Gerken
Abstract: Equivariant neural networks have in recent years become an important technique for guiding architecture selection for neural networks with many applications in domains ranging from medical image analysis to quantum chemistry. In particular, as the most general linear equivariant layers with respect to the regular representation, group convolutions have been highly impactful in numerous applications. Although equivariant architectures have been studied extensively, much less is known about the training dynamics of equivariant neural networks. Concurrently, neural tangent kernels (NTKs) have emerged as a powerful tool to analytically understand the training dynamics of wide neural networks. In this work, we combine these two fields for the first time by giving explicit expressions for NTKs of group convolutional neural networks. In numerical experiments, we demonstrate superior performance for equivariant NTKs over non-equivariant NTKs on a classification task for medical images.
Authors: Davide Corsi, Guy Amir, Andoni Rodriguez, Cesar Sanchez, Guy Katz, Roy Fox
Abstract: In recent years, Deep Reinforcement Learning (DRL) has emerged as an effective approach to solving real-world tasks. However, despite their successes, DRL-based policies suffer from poor reliability, which limits their deployment in safety-critical domains. As a result, various methods have been put forth to address this issue by providing formal safety guarantees. Two main approaches include shielding and verification. While shielding ensures the safe behavior of the policy by employing an external online component (i.e., a ``shield'') that overruns potentially dangerous actions, this approach has a significant computational cost as the shield must be invoked at runtime to validate every decision. On the other hand, verification is an offline process that can identify policies that are unsafe, prior to their deployment, yet, without providing alternative actions when such a policy is deemed unsafe. In this work, we present verification-guided shielding -- a novel approach that bridges the DRL reliability gap by integrating these two methods. Our approach combines both formal and probabilistic verification tools to partition the input domain into safe and unsafe regions. In addition, we employ clustering and symbolic representation procedures that compress the unsafe regions into a compact representation. This, in turn, allows to temporarily activate the shield solely in (potentially) unsafe regions, in an efficient manner. Our novel approach allows to significantly reduce runtime overhead while still preserving formal safety guarantees. We extensively evaluate our approach on two benchmarks from the robotic navigation domain, as well as provide an in-depth analysis of its scalability and completeness.
Authors: Sloan Nietert, Ziv Goldfeld, Soroosh Shafiee
Abstract: We consider learning in an adversarial environment, where an $\varepsilon$-fraction of samples from a distribution $P$ are arbitrarily modified (*global* corruptions) and the remaining perturbations have average magnitude bounded by $\rho$ (*local* corruptions). Given access to $n$ such corrupted samples, we seek a computationally efficient estimator $\hat{P}_n$ that minimizes the Wasserstein distance $\mathsf{W}_1(\hat{P}_n,P)$. In fact, we attack the fine-grained task of minimizing $\mathsf{W}_1(\Pi_\# \hat{P}_n, \Pi_\# P)$ for all orthogonal projections $\Pi \in \mathbb{R}^{d \times d}$, with performance scaling with $\mathrm{rank}(\Pi) = k$. This allows us to account simultaneously for mean estimation ($k=1$), distribution estimation ($k=d$), as well as the settings interpolating between these two extremes. We characterize the optimal population-limit risk for this task and then develop an efficient finite-sample algorithm with error bounded by $\sqrt{\varepsilon k} + \rho + d^{O(1)}\tilde{O}(n^{-1/k})$ when $P$ has bounded moments of order $2+\delta$, for constant $\delta > 0$. For data distributions with bounded covariance, our finite-sample bounds match the minimax population-level optimum for large sample sizes. Our efficient procedure relies on a novel trace norm approximation of an ideal yet intractable 2-Wasserstein projection estimator. We apply this algorithm to robust stochastic optimization, and, in the process, uncover a new method for overcoming the curse of dimensionality in Wasserstein distributionally robust optimization.
Authors: Kimia Kazemian, Yahya Sattar, Sarah Dean
Abstract: Modern data-driven control applications call for flexible nonlinear models that are amenable to principled controller synthesis and realtime feedback. Many nonlinear dynamical systems of interest are control affine. We propose two novel classes of nonlinear feature representations which capture control affine structure while allowing for arbitrary complexity in the state dependence. Our methods make use of random features (RF) approximations, inheriting the expressiveness of kernel methods at a lower computational cost. We formalize the representational capabilities of our methods by showing their relationship to the Affine Dot Product (ADP) kernel proposed by Casta\~neda et al. (2021) and a novel Affine Dense (AD) kernel that we introduce. We further illustrate the utility by presenting a case study of data-driven optimization-based control using control certificate functions (CCF). Simulation experiments on a double pendulum empirically demonstrate the advantages of our methods.
Authors: Romain Ilbert, Thai V. Hoang, Zonghua Zhang
Abstract: Our study investigates the impact of data augmentation on the performance of multivariate time series models, focusing on datasets from the UCR archive. Despite the limited size of these datasets, we achieved classification accuracy improvements in 10 out of 13 datasets using the Rocket and InceptionTime models. This highlights the essential role of sufficient data in training effective models, paralleling the advancements seen in computer vision. Our work delves into adapting and applying existing methods in innovative ways to the domain of multivariate time series classification. Our comprehensive exploration of these techniques sets a new standard for addressing data scarcity in time series analysis, emphasizing that diverse augmentation strategies are crucial for unlocking the potential of both traditional and deep learning models. Moreover, by meticulously analyzing and applying a variety of augmentation techniques, we demonstrate that strategic data enrichment can enhance model accuracy. This not only establishes a benchmark for future research in time series analysis but also underscores the importance of adopting varied augmentation approaches to improve model performance in the face of limited data availability.
Authors: Salma Kharrat, Marco Canini, Samuel Horvath
Abstract: This work tackles the challenges of data heterogeneity and communication limitations in decentralized federated learning. We focus on creating a collaboration graph that guides each client in selecting suitable collaborators for training personalized models that leverage their local data effectively. Our approach addresses these issues through a novel, communication-efficient strategy that enhances resource efficiency. Unlike traditional methods, our formulation identifies collaborators at a granular level by considering combinatorial relations of clients, enhancing personalization while minimizing communication overhead. We achieve this through a bi-level optimization framework that employs a constrained greedy algorithm, resulting in a resource-efficient collaboration graph for personalized learning. Extensive evaluation against various baselines across diverse datasets demonstrates the superiority of our method, named DPFL. DPFL consistently outperforms other approaches, showcasing its effectiveness in handling real-world data heterogeneity, minimizing communication overhead, enhancing resource efficiency, and building personalized models in decentralized federated learning scenarios.
Authors: Ziyan Luo, Xujie Si
Abstract: Solving Constrained Horn Clauses (CHCs) is a fundamental challenge behind a wide range of verification and analysis tasks. Data-driven approaches show great promise in improving CHC solving without the painstaking manual effort of creating and tuning various heuristics. However, a large performance gap exists between data-driven CHC solvers and symbolic reasoning-based solvers. In this work, we develop a simple but effective framework, "Chronosymbolic Learning", which unifies symbolic information and numerical data points to solve a CHC system efficiently. We also present a simple instance of Chronosymbolic Learning with a data-driven learner and a BMC-styled reasoner. Despite its relative simplicity, experimental results show the efficacy and robustness of our tool. It outperforms state-of-the-art CHC solvers on a dataset consisting of 288 benchmarks, including many instances with non-linear integer arithmetics.
Authors: Akshansh Mishra
Abstract: The research paper presents a novel approach to optimizing the tensile stress of Triply Periodic Minimal Surface (TPMS) structures through machine learning and Simulated Annealing (SA). The study evaluates the performance of Random Forest, Decision Tree, and XGBoost models in predicting tensile stress, using a dataset generated from finite element analysis of TPMS models. The objective function minimized the negative R-squared value on the validation set to enhance model accuracy. The SA-XGBoost model outperformed the others, achieving an R-squared value of 0.96. In contrast, the SA-Random Forest model achieved an R squared value of 0.89 while the SA-Decision Tree model exhibited greater fluctuations in validation scores. This demonstrates that the SA-XGBoost model is most effective in capturing the complex relationships within the data. The integration of SA helps in optimizing the hyperparameters of these machine learning models, thereby enhancing their predictive capabilities.
Authors: Lane E. Schultz, Yiqi Wang, Ryan Jacobs, Dane Morgan
Abstract: Knowledge of the domain of applicability of a machine learning model is essential to ensuring accurate and reliable model predictions. In this work, we develop a new approach of assessing model domain and demonstrate that our approach provides accurate and meaningful designation of in-domain versus out-of-domain when applied across multiple model types and material property data sets. Our approach assesses the distance between a test and training data point in feature space by using kernel density estimation and shows that this distance provides an effective tool for domain determination. We show that chemical groups considered unrelated based on established chemical knowledge exhibit significant dissimilarities by our measure. We also show that high measures of dissimilarity are associated with poor model performance (i.e., high residual magnitudes) and poor estimates of model uncertainty (i.e., unreliable uncertainty estimation). Automated tools are provided to enable researchers to establish acceptable dissimilarity thresholds to identify whether new predictions of their own machine learning models are in-domain versus out-of-domain.
Authors: Ryan Benkert, Mohit Prabhushankar, Ghassan AlRegib
Abstract: This paper presents a discussion on data selection for deep learning in the field of seismic interpretation. In order to achieve a robust generalization to the target volume, it is crucial to identify the specific samples are the most informative to the training process. The selection of the training set from a target volume is a critical factor in determining the effectiveness of the deep learning algorithm for interpreting seismic volumes. This paper proposes the inclusion of interpretation disagreement as a valuable and intuitive factor in the process of selecting training sets. The development of a novel data selection framework is inspired by established practices in seismic interpretation. The framework we have developed utilizes representation shifts to effectively model interpretation disagreement within neural networks. Additionally, it incorporates the disagreement measure to enhance attention towards geologically interesting regions throughout the data selection workflow. By combining this approach with active learning, a well-known machine learning paradigm for data selection, we arrive at a comprehensive and innovative framework for training set selection in seismic interpretation. In addition, we offer a specific implementation of our proposed framework, which we have named ATLAS. This implementation serves as a means for data selection. In this study, we present the results of our comprehensive experiments, which clearly indicate that ATLAS consistently surpasses traditional active learning frameworks in the field of seismic interpretation. Our findings reveal that ATLAS achieves improvements of up to 12% in mean intersection-over-union.
Authors: Aryan Mathur
Abstract: In this paper of a research based project, using Bidirectional Long Short-Term Memory (BiLSTM) networks, we provide a novel Fight Scene Detection (FSD) model which can be used for Movie Highlight Generation Systems (MHGS) based on deep learning and Neural Networks . Movies usually have Fight Scenes to keep the audience amazed. For trailer generation, or any other application of Highlight generation, it is very tidious to first identify all such scenes manually and then compile them to generate a highlight serving the purpose. Our proposed FSD system utilises temporal characteristics of the movie scenes and thus is capable to automatically identify fight scenes. Thereby helping in the effective production of captivating movie highlights. We observe that the proposed solution features 93.5% accuracy and is higher than 2D CNN with Hough Forests which being 92% accurate and is significantly higher than 3D CNN which features an accuracy of 65%.
Authors: Vahid Negahdari, Seyed Reza Moghadasi, Mohammad Reza Razvan
Abstract: Full-Waveform Inversion (FWI) is a nonlinear iterative seismic imaging technique that, by reducing the misfit between recorded and predicted seismic waveforms, can produce detailed estimates of subsurface geophysical properties. Nevertheless, the strong nonlinearity of FWI can trap the optimization in local minima. This issue arises due to factors such as improper initial values, the absence of low frequencies in the measurements, noise, and other related considerations. To address this challenge and with the advent of advanced machine-learning techniques, data-driven methods, such as deep learning, have attracted significantly increasing attention in the geophysical community. Furthermore, the elastic wave equation should be included in FWI to represent elastic effects accurately. The intersection of data-driven techniques and elastic scattering theories presents opportunities and challenges. In this paper, by using the knowledge of elastic scattering (Physics of problem) and integrating it with deep learning techniques, we propose methods for the solution of time-harmonic FWI to enhance accuracy compared to pure data-driven approaches. Moreover, by modifying the structure of the Variational Autoencoder, we introduce a probabilistic deep learning method based on the physics of the problem that enables us to explore the uncertainties of the solution. According to the limited availability of datasets in this field and to assess the performance and accuracy of the proposed methods, we create a comprehensive dataset close to reality and conduct a comparative analysis of the presented approaches to it.
Authors: Victor Yon, Bastien Galaup, Claude Rohrbacher, Joffrey Rivard, Cl\'ement Godfrin, Roy Li, Stefan Kubicek, Kristiaan De Greve, Louis Gaudreau, Eva Dupont-Ferrier, Yann Beilliard, Roger G. Melko, Dominique Drouin
Abstract: This study presents a machine-learning-based procedure to automate the charge tuning of semiconductor spin qubits with minimal human intervention, addressing one of the significant challenges in scaling up quantum dot technologies. This method exploits artificial neural networks to identify noisy transition lines in stability diagrams, guiding a robust exploration strategy leveraging neural networks' uncertainty estimations. Tested across three distinct offline experimental datasets representing different single quantum dot technologies, the approach achieves over 99% tuning success rate in optimal cases, where more than 10% of the success is directly attributable to uncertainty exploitation. The challenging constraints of small training sets containing high diagram-to-diagram variability allowed us to evaluate the capabilities and limits of the proposed procedure.
Authors: Seyed A. Esmaeili, Kshipra Bhawalkar, Zhe Feng, Di Wang, Haifeng Xu
Abstract: Generative AI (GenAI) will have significant impact on content creation platforms. In this paper, we study the dynamic competition between a GenAI and a human contributor. Unlike the human, the GenAI's content only improves when more contents are created by human over the time; however, GenAI has the advantage of generating content at a lower cost. We study the algorithmic problem in this dynamic competition model about how the human contributor can maximize her utility when competing against the GenAI for content generation over a set of topics. In time-sensitive content domains (e.g., news or pop music creation) where contents' value diminishes over time, we show that there is no polynomial time algorithm for finding the human's optimal (dynamic) strategy, unless the randomized exponential time hypothesis is false. Fortunately, we are able to design a polynomial time algorithm that naturally cycles between myopically optimizing over a short time window and pausing and provably guarantees an approximation ratio of $\frac{1}{2}$. We then turn to time-insensitive content domains where contents do not lose their value (e.g., contents on history facts). Interestingly, we show that this setting permits a polynomial time algorithm that maximizes the human's utility in the long run.
Authors: Arash Gholami Davoodi, Seyed Pouyan Mousavi Davoudi, Pouya Pezeshkpour
Abstract: Large language models (LLMs) demonstrate impressive capabilities in mathematical reasoning. However, despite these achievements, current evaluations are mostly limited to specific mathematical topics, and it remains unclear whether LLMs are genuinely engaging in reasoning. To address these gaps, we present the Mathematical Topics Tree (MaTT) benchmark, a challenging and structured benchmark that offers 1,958 questions across a wide array of mathematical subjects, each paired with a detailed hierarchical chain of topics. Upon assessing different LLMs using the MaTT benchmark, we find that the most advanced model, GPT-4, achieved a mere 54\% accuracy in a multiple-choice scenario. Interestingly, even when employing Chain-of-Thought prompting, we observe mostly no notable improvement. Moreover, LLMs accuracy dramatically reduced by up to 24.2 percentage point when the questions were presented without providing choices. Further detailed analysis of the LLMs' performance across a range of topics showed significant discrepancy even for closely related subtopics within the same general mathematical area. In an effort to pinpoint the reasons behind LLMs performances, we conducted a manual evaluation of the completeness and correctness of the explanations generated by GPT-4 when choices were available. Surprisingly, we find that in only 53.3\% of the instances where the model provided a correct answer, the accompanying explanations were deemed complete and accurate, i.e., the model engaged in genuine reasoning.
Authors: Sajid Javed, Arif Mahmood, Iyyakutti Iyappan Ganapathi, Fayaz Ali Dharejo, Naoufel Werghi, Mohammed Bennamoun
Abstract: This paper proposes Comprehensive Pathology Language Image Pre-training (CPLIP), a new unsupervised technique designed to enhance the alignment of images and text in histopathology for tasks such as classification and segmentation. This methodology enriches vision-language models by leveraging extensive data without needing ground truth annotations. CPLIP involves constructing a pathology-specific dictionary, generating textual descriptions for images using language models, and retrieving relevant images for each text snippet via a pre-trained model. The model is then fine-tuned using a many-to-many contrastive learning method to align complex interrelated concepts across both modalities. Evaluated across multiple histopathology tasks, CPLIP shows notable improvements in zero-shot learning scenarios, outperforming existing methods in both interpretability and robustness and setting a higher benchmark for the application of vision-language models in the field. To encourage further research and replication, the code for CPLIP is available on GitHub at https://cplip.github.io/
URLs: https://cplip.github.io/
Authors: Ziyu Wang, Chris Holmes
Abstract: Applications of large language models often involve the generation of free-form responses, in which case uncertainty quantification becomes challenging. This is due to the need to identify task-specific uncertainties (e.g., about the semantics) which appears difficult to define in general cases. This work addresses these challenges from a perspective of Bayesian decision theory, starting from the assumption that our utility is characterized by a similarity measure that compares a generated response with a hypothetical true response. We discuss how this assumption enables principled quantification of the model's subjective uncertainty and its calibration. We further derive a measure for epistemic uncertainty, based on a missing data perspective and its characterization as an excess risk. The proposed measures can be applied to black-box language models. We demonstrate the proposed methods on question answering and machine translation tasks, where they extract broadly meaningful uncertainty estimates from GPT and Gemini models and quantify their calibration.
Authors: M. J. J. de Grauw, E. Th. Scholten, E. J. Smit, M. J. C. M. Rutten, M. Prokop, B. van Ginneken, A. Hering
Abstract: Size measurements of tumor manifestations on follow-up CT examinations are crucial for evaluating treatment outcomes in cancer patients. Efficient lesion segmentation can speed up these radiological workflows. While numerous benchmarks and challenges address lesion segmentation in specific organs like the liver, kidneys, and lungs, the larger variety of lesion types encountered in clinical practice demands a more universal approach. To address this gap, we introduced the ULS23 benchmark for 3D universal lesion segmentation in chest-abdomen-pelvis CT examinations. The ULS23 training dataset contains 38,693 lesions across this region, including challenging pancreatic, colon and bone lesions. For evaluation purposes, we curated a dataset comprising 775 lesions from 284 patients. Each of these lesions was identified as a target lesion in a clinical context, ensuring diversity and clinical relevance within this dataset. The ULS23 benchmark is publicly accessible via uls23.grand-challenge.org, enabling researchers worldwide to assess the performance of their segmentation methods. Furthermore, we have developed and publicly released our baseline semi-supervised 3D lesion segmentation model. This model achieved an average Dice coefficient of 0.703 $\pm$ 0.240 on the challenge test set. We invite ongoing submissions to advance the development of future ULS models.
Authors: Cong Zeng, Shengkun Tang, Xianjun Yang, Yuanzhou Chen, Yiyou Sun, zhiqiang xu, Yao Li, Haifeng Chen, Wei Cheng, Dongkuan Xu
Abstract: The advent of Large Language Models (LLMs) has revolutionized text generation, producing outputs that closely mimic human writing. This blurring of lines between machine- and human-written text presents new challenges in distinguishing one from the other a task further complicated by the frequent updates and closed nature of leading proprietary LLMs. Traditional logits-based detection methods leverage surrogate models for identifying LLM-generated content when the exact logits are unavailable from black-box LLMs. However, these methods grapple with the misalignment between the distributions of the surrogate and the often undisclosed target models, leading to performance degradation, particularly with the introduction of new, closed-source models. Furthermore, while current methodologies are generally effective when the source model is identified, they falter in scenarios where the model version remains unknown, or the test set comprises outputs from various source models. To address these limitations, we present Distribution-Aligned LLMs Detection (DALD), an innovative framework that redefines the state-of-the-art performance in black-box text detection even without logits from source LLMs. DALD is designed to align the surrogate model's distribution with that of unknown target LLMs, ensuring enhanced detection capability and resilience against rapid model iterations with minimal training investment. By leveraging corpus samples from publicly accessible outputs of advanced models such as ChatGPT, GPT-4 and Claude-3, DALD fine-tunes surrogate models to synchronize with unknown source model distributions effectively.
Authors: Guojin Chen, Keren Zhu, Seunggeun Kim, Hanqing Zhu, Yao Lai, Bei Yu, David Z. Pan
Abstract: Analog layout synthesis faces significant challenges due to its dependence on manual processes, considerable time requirements, and performance instability. Current Bayesian Optimization (BO)-based techniques for analog layout synthesis, despite their potential for automation, suffer from slow convergence and extensive data needs, limiting their practical application. This paper presents the \texttt{LLANA} framework, a novel approach that leverages Large Language Models (LLMs) to enhance BO by exploiting the few-shot learning abilities of LLMs for more efficient generation of analog design-dependent parameter constraints. Experimental results demonstrate that \texttt{LLANA} not only achieves performance comparable to state-of-the-art (SOTA) BO methods but also enables a more effective exploration of the analog circuit design space, thanks to LLM's superior contextual understanding and learning efficiency. The code is available at \url{https://github.com/dekura/LLANA}.
Authors: Zhuoqun Wang, Naoki Awaya, Li Ma
Abstract: A common objective in the analysis of tabular data is estimating the conditional distribution (in contrast to only producing predictions) of a set of "outcome" variables given a set of "covariates", which is sometimes referred to as the "density regression" problem. Beyond estimation on the conditional distribution, the generative ability of drawing synthetic samples from the learned conditional distribution is also desired as it further widens the range of applications. We propose a flow-based generative model tailored for the density regression task on tabular data. Our flow applies a sequence of tree-based piecewise-linear transforms on initial uniform noise to eventually generate samples from complex conditional densities of (univariate or multivariate) outcomes given the covariates and allows efficient analytical evaluation of the fitted conditional density on any point in the sample space. We introduce a training algorithm for fitting the tree-based transforms using a divide-and-conquer strategy that transforms maximum likelihood training of the tree-flow into training a collection of binary classifiers--one at each tree split--under cross-entropy loss. We assess the performance of our method under out-of-sample likelihood evaluation and compare it with a variety of state-of-the-art conditional density learners on a range of simulated and real benchmark tabular datasets. Our method consistently achieves comparable or superior performance at a fraction of the training and sampling budget. Finally, we demonstrate the utility of our method's generative ability through an application to generating synthetic longitudinal microbiome compositional data based on training our flow on a publicly available microbiome study.
Authors: Eddy Solomon, Patricia M. Johnson, Zhengguo Tan, Radhika Tibrewala, Yvonne W. Lui, Florian Knoll, Linda Moy, Sungheon Gene Kim, Laura Heacock
Abstract: This data curation work introduces the first large-scale dataset of radial k-space and DICOM data for breast DCE-MRI acquired in diagnostic breast MRI exams. Our dataset includes case-level labels indicating patient age, menopause status, lesion status (negative, benign, and malignant), and lesion type for each case. The public availability of this dataset and accompanying reconstruction code will support research and development of fast and quantitative breast image reconstruction and machine learning methods.
Authors: Oleg Smirnov, Labinot Polisi
Abstract: In this paper, we introduce the Behavior Structformer, a method for modeling user behavior using structured tokenization within a Transformer-based architecture. By converting tracking events into dense tokens, this approach enhances model training efficiency and effectiveness. We demonstrate its superior performance through ablation studies and benchmarking against traditional tabular and semi-structured baselines. The results indicate that structured tokenization with sequential processing significantly improves behavior modeling.
Authors: Avrajit Ghosh, Xitong Zhang, Kenneth K. Sun, Qing Qu, Saiprasad Ravishankar, Rongrong Wang
Abstract: We introduce Optimal Eye Surgeon (OES), a framework for pruning and training deep image generator networks. Typically, untrained deep convolutional networks, which include image sampling operations, serve as effective image priors (Ulyanov et al., 2018). However, they tend to overfit to noise in image restoration tasks due to being overparameterized. OES addresses this by adaptively pruning networks at random initialization to a level of underparameterization. This process effectively captures low-frequency image components even without training, by just masking. When trained to fit noisy images, these pruned subnetworks, which we term Sparse-DIP, resist overfitting to noise. This benefit arises from underparameterization and the regularization effect of masking, constraining them in the manifold of image priors. We demonstrate that subnetworks pruned through OES surpass other leading pruning methods, such as the Lottery Ticket Hypothesis, which is known to be suboptimal for image recovery tasks (Wu et al., 2023). Our extensive experiments demonstrate the transferability of OES-masks and the characteristics of sparse-subnetworks for image generation. Code is available at https://github.com/Avra98/Optimal-Eye-Surgeon.git.
Authors: Mehrdad Khatir, Chandan K. Reddy
Abstract: This paper explores the concept formation and alignment within the realm of language models (LMs). We propose a mechanism for identifying concepts and their hierarchical organization within the semantic representations learned by various LMs, encompassing a spectrum from early models like Glove to the transformer-based language models like ALBERT and T5. Our approach leverages the inherent structure present in the semantic embeddings generated by these models to extract a taxonomy of concepts and their hierarchical relationships. This investigation sheds light on how LMs develop conceptual understanding and opens doors to further research to improve their ability to reason and leverage real-world knowledge. We further conducted experiments and observed the possibility of isolating these extracted conceptual representations from the reasoning modules of the transformer-based LMs. The observed concept formation along with the isolation of conceptual representations from the reasoning modules can enable targeted token engineering to open the door for potential applications in knowledge transfer, explainable AI, and the development of more modular and conceptually grounded language models.
Authors: Hao Liu, Jiahui Cheng, Wenjing Liao
Abstract: Deep learning has exhibited remarkable results across diverse areas. To understand its success, substantial research has been directed towards its theoretical foundations. Nevertheless, the majority of these studies examine how well deep neural networks can model functions with uniform regularity. In this paper, we explore a different angle: how deep neural networks can adapt to different regularity in functions across different locations and scales and nonuniform data distributions. More precisely, we focus on a broad class of functions defined by nonlinear tree-based approximation. This class encompasses a range of function types, such as functions with uniform regularity and discontinuous functions. We develop nonparametric approximation and estimation theories for this function class using deep ReLU networks. Our results show that deep neural networks are adaptive to different regularity of functions and nonuniform data distributions at different locations and scales. We apply our results to several function classes, and derive the corresponding approximation and generalization errors. The validity of our results is demonstrated through numerical experiments.
Authors: Yanling Wang, Haoyang Li, Hao Zou, Jing Zhang, Xinlei He, Qi Li, Ke Xu
Abstract: Despite the remarkable advance of large language models (LLMs), the prevalence of non-factual responses remains a common issue. This work studies non-factuality prediction (NFP), which predicts whether an LLM will generate non-factual responses to a question before the generation process. Previous efforts on NFP usually rely on extensive computation. In this work, we conduct extensive analysis to explore the capabilities of using a lightweight probe to elicit ``whether an LLM knows'' from the hidden representations of questions. Additionally, we discover that the non-factuality probe employs similar patterns for NFP across multiple LLMs. Motivated by the intriguing finding, we conduct effective transfer learning for cross-LLM NFP and propose a question-aligned strategy to ensure the efficacy of mini-batch based training.
Authors: Kai Nakaishi, Yoshihiko Nishikawa, Koji Hukushima
Abstract: The performance of large language models (LLMs) strongly depends on the \textit{temperature} parameter. Empirically, at very low temperatures, LLMs generate sentences with clear repetitive structures, while at very high temperatures, generated sentences are often incomprehensible. In this study, using GPT-2, we numerically demonstrate that the difference between the two regimes is not just a smooth change but a phase transition with singular, divergent statistical quantities. Our extensive analysis shows that critical behaviors, such as a power-law decay of correlation in a text, emerge in the LLM at the transition temperature as well as in a natural language dataset. We also discuss that several statistical quantities characterizing the criticality should be useful to evaluate the performance of LLMs.
Authors: Bo Chen, Zhilei Bei, Xingyi Cheng, Pan Li, Jie Tang, Le Song
Abstract: Multiple Sequence Alignment (MSA) plays a pivotal role in unveiling the evolutionary trajectories of protein families. The accuracy of protein structure predictions is often compromised for protein sequences that lack sufficient homologous information to construct high quality MSA. Although various methods have been proposed to generate virtual MSA under these conditions, they fall short in comprehensively capturing the intricate coevolutionary patterns within MSA or require guidance from external oracle models. Here we introduce MSAGPT, a novel approach to prompt protein structure predictions via MSA generative pretraining in the low MSA regime. MSAGPT employs a simple yet effective 2D evolutionary positional encoding scheme to model complex evolutionary patterns. Endowed by this, its flexible 1D MSA decoding framework facilitates zero or few shot learning. Moreover, we demonstrate that leveraging the feedback from AlphaFold2 can further enhance the model capacity via Rejective Fine tuning (RFT) and Reinforcement Learning from AF2 Feedback (RLAF). Extensive experiments confirm the efficacy of MSAGPT in generating faithful virtual MSA to enhance the structure prediction accuracy. The transfer learning capabilities also highlight its great potential for facilitating other protein tasks.
Authors: Yonatan Amaru, Prasanna Wudali, Yuval Elovici, Asaf Shabtai
Abstract: Advanced persistent threats (APTs) pose significant challenges for organizations, leading to data breaches, financial losses, and reputational damage. Existing provenance-based approaches for APT detection often struggle with high false positive rates, a lack of interpretability, and an inability to adapt to evolving system behavior. We introduce RAPID, a novel deep learning-based method for robust APT detection and investigation, leveraging context-aware anomaly detection and alert tracing. By utilizing self-supervised sequence learning and iteratively learned embeddings, our approach effectively adapts to dynamic system behavior. The use of provenance tracing both enriches the alerts and enhances the detection capabilities of our approach. Our extensive evaluation demonstrates RAPID's effectiveness and computational efficiency in real-world scenarios. In addition, RAPID achieves higher precision and recall than state-of-the-art methods, significantly reducing false positives. RAPID integrates contextual information and facilitates a smooth transition from detection to investigation, providing security teams with detailed insights to efficiently address APT threats.
Authors: I-Hung Hsu, Zifeng Wang, Long T. Le, Lesly Miculicich, Nanyun Peng, Chen-Yu Lee, Tomas Pfister
Abstract: Grounded generation aims to equip language models (LMs) with the ability to produce more credible and accountable responses by accurately citing verifiable sources. However, existing methods, by either feeding LMs with raw or preprocessed materials, remain prone to errors. To address this, we introduce CaLM, a novel verification framework. CaLM leverages the insight that a robust grounded response should be consistent with information derived solely from its cited sources. Our framework empowers smaller LMs, which rely less on parametric memory and excel at processing relevant information given a query, to validate the output of larger LMs. Larger LM responses that closely align with the smaller LMs' output, which relies exclusively on cited documents, are verified. Responses showing discrepancies are iteratively refined through a feedback loop. Experiments on three open-domain question-answering datasets demonstrate significant performance gains of 1.5% to 7% absolute average without any required model fine-tuning.
Authors: Sakshi Mahendru, Tejul Pandit
Abstract: We introduce Venn Diagram (VD) Prompting, an innovative prompting technique which allows Large Language Models (LLMs) to combine and synthesize information across complex, diverse and long-context documents in knowledge-intensive question-answering tasks. Generating answers from multiple documents involves numerous steps to extract relevant and unique information and amalgamate it into a cohesive response. To improve the quality of the final answer, multiple LLM calls or pretrained models are used to perform different tasks such as summarization, reorganization and customization. The approach covered in the paper focuses on replacing the multi-step strategy via a single LLM call using VD prompting. Our proposed technique also aims to eliminate the inherent position bias in the LLMs, enhancing consistency in answers by removing sensitivity to the sequence of input information. It overcomes the challenge of inconsistency traditionally associated with varying input sequences. We also explore the practical applications of the VD prompt based on our examination of the prompt's outcomes. In the experiments performed on four public benchmark question-answering datasets, VD prompting continually matches or surpasses the performance of a meticulously crafted instruction prompt which adheres to optimal guidelines and practices.
Authors: Jiancong Xiao, Ruoyu Sun, Qi Long, Weijie J. Su
Abstract: Training Deep Neural Networks (DNNs) with adversarial examples often results in poor generalization to test-time adversarial data. This paper investigates this issue, known as adversarially robust generalization, through the lens of Rademacher complexity. Building upon the studies by Khim and Loh (2018); Yin et al. (2019), numerous works have been dedicated to this problem, yet achieving a satisfactory bound remains an elusive goal. Existing works on DNNs either apply to a surrogate loss instead of the robust loss or yield bounds that are notably looser compared to their standard counterparts. In the latter case, the bounds have a higher dependency on the width $m$ of the DNNs or the dimension $d$ of the data, with an extra factor of at least $\mathcal{O}(\sqrt{m})$ or $\mathcal{O}(\sqrt{d})$. This paper presents upper bounds for adversarial Rademacher complexity of DNNs that match the best-known upper bounds in standard settings, as established in the work of Bartlett et al. (2017), with the dependency on width and dimension being $\mathcal{O}(\ln(dm))$. The central challenge addressed is calculating the covering number of adversarial function classes. We aim to construct a new cover that possesses two properties: 1) compatibility with adversarial examples, and 2) precision comparable to covers used in standard settings. To this end, we introduce a new variant of covering number called the \emph{uniform covering number}, specifically designed and proven to reconcile these two properties. Consequently, our method effectively bridges the gap between Rademacher complexity in robust and standard generalization.
Authors: Lecheng Zheng, Zhengzhang Chen, Dongjie Wang, Chengyuan Deng, Reon Matsuoka, Haifeng Chen
Abstract: Root cause analysis (RCA) is crucial for enhancing the reliability and performance of complex systems. However, progress in this field has been hindered by the lack of large-scale, open-source datasets tailored for RCA. To bridge this gap, we introduce LEMMA-RCA, a large dataset designed for diverse RCA tasks across multiple domains and modalities. LEMMA-RCA features various real-world fault scenarios from IT and OT operation systems, encompassing microservices, water distribution, and water treatment systems, with hundreds of system entities involved. We evaluate the quality of LEMMA-RCA by testing the performance of eight baseline methods on this dataset under various settings, including offline and online modes as well as single and multiple modalities. Our experimental results demonstrate the high quality of LEMMA-RCA. The dataset is publicly available at https://lemma-rca.github.io/.
Authors: Chengyuan Deng, Yiqun Duan, Xin Jin, Heng Chang, Yijun Tian, Han Liu, Henry Peng Zou, Yiqiao Jin, Yijia Xiao, Yichen Wang, Shenghao Wu, Zongxing Xie, Kuofeng Gao, Sihong He, Jun Zhuang, Lu Cheng, Haohan Wang
Abstract: Large Language Models (LLMs) have achieved unparalleled success across diverse language modeling tasks in recent years. However, this progress has also intensified ethical concerns, impacting the deployment of LLMs in everyday contexts. This paper provides a comprehensive survey of ethical challenges associated with LLMs, from longstanding issues such as copyright infringement, systematic bias, and data privacy, to emerging problems like truthfulness and social norms. We critically analyze existing research aimed at understanding, examining, and mitigating these ethical risks. Our survey underscores integrating ethical standards and societal values into the development of LLMs, thereby guiding the development of responsible and ethically aligned language models.
Authors: Jatayu Baxi, Brijesh Bhatt
Abstract: Computational morphology handles the language processing at the word level. It is one of the foundational tasks in the NLP pipeline for the development of higher level NLP applications. It mainly deals with the processing of words and word forms. Computational Morphology addresses various sub problems such as morpheme boundary detection, lemmatization, morphological feature tagging, morphological reinflection etc. In this paper, we present exhaustive survey of the methods for developing computational morphology related tools. We survey the literature in the chronological order starting from the conventional methods till the recent evolution of deep neural network based approaches. We also review the existing datasets available for this task across the languages. We discuss about the effectiveness of neural model compared with the traditional models and present some unique challenges associated with building the computational morphology tools. We conclude by discussing some recent and open research issues in this field.
Authors: Tzu-Quan Lin, Hung-yi Lee, Hao Tang
Abstract: Self-supervised speech models have shown to be useful for various tasks, but their large size limits the use in devices with low computing power and memory. In this work, we explore early exit, an approach for reducing latency by exiting the forward process of a network early. Most approaches of early exit need a separate early exit model for each task, with some even requiring fine-tuning of the entire pretrained model. We introduce Data Adaptive Self-Supervised Early Exit (DAISY), an approach that decides when to exit based on the self-supervised loss, eliminating the need for multiple round of training and fine-tuning. DAISY matches the performance of HuBERT on the MiniSUPERB benchmark, but with much faster inference times. Our analysis on the adaptivity of DAISY shows that the model exits early (using fewer layers) on clean data while exits late (using more layers) on noisy data, dynamically adjusting the computational cost of inference based on the noise level of each sample.
Authors: Susu Sun, Stefano Woerner, Andreas Maier, Lisa M. Koch, Christian F. Baumgartner
Abstract: Interpretability is crucial for machine learning algorithms in high-stakes medical applications. However, high-performing neural networks typically cannot explain their predictions. Post-hoc explanation methods provide a way to understand neural networks but have been shown to suffer from conceptual problems. Moreover, current research largely focuses on providing local explanations for individual samples rather than global explanations for the model itself. In this paper, we propose Attri-Net, an inherently interpretable model for multi-label classification that provides local and global explanations. Attri-Net first counterfactually generates class-specific attribution maps to highlight the disease evidence, then performs classification with logistic regression classifiers based solely on the attribution maps. Local explanations for each prediction can be obtained by interpreting the attribution maps weighted by the classifiers' weights. Global explanation of whole model can be obtained by jointly considering learned average representations of the attribution maps for each class (called the class centers) and the weights of the linear classifiers. To ensure the model is ``right for the right reason", we further introduce a mechanism to guide the model's explanations to align with human knowledge. Our comprehensive evaluations show that Attri-Net can generate high-quality explanations consistent with clinical knowledge while not sacrificing classification performance.
Authors: Shichong Peng, Yanshu Zhang, Ke Li
Abstract: We propose the problem of point-level 3D scene interpolation, which aims to simultaneously reconstruct a 3D scene in two states from multiple views, synthesize smooth point-level interpolations between them, and render the scene from novel viewpoints, all without any supervision between the states. The primary challenge is on achieving a smooth transition between states that may involve significant and non-rigid changes. To address these challenges, we introduce "PAPR in Motion", a novel approach that builds upon the recent Proximity Attention Point Rendering (PAPR) technique, which can deform a point cloud to match a significantly different shape and render a visually coherent scene even after non-rigid deformations. Our approach is specifically designed to maintain the temporal consistency of the geometric structure by introducing various regularization techniques for PAPR. The result is a method that can effectively bridge large scene changes and produce visually coherent and temporally smooth interpolations in both geometry and appearance. Evaluation across diverse motion types demonstrates that "PAPR in Motion" outperforms the leading neural renderer for dynamic scenes. For more results and code, please visit our project website at https://niopeng.github.io/PAPR-in-Motion/ .
Authors: Biqing Qi, Pengfei Li, Fangyuan Li, Junqi Gao, Kaiyan Zhang, Bowen Zhou
Abstract: Direct Preference Optimization (DPO) improves the alignment of large language models (LLMs) with human values by training directly on human preference datasets, eliminating the need for reward models. However, due to the presence of cross-domain human preferences, direct continual training can lead to catastrophic forgetting, limiting DPO's performance and efficiency. Inspired by intraspecific competition driving species evolution, we propose a Online Fast-Slow chasing DPO (OFS-DPO) for preference alignment, simulating competition through fast and slow chasing among models to facilitate rapid adaptation. Specifically, we first derive the regret upper bound for online learning, validating our motivation with a min-max optimization pattern. Based on this, we introduce two identical modules using Low-rank Adaptive (LoRA) with different optimization speeds to simulate intraspecific competition, and propose a new regularization term to guide their learning. To further mitigate catastrophic forgetting in cross-domain scenarios, we extend the OFS-DPO with LoRA modules combination strategy, resulting in the Cross domain Online Fast-Slow chasing DPO (COFS-DPO). This method leverages linear combinations of fast modules parameters from different task domains, fully utilizing historical information to achive continual value alignment. Experimental results show that OFS-DPO outperforms DPO in in-domain alignment, while COFS-DPO excels in cross-domain continual learning scenarios.
Authors: Yiqing Shen, Zan Chen, Michail Mamalakis, Luhan He, Haiyang Xia, Tianbin Li, Yanzhou Su, Junjun He, Yu Guang Wang
Abstract: The parallels between protein sequences and natural language in their sequential structures have inspired the application of large language models (LLMs) to protein understanding. Despite the success of LLMs in NLP, their effectiveness in comprehending protein sequences remains an open question, largely due to the absence of datasets linking protein sequences to descriptive text. Researchers have then attempted to adapt LLMs for protein understanding by integrating a protein sequence encoder with a pre-trained LLM. However, this adaptation raises a fundamental question: "Can LLMs, originally designed for NLP, effectively comprehend protein sequences as a form of language?" Current datasets fall short in addressing this question due to the lack of a direct correlation between protein sequences and corresponding text descriptions, limiting the ability to train and evaluate LLMs for protein understanding effectively. To bridge this gap, we introduce ProteinLMDataset, a dataset specifically designed for further self-supervised pretraining and supervised fine-tuning (SFT) of LLMs to enhance their capability for protein sequence comprehension. Specifically, ProteinLMDataset includes 17.46 billion tokens for pretraining and 893,000 instructions for SFT. Additionally, we present ProteinLMBench, the first benchmark dataset consisting of 944 manually verified multiple-choice questions for assessing the protein understanding capabilities of LLMs. ProteinLMBench incorporates protein-related details and sequences in multiple languages, establishing a new standard for evaluating LLMs' abilities in protein comprehension. The large language model InternLM2-7B, pretrained and fine-tuned on the ProteinLMDataset, outperforms GPT-4 on ProteinLMBench, achieving the highest accuracy score. The dataset and the benchmark are available at https://huggingface.co/datasets/tsynbio/ProteinLMBench.
URLs: https://huggingface.co/datasets/tsynbio/ProteinLMBench.
Authors: Zhijun Liu, Shuai Wang, Sho Inoue, Qibing Bai, Haizhou Li
Abstract: Audio language models have recently emerged as a promising approach for various audio generation tasks, relying on audio tokenizers to encode waveforms into sequences of discrete symbols. Audio tokenization often poses a necessary compromise between code bitrate and reconstruction accuracy. When dealing with low-bitrate audio codes, language models are constrained to process only a subset of the information embedded in the audio, which in turn restricts their generative capabilities. To circumvent these issues, we propose encoding audio as vector sequences in continuous space $\mathbb R^d$ and autoregressively generating these sequences using a decoder-only diffusion transformer (ARDiT). Our findings indicate that ARDiT excels in zero-shot text-to-speech and exhibits performance that compares to or even surpasses that of state-of-the-art models. High-bitrate continuous speech representation enables almost flawless reconstruction, allowing our model to achieve nearly perfect speech editing. Our experiments reveal that employing Integral Kullback-Leibler (IKL) divergence for distillation at each autoregressive step significantly boosts the perceived quality of the samples. Simultaneously, it condenses the iterative sampling process of the diffusion model into a single step. Furthermore, ARDiT can be trained to predict several continuous vectors in one step, significantly reducing latency during sampling. Impressively, one of our models can generate $170$ ms of $24$ kHz speech per evaluation step with minimal degradation in performance. Audio samples are available at http://ardit-tts.github.io/ .
Authors: Jason Cai, Hang Su, Monica Sunkara, Igor Shalyminov, Saab Mansour
Abstract: Large Language Models (LLMs) are powerful models for generation tasks, but they may not generate good quality outputs in their first attempt. Apart from model fine-tuning, existing approaches to improve prediction accuracy and quality typically involve LLM self-improvement / self-reflection that incorporate feedback from models themselves. Despite their effectiveness, these methods are hindered by their high computational cost and lack of scalability. In this work, we propose CERET, a method for refining text generations by considering semantic stability, entailment and inter-sample uncertainty measures. Experimental results show that CERET outperforms Self-consistency and Self-rerank baselines consistently under various task setups, by ~1.6% in Rouge-1 for abstractive summarization and ~3.5% in hit rate for question answering. Compared to LLM Self-rerank method, our approach only requires 9.4% of its latency and is more cost-effective.
Authors: Minghao Shao, Sofija Jancheska, Meet Udeshi, Brendan Dolan-Gavitt, Haoran Xi, Kimberly Milner, Boyuan Chen, Max Yin, Siddharth Garg, Prashanth Krishnamurthy, Farshad Khorrami, Ramesh Karri, Muhammad Shafique
Abstract: Large Language Models (LLMs) are being deployed across various domains today. However, their capacity to solve Capture the Flag (CTF) challenges in cybersecurity has not been thoroughly evaluated. To address this, we develop a novel method to assess LLMs in solving CTF challenges by creating a scalable, open-source benchmark database specifically designed for these applications. This database includes metadata for LLM testing and adaptive learning, compiling a diverse range of CTF challenges from popular competitions. Utilizing the advanced function calling capabilities of LLMs, we build a fully automated system with an enhanced workflow and support for external tool calls. Our benchmark dataset and automated framework allow us to evaluate the performance of five LLMs, encompassing both black-box and open-source models. This work lays the foundation for future research into improving the efficiency of LLMs in interactive cybersecurity tasks and automated task planning. By providing a specialized dataset, our project offers an ideal platform for developing, testing, and refining LLM-based approaches to vulnerability detection and resolution. Evaluating LLMs on these challenges and comparing with human performance yields insights into their potential for AI-driven cybersecurity solutions to perform real-world threat management. We make our dataset open source to public https://github.com/NYU-LLM-CTF/LLM_CTF_Database along with our playground automated framework https://github.com/NYU-LLM-CTF/llm_ctf_automation.
URLs: https://github.com/NYU-LLM-CTF/LLM_CTF_Database, https://github.com/NYU-LLM-CTF/llm_ctf_automation.
Authors: Yunhe Gao, Difei Gu, Mu Zhou, Dimitris Metaxas
Abstract: Although explainability is essential in the clinical diagnosis, most deep learning models still function as black boxes without elucidating their decision-making process. In this study, we investigate the explainable model development that can mimic the decision-making process of human experts by fusing the domain knowledge of explicit diagnostic criteria. We introduce a simple yet effective framework, Explicd, towards Explainable language-informed criteria-based diagnosis. Explicd initiates its process by querying domain knowledge from either large language models (LLMs) or human experts to establish diagnostic criteria across various concept axes (e.g., color, shape, texture, or specific patterns of diseases). By leveraging a pretrained vision-language model, Explicd injects these criteria into the embedding space as knowledge anchors, thereby facilitating the learning of corresponding visual concepts within medical images. The final diagnostic outcome is determined based on the similarity scores between the encoded visual concepts and the textual criteria embeddings. Through extensive evaluation of five medical image classification benchmarks, Explicd has demonstrated its inherent explainability and extends to improve classification performance compared to traditional black-box models.
Authors: Sayan Mandal
Abstract: Glaucoma is a chronic eye disease characterized by optic neuropathy, leading to irreversible vision loss. It progresses gradually, often remaining undiagnosed until advanced stages. Early detection is crucial to monitor atrophy and develop treatment strategies to prevent further vision impairment. Data-centric methods have enabled computer-aided algorithms for precise glaucoma diagnosis. In this study, we use deep learning models to identify complex disease traits and progression criteria, detecting subtle changes indicative of glaucoma. We explore the structure-function relationship in glaucoma progression and predict functional impairment from structural eye deterioration. We analyze statistical and machine learning methods, including deep learning techniques with optical coherence tomography (OCT) scans for accurate progression prediction. Addressing challenges like age variability, data imbalances, and noisy labels, we develop novel semi-supervised time-series algorithms: 1. Weakly-Supervised Time-Series Learning: We create a CNN-LSTM model to encode spatiotemporal features from OCT scans. This approach uses age-related progression and positive-unlabeled data to establish robust pseudo-progression criteria, bypassing gold-standard labels. 2. Semi-Supervised Time-Series Learning: Using labels from Guided Progression Analysis (GPA) in a contrastive learning scheme, the CNN-LSTM architecture learns from potentially mislabeled data to improve prediction accuracy. Our methods outperform conventional and state-of-the-art techniques.
Authors: Pranav Jeevan, Amit Sethi
Abstract: In contemporary computer vision applications, particularly image classification, architectural backbones pre-trained on large datasets like ImageNet are commonly employed as feature extractors. Despite the widespread use of these pre-trained convolutional neural networks (CNNs), there remains a gap in understanding the performance of various resource-efficient backbones across diverse domains and dataset sizes. Our study systematically evaluates multiple lightweight, pre-trained CNN backbones under consistent training settings across a variety of datasets, including natural images, medical images, galaxy images, and remote sensing images. This comprehensive analysis aims to aid machine learning practitioners in selecting the most suitable backbone for their specific problem, especially in scenarios involving small datasets where fine-tuning a pre-trained network is crucial. Even though attention-based architectures are gaining popularity, we observed that they tend to perform poorly under low data finetuning tasks compared to CNNs. We also observed that some CNN architectures such as ConvNeXt, RegNet and EfficientNet performs well compared to others on a diverse set of domains consistently. Our findings provide actionable insights into the performance trade-offs and effectiveness of different backbones, facilitating informed decision-making in model selection for a broad spectrum of computer vision domains. Our code is available here: https://github.com/pranavphoenix/Backbones
Authors: Mark Hamilton, Andrew Zisserman, John R. Hershey, William T. Freeman
Abstract: We present DenseAV, a novel dual encoder grounding architecture that learns high-resolution, semantically meaningful, and audio-visually aligned features solely through watching videos. We show that DenseAV can discover the ``meaning'' of words and the ``location'' of sounds without explicit localization supervision. Furthermore, it automatically discovers and distinguishes between these two types of associations without supervision. We show that DenseAV's localization abilities arise from a new multi-head feature aggregation operator that directly compares dense image and audio representations for contrastive learning. In contrast, many other systems that learn ``global'' audio and video representations cannot localize words and sound. Finally, we contribute two new datasets to improve the evaluation of AV representations through speech and sound prompted semantic segmentation. On these and other datasets we show DenseAV dramatically outperforms the prior art on speech and sound prompted semantic segmentation. DenseAV outperforms the previous state-of-the-art, ImageBind, on cross-modal retrieval using fewer than half of the parameters. Project Page: \href{https://aka.ms/denseav}{https://aka.ms/denseav}
Authors: Retsef Levi, Elisabeth Paulson, Georgia Perakis, Emily Zhang
Abstract: We address a core problem in causal inference: estimating heterogeneous treatment effects using panel data with general treatment patterns. Many existing methods either do not utilize the potential underlying structure in panel data or have limitations in the allowable treatment patterns. In this work, we propose and evaluate a new method that first partitions observations into disjoint clusters with similar treatment effects using a regression tree, and then leverages the (assumed) low-rank structure of the panel data to estimate the average treatment effect for each cluster. Our theoretical results establish the convergence of the resulting estimates to the true treatment effects. Computation experiments with semi-synthetic data show that our method achieves superior accuracy compared to alternative approaches, using a regression tree with no more than 40 leaves. Hence, our method provides more accurate and interpretable estimates than alternative methods.
Authors: Daniel Hothem, Ashe Miller, Timothy Proctor
Abstract: Quantum computers have the potential to revolutionize diverse fields, including quantum chemistry, materials science, and machine learning. However, contemporary quantum computers experience errors that often cause quantum programs run on them to fail. Until quantum computers can reliably execute large quantum programs, stakeholders will need fast and reliable methods for assessing a quantum computer's capability-i.e., the programs it can run and how well it can run them. Previously, off-the-shelf neural network architectures have been used to model quantum computers' capabilities, but with limited success, because these networks fail to learn the complex quantum physics that determines real quantum computers' errors. We address this shortcoming with a new quantum-physics-aware neural network architecture for learning capability models. Our architecture combines aspects of graph neural networks with efficient approximations to the physics of errors in quantum programs. This approach achieves up to $\sim50\%$ reductions in mean absolute error on both experimental and simulated data, over state-of-the-art models based on convolutional neural networks.
Authors: Li Jiang, Xiao Li, Andre Milzarek, Junwen Qiu
Abstract: Chung's lemma is a classical tool for establishing asymptotic convergence rates of (stochastic) optimization methods under strong convexity-type assumptions and appropriate polynomial diminishing step sizes. In this work, we develop a generalized version of Chung's lemma, which provides a simple non-asymptotic convergence framework for a more general family of step size rules. We demonstrate broad applicability of the proposed generalized Chung's lemma by deriving tight non-asymptotic convergence rates for a large variety of stochastic methods. In particular, we obtain partially new non-asymptotic complexity results for stochastic optimization methods, such as stochastic gradient descent and random reshuffling, under a general $(\theta,\mu)$-Polyak-Lojasiewicz (PL) condition and for various step sizes strategies, including polynomial, constant, exponential, and cosine step sizes rules. Notably, as a by-product of our analysis, we observe that exponential step sizes can adapt to the objective function's geometry, achieving the optimal convergence rate without requiring exact knowledge of the underlying landscape. Our results demonstrate that the developed variant of Chung's lemma offers a versatile, systematic, and streamlined approach to establish non-asymptotic convergence rates under general step size rules.
Authors: Cheng Tan, Dongxin Lyu, Siyuan Li, Zhangyang Gao, Jingxuan Wei, Siqi Ma, Zicheng Liu, Stan Z. Li
Abstract: Large Language Models (LLMs) have demonstrated wide-ranging applications across various fields and have shown significant potential in the academic peer-review process. However, existing applications are primarily limited to static review generation based on submitted papers, which fail to capture the dynamic and iterative nature of real-world peer reviews. In this paper, we reformulate the peer-review process as a multi-turn, long-context dialogue, incorporating distinct roles for authors, reviewers, and decision makers. We construct a comprehensive dataset containing over 26,841 papers with 92,017 reviews collected from multiple sources, including the top-tier conference and prestigious journal. This dataset is meticulously designed to facilitate the applications of LLMs for multi-turn dialogues, effectively simulating the complete peer-review process. Furthermore, we propose a series of metrics to evaluate the performance of LLMs for each role under this reformulated peer-review setting, ensuring fair and comprehensive evaluations. We believe this work provides a promising perspective on enhancing the LLM-driven peer-review process by incorporating dynamic, role-based interactions. It aligns closely with the iterative and interactive nature of real-world academic peer review, offering a robust foundation for future research and development in this area. We open-source the dataset at https://github.com/chengtan9907/ReviewMT.
Authors: Donsub Rim, Gerrit Welper
Abstract: We construct a new representation of entropy solutions to nonlinear scalar conservation laws with a smooth convex flux function in a single spatial dimension. The representation is a generalization of the method of characteristics and posseses a compositional form. While it is a nonlinear representation, the embedded dynamics of the solution in the time variable is linear. This representation is then discretized as a manifold of implicit neural representations where the feedforward neural network architecture has a low rank structure. Finally, we show that the low rank neural representation with a fixed number of layers and a small number of coefficients can approximate any entropy solution regardless of the complexity of the shock topology, while retaining the linearity of the embedded dynamics.
Authors: Kumar Manas, Stefan Zwicklbauer, Adrian Paschke
Abstract: Traffic rules formalization is crucial for verifying the compliance and safety of autonomous vehicles (AVs). However, manual translation of natural language traffic rules as formal specification requires domain knowledge and logic expertise, which limits its adaptation. This paper introduces TR2MTL, a framework that employs large language models (LLMs) to automatically translate traffic rules (TR) into metric temporal logic (MTL). It is envisioned as a human-in-loop system for AV rule formalization. It utilizes a chain-of-thought in-context learning approach to guide the LLM in step-by-step translation and generating valid and grammatically correct MTL formulas. It can be extended to various forms of temporal logic and rules. We evaluated the framework on a challenging dataset of traffic rules we created from various sources and compared it against LLMs using different in-context learning methods. Results show that TR2MTL is domain-agnostic, achieving high accuracy and generalization capability even with a small dataset. Moreover, the method effectively predicts formulas with varying degrees of logical and semantic structure in unstructured traffic rules.
Authors: Arya Akhavan, Karim Lounici, Massimiliano Pontil, Alexandre B. Tsybakov
Abstract: We study the contextual continuum bandits problem, where the learner sequentially receives a side information vector and has to choose an action in a convex set, minimizing a function associated to the context. The goal is to minimize all the underlying functions for the received contexts, leading to a dynamic (contextual) notion of regret, which is stronger than the standard static regret. Assuming that the objective functions are H\"older with respect to the contexts, we demonstrate that any algorithm achieving a sub-linear static regret can be extended to achieve a sub-linear dynamic regret. We further study the case of strongly convex and smooth functions when the observations are noisy. Inspired by the interior point method and employing self-concordant barriers, we propose an algorithm achieving a sub-linear dynamic regret. Lastly, we present a minimax lower bound, implying two key facts. First, no algorithm can achieve sub-linear dynamic regret over functions that are not continuous with respect to the context. Second, for strongly convex and smooth functions, the algorithm that we propose achieves, up to a logarithmic factor, the minimax optimal rate of dynamic regret as a function of the number of queries.
Authors: Christoph Liebender, Ranulfo Bezerra, Kazunori Ohno, Satoshi Tadokoro
Abstract: The use of AI in public spaces continually raises concerns about privacy and the protection of sensitive data. An example is the deployment of detection and recognition methods on humans, where images are provided by surveillance cameras. This results in the acquisition of great amounts of sensitive data, since the capture and transmission of images taken by such cameras happens unaltered, for them to be received by a server on the network. However, many applications do not explicitly require the identity of a given person in a scene; An anonymized representation containing information of the person's position while preserving the context of them in the scene suffices. We show how using a customized loss function on region of interests (ROI) can achieve sufficient anonymization such that human faces become unrecognizable while persons are kept detectable, by training an end-to-end optimized autoencoder for learned image compression that utilizes the flexibility of the learned analysis and reconstruction transforms for the task of mutating parts of the compression result. This approach enables compression and anonymization in one step on the capture device, instead of transmitting sensitive, nonanonymized data over the network. Additionally, we evaluate how this anonymization impacts the average precision of pre-trained foundation models on detecting faces (MTCNN) and humans (YOLOv8) in comparison to non-ANN based methods, while considering compression rate and latency.
Authors: Thomas Le Menestrel, Manuel Rivas
Abstract: Docking is a crucial component in drug discovery aimed at predicting the binding conformation and affinity between small molecules and target proteins. ML-based docking has recently emerged as a prominent approach, outpacing traditional methods like DOCK and AutoDock Vina in handling the growing scale and complexity of molecular libraries. However, the availability of comprehensive and user-friendly datasets for training and benchmarking ML-based docking algorithms remains limited. We introduce Smiles2Dock, an open large-scale multi-task dataset for molecular docking. We created a framework combining P2Rank and AutoDock Vina to dock 1.7 million ligands from the ChEMBL database against 15 AlphaFold proteins, giving us more than 25 million protein-ligand binding scores. The dataset leverages a wide range of high-accuracy AlphaFold protein models, encompasses a diverse set of biologically relevant compounds and enables researchers to benchmark all major approaches for ML-based docking such as Graph, Transformer and CNN-based methods. We also introduce a novel Transformer-based architecture for docking scores prediction and set it as an initial benchmark for our dataset. Our dataset and code are publicly available to support the development of novel ML-based methods for molecular docking to advance scientific research in this field.
Authors: Masahiro Watanabe, Naoshi Uchihira
Abstract: Digital transformation (DX) has recently become a pressing issue for many companies as the latest digital technologies, such as artificial intelligence and the Internet of Things, can be easily utilized. However, devising new business models is not easy for compa-nies, though they can improve their operations through digital technologies. Thus, business model design support methods are needed by people who lack digital tech-nology expertise. In contrast, large language models (LLMs) represented by ChatGPT and natural language processing utilizing LLMs have been developed revolutionarily. A business model design support system that utilizes these technologies has great potential. However, research on this area is scant. Accordingly, this study proposes an LLM-based method for comparing and analyzing similar companies from different business do-mains as a first step toward business model design support utilizing LLMs. This method can support idea generation in digital business model design.
Authors: Jialin Yu, Andreas Koukorinis, Nicol\`o Colombo, Yuchen Zhu, Ricardo Silva
Abstract: We consider sequential treatment regimes where each unit is exposed to combinations of interventions over time. When interventions are described by qualitative labels, such as ``close schools for a month due to a pandemic'' or ``promote this podcast to this user during this week'', it is unclear which appropriate structural assumptions allow us to generalize behavioral predictions to previously unseen combinatorial sequences. Standard black-box approaches mapping sequences of categorical variables to outputs are applicable, but they rely on poorly understood assumptions on how reliable generalization can be obtained, and may underperform under sparse sequences, temporal variability, and large action spaces. To approach that, we pose an explicit model for \emph{composition}, that is, how the effect of sequential interventions can be isolated into modules, clarifying which data conditions allow for the identification of their combined effect at different units and time steps. We show the identification properties of our compositional model, inspired by advances in causal matrix factorization methods but focusing on predictive models for novel compositions of interventions instead of matrix completion tasks and causal effect estimation. We compare our approach to flexible but generic black-box models to illustrate how structure aids prediction in sparse data conditions.
Authors: Zhan Zhang, Qin Zhang, Yang Jiao, Lin Lu, Lin Ma, Aihua Liu, Xiao Liu, Juan Zhao, Yajun Xue, Bing Wei, Mingxia Zhang, Ru Gao, Hong Zhao, Jie Lu, Fan Li, Yang Zhang, Yiming Wang, Lei Zhang, Fengwei Tian, Jie Hu, Xin Gou
Abstract: AI-aided clinical diagnosis is desired in medical care. Existing deep learning models lack explainability and mainly focus on image analysis. The recently developed Dynamic Uncertain Causality Graph (DUCG) approach is causality-driven, explainable, and invariant across different application scenarios, without problems of data collection, labeling, fitting, privacy, bias, generalization, high cost and high energy consumption. Through close collaboration between clinical experts and DUCG technicians, 46 DUCG models covering 54 chief complaints were constructed. Over 1,000 diseases can be diagnosed without triage. Before being applied in real-world, the 46 DUCG models were retrospectively verified by third-party hospitals. The verified diagnostic precisions were no less than 95%, in which the diagnostic precision for every disease including uncommon ones was no less than 80%. After verifications, the 46 DUCG models were applied in the real-world in China. Over one million real diagnosis cases have been performed, with only 17 incorrect diagnoses identified. Due to DUCG's transparency, the mistakes causing the incorrect diagnoses were found and corrected. The diagnostic abilities of the clinicians who applied DUCG frequently were improved significantly. Following the introduction to the earlier presented DUCG methodology, the recommendation algorithm for potential medical checks is presented and the key idea of DUCG is extracted.
Authors: Jeff Calder, Nadejda Drenska, Drisana Mosaphir
Abstract: This work investigates the online machine learning problem of prediction with expert advice in an adversarial setting through numerical analysis of, and experiments with, a related partial differential equation. The problem is a repeated two-person game involving decision-making at each step informed by $n$ experts in an adversarial environment. The continuum limit of this game over a large number of steps is a degenerate elliptic equation whose solution encodes the optimal strategies for both players. We develop numerical methods for approximating the solution of this equation in relatively high dimensions ($n\leq 10$) by exploiting symmetries in the equation and the solution to drastically reduce the size of the computational domain. Based on our numerical results we make a number of conjectures about the optimality of various adversarial strategies, in particular about the non-optimality of the COMB strategy.
Authors: Muthukumar K A, Amit Gurung, Priya Ranjan
Abstract: Classifying 3D MRI images for early detection of Alzheimer's disease is a critical task in medical imaging. Traditional approaches using Convolutional Neural Networks (CNNs) and Transformers face significant challenges in this domain. CNNs, while effective in capturing local spatial features, struggle with long-range dependencies and often require extensive computational resources for high-resolution 3D data. Transformers, on the other hand, excel in capturing global context but suffer from quadratic complexity in inference time and require substantial memory, making them less efficient for large-scale 3D MRI data. To address these limitations, we propose the use of Vision Mamba, an advanced model based on State Space Models (SSMs), for the classification of 3D MRI images to detect Alzheimer's disease. Vision Mamba leverages dynamic state representations and the selective scan algorithm, allowing it to efficiently capture and retain important spatial information across 3D volumes. By dynamically adjusting state transitions based on input features, Vision Mamba can selectively retain relevant information, leading to more accurate and computationally efficient processing of 3D MRI data. Our approach combines the parallelizable nature of convolutional operations during training with the efficient, recurrent processing of states during inference. This architecture not only improves computational efficiency but also enhances the model's ability to handle long-range dependencies within 3D medical images. Experimental results demonstrate that Vision Mamba outperforms traditional CNN and Transformer models accuracy, making it a promising tool for the early detection of Alzheimer's disease using 3D MRI data.
Authors: Huma Ameer, Seemab Latif, Rabia Latif
Abstract: The automated classification of stuttered speech has significant implications for timely assessments providing assistance to speech language pathologists. Despite notable advancements in the field, the cases in which multiple disfluencies occur in speech require attention. We have taken a progressive approach to fill this gap by classifying multi-stuttered speech more efficiently. The problem has been addressed by firstly curating a dataset of multi-stuttered disfluencies from SEP-28k audio clips. Secondly, employing Whisper, a state-of-the-art speech recognition model has been leveraged by using its encoder and taking the problem as multi-label classification. Thirdly, using a 6 encoder layer Whisper and experimenting with various layer freezing strategies, a computationally efficient configuration of the model was identified. The proposed configuration achieved micro, macro, and weighted F1- scores of 0.88, 0.85, and 0.87, correspondingly on an external test dataset i.e. Fluency-Bank. In addition, through layer freezing strategies, we were able to achieve the aforementioned results by fine-tuning a single encoder layer, consequently, reducing the model's trainable parameters from 20.27 million to 3.29 million. This research study unveils the contribution of the last encoder layer in the identification of disfluencies in stuttered speech. Consequently, it has led to a computationally efficient approach which makes the model more adaptable for various dialects and languages.
Authors: Qizhi Pei, Lijun Wu, Kaiyuan Gao, Jinhua Zhu, Rui Yan
Abstract: The integration of molecule and language has garnered increasing attention in molecular science. Recent advancements in Language Models (LMs) have demonstrated potential for the comprehensive modeling of molecule and language. However, existing works exhibit notable limitations. Most existing works overlook the modeling of 3D information, which is crucial for understanding molecular structures and also functions. While some attempts have been made to leverage external structure encoding modules to inject the 3D molecular information into LMs, there exist obvious difficulties that hinder the integration of molecular structure and language text, such as modality alignment and separate tuning. To bridge this gap, we propose 3D-MolT5, a unified framework designed to model both 1D molecular sequence and 3D molecular structure. The key innovation lies in our methodology for mapping fine-grained 3D substructure representations (based on 3D molecular fingerprints) to a specialized 3D token vocabulary for 3D-MolT5. This 3D structure token vocabulary enables the seamless combination of 1D sequence and 3D structure representations in a tokenized format, allowing 3D-MolT5 to encode molecular sequence (SELFIES), molecular structure, and text sequences within a unified architecture. Alongside, we further introduce 1D and 3D joint pre-training to enhance the model's comprehension of these diverse modalities in a joint representation space and better generalize to various tasks for our foundation model. Through instruction tuning on multiple downstream datasets, our proposed 3D-MolT5 shows superior performance than existing methods in molecular property prediction, molecule captioning, and text-based molecule generation tasks. Our code will be available on GitHub soon.
Authors: Leigang Qu, Haochuan Li, Tan Wang, Wenjie Wang, Yongqi Li, Liqiang Nie, Tat-Seng Chua
Abstract: How humans can efficiently and effectively acquire images has always been a perennial question. A typical solution is text-to-image retrieval from an existing database given the text query; however, the limited database typically lacks creativity. By contrast, recent breakthroughs in text-to-image generation have made it possible to produce fancy and diverse visual content, but it faces challenges in synthesizing knowledge-intensive images. In this work, we rethink the relationship between text-to-image generation and retrieval and propose a unified framework in the context of Multimodal Large Language Models (MLLMs). Specifically, we first explore the intrinsic discriminative abilities of MLLMs and introduce a generative retrieval method to perform retrieval in a training-free manner. Subsequently, we unify generation and retrieval in an autoregressive generation way and propose an autonomous decision module to choose the best-matched one between generated and retrieved images as the response to the text query. Additionally, we construct a benchmark called TIGeR-Bench, including creative and knowledge-intensive domains, to standardize the evaluation of unified text-to-image generation and retrieval. Extensive experimental results on TIGeR-Bench and two retrieval benchmarks, i.e., Flickr30K and MS-COCO, demonstrate the superiority and effectiveness of our proposed method.
Authors: Ahmed Attia
Abstract: We present a fully probabilistic approach for solving binary optimization problems with black-box objective functions and with budget constraints. In the probabilistic approach, the optimization variable is viewed as a random variable and is associated with a parametric probability distribution. The original optimization problem is replaced with an optimization over the expected value of the original objective, which is then optimized over the probability distribution parameters. The resulting optimal parameter (optimal policy) is used to sample the binary space to produce estimates of the optimal solution(s) of the original binary optimization problem. The probability distribution is chosen from the family of Bernoulli models because the optimization variable is binary. The optimization constraints generally restrict the feasibility region. This can be achieved by modeling the random variable with a conditional distribution given satisfiability of the constraints. Thus, in this work we develop conditional Bernoulli distributions to model the random variable conditioned by the total number of nonzero entries, that is, the budget constraint. This approach (a) is generally applicable to binary optimization problems with nonstochastic black-box objective functions and budget constraints; (b) accounts for budget constraints by employing conditional probabilities that sample only the feasible region and thus considerably reduces the computational cost compared with employing soft constraints; and (c) does not employ soft constraints and thus does not require tuning of a regularization parameter, for example to promote sparsity, which is challenging in sensor placement optimization problems. The proposed approach is verified numerically by using an idealized bilinear binary optimization problem and is validated by using a sensor placement experiment in a parameter identification setup.
Authors: Paulina Kulyt\.e, Francisco Vargas, Simon Valentin Mathis, Yu Guang Wang, Jos\'e Miguel Hern\'andez-Lobato, Pietro Li\`o
Abstract: Antibodies, crucial for immune defense, primarily rely on complementarity-determining regions (CDRs) to bind and neutralize antigens, such as viruses. The design of these CDRs determines the antibody's affinity and specificity towards its target. Generative models, particularly denoising diffusion probabilistic models (DDPMs), have shown potential to advance the structure-based design of CDR regions. However, only a limited dataset of bound antibody-antigen structures is available, and generalization to out-of-distribution interfaces remains a challenge. Physics based force-fields, which approximate atomic interactions, offer a coarse but universal source of information to better mold designs to target interfaces. Integrating this foundational information into diffusion models is, therefore, highly desirable. Here, we propose a novel approach to enhance the sampling process of diffusion models by integrating force field energy-based feedback. Our model, DiffForce, employs forces to guide the diffusion sampling process, effectively blending the two distributions. Through extensive experiments, we demonstrate that our method guides the model to sample CDRs with lower energy, enhancing both the structure and sequence of the generated antibodies.
Authors: William Avery, Mustafa Munir, Radu Marculescu
Abstract: To compete with existing mobile architectures, MobileViG introduces Sparse Vision Graph Attention (SVGA), a fast token-mixing operator based on the principles of GNNs. However, MobileViG scales poorly with model size, falling at most 1% behind models with similar latency. This paper introduces Mobile Graph Convolution (MGC), a new vision graph neural network (ViG) module that solves this scaling problem. Our proposed mobile vision architecture, MobileViGv2, uses MGC to demonstrate the effectiveness of our approach. MGC improves on SVGA by increasing graph sparsity and introducing conditional positional encodings to the graph operation. Our smallest model, MobileViGv2-Ti, achieves a 77.7% top-1 accuracy on ImageNet-1K, 2% higher than MobileViG-Ti, with 0.9 ms inference latency on the iPhone 13 Mini NPU. Our largest model, MobileViGv2-B, achieves an 83.4% top-1 accuracy, 0.8% higher than MobileViG-B, with 2.7 ms inference latency. Besides image classification, we show that MobileViGv2 generalizes well to other tasks. For object detection and instance segmentation on MS COCO 2017, MobileViGv2-M outperforms MobileViG-M by 1.2 $AP^{box}$ and 0.7 $AP^{mask}$, and MobileViGv2-B outperforms MobileViG-B by 1.0 $AP^{box}$ and 0.7 $AP^{mask}$. For semantic segmentation on ADE20K, MobileViGv2-M achieves 42.9% $mIoU$ and MobileViGv2-B achieves 44.3% $mIoU$. Our code can be found at \url{https://github.com/SLDGroup/MobileViGv2}.
Authors: Shlomo Salo Elia, Aviad Malachi, Vered Aharonson, Gadi Pinkas
Abstract: Domain adaptation is often hampered by exceedingly small target datasets and inaccessible source data. These conditions are prevalent in speech verification, where privacy policies and/or languages with scarce speech resources limit the availability of sufficient data. This paper explored techniques of sourcefree domain adaptation unto a limited target speech dataset for speaker verificationin data-scarce languages. Both language and channel mis-match between source and target were investigated. Fine-tuning methods were evaluated and compared across different sizes of labeled target data. A novel iterative cluster-learn algorithm was studied for unlabeled target datasets.
Authors: Avital Shafran, Roei Schuster, Vitaly Shmatikov
Abstract: Retrieval-augmented generation (RAG) systems respond to queries by retrieving relevant documents from a knowledge database, then generating an answer by applying an LLM to the retrieved documents. We demonstrate that RAG systems that operate on databases with potentially untrusted content are vulnerable to a new class of denial-of-service attacks we call jamming. An adversary can add a single ``blocker'' document to the database that will be retrieved in response to a specific query and, furthermore, result in the RAG system not answering the query - ostensibly because it lacks the information or because the answer is unsafe. We describe and analyze several methods for generating blocker documents, including a new method based on black-box optimization that does not require the adversary to know the embedding or LLM used by the target RAG system, nor access to an auxiliary LLM to generate blocker documents. We measure the efficacy of the considered methods against several LLMs and embeddings, and demonstrate that the existing safety metrics for LLMs do not capture their vulnerability to jamming. We then discuss defenses against blocker documents.
Authors: Yilin Wang, Haiyang Xu, Xiang Zhang, Zeyuan Chen, Zhizhou Sha, Zirui Wang, Zhuowen Tu
Abstract: We provide a two-way integration for the widely adopted ControlNet by integrating external condition generation algorithms into a single dense prediction method and incorporating its individually trained image generation processes into a single model. Despite its tremendous success, the ControlNet of a two-stage pipeline bears limitations in being not self-contained (e.g. calls the external condition generation algorithms) with a large model redundancy (separately trained models for different types of conditioning inputs). Our proposed OmniControlNet consolidates 1) the condition generation (e.g., HED edges, depth maps, user scribble, and animal pose) by a single multi-tasking dense prediction algorithm under the task embedding guidance and 2) the image generation process for different conditioning types under the textual embedding guidance. OmniControlNet achieves significantly reduced model complexity and redundancy while capable of producing images of comparable quality for conditioned text-to-image generation.
Authors: Yanis Labrak, Adel Moumen, Richard Dufour, Mickael Rouvier
Abstract: In the rapidly evolving landscape of spoken question-answering (SQA), the integration of large language models (LLMs) has emerged as a transformative development. Conventional approaches often entail the use of separate models for question audio transcription and answer selection, resulting in significant resource utilization and error accumulation. To tackle these challenges, we explore the effectiveness of end-to-end (E2E) methodologies for SQA in the medical domain. Our study introduces a novel zero-shot SQA approach, compared to traditional cascade systems. Through a comprehensive evaluation conducted on a new open benchmark of 8 medical tasks and 48 hours of synthetic audio, we demonstrate that our approach requires up to 14.7 times fewer resources than a combined 1.3B parameters LLM with a 1.55B parameters ASR model while improving average accuracy by 0.5\%. These findings underscore the potential of E2E methodologies for SQA in resource-constrained contexts.
Authors: Khaled Alrfou, Tian Zhao
Abstract: Medical image segmentation is crucial for disease diagnosis and monitoring. Though effective, the current segmentation networks such as UNet struggle with capturing long-range features. More accurate models such as TransUNet, Swin-UNet, and CS-UNet have higher computation complexity. To address this problem, we propose GCtx-UNet, a lightweight segmentation architecture that can capture global and local image features with accuracy better or comparable to the state-of-the-art approaches. GCtx-UNet uses vision transformer that leverages global context self-attention modules joined with local self-attention to model long and short range spatial dependencies. GCtx-UNet is evaluated on the Synapse multi-organ abdominal CT dataset, the ACDC cardiac MRI dataset, and several polyp segmentation datasets. In terms of Dice Similarity Coefficient (DSC) and Hausdorff Distance (HD) metrics, GCtx-UNet outperformed CNN-based and Transformer-based approaches, with notable gains in the segmentation of complex and small anatomical structures. Moreover, GCtx-UNet is much more efficient than the state-of-the-art approaches with smaller model size, lower computation workload, and faster training and inference speed, making it a practical choice for clinical applications.
Authors: Aidan Z. H. Yang, Haoye Tian, He Ye, Ruben Martins, Claire Le Goues
Abstract: Software security vulnerabilities allow attackers to perform malicious activities to disrupt software operations. Recent Transformer-based language models have significantly advanced vulnerability detection, surpassing the capabilities of static analysis based deep learning models. However, language models trained solely on code tokens do not capture either the explanation of vulnerability type or the data flow structure information of code, both of which are crucial for vulnerability detection. We propose a novel technique that integrates a multitask sequence-to-sequence LLM with pro-gram control flow graphs encoded as a graph neural network to achieve sequence-to-classification vulnerability detection. We introduce MSIVD, multitask self-instructed fine-tuning for vulnerability detection, inspired by chain-of-thought prompting and LLM self-instruction. Our experiments demonstrate that MSIVD achieves superior performance, outperforming the highest LLM-based vulnerability detector baseline (LineVul), with a F1 score of 0.92 on the BigVul dataset, and 0.48 on the PreciseBugs dataset. By training LLMs and GNNs simultaneously using a combination of code and explanatory metrics of a vulnerable program, MSIVD represents a promising direction for advancing LLM-based vulnerability detection that generalizes to unseen data. Based on our findings, we further discuss the necessity for new labelled security vulnerability datasets, as recent LLMs have seen or memorized prior datasets' held-out evaluation data.
Authors: Mingwei Tang, Meng Liu, Hong Li, Junjie Yang, Chenglin Wei, Boyang Li, Dai Li, Rengan Xu, Yifan Xu, Zehua Zhang, Xiangyu Wang, Linfeng Liu, Yuelei Xie, Chengye Liu, Labib Fawaz, Li Li, Hongnan Wang, Bill Zhu, Sri Reddy
Abstract: User representation is crucial for recommendation systems as it helps to deliver personalized recommendations by capturing user preferences and behaviors in low-dimensional vectors. High-quality user embeddings can capture subtle preferences, enable precise similarity calculations, and adapt to changing preferences over time to maintain relevance. The effectiveness of recommendation systems depends significantly on the quality of user embedding. We propose to asynchronously learn high fidelity user embeddings for billions of users each day from sequence based multimodal user activities in Meta platforms through a Transformer-like large scale feature learning module. The async learned user representations embeddings (ALURE) are further converted to user similarity graphs through graph learning and then combined with user realtime activities to retrieval highly related ads candidates for the entire ads delivery system. Our method shows significant gains in both offline and online experiments.
Authors: Abdelrahman Zayed, Goncalo Mordido, Ioana Baldini, Sarath Chandar
Abstract: The widespread use of large language models has brought up essential questions about the potential biases these models might learn. This led to the development of several metrics aimed at evaluating and mitigating these biases. In this paper, we first demonstrate that prompt-based fairness metrics exhibit poor agreement, as measured by correlation, raising important questions about the reliability of fairness assessment using prompts. Then, we outline six relevant reasons why such a low correlation is observed across existing metrics. Based on these insights, we propose a method called Correlated Fairness Output (CAIRO) to enhance the correlation between fairness metrics. CAIRO augments the original prompts of a given fairness metric by using several pre-trained language models and then selects the combination of the augmented prompts that achieves the highest correlation across metrics. We show a significant improvement in Pearson correlation from 0.3 and 0.18 to 0.90 and 0.98 across metrics for gender and religion biases, respectively. Our code is available at https://github.com/chandar-lab/CAIRO.
Authors: Manuel Cherep, Nikhil Singh
Abstract: Learning robust audio representations currently demands extensive datasets of real-world sound recordings. By applying artificial transformations to these recordings, models can learn to recognize similarities despite subtle variations through techniques like contrastive learning. However, these transformations are only approximations of the true diversity found in real-world sounds, which are generated by complex interactions of physical processes, from vocal cord vibrations to the resonance of musical instruments. We propose a solution to both the data scale and transformation limitations, leveraging synthetic audio. By randomly perturbing the parameters of a sound synthesizer, we generate audio doppelg\"angers-synthetic positive pairs with causally manipulated variations in timbre, pitch, and temporal envelopes. These variations, difficult to achieve through transformations of existing audio, provide a rich source of contrastive information. Despite the shift to randomly generated synthetic data, our method produces strong representations, competitive with real data on standard audio classification benchmarks. Notably, our approach is lightweight, requires no data storage, and has only a single hyperparameter, which we extensively analyze. We offer this method as a complement to existing strategies for contrastive learning in audio, using synthesized sounds to reduce the data burden on practitioners.
Authors: Sajjad Amini, Mohammadreza Teymoorianfard, Shiqing Ma, Amir Houmansadr
Abstract: We present a simple yet effective method to improve the robustness of Convolutional Neural Networks (CNNs) against adversarial examples by post-processing an adversarially trained model. Our technique, MeanSparse, cascades the activation functions of a trained model with novel operators that sparsify mean-centered feature vectors. This is equivalent to reducing feature variations around the mean, and we show that such reduced variations merely affect the model's utility, yet they strongly attenuate the adversarial perturbations and decrease the attacker's success rate. Our experiments show that, when applied to the top models in the RobustBench leaderboard, it achieves a new robustness record of 72.08% (from 71.07%) and 59.64% (from 59.56%) on CIFAR-10 and ImageNet, respectively, in term of AutoAttack accuracy. Code is available at https://github.com/SPIN-UMass/MeanSparse
Authors: Lingkai Kong, Haorui Wang, Wenhao Mu, Yuanqi Du, Yuchen Zhuang, Yifei Zhou, Yue Song, Rongzhi Zhang, Kai Wang, Chao Zhang
Abstract: Aligning large language models (LLMs) with human objectives is crucial for real-world applications. However, fine-tuning LLMs for alignment often suffers from unstable training and requires substantial computing resources. Test-time alignment techniques, such as prompting and guided decoding, do not modify the underlying model, and their performance remains dependent on the original model's capabilities. To address these challenges, we propose aligning LLMs through representation editing. The core of our method is to view a pre-trained autoregressive LLM as a discrete-time stochastic dynamical system. To achieve alignment for specific objectives, we introduce external control signals into the state space of this language dynamical system. We train a value function directly on the hidden states according to the Bellman equation, enabling gradient-based optimization to obtain the optimal control signals at test time. Our experiments demonstrate that our method outperforms existing test-time alignment techniques while requiring significantly fewer resources compared to fine-tuning methods.
Authors: Hiroyuki Hanada, Aoyama Tatsuya, Akahane Satoshi, Tomonari Tanaka, Yoshito Okura, Yu Inatsu, Noriaki Hashimoto, Shion Takeno, Taro Murayama, Hanju Lee, Shinya Kojima, Ichiro Takeuchi
Abstract: In this study, we propose a machine learning method called Distributionally Robust Safe Sample Screening (DRSSS). DRSSS aims to identify unnecessary training samples, even when the distribution of the training samples changes in the future. To achieve this, we effectively combine the distributionally robust (DR) paradigm, which aims to enhance model robustness against variations in data distribution, with the safe sample screening (SSS), which identifies unnecessary training samples prior to model training. Since we need to consider an infinite number of scenarios regarding changes in the distribution, we applied SSS because it does not require model training after the change of the distribution. In this paper, we employed the covariate shift framework to represent the distribution of training samples and reformulated the DR covariate-shift problem as a weighted empirical risk minimization problem, where the weights are subject to uncertainty within a predetermined range. By extending the existing SSS technique to accommodate this weight uncertainty, the DRSSS method is capable of reliably identifying unnecessary samples under any future distribution within a specified range. We provide a theoretical guarantee for the DRSSS method and validate its performance through numerical experiments on both synthetic and real-world datasets.
Authors: David Romero, Chenyang Lyu, Haryo Akbarianto Wibowo, Teresa Lynn, Injy Hamed, Aditya Nanda Kishore, Aishik Mandal, Alina Dragonetti, Artem Abzaliev, Atnafu Lambebo Tonja, Bontu Fufa Balcha, Chenxi Whitehouse, Christian Salamea, Dan John Velasco, David Ifeoluwa Adelani, David Le Meur, Emilio Villa-Cueva, Fajri Koto, Fauzan Farooqui, Frederico Belcavello, Ganzorig Batnasan, Gisela Vallejo, Grainne Caulfield, Guido Ivetta, Haiyue Song, Henok Biadglign Ademtew, Hern\'an Maina, Holy Lovenia, Israel Abebe Azime, Jan Christian Blaise Cruz, Jay Gala, Jiahui Geng, Jesus-German Ortiz-Barajas, Jinheon Baek, Jocelyn Dunstan, Laura Alonso Alemany, Kumaranage Ravindu Yasas Nagasinghe, Luciana Benotti, Luis Fernando D'Haro, Marcelo Viridiano, Marcos Estecha-Garitagoitia, Maria Camila Buitrago Cabrera, Mario Rodr\'iguez-Cantelar, M\'elanie Jouitteau, Mihail Mihaylov, Mohamed Fazli Mohamed Imam, Muhammad Farid Adilazuarda, Munkhjargal Gochoo, Munkh-Erdene Otgonbold, Naome Etori, Olivier Niyomugisha, Paula M\'onica Silva, Pranjal Chitale, Raj Dabre, Rendi Chevi, Ruochen Zhang, Ryandito Diandaru, Samuel Cahyawijaya, Santiago G\'ongora, Soyeong Jeong, Sukannya Purkayastha, Tatsuki Kuribayashi, Thanmay Jayakumar, Tiago Timponi Torrent, Toqeer Ehsan, Vladimir Araujo, Yova Kementchedjhieva, Zara Burzo, Zheng Wei Lim, Zheng Xin Yong, Oana Ignat, Joan Nwatu, Rada Mihalcea, Thamar Solorio, Alham Fikri Aji
Abstract: Visual Question Answering (VQA) is an important task in multimodal AI, and it is often used to test the ability of vision-language models to understand and reason on knowledge present in both visual and textual data. However, most of the current VQA models use datasets that are primarily focused on English and a few major world languages, with images that are typically Western-centric. While recent efforts have tried to increase the number of languages covered on VQA datasets, they still lack diversity in low-resource languages. More importantly, although these datasets often extend their linguistic range via translation or some other approaches, they usually keep images the same, resulting in narrow cultural representation. To address these limitations, we construct CVQA, a new Culturally-diverse multilingual Visual Question Answering benchmark, designed to cover a rich set of languages and cultures, where we engage native speakers and cultural experts in the data collection process. As a result, CVQA includes culturally-driven images and questions from across 28 countries on four continents, covering 26 languages with 11 scripts, providing a total of 9k questions. We then benchmark several Multimodal Large Language Models (MLLMs) on CVQA, and show that the dataset is challenging for the current state-of-the-art models. This benchmark can serve as a probing evaluation suite for assessing the cultural capability and bias of multimodal models and hopefully encourage more research efforts toward increasing cultural awareness and linguistic diversity in this field.
Authors: Jingru Jia, Zehua Yuan, Junhao Pan, Paul McNamara, Deming Chen
Abstract: When making decisions under uncertainty, individuals often deviate from rational behavior, which can be evaluated across three dimensions: risk preference, probability weighting, and loss aversion. Given the widespread use of large language models (LLMs) in decision-making processes, it is crucial to assess whether their behavior aligns with human norms and ethical expectations or exhibits potential biases. Several empirical studies have investigated the rationality and social behavior performance of LLMs, yet their internal decision-making tendencies and capabilities remain inadequately understood. This paper proposes a framework, grounded in behavioral economics, to evaluate the decision-making behaviors of LLMs. Through a multiple-choice-list experiment, we estimate the degree of risk preference, probability weighting, and loss aversion in a context-free setting for three commercial LLMs: ChatGPT-4.0-Turbo, Claude-3-Opus, and Gemini-1.0-pro. Our results reveal that LLMs generally exhibit patterns similar to humans, such as risk aversion and loss aversion, with a tendency to overweight small probabilities. However, there are significant variations in the degree to which these behaviors are expressed across different LLMs. We also explore their behavior when embedded with socio-demographic features, uncovering significant disparities. For instance, when modeled with attributes of sexual minority groups or physical disabilities, Claude-3-Opus displays increased risk aversion, leading to more conservative choices. These findings underscore the need for careful consideration of the ethical implications and potential biases in deploying LLMs in decision-making scenarios. Therefore, this study advocates for developing standards and guidelines to ensure that LLMs operate within ethical boundaries while enhancing their utility in complex decision-making environments.
Authors: Hongjia Yang, Guanhua Wang, Ziyu Li, Haoxiang Li, Jialan Zheng, Yuxin Hu, Xiaozhi Cao, Congyu Liao, Huihui Ye, Qiyuan Tian
Abstract: Magnetic resonance imaging (MRI) has significantly benefited from the resurgence of artificial intelligence (AI). By leveraging AI's capabilities in large-scale optimization and pattern recognition, innovative methods are transforming the MRI acquisition workflow, including planning, sequence design, and correction of acquisition artifacts. These emerging algorithms demonstrate substantial potential in enhancing the efficiency and throughput of acquisition steps. This review discusses several pivotal AI-based methods in neuro MRI acquisition, focusing on their technological advances, impact on clinical practice, and potential risks.
Authors: Shijie Wang, Saptarshi Chakraborty, Qian Qin, Ray Bai
Abstract: Mixing (or prior) density estimation is an important problem in machine learning and statistics, especially in empirical Bayes $g$-modeling where accurately estimating the prior is necessary for making good posterior inferences. In this paper, we propose neural-$g$, a new neural network-based estimator for $g$-modeling. Neural-$g$ uses a softmax output layer to ensure that the estimated prior is a valid probability density. Under default hyperparameters, we show that neural-$g$ is very flexible and capable of capturing many unknown densities, including those with flat regions, heavy tails, and/or discontinuities. In contrast, existing methods struggle to capture all of these prior shapes. We provide justification for neural-$g$ by establishing a new universal approximation theorem regarding the capability of neural networks to learn arbitrary probability mass functions. To accelerate convergence of our numerical implementation, we utilize a weighted average gradient descent approach to update the network parameters. Finally, we extend neural-$g$ to multivariate prior density estimation. We illustrate the efficacy of our approach through simulations and analyses of real datasets. A software package to implement neural-$g$ is publicly available at https://github.com/shijiew97/neuralG.
Authors: Yutong Han, Yan Yuan, Lili Mou
Abstract: Radiology report analysis provides valuable information that can aid with public health initiatives, and has been attracting increasing attention from the research community. In this work, we present a novel insight that the structure of a radiology report (namely, the Findings and Impression sections) offers different views of a radiology scan. Based on this intuition, we further propose a co-training approach, where two machine learning models are built upon the Findings and Impression sections, respectively, and use each other's information to boost performance with massive unlabeled data in a semi-supervised manner. We conducted experiments in a public health surveillance study, and results show that our co-training approach is able to improve performance using the dual views and surpass competing supervised and semi-supervised methods.
Authors: Binglei Lou, David Boland, Philip H. W. Leong
Abstract: Machine learning ensembles combine multiple base models to produce a more accurate output. They can be applied to a range of machine learning problems, including anomaly detection. In this paper, we investigate how to maximize the composability and scalability of an FPGA-based streaming ensemble anomaly detector (fSEAD). To achieve this, we propose a flexible computing architecture consisting of multiple partially reconfigurable regions, pblocks, which each implement anomaly detectors. Our proof-of-concept design supports three state-of-the-art anomaly detection algorithms: Loda, RS-Hash and xStream. Each algorithm is scalable, meaning multiple instances can be placed within a pblock to improve performance. Moreover, fSEAD is implemented using High-level synthesis (HLS), meaning further custom anomaly detectors can be supported. Pblocks are interconnected via an AXI-switch, enabling them to be composed in an arbitrary fashion before combining and merging results at run-time to create an ensemble that maximizes the use of FPGA resources and accuracy. Through utilizing reconfigurable Dynamic Function eXchange (DFX), the detector can be modified at run-time to adapt to changing environmental conditions. We compare fSEAD to an equivalent central processing unit (CPU) implementation using four standard datasets, with speed-ups ranging from $3\times$ to $8\times$.
Authors: Jiawei Liu, Jia Le Tian, Vijay Daita, Yuxiang Wei, Yifeng Ding, Yuhan Katherine Wang, Jun Yang, Lingming Zhang
Abstract: Recent advances have been improving the context windows of Large Language Models (LLMs). To quantify the real long-context capabilities of LLMs, evaluators such as the popular Needle in a Haystack have been developed to test LLMs over a large chunk of raw texts. While effective, current evaluations overlook the insight of how LLMs work with long-context code, i.e., repositories. To this end, we initiate the RepoQA benchmark to evaluate LLMs on long-context code understanding. Traditional needle testers ask LLMs to directly retrieve the answer from the context without necessary deep understanding. In RepoQA, we built our initial task, namely Searching Needle Function (SNF), which exercises LLMs to search functions given their natural-language description, i.e., LLMs cannot find the desired function if they cannot understand the description and code. RepoQA is multilingual and comprehensive: it includes 500 code search tasks gathered from 50 popular repositories across 5 modern programming languages. By evaluating 26 general and code-specific LLMs on RepoQA, we show (i) there is still a small gap between the best open and proprietary models; (ii) different models are good at different languages; and (iii) models may understand code better without comments.
Authors: Hafez Ghaemi, Shirin Jamshidi, Mohammad Mashreghi, Majid Nili Ahmadabadi, Hamed Kebriaei
Abstract: Markov games (MGs) and multi-agent reinforcement learning (MARL) are studied to model decision making in multi-agent systems. Traditionally, the objective in MG and MARL has been risk-neutral, i.e., agents are assumed to optimize a performance metric such as expected return, without taking into account subjective or cognitive preferences of themselves or of other agents. However, ignoring such preferences leads to inaccurate models of decision making in many real-world scenarios in finance, operations research, and behavioral economics. Therefore, when these preferences are present, it is necessary to incorporate a suitable measure of risk into the optimization objective of agents, which opens the door to risk-sensitive MG and MARL. In this paper, we systemically review the literature on risk sensitivity in MG and MARL that has been growing in recent years alongside other areas of reinforcement learning and game theory. We define and mathematically describe different risk measures used in MG and MARL and individually for each measure, discuss articles that incorporate it. Finally, we identify recent trends in theoretical and applied works in the field and discuss possible directions of future research.
Authors: Zichun Yu, Spandan Das, Chenyan Xiong
Abstract: Pretraining data selection has the potential to improve language model pretraining efficiency by utilizing higher-quality data from massive web data corpora. Current data selection methods, which rely on either hand-crafted rules or larger reference models, are conducted statically and do not capture the evolving data preferences during pretraining. In this paper, we introduce model-aware data selection with data influence models (MATES), where a data influence model continuously adapts to the evolving data preferences of the pretraining model and then selects the data most effective for the current pretraining progress. Specifically, we fine-tune a small data influence model to approximate oracle data preference signals collected by locally probing the pretraining model and to select data accordingly for the next pretraining stage. Experiments on Pythia and the C4 dataset demonstrate that MATES significantly outperforms random data selection on extensive downstream tasks in both zero- and few-shot settings. It doubles the gains achieved by recent data selection approaches that leverage larger reference models and reduces the total FLOPs required to reach certain performances by half. Further analysis validates the ever-changing data preferences of pretraining models and the effectiveness of our data influence models to capture them. Our code is open-sourced at https://github.com/cxcscmu/MATES.
Authors: Guanghui Yu, Robert Kasumba, Chien-Ju Ho, William Yeoh
Abstract: To enable effective human-AI collaboration, merely optimizing AI performance while ignoring humans is not sufficient. Recent research has demonstrated that designing AI agents to account for human behavior leads to improved performance in human-AI collaboration. However, a limitation of most existing approaches is their assumption that human behavior is static, irrespective of AI behavior. In reality, humans may adjust their action plans based on their observations of AI behavior. In this paper, we address this limitation by enabling a collaborative AI agent to consider the beliefs of its human partner, i.e., what the human partner thinks the AI agent is doing, and design its action plan to facilitate easier collaboration with its human partner. Specifically, we developed a model of human beliefs that accounts for how humans reason about the behavior of their AI partners. Based on this belief model, we then developed an AI agent that considers both human behavior and human beliefs in devising its strategy for working with humans. Through extensive real-world human-subject experiments, we demonstrated that our belief model more accurately predicts humans' beliefs about AI behavior. Moreover, we showed that our design of AI agents that accounts for human beliefs enhances performance in human-AI collaboration.
Authors: Dongyoon Hwang, Byungkun Lee, Hojoon Lee, Hyunseung Kim, Jaegul Choo
Abstract: Vision Transformers (ViT), when paired with large-scale pretraining, have shown remarkable performance across various computer vision tasks, primarily due to their weak inductive bias. However, while such weak inductive bias aids in pretraining scalability, this may hinder the effective adaptation of ViTs for visuo-motor control tasks as a result of the absence of control-centric inductive biases. Such absent inductive biases include spatial locality and translation equivariance bias which convolutions naturally offer. To this end, we introduce Convolution Injector (CoIn), an add-on module that injects convolutions which are rich in locality and equivariance biases into a pretrained ViT for effective adaptation in visuo-motor control. We evaluate CoIn with three distinct types of pretrained ViTs (CLIP, MVP, VC-1) across 12 varied control tasks within three separate domains (Adroit, MetaWorld, DMC), and demonstrate that CoIn consistently enhances control task performance across all experimented environments and models, validating the effectiveness of providing pretrained ViTs with control-centric biases.
Authors: Ishan Chokshi, Shrihari Vasudevan, Nachiappan Sundaram, Raaghul Ranganathan
Abstract: Software-Defined Networks (SDN) are the standard architecture for network deployment. Intrusion Detection Systems (IDS) are a pivotal part of this technology as networks become more vulnerable to new and sophisticated attacks. Machine Learning (ML)-based IDS are increasingly seen as the most effective approach to handle this issue. However, IDS datasets suffer from high class imbalance, which impacts the performance of standard ML models. We propose Sequential Binary Classification (SBC) - an algorithm for multi-class classification to address this issue. SBC is a hierarchical cascade of base classifiers, each of which can be modelled on any general binary classifier. Extensive experiments are reported on benchmark datasets that evaluate the performance of SBC under different scenarios.
Authors: Donggeun Ko, Sangwoo Jo, Dongjun Lee, Namjun Park, Jaekwang Kim
Abstract: Dataset bias is a significant challenge in machine learning, where specific attributes, such as texture or color of the images are unintentionally learned resulting in detrimental performance. To address this, previous efforts have focused on debiasing models either by developing novel debiasing algorithms or by generating synthetic data to mitigate the prevalent dataset biases. However, generative approaches to date have largely relied on using bias-specific samples from the dataset, which are typically too scarce. In this work, we propose, DiffInject, a straightforward yet powerful method to augment synthetic bias-conflict samples using a pretrained diffusion model. This approach significantly advances the use of diffusion models for debiasing purposes by manipulating the latent space. Our framework does not require any explicit knowledge of the bias types or labelling, making it a fully unsupervised setting for debiasing. Our methodology demonstrates substantial result in effectively reducing dataset bias.
Authors: Leonardo Scabini, Andre Sacilotti, Kallil M. Zielinski, Lucas C. Ribas, Bernard De Baets, Odemir M. Bruno
Abstract: Texture, a significant visual attribute in images, has been extensively investigated across various image recognition applications. Convolutional Neural Networks (CNNs), which have been successful in many computer vision tasks, are currently among the best texture analysis approaches. On the other hand, Vision Transformers (ViTs) have been surpassing the performance of CNNs on tasks such as object recognition, causing a paradigm shift in the field. However, ViTs have so far not been scrutinized for texture recognition, hindering a proper appreciation of their potential in this specific setting. For this reason, this work explores various pre-trained ViT architectures when transferred to tasks that rely on textures. We review 21 different ViT variants and perform an extensive evaluation and comparison with CNNs and hand-engineered models on several tasks, such as assessing robustness to changes in texture rotation, scale, and illumination, and distinguishing color textures, material textures, and texture attributes. The goal is to understand the potential and differences among these models when directly applied to texture recognition, using pre-trained ViTs primarily for feature extraction and employing linear classifiers for evaluation. We also evaluate their efficiency, which is one of the main drawbacks in contrast to other methods. Our results show that ViTs generally outperform both CNNs and hand-engineered models, especially when using stronger pre-training and tasks involving in-the-wild textures (images from the internet). We highlight the following promising models: ViT-B with DINO pre-training, BeiTv2, and the Swin architecture, as well as the EfficientFormer as a low-cost alternative. In terms of efficiency, although having a higher number of GFLOPs and parameters, ViT-B and BeiT(v2) can achieve a lower feature extraction time on GPUs compared to ResNet50.
Authors: Zhiquan Tan, Lai Wei, Jindong Wang, Xing Xie, Weiran Huang
Abstract: Large language models (LLMs) have achieved remarkable progress in linguistic tasks, necessitating robust evaluation frameworks to understand their capabilities and limitations. Inspired by Feynman's principle of understanding through creation, we introduce a self-knowledge evaluation framework that is easy to implement, evaluating models on their ability to comprehend and respond to self-generated questions. Our findings, based on testing multiple models across diverse tasks, reveal significant gaps in the model's self-knowledge ability. Further analysis indicates these gaps may be due to misalignment with human attention mechanisms. Additionally, fine-tuning on self-generated math task may enhance the model's math performance, highlighting the potential of the framework for efficient and insightful model evaluation and may also contribute to the improvement of LLMs.
Authors: Ra\"ul P\'erez-Gonzalo, Andreas Espersen, Antonio Agudo
Abstract: Rate-distortion optimization through neural networks has accomplished competitive results in compression efficiency and image quality. This learning-based approach seeks to minimize the compromise between compression rate and reconstructed image quality by automatically extracting and retaining crucial information, while discarding less critical details. A successful technique consists in introducing a deep hyperprior that operates within a 2-level nested latent variable model, enhancing compression by capturing complex data dependencies. This paper extends this concept by designing a generalized L-level nested generative model with a Markov chain structure. We demonstrate as L increases that a trainable prior is detrimental and explore a common dimensionality along the distinct latent variables to boost compression performance. As this structured framework can represent autoregressive coders, we outperform the hyperprior model and achieve state-of-the-art performance while reducing substantially the computational cost. Our experimental evaluation is performed on wind turbine scenarios to study its application on visual inspections
Authors: Jesse van Remmerden, Maurice Kenter, Diederik M. Roijers, Charalampos Andriotis, Yingqian Zhang, Zaharah Bukhsh
Abstract: In this paper, we introduce Multi-Objective Deep Centralized Multi-Agent Actor-Critic (MO- DCMAC), a multi-objective reinforcement learning (MORL) method for infrastructural maintenance optimization, an area traditionally dominated by single-objective reinforcement learning (RL) approaches. Previous single-objective RL methods combine multiple objectives, such as probability of collapse and cost, into a singular reward signal through reward-shaping. In contrast, MO-DCMAC can optimize a policy for multiple objectives directly, even when the utility function is non-linear. We evaluated MO-DCMAC using two utility functions, which use probability of collapse and cost as input. The first utility function is the Threshold utility, in which MO-DCMAC should minimize cost so that the probability of collapse is never above the threshold. The second is based on the Failure Mode, Effects, and Criticality Analysis (FMECA) methodology used by asset managers to asses maintenance plans. We evaluated MO-DCMAC, with both utility functions, in multiple maintenance environments, including ones based on a case study of the historical quay walls of Amsterdam. The performance of MO-DCMAC was compared against multiple rule-based policies based on heuristics currently used for constructing maintenance plans. Our results demonstrate that MO-DCMAC outperforms traditional rule-based policies across various environments and utility functions.
Authors: Julius Richter, Yi-Chiao Wu, Steven Krenn, Simon Welker, Bunlong Lay, Shinji Watanabe, Alexander Richard, Timo Gerkmann
Abstract: We release the EARS (Expressive Anechoic Recordings of Speech) dataset, a high-quality speech dataset comprising 107 speakers from diverse backgrounds, totaling in 100 hours of clean, anechoic speech data. The dataset covers a large range of different speaking styles, including emotional speech, different reading styles, non-verbal sounds, and conversational freeform speech. We benchmark various methods for speech enhancement and dereverberation on the dataset and evaluate their performance through a set of instrumental metrics. In addition, we conduct a listening test with 20 participants for the speech enhancement task, where a generative method is preferred. We introduce a blind test set that allows for automatic online evaluation of uploaded data. Dataset download links and automatic evaluation server can be found online.
Authors: Vladimir Bataev, Hainan Xu, Daniel Galvez, Vitaly Lavrukhin, Boris Ginsburg
Abstract: This paper introduces a highly efficient greedy decoding algorithm for Transducer inference. We propose a novel data structure using CUDA tensors to represent partial hypotheses in a batch that supports parallelized hypothesis manipulations. During decoding, our algorithm maximizes GPU parallelism by adopting a nested-loop design, where the inner loop consumes all blank predictions, while non-blank predictions are handled in the outer loop. Our algorithm is general-purpose and can work with both conventional Transducers and Token-and-Duration Transducers. Experiments show that the label-looping algorithm can bring a speedup up to 2.0X compared to conventional batched decoding algorithms when using batch size 32, and can be combined with other compiler or GPU call-related techniques to bring more speedup. We will open-source our implementation to benefit the research community.
Authors: Girish Kulathumani, Samruth Ananthanarayanan, Ganesh Narayanan
Abstract: Siren represents a pioneering research effort aimed at fortifying cybersecurity through strategic integration of deception, machine learning, and proactive threat analysis. Drawing inspiration from mythical sirens, this project employs sophisticated methods to lure potential threats into controlled environments. The system features a dynamic machine learning model for real-time analysis and classification, ensuring continuous adaptability to emerging cyber threats. The architectural framework includes a link monitoring proxy, a purpose-built machine learning model for dynamic link analysis, and a honeypot enriched with simulated user interactions to intensify threat engagement. Data protection within the honeypot is fortified with probabilistic encryption. Additionally, the incorporation of simulated user activity extends the system's capacity to capture and learn from potential attackers even after user disengagement. Siren introduces a paradigm shift in cybersecurity, transforming traditional defense mechanisms into proactive systems that actively engage and learn from potential adversaries. The research strives to enhance user protection while yielding valuable insights for ongoing refinement in response to the evolving landscape of cybersecurity threats.
Authors: Seungchan Ko, Sang Hyeon Park
Abstract: Physics-informed neural networks (PINNs) have recently emerged as a promising way to compute the solutions of partial differential equations (PDEs) using deep neural networks. However, despite their significant success in various fields, it remains unclear in many aspects how to effectively train PINNs if the solutions of PDEs exhibit stiff behaviors or high frequencies. In this paper, we propose a new method for training PINNs using variable-scaling techniques. This method is simple and it can be applied to a wide range of problems including PDEs with rapidly-varying solutions. Throughout various numerical experiments, we will demonstrate the effectiveness of the proposed method for these problems and confirm that it can significantly improve the training efficiency and performance of PINNs. Furthermore, based on the analysis of the neural tangent kernel (NTK), we will provide theoretical evidence for this phenomenon and show that our methods can indeed improve the performance of PINNs.
Authors: Alona Sakhnenko, Julian Sikora, Jeanette Miriam Lorenz
Abstract: In this work, we are introducing a Quantum-Classical Bayesian Neural Network (QCBNN) that is capable to perform uncertainty-aware classification of classical medical dataset. This model is a symbiosis of a classical Convolutional NN that performs ultra-sound image processing and a quantum circuit that generates its stochastic weights, within a Bayesian learning framework. To test the utility of this idea for the possible future deployment in the medical sector we track multiple behavioral metrics that capture both predictive performance as well as model's uncertainty. It is our ambition to create a hybrid model that is capable to classify samples in a more uncertainty aware fashion, which will advance the trustworthiness of these models and thus bring us step closer to utilizing them in the industry. We test multiple setups for quantum circuit for this task, and our best architectures display bigger uncertainty gap between correctly and incorrectly identified samples than its classical benchmark at an expense of a slight drop in predictive performance. The innovation of this paper is two-fold: (1) combining of different approaches that allow the stochastic weights from the quantum circuit to be continues thus allowing the model to classify application-driven dataset; (2) studying architectural features of quantum circuit that make-or-break these models, which pave the way into further investigation of more informed architectural designs.
Authors: Juan Manuel Zambrano Chaves, Eric Wang, Tao Tu, Eeshit Dhaval Vaishnav, Byron Lee, S. Sara Mahdavi, Christopher Semturs, David Fleet, Vivek Natarajan, Shekoofeh Azizi
Abstract: Developing therapeutics is a lengthy and expensive process that requires the satisfaction of many different criteria, and AI models capable of expediting the process would be invaluable. However, the majority of current AI approaches address only a narrowly defined set of tasks, often circumscribed within a particular domain. To bridge this gap, we introduce Tx-LLM, a generalist large language model (LLM) fine-tuned from PaLM-2 which encodes knowledge about diverse therapeutic modalities. Tx-LLM is trained using a collection of 709 datasets that target 66 tasks spanning various stages of the drug discovery pipeline. Using a single set of weights, Tx-LLM simultaneously processes a wide variety of chemical or biological entities(small molecules, proteins, nucleic acids, cell lines, diseases) interleaved with free-text, allowing it to predict a broad range of associated properties, achieving competitive with state-of-the-art (SOTA) performance on 43 out of 66 tasks and exceeding SOTA on 22. Among these, Tx-LLM is particularly powerful and exceeds best-in-class performance on average for tasks combining molecular SMILES representations with text such as cell line names or disease names, likely due to context learned during pretraining. We observe evidence of positive transfer between tasks with diverse drug types (e.g.,tasks involving small molecules and tasks involving proteins), and we study the impact of model size, domain finetuning, and prompting strategies on performance. We believe Tx-LLM represents an important step towards LLMs encoding biochemical knowledge and could have a future role as an end-to-end tool across the drug discovery development pipeline.
Authors: Yianxia Qian, Yongchao Zhang, Suchuan Dong
Abstract: We present the hidden-layer concatenated physics informed neural network (HLConcPINN) method, which combines hidden-layer concatenated feed-forward neural networks, a modified block time marching strategy, and a physics informed approach for approximating partial differential equations (PDEs). We analyze the convergence properties and establish the error bounds of this method for two types of PDEs: parabolic (exemplified by the heat and Burgers' equations) and hyperbolic (exemplified by the wave and nonlinear Klein-Gordon equations). We show that its approximation error of the solution can be effectively controlled by the training loss for dynamic simulations with long time horizons. The HLConcPINN method in principle allows an arbitrary number of hidden layers not smaller than two and any of the commonly-used smooth activation functions for the hidden layers beyond the first two, with theoretical guarantees. This generalizes several recent neural-network techniques, which have theoretical guarantees but are confined to two hidden layers in the network architecture and the $\tanh$ activation function. Our theoretical analyses subsequently inform the formulation of appropriate training loss functions for these PDEs, leading to physics informed neural network (PINN) type computational algorithms that differ from the standard PINN formulation. Ample numerical experiments are presented based on the proposed algorithm to validate the effectiveness of this method and confirm aspects of the theoretical analyses.
Authors: Daniel Brignac, Abhijit Mahalanobis
Abstract: Deep learners tend to perform well when trained under the closed set assumption but struggle when deployed under open set conditions. This motivates the field of Open Set Recognition in which we seek to give deep learners the ability to recognize whether a data sample belongs to the known classes trained on or comes from the surrounding infinite world. Existing open set recognition methods typically rely upon a single function for the dual task of distinguishing between knowns and unknowns as well as making known class distinction. This dual process leaves performance on the table as the function is not specialized for either task. In this work, we introduce Cascading Unknown Detection with Known Classification (Cas-DC), where we instead learn specialized functions in a cascading fashion for both known/unknown detection and fine class classification amongst the world of knowns. Our experiments and analysis demonstrate that Cas-DC handily outperforms modern methods in open set recognition when compared using AUROC scores and correct classification rate at various true positive rates.
Authors: Yi Gu, Zhendong Wang, Yueqin Yin, Yujia Xie, Mingyuan Zhou
Abstract: Aligning large language models with human preferences has emerged as a critical focus in language modeling research. Yet, integrating preference learning into Text-to-Image (T2I) generative models is still relatively uncharted territory. The Diffusion-DPO technique made initial strides by employing pairwise preference learning in diffusion models tailored for specific text prompts. We introduce Diffusion-RPO, a new method designed to align diffusion-based T2I models with human preferences more effectively. This approach leverages both prompt-image pairs with identical prompts and those with semantically related content across various modalities. Furthermore, we have developed a new evaluation metric, style alignment, aimed at overcoming the challenges of high costs, low reproducibility, and limited interpretability prevalent in current evaluations of human preference alignment. Our findings demonstrate that Diffusion-RPO outperforms established methods such as Supervised Fine-Tuning and Diffusion-DPO in tuning Stable Diffusion versions 1.5 and XL-1.0, achieving superior results in both automated evaluations of human preferences and style alignment. Our code is available at https://github.com/yigu1008/Diffusion-RPO
Authors: Yuta Nagano, Andrew Pyo, Martina Milighetti, James Henderson, John Shawe-Taylor, Benny Chain, Andreas Tiffeau-Mayer
Abstract: Computational prediction of the interaction of T cell receptors (TCRs) and their ligands is a grand challenge in immunology. Despite advances in high-throughput assays, specificity-labelled TCR data remains sparse. In other domains, the pre-training of language models on unlabelled data has been successfully used to address data bottlenecks. However, it is unclear how to best pre-train protein language models for TCR specificity prediction. Here we introduce a TCR language model called SCEPTR (Simple Contrastive Embedding of the Primary sequence of T cell Receptors), capable of data-efficient transfer learning. Through our model, we introduce a novel pre-training strategy combining autocontrastive learning and masked-language modelling, which enables SCEPTR to achieve its state-of-the-art performance. In contrast, existing protein language models and a variant of SCEPTR pre-trained without autocontrastive learning are outperformed by sequence alignment-based methods. We anticipate that contrastive learning will be a useful paradigm to decode the rules of TCR specificity.
Authors: Florian Lux, Sarina Meyer, Lyonel Behringer, Frank Zalkow, Phat Do, Matt Coler, Emanu\"el A. P. Habets, Ngoc Thang Vu
Abstract: In this work, we take on the challenging task of building a single text-to-speech synthesis system that is capable of generating speech in over 7000 languages, many of which lack sufficient data for traditional TTS development. By leveraging a novel integration of massively multilingual pretraining and meta learning to approximate language representations, our approach enables zero-shot speech synthesis in languages without any available data. We validate our system's performance through objective measures and human evaluation across a diverse linguistic landscape. By releasing our code and models publicly, we aim to empower communities with limited linguistic resources and foster further innovation in the field of speech technology.
Authors: Achraf Azize, Marc Jourdan, Aymen Al Marjani, Debabrota Basu
Abstract: Best Arm Identification (BAI) problems are progressively used for data-sensitive applications, such as designing adaptive clinical trials, tuning hyper-parameters, and conducting user studies. Motivated by the data privacy concerns invoked by these applications, we study the problem of BAI with fixed confidence in both the local and central models, i.e. $\epsilon$-local and $\epsilon$-global Differential Privacy (DP). First, to quantify the cost of privacy, we derive lower bounds on the sample complexity of any $\delta$-correct BAI algorithm satisfying $\epsilon$-global DP or $\epsilon$-local DP. Our lower bounds suggest the existence of two privacy regimes. In the high-privacy regime, the hardness depends on a coupled effect of privacy and novel information-theoretic quantities involving the Total Variation. In the low-privacy regime, the lower bounds reduce to the non-private lower bounds. We propose $\epsilon$-local DP and $\epsilon$-global DP variants of a Top Two algorithm, namely CTB-TT and AdaP-TT*, respectively. For $\epsilon$-local DP, CTB-TT is asymptotically optimal by plugging in a private estimator of the means based on Randomised Response. For $\epsilon$-global DP, our private estimator of the mean runs in arm-dependent adaptive episodes and adds Laplace noise to ensure a good privacy-utility trade-off. By adapting the transportation costs, the expected sample complexity of AdaP-TT* reaches the asymptotic lower bound up to multiplicative constants.
Authors: Gabriel Rioux, Apoorva Nitsure, Mattia Rigotti, Kristjan Greenewald, Youssef Mroueh
Abstract: Stochastic dominance is an important concept in probability theory, econometrics and social choice theory for robustly modeling agents' preferences between random outcomes. While many works have been dedicated to the univariate case, little has been done in the multivariate scenario, wherein an agent has to decide between different multivariate outcomes. By exploiting a characterization of multivariate first stochastic dominance in terms of couplings, we introduce a statistic that assesses multivariate almost stochastic dominance under the framework of Optimal Transport with a smooth cost. Further, we introduce an entropic regularization of this statistic, and establish a central limit theorem (CLT) and consistency of the bootstrap procedure for the empirical statistic. Armed with this CLT, we propose a hypothesis testing framework as well as an efficient implementation using the Sinkhorn algorithm. We showcase our method in comparing and benchmarking Large Language Models that are evaluated on multiple metrics. Our multivariate stochastic dominance test allows us to capture the dependencies between the metrics in order to make an informed and statistically significant decision on the relative performance of the models.
Authors: Jordan Voas, Raymond Mooney, David Harwath
Abstract: We introduce Semantic Parsing in Contextual Environments (SPICE), a task designed to enhance artificial agents' contextual awareness by integrating multimodal inputs with prior contexts. SPICE goes beyond traditional semantic parsing by offering a structured, interpretable framework for dynamically updating an agent's knowledge with new information, mirroring the complexity of human communication. We develop the VG-SPICE dataset, crafted to challenge agents with visual scene graph construction from spoken conversational exchanges, highlighting speech and visual data integration. We also present the Audio-Vision Dialogue Scene Parser (AViD-SP) developed for use on VG-SPICE. These innovations aim to improve multimodal information processing and integration. Both the VG-SPICE dataset and the AViD-SP model are publicly available.
Authors: Jincheng Dai, Xiaoqi Qin, Sixian Wang, Lexi Xu, Kai Niu, Ping Zhang
Abstract: Information theory and machine learning are inextricably linked and have even been referred to as "two sides of the same coin". One particularly elegant connection is the essential equivalence between probabilistic generative modeling and data compression or transmission. In this article, we reveal the dual-functionality of deep generative models that reshapes both data compression for efficiency and transmission error concealment for resiliency. We present how the contextual predictive capabilities of powerful generative models can be well positioned to be strong compressors and estimators. In this sense, we advocate for viewing the deep generative modeling problem through the lens of end-to-end communications, and evaluate the compression and error restoration capabilities of foundation generative models. We show that the kernel of many large generative models is powerful predictor that can capture complex relationships among semantic latent variables, and the communication viewpoints provide novel insights into semantic feature tokenization, contextual learning, and usage of deep generative models. In summary, our article highlights the essential connections of generative AI to source and channel coding techniques, and motivates researchers to make further explorations in this emerging topic.
Authors: Miruna Oprescu, Nathan Kallus
Abstract: Accurately predicting conditional average treatment effects (CATEs) is crucial in personalized medicine and digital platform analytics. Since often the treatments of interest cannot be directly randomized, observational data is leveraged to learn CATEs, but this approach can incur significant bias from unobserved confounding. One strategy to overcome these limitations is to seek latent quasi-experiments in instrumental variables (IVs) for the treatment, for example, a randomized intent to treat or a randomized product recommendation. This approach, on the other hand, can suffer from low compliance, i.e., IV weakness. Some subgroups may even exhibit zero compliance meaning we cannot instrument for their CATEs at all. In this paper we develop a novel approach to combine IV and observational data to enable reliable CATE estimation in the presence of unobserved confounding in the observational data and low compliance in the IV data, including no compliance for some subgroups. We propose a two-stage framework that first learns biased CATEs from the observational data, and then applies a compliance-weighted correction using IV data, effectively leveraging IV strength variability across covariates. We characterize the convergence rates of our method and validate its effectiveness through a simulation study. Additionally, we demonstrate its utility with real data by analyzing the heterogeneous effects of 401(k) plan participation on wealth.
Authors: Tianyu Zhang, Suyuchen Wang, Lu Li, Ge Zhang, Perouz Taslakian, Sai Rajeswar, Jie Fu, Bang Liu, Yoshua Bengio
Abstract: We introduce Visual Caption Restoration (VCR), a novel vision-language task that challenges models to accurately restore partially obscured texts using pixel-level hints within images. This task stems from the observation that text embedded in images is intrinsically different from common visual elements and natural language due to the need to align the modalities of vision, text, and text embedded in images. While numerous works have integrated text embedded in images into visual question-answering tasks, approaches to these tasks generally rely on optical character recognition or masked language modeling, thus reducing the task to mainly text-based processing. However, text-based processing becomes ineffective in VCR as accurate text restoration depends on the combined information from provided images, context, and subtle cues from the tiny exposed areas of masked texts. We develop a pipeline to generate synthetic images for the VCR task using image-caption pairs, with adjustable caption visibility to control the task difficulty. With this pipeline, we construct a dataset for VCR called VCR-Wiki using images with captions from Wikipedia, comprising 2.11M English and 346K Chinese entities in both easy and hard split variants. Our results reveal that current vision language models significantly lag behind human performance in the VCR task, and merely fine-tuning the models on our dataset does not lead to notable improvements. We release VCR-Wiki and the data construction code to facilitate future research.
Authors: Zhen Xing, Qi Dai, Zejia Weng, Zuxuan Wu, Yu-Gang Jiang
Abstract: Text-guided video prediction (TVP) involves predicting the motion of future frames from the initial frame according to an instruction, which has wide applications in virtual reality, robotics, and content creation. Previous TVP methods make significant breakthroughs by adapting Stable Diffusion for this task. However, they struggle with frame consistency and temporal stability primarily due to the limited scale of video datasets. We observe that pretrained Image2Video diffusion models possess good priors for video dynamics but they lack textual control. Hence, transferring Image2Video models to leverage their video dynamic priors while injecting instruction control to generate controllable videos is both a meaningful and challenging task. To achieve this, we introduce the Multi-Modal Large Language Model (MLLM) to predict future video states based on initial frames and text instructions. More specifically, we design a dual query transformer (DQFormer) architecture, which integrates the instructions and frames into the conditional embeddings for future frame prediction. Additionally, we develop Long-Short Term Temporal Adapters and Spatial Adapters that can quickly transfer general video diffusion models to specific scenarios with minimal training costs. Experimental results show that our method significantly outperforms state-of-the-art techniques on four datasets: Something Something V2, Epic Kitchen-100, Bridge Data, and UCF-101. Notably, AID achieves 91.2% and 55.5% FVD improvements on Bridge and SSv2 respectively, demonstrating its effectiveness in various domains. More examples can be found at our website https://chenhsing.github.io/AID.
Authors: Joongwon Kim, Bhargavi Paranjape, Tushar Khot, Hannaneh Hajishirzi
Abstract: Language agents perform complex tasks by using tools to execute each step precisely. However, most existing agents are based on proprietary models or designed to target specific tasks, such as mathematics or multi-hop question answering. We introduce Husky, a holistic, open-source language agent that learns to reason over a unified action space to address a diverse set of complex tasks involving numerical, tabular, and knowledge-based reasoning. Husky iterates between two stages: 1) generating the next action to take towards solving a given task and 2) executing the action using expert models and updating the current solution state. We identify a thorough ontology of actions for addressing complex tasks and curate high-quality data to train expert models for executing these actions. Our experiments show that Husky outperforms prior language agents across 14 evaluation datasets. Moreover, we introduce HuskyQA, a new evaluation set which stress tests language agents for mixed-tool reasoning, with a focus on retrieving missing knowledge and performing numerical reasoning. Despite using 7B models, Husky matches or even exceeds frontier LMs such as GPT-4 on these tasks, showcasing the efficacy of our holistic approach in addressing complex reasoning problems. Our code and models are available at https://github.com/agent-husky/Husky-v1.
Authors: Songze Li, Mingzhe Liu, Mengqi Chen
Abstract: The rapid development of large machine learning (ML) models requires a massive amount of training data, resulting in booming demands of data sharing and trading through data markets. Traditional centralized data markets suffer from low level of security, and emerging decentralized platforms are faced with efficiency and privacy challenges. In this paper, we propose OmniLytics+, the first decentralized data market, built upon blockchain and smart contract technologies, to simultaneously achieve 1) data (resp., model) privacy for the data (resp. model) owner; 2) robustness against malicious data owners; 3) efficient data validation and aggregation. Specifically, adopting the zero-knowledge (ZK) rollup paradigm, OmniLytics+ proposes to secret share encrypted local gradients, computed from the encrypted global model, with a set of untrusted off-chain servers, who collaboratively generate a ZK proof on the validity of the gradient. In this way, the storage and processing overheads are securely offloaded from blockchain verifiers, significantly improving the privacy, efficiency, and affordability over existing rollup solutions. We implement the proposed OmniLytics+ data market as an Ethereum smart contract [41]. Extensive experiments demonstrate the effectiveness of OmniLytics+ in training large ML models in presence of malicious data owner, and the substantial advantages of OmniLytics+ in gas cost and execution time over baselines.
Authors: Clara C. Wanjura, Florian Marquardt
Abstract: The widespread adoption of machine learning and artificial intelligence in all branches of science and technology has created a need for energy-efficient, alternative hardware platforms. While such neuromorphic approaches have been proposed and realised for a wide range of platforms, physically extracting the gradients required for training remains challenging as generic approaches only exist in certain cases. Equilibrium propagation (EP) is such a procedure that has been introduced and applied to classical energy-based models which relax to an equilibrium. Here, we show a direct connection between EP and Onsager reciprocity and exploit this to derive a quantum version of EP. This can be used to optimize loss functions that depend on the expectation values of observables of an arbitrary quantum system. Specifically, we illustrate this new concept with supervised and unsupervised learning examples in which the input or the solvable task is of quantum mechanical nature, e.g., the recognition of quantum many-body ground states, quantum phase exploration, sensing and phase boundary exploration. We propose that in the future quantum EP may be used to solve tasks such as quantum phase discovery with a quantum simulator even for Hamiltonians which are numerically hard to simulate or even partially unknown. Our scheme is relevant for a variety of quantum simulation platforms such as ion chains, superconducting qubit arrays, neutral atom Rydberg tweezer arrays and strongly interacting atoms in optical lattices.
Authors: Mohidul Haque Mridul, Mohammad Foysal Khan, Redwan Ahmed Rizvee, Md Mosaddek Khan
Abstract: In Multi-agent Reinforcement Learning (MARL), accurately perceiving opponents' strategies is essential for both cooperative and adversarial contexts, particularly within dynamic environments. While Proximal Policy Optimization (PPO) and related algorithms such as Actor-Critic with Experience Replay (ACER), Trust Region Policy Optimization (TRPO), and Deep Deterministic Policy Gradient (DDPG) perform well in single-agent, stationary environments, they suffer from high variance in MARL due to non-stationary and hidden policies of opponents, leading to diminished reward performance. Additionally, existing methods in MARL face significant challenges, including the need for inter-agent communication, reliance on explicit reward information, high computational demands, and sampling inefficiencies. These issues render them less effective in continuous environments where opponents may abruptly change their policies without prior notice. Against this background, we present OPS-DeMo (Online Policy Switch-Detection Model), an online algorithm that employs dynamic error decay to detect changes in opponents' policies. OPS-DeMo continuously updates its beliefs using an Assumed Opponent Policy (AOP) Bank and selects corresponding responses from a pre-trained Response Policy Bank. Each response policy is trained against consistently strategizing opponents, reducing training uncertainty and enabling the effective use of algorithms like PPO in multi-agent environments. Comparative assessments show that our approach outperforms PPO-trained models in dynamic scenarios like the Predator-Prey setting, providing greater robustness to sudden policy shifts and enabling more informed decision-making through precise opponent policy insights.
Authors: Hidde Fokkema, Dirk van der Hoeven, Tor Lattimore, Jack J. Mayo
Abstract: We introduce a computationally efficient algorithm for zeroth-order bandit convex optimisation and prove that in the adversarial setting its regret is at most $d^{3.5} \sqrt{n} \mathrm{polylog}(n, d)$ with high probability where $d$ is the dimension and $n$ is the time horizon. In the stochastic setting the bound improves to $M d^{2} \sqrt{n} \mathrm{polylog}(n, d)$ where $M \in [d^{-1/2}, d^{-1 / 4}]$ is a constant that depends on the geometry of the constraint set and the desired computational properties.
Authors: Elise Han, Chengpiao Huang, Kaizheng Wang
Abstract: Distribution-free prediction sets play a pivotal role in uncertainty quantification for complex statistical models. Their validity hinges on reliable calibration data, which may not be readily available as real-world environments often undergo unknown changes over time. In this paper, we propose a strategy for choosing an adaptive window and use the data therein to construct prediction sets. The window is selected by optimizing an estimated bias-variance tradeoff. We provide sharp coverage guarantees for our method, showing its adaptivity to the underlying temporal drift. We also illustrate its efficacy through numerical experiments on synthetic and real data.
Authors: Franz M. Rohrhofer, Stefan Posch, Clemens G\"o{\ss}nitzer, Bernhard C. Geiger
Abstract: Physics-informed neural networks (PINNs) have emerged as a promising deep learning method, capable of solving forward and inverse problems governed by differential equations. Despite their recent advance, it is widely acknowledged that PINNs are difficult to train and often require a careful tuning of loss weights when data and physics loss functions are combined by scalarization of a multi-objective (MO) problem. In this paper, we aim to understand how parameters of the physical system, such as characteristic length and time scales, the computational domain, and coefficients of differential equations affect MO optimization and the optimal choice of loss weights. Through a theoretical examination of where these system parameters appear in PINN training, we find that they effectively and individually scale the loss residuals, causing imbalances in MO optimization with certain choices of system parameters. The immediate effects of this are reflected in the apparent Pareto front, which we define as the set of loss values achievable with gradient-based training and visualize accordingly. We empirically verify that loss weights can be used successfully to compensate for the scaling of system parameters, and enable the selection of an optimal solution on the apparent Pareto front that aligns well with the physically valid solution. We further demonstrate that by altering the system parameterization, the apparent Pareto front can shift and exhibit locally convex parts, resulting in a wider range of loss weights for which gradient-based training becomes successful. This work explains the effects of system parameters on MO optimization in PINNs, and highlights the utility of proposed loss weighting schemes.
Authors: Brian Chmiel, Ron Banner, Elad Hoffer, Hilla Ben Yaacov, Daniel Soudry
Abstract: Quantization of the weights and activations is one of the main methods to reduce the computational footprint of Deep Neural Networks (DNNs) training. Current methods enable 4-bit quantization of the forward phase. However, this constitutes only a third of the training process. Reducing the computational footprint of the entire training process requires the quantization of the neural gradients, i.e., the loss gradients with respect to the outputs of intermediate neural layers. Previous works separately showed that accurate 4-bit quantization of the neural gradients needs to (1) be unbiased and (2) have a log scale. However, no previous work aimed to combine both ideas, as we do in this work. Specifically, we examine the importance of having unbiased quantization in quantized neural network training, where to maintain it, and how to combine it with logarithmic quantization. Based on this, we suggest a $\textit{logarithmic unbiased quantization}$ (LUQ) method to quantize both the forward and backward phases to 4-bit, achieving state-of-the-art results in 4-bit training without the overhead. For example, in ResNet50 on ImageNet, we achieved a degradation of 1.1%. We further improve this to a degradation of only 0.32% after three epochs of high precision fine-tuning, combined with a variance reduction method -- where both these methods add overhead comparable to previously suggested methods.
Authors: Tiancheng Qin, S. Rasoul Etesami, C\'esar A. Uribe
Abstract: Modern machine learning architectures are often highly expressive. They are usually over-parameterized and can interpolate the data by driving the empirical loss close to zero. We analyze the convergence of Local SGD (or FedAvg) for such over-parameterized models in the heterogeneous data setting and improve upon the existing literature by establishing the following convergence rates. For general convex loss functions, we establish an error bound of $\O(1/T)$ under a mild data similarity assumption and an error bound of $\O(K/T)$ otherwise, where $K$ is the number of local steps and $T$ is the total number of iterations. For non-convex loss functions we prove an error bound of $\O(K/T)$. These bounds improve upon the best previous bound of $\O(1/\sqrt{nT})$ in both cases, where $n$ is the number of nodes, when no assumption on the model being over-parameterized is made. We complete our results by providing problem instances in which our established convergence rates are tight to a constant factor with a reasonably small stepsize. Finally, we validate our theoretical results by performing large-scale numerical experiments that reveal the convergence behavior of Local SGD for practical over-parameterized deep learning models, in which the $\O(1/T)$ convergence rate of Local SGD is clearly shown.
Authors: Catarina Moreira, Yu-Liang Chou, Chihcheng Hsieh, Chun Ouyang, Joaquim Jorge, Jo\~ao Madeiras Pereira
Abstract: This study investigates the impact of machine learning models on the generation of counterfactual explanations by conducting a benchmark evaluation over three different types of models: a decision tree (fully transparent, interpretable, white-box model), a random forest (semi-interpretable, grey-box model), and a neural network (fully opaque, black-box model). We tested the counterfactual generation process using four algorithms (DiCE, WatcherCF, prototype, and GrowingSpheresCF) in the literature in 25 different datasets. Our findings indicate that: (1) Different machine learning models have little impact on the generation of counterfactual explanations; (2) Counterfactual algorithms based uniquely on proximity loss functions are not actionable and will not provide meaningful explanations; (3) One cannot have meaningful evaluation results without guaranteeing plausibility in the counterfactual generation. Algorithms that do not consider plausibility in their internal mechanisms will lead to biased and unreliable conclusions if evaluated with the current state-of-the-art metrics; (4) A counterfactual inspection analysis is strongly recommended to ensure a robust examination of counterfactual explanations and the potential identification of biases.
Authors: Brian Chmiel, Itay Hubara, Ron Banner, Daniel Soudry
Abstract: In deep learning, fine-grained N:M sparsity reduces the data footprint and bandwidth of a General Matrix multiply (GEMM) up to x2, and doubles throughput by skipping computation of zero values. So far, it was mainly only used to prune weights to accelerate the forward and backward phases. We examine how this method can be used also for the neural gradients (i.e., loss gradients with respect to the intermediate neural layer outputs). To this end, we first establish a tensor-level optimality criteria. Previous works aimed to minimize the mean-square-error (MSE) of each pruned block. We show that while minimization of the MSE works fine for pruning the weights and activations, it catastrophically fails for the neural gradients. Instead, we show that accurate pruning of the neural gradients requires an unbiased minimum-variance pruning mask. We design such specialized masks, and find that in most cases, 1:2 sparsity is sufficient for training, and 2:4 sparsity is usually enough when this is not the case. Further, we suggest combining several such methods together in order to potentially speed up training even more.
Authors: Huy Q. Le, Minh N. H. Nguyen, Shashi Raj Pandey, Chaoning Zhang, Choong Seon Hong
Abstract: In a practical setting, how to enable robust Federated Learning (FL) systems, both in terms of generalization and personalization abilities, is one important research question. It is a challenging issue due to the consequences of non-i.i.d. properties of client's data, often referred to as statistical heterogeneity, and small local data samples from the various data distributions. Therefore, to develop robust generalized global and personalized models, conventional FL methods need to redesign the knowledge aggregation from biased local models while considering huge divergence of learning parameters due to skewed client data. In this work, we demonstrate that the knowledge transfer mechanism achieves these objectives and develop a novel knowledge distillation-based approach to study the extent of knowledge transfer between the global model and local models. Henceforth, our method considers the suitability of transferring the outcome distribution and (or) the embedding vector of representation from trained models during cross-device knowledge transfer using a small proxy dataset in heterogeneous FL. In doing so, we alternatively perform cross-device knowledge transfer following general formulations as 1) global knowledge transfer and 2) on-device knowledge transfer. Through simulations on three federated datasets, we show the proposed method achieves significant speedups and high personalized performance of local models. Furthermore, the proposed approach offers a more stable algorithm than other baselines during the training, with minimal communication data load when exchanging the trained model's outcomes and representation.
Authors: Moshe Kimhi, Tal Rozen, Avi Mendelson, Chaim Baskin
Abstract: Quantized neural networks are well known for reducing the latency, power consumption, and model size without significant harm to the performance. This makes them highly appropriate for systems with limited resources and low power capacity. Mixed-precision quantization offers better utilization of customized hardware that supports arithmetic operations at different bitwidths. Quantization methods either aim to minimize the compression loss given a desired reduction or optimize a dependent variable for a specified property of the model (such as FLOPs or model size); both make the performance inefficient when deployed on specific hardware, but more importantly, quantization methods assume that the loss manifold holds a global minimum for a quantized model that copes with the global minimum of the full precision counterpart. Challenging this assumption, we argue that the optimal minimum changes as the precision changes, and thus, it is better to look at quantization as a random process, placing the foundation for a different approach to quantize neural networks, which, during the training procedure, quantizes the model to a different precision, looks at the bit allocation as a Markov Decision Process, and then, finds an optimal bitwidth allocation for measuring specified behaviors on a specific device via direct signals from the particular hardware architecture. By doing so, we avoid the basic assumption that the loss behaves the same way for a quantized model. Automatic Mixed-Precision Quantization for Edge Devices (dubbed AMED) demonstrates its superiority over current state-of-the-art schemes in terms of the trade-off between neural network accuracy and hardware efficiency, backed by a comprehensive evaluation.
Authors: Pierre Marion, Adeline Fermanian, G\'erard Biau, Jean-Philippe Vert
Abstract: Deep ResNets are recognized for achieving state-of-the-art results in complex machine learning tasks. However, the remarkable performance of these architectures relies on a training procedure that needs to be carefully crafted to avoid vanishing or exploding gradients, particularly as the depth $L$ increases. No consensus has been reached on how to mitigate this issue, although a widely discussed strategy consists in scaling the output of each layer by a factor $\alpha_L$. We show in a probabilistic setting that with standard i.i.d.~initializations, the only non-trivial dynamics is for $\alpha_L = \frac{1}{\sqrt{L}}$; other choices lead either to explosion or to identity mapping. This scaling factor corresponds in the continuous-time limit to a neural stochastic differential equation, contrarily to a widespread interpretation that deep ResNets are discretizations of neural ordinary differential equations. By contrast, in the latter regime, stability is obtained with specific correlated initializations and $\alpha_L = \frac{1}{L}$. Our analysis suggests a strong interplay between scaling and regularity of the weights as a function of the layer index. Finally, in a series of experiments, we exhibit a continuous range of regimes driven by these two parameters, which jointly impact performance before and after training.
Authors: Vikram S Chundawat, Ayush K Tarun, Murari Mandal, Mukund Lahoti, Pratik Narang
Abstract: Synthetic tabular data generation becomes crucial when real data is limited, expensive to collect, or simply cannot be used due to privacy concerns. However, producing good quality synthetic data is challenging. Several probabilistic, statistical, generative adversarial networks (GANs), and variational auto-encoder (VAEs) based approaches have been presented for synthetic tabular data generation. Once generated, evaluating the quality of the synthetic data is quite challenging. Some of the traditional metrics have been used in the literature but there is lack of a common, robust, and single metric. This makes it difficult to properly compare the effectiveness of different synthetic tabular data generation methods. In this paper we propose a new universal metric, TabSynDex, for robust evaluation of synthetic data. The proposed metric assesses the similarity of synthetic data with real data through different component scores which evaluate the characteristics that are desirable for ``high quality'' synthetic data. Being a single score metric and having an implicit bound, TabSynDex can also be used to observe and evaluate the training of neural network based approaches. This would help in obtaining insights that was not possible earlier. We present several baseline models for comparative analysis of the proposed evaluation metric with existing generative models. We also give a comparative analysis between TabSynDex and existing synthetic tabular data evaluation metrics. This shows the effectiveness and universality of our metric over the existing metrics. Source Code: \url{https://github.com/vikram2000b/tabsyndex}
Authors: Benjamin Leblanc, Pascal Germain
Abstract: We study the use of binary activated neural networks as interpretable and explainable predictors in the context of regression tasks on tabular data; more specifically, we provide guarantees on their expressiveness, present an approach based on the efficient computation of SHAP values for quantifying the relative importance of the features, hidden neurons and even weights. As the model's simplicity is instrumental in achieving interpretability, we propose a greedy algorithm for building compact binary activated networks. This approach doesn't need to fix an architecture for the network in advance: it is built one layer at a time, one neuron at a time, leading to predictors that aren't needlessly complex for a given task.
Authors: Jinxiong Zhang
Abstract: We will show how to implement binary decision tree traversal in the language of matrix computation. Our main contribution is to propose some equivalent algorithms of binary tree traversal based on a novel matrix representation of the hierarchical structure of the decision tree. Our key idea is to travel the binary decision tree by maximum inner product search. We not only implement decision tree methods without the recursive traverse but also delve into the partitioning nature of tree-based methods.
Authors: Shakir Showkat Sofi, Ivan Oseledets
Abstract: The majority of real-world processes are spatiotemporal, and the data generated by them exhibits both spatial and temporal evolution. Weather is one of the most essential processes in this domain, and weather forecasting has become a crucial part of our daily routine. Weather data analysis is considered the most complex and challenging task. Although numerical weather prediction models are currently state-of-the-art, they are resource-intensive and time-consuming. Numerous studies have proposed time series-based models as a viable alternative to numerical forecasts. Recent research in the area of time series analysis indicates significant advancements, particularly regarding the use of state-space-based models (white box) and, more recently, the integration of machine learning and deep neural network-based models (black box). The most famous examples of such models are RNNs and transformers. These models have demonstrated remarkable results in the field of time-series analysis and have demonstrated effectiveness in modelling temporal correlations. It is crucial to capture both temporal and spatial correlations for a spatiotemporal process, as the values at nearby locations and time affect the values of a spatiotemporal process at a specific point. This self-contained paper explores various regional data-driven weather forecasting methods, i.e., forecasting over multiple latitude-longitude points (matrix-shaped spatial grid) to capture spatiotemporal correlations. The results showed that spatiotemporal prediction models reduced computational costs while improving accuracy. In particular, the proposed tensor train dynamic mode decomposition-based forecasting model has comparable accuracy to the state-of-the-art models without the need for training. We provide convincing numerical experiments to show that the proposed approach is practical.
Authors: Xinliang Liu, Bo Xu, Shuhao Cao, Lei Zhang
Abstract: Neural operators have emerged as a powerful tool for learning the mapping between infinite-dimensional parameter and solution spaces of partial differential equations (PDEs). In this work, we focus on multiscale PDEs that have important applications such as reservoir modeling and turbulence prediction. We demonstrate that for such PDEs, the spectral bias towards low-frequency components presents a significant challenge for existing neural operators. To address this challenge, we propose a hierarchical attention neural operator (HANO) inspired by the hierarchical matrix approach. HANO features a scale-adaptive interaction range and self-attentions over a hierarchy of levels, enabling nested feature computation with controllable linear cost and encoding/decoding of multiscale solution space. We also incorporate an empirical $H^1$ loss function to enhance the learning of high-frequency components. Our numerical experiments demonstrate that HANO outperforms state-of-the-art (SOTA) methods for representative multiscale problems.
Authors: Ziqiao Wang, Yongyi Mao
Abstract: Stochastic differential equations (SDEs) have been shown recently to characterize well the dynamics of training machine learning models with SGD. When the generalization error of the SDE approximation closely aligns with that of SGD in expectation, it provides two opportunities for understanding better the generalization behaviour of SGD through its SDE approximation. Firstly, viewing SGD as full-batch gradient descent with Gaussian gradient noise allows us to obtain trajectory-based generalization bound using the information-theoretic bound from Xu and Raginsky [2017]. Secondly, assuming mild conditions, we estimate the steady-state weight distribution of SDE and use information-theoretic bounds from Xu and Raginsky [2017] and Negrea et al. [2019] to establish terminal-state-based generalization bounds. Our proposed bounds have some advantages, notably the trajectory-based bound outperforms results in Wang and Mao [2022], and the terminal-state-based bound exhibits a fast decay rate comparable to stability-based bounds.
Authors: Kai-Chieh Hsu, Duy Phuong Nguyen, Jaime Fern\'andez Fisac
Abstract: The deployment of robots in uncontrolled environments requires them to operate robustly under previously unseen scenarios, like irregular terrain and wind conditions. Unfortunately, while rigorous safety frameworks from robust optimal control theory scale poorly to high-dimensional nonlinear dynamics, control policies computed by more tractable "deep" methods lack guarantees and tend to exhibit little robustness to uncertain operating conditions. This work introduces a novel approach enabling scalable synthesis of robust safety-preserving controllers for robotic systems with general nonlinear dynamics subject to bounded modeling error by combining game-theoretic safety analysis with adversarial reinforcement learning in simulation. Following a soft actor-critic scheme, a safety-seeking fallback policy is co-trained with an adversarial "disturbance" agent that aims to invoke the worst-case realization of model error and training-to-deployment discrepancy allowed by the designer's uncertainty. While the learned control policy does not intrinsically guarantee safety, it is used to construct a real-time safety filter (or shield) with robust safety guarantees based on forward reachability rollouts. This shield can be used in conjunction with a safety-agnostic control policy, precluding any task-driven actions that could result in loss of safety. We evaluate our learning-based safety approach in a 5D race car simulator, compare the learned safety policy to the numerically obtained optimal solution, and empirically validate the robust safety guarantee of our proposed safety shield against worst-case model discrepancy.
Authors: Yongxin Guo, Xiaoying Tang, Tao Lin
Abstract: Federated Learning (FL) is a machine learning paradigm that safeguards privacy by retaining client data on edge devices. However, optimizing FL in practice can be challenging due to the diverse and heterogeneous nature of the learning system. Though recent research has focused on improving the optimization of FL when distribution shifts occur among clients, ensuring global performance when multiple types of distribution shifts occur simultaneously among clients -- such as feature distribution shift, label distribution shift, and concept shift -- remain under-explored. In this paper, we identify the learning challenges posed by the simultaneous occurrence of diverse distribution shifts and propose a clustering principle to overcome these challenges. Through our research, we find that existing methods fail to address the clustering principle. Therefore, we propose a novel clustering algorithm framework, dubbed as FedRC, which adheres to our proposed clustering principle by incorporating a bi-level optimization problem and a novel objective function. Extensive experiments demonstrate that FedRC significantly outperforms other SOTA cluster-based FL methods. Our code is available at \url{https://github.com/LINs-lab/FedRC}.
Authors: Bahareh Tasdighi, Abdullah Akg\"ul, Manuel Haussmann, Kenny Kazimirzak Brink, Melih Kandemir
Abstract: Actor-critic algorithms address the dual goals of reinforcement learning (RL), policy evaluation and improvement via two separate function approximators. The practicality of this approach comes at the expense of training instability, caused mainly by the destructive effect of the approximation errors of the critic on the actor. We tackle this bottleneck by employing an existing Probably Approximately Correct (PAC) Bayesian bound for the first time as the critic training objective of the Soft Actor-Critic (SAC) algorithm. We further demonstrate that online learning performance improves significantly when a stochastic actor explores multiple futures by critic-guided random search. We observe our resulting algorithm to compare favorably against the state-of-the-art SAC implementation on multiple classical control and locomotion tasks in terms of both sample efficiency and regret.
Authors: Hankyul Koh, Joon-hyuk Ko, Wonho Jhe
Abstract: We propose a new high-performance activation function, Moderate Adaptive Linear Units (MoLU), for the deep neural network. The MoLU is a simple, beautiful and powerful activation function that can be a good main activation function among hundreds of activation functions. Because the MoLU is made up of the elementary functions, not only it is a infinite diffeomorphism (i.e. smooth and infinitely differentiable over whole domains), but also it decreases training time.
Authors: Keller Jordan
Abstract: Typical neural network trainings have substantial variance in test-set performance between repeated runs, impeding hyperparameter comparison and training reproducibility. In this work we present the following results towards understanding this variation. (1) Despite having significant variance on their test-sets, we demonstrate that standard CIFAR-10 and ImageNet trainings have little variance in performance on the underlying test-distributions from which their test-sets are sampled. (2) We show that these trainings make approximately independent errors on their test-sets. That is, the event that a trained network makes an error on one particular example does not affect its chances of making errors on other examples, relative to their average rates over repeated runs of training with the same hyperparameters. (3) We prove that the variance of neural network trainings on their test-sets is a downstream consequence of the class-calibration property discovered by Jiang et al. (2021). Our analysis yields a simple formula which accurately predicts variance for the binary classification case. (4) We conduct preliminary studies of data augmentation, learning rate, finetuning instability and distribution-shift through the lens of variance between runs.
Authors: Micah Goldblum, Marc Finzi, Keefer Rowan, Andrew Gordon Wilson
Abstract: No free lunch theorems for supervised learning state that no learner can solve all problems or that all learners achieve exactly the same accuracy on average over a uniform distribution on learning problems. Accordingly, these theorems are often referenced in support of the notion that individual problems require specially tailored inductive biases. While virtually all uniformly sampled datasets have high complexity, real-world problems disproportionately generate low-complexity data, and we argue that neural network models share this same preference, formalized using Kolmogorov complexity. Notably, we show that architectures designed for a particular domain, such as computer vision, can compress datasets on a variety of seemingly unrelated domains. Our experiments show that pre-trained and even randomly initialized language models prefer to generate low-complexity sequences. Whereas no free lunch theorems seemingly indicate that individual problems require specialized learners, we explain how tasks that often require human intervention such as picking an appropriately sized model when labeled data is scarce or plentiful can be automated into a single learning algorithm. These observations justify the trend in deep learning of unifying seemingly disparate problems with an increasingly small set of machine learning models.
Authors: Elias Werner, Nishant Kumar, Matthias Lieber, Sunna Torge, Stefan Gumhold, Wolfgang E. Nagel
Abstract: Concept drift detection is crucial for many AI systems to ensure the system's reliability. These systems often have to deal with large amounts of data or react in real-time. Thus, drift detectors must meet computational requirements or constraints with a comprehensive performance evaluation. However, so far, the focus of developing drift detectors is on inference quality, e.g. accuracy, but not on computational performance, such as runtime. Many of the previous works consider computational performance only as a secondary objective and do not have a benchmark for such evaluation. Hence, we propose and explain performance engineering for unsupervised concept drift detection that reflects on computational complexities, benchmarking, and performance analysis. We provide the computational complexities of existing unsupervised drift detectors and discuss why further computational performance investigations are required. Hence, we state and substantiate the aspects of a benchmark for unsupervised drift detection reflecting on inference quality and computational performance. Furthermore, we demonstrate performance analysis practices that have proven their effectiveness in High-Performance Computing, by tracing two drift detectors and displaying their performance data.
Authors: Brian Belgodere, Pierre Dognin, Adam Ivankay, Igor Melnyk, Youssef Mroueh, Aleksandra Mojsilovic, Jiri Navratil, Apoorva Nitsure, Inkit Padhi, Mattia Rigotti, Jerret Ross, Yair Schiff, Radhika Vedpathak, Richard A. Young
Abstract: Real-world data often exhibits bias, imbalance, and privacy risks. Synthetic datasets have emerged to address these issues. This paradigm relies on generative AI models to generate unbiased, privacy-preserving data while maintaining fidelity to the original data. However, assessing the trustworthiness of synthetic datasets and models is a critical challenge. We introduce a holistic auditing framework that comprehensively evaluates synthetic datasets and AI models. It focuses on preventing bias and discrimination, ensures fidelity to the source data, assesses utility, robustness, and privacy preservation. We demonstrate the framework's effectiveness by auditing various generative models across diverse use cases like education, healthcare, banking, and human resources, spanning different data modalities such as tabular, time-series, vision, and natural language. This holistic assessment is essential for compliance with regulatory safeguards. We introduce a trustworthiness index to rank synthetic datasets based on their safeguards trade-offs. Furthermore, we present a trustworthiness-driven model selection and cross-validation process during training, exemplified with "TrustFormers" across various data types. This approach allows for controllable trustworthiness trade-offs in synthetic data creation. Our auditing framework fosters collaboration among stakeholders, including data scientists, governance experts, internal reviewers, external certifiers, and regulators. This transparent reporting should become a standard practice to prevent bias, discrimination, and privacy violations, ensuring compliance with policies and providing accountability, safety, and performance guarantees.
Authors: Jiangyi Fang, Liyue Chen, Di Chai, Yayao Hong, Xiuhuai Xie, Longbiao Chen, Leye Wang
Abstract: Spatiotemporal crowd flow prediction is one of the key technologies in smart cities. Currently, there are two major pain points that plague related research and practitioners. Firstly, crowd flow is related to multiple domain knowledge factors; however, due to the diversity of application scenarios, it is difficult for subsequent work to make reasonable and comprehensive use of domain knowledge. Secondly, with the development of deep learning technology, the implementation of relevant techniques has become increasingly complex; reproducing advanced models has become a time-consuming and increasingly cumbersome task. To address these issues, we design and implement a spatiotemporal crowd flow prediction toolbox called UCTB (Urban Computing Tool Box), which integrates multiple spatiotemporal domain knowledge and state-of-the-art models simultaneously. The relevant code and supporting documents have been open-sourced at https://github.com/uctb/UCTB.
Authors: Jonathan Svirsky, Ofir Lindenbaum
Abstract: Clustering is a fundamental learning task widely used as a first step in data analysis. For example, biologists use cluster assignments to analyze genome sequences, medical records, or images. Since downstream analysis is typically performed at the cluster level, practitioners seek reliable and interpretable clustering models. We propose a new deep-learning framework for general domain tabular data that predicts interpretable cluster assignments at the instance and cluster levels. First, we present a self-supervised procedure to identify the subset of the most informative features from each data point. Then, we design a model that predicts cluster assignments and a gate matrix that provides cluster-level feature selection. Overall, our model provides cluster assignments with an indication of the driving feature for each sample and each cluster. We show that the proposed method can reliably predict cluster assignments in biological, text, image, and physics tabular datasets. Furthermore, using previously proposed metrics, we verify that our model leads to interpretable results at a sample and cluster level. Our code is available at https://github.com/jsvir/idc.
Authors: Tim Schneider, Amin Totounferoush, Wolfgang Nowak, Steffen Staab
Abstract: Symbolic Regression (SR) allows for the discovery of scientific equations from data. To limit the large search space of possible equations, prior knowledge has been expressed in terms of formal grammars that characterize subsets of arbitrary strings. However, there is a mismatch between context-free grammars required to express the set of syntactically correct equations, missing closure properties of the former, and a tree structure of the latter. Our contributions are to (i) compactly express experts' prior beliefs about which equations are more likely to be expected by probabilistic Regular Tree Expressions (pRTE), and (ii) adapt Bayesian inference to make such priors efficiently available for symbolic regression encoded as finite state machines. Our scientific case studies show its effectiveness in soil science to find sorption isotherms and for modeling hyper-elastic materials.
Authors: Rishub Tamirisa, John Won, Chengjun Lu, Ron Arel, Andy Zhou
Abstract: Recent advancements in federated learning (FL) seek to increase client-level performance by fine-tuning client parameters on local data or personalizing architectures for the local task. Existing methods for such personalization either prune a global model or fine-tune a global model on a local client distribution. However, these existing methods either personalize at the expense of retaining important global knowledge, or predetermine network layers for fine-tuning, resulting in suboptimal storage of global knowledge within client models. Enlightened by the lottery ticket hypothesis, we first introduce a hypothesis for finding optimal client subnetworks to locally fine-tune while leaving the rest of the parameters frozen. We then propose a novel FL framework, FedSelect, using this procedure that directly personalizes both client subnetwork structure and parameters, via the simultaneous discovery of optimal parameters for personalization and the rest of parameters for global aggregation during training. We show that this method achieves promising results on CIFAR-10.
Authors: Frank Nielsen
Abstract: Data sets of multivariate normal distributions abound in many scientific areas like diffusion tensor imaging, structure tensor computer vision, radar signal processing, machine learning, just to name a few. In order to process those normal data sets for downstream tasks like filtering, classification or clustering, one needs to define proper notions of dissimilarities between normals and paths joining them. The Fisher-Rao distance defined as the Riemannian geodesic distance induced by the Fisher information metric is such a principled metric distance which however is not known in closed-form excepts for a few particular cases. In this work, we first report a fast and robust method to approximate arbitrarily finely the Fisher-Rao distance between multivariate normal distributions. Second, we introduce a class of distances based on diffeomorphic embeddings of the normal manifold into a submanifold of the higher-dimensional symmetric positive-definite cone corresponding to the manifold of centered normal distributions. We show that the projective Hilbert distance on the cone yields a metric on the embedded normal submanifold and we pullback that cone distance with its associated straight line Hilbert cone geodesics to obtain a distance and smooth paths between normal distributions. Compared to the Fisher-Rao distance approximation, the pullback Hilbert cone distance is computationally light since it requires to compute only the extreme minimal and maximal eigenvalues of matrices. Finally, we show how to use those distances in clustering tasks.
Authors: Alireza Rafiei, Ronald Moore, Sina Jahromi, Farshid Hajati, Rishikesan Kamaleswaran
Abstract: As a subset of machine learning, meta-learning, or learning to learn, aims at improving the model's capabilities by employing prior knowledge and experience. A meta-learning paradigm can appropriately tackle the conventional challenges of traditional learning approaches, such as insufficient number of samples, domain shifts, and generalization. These unique characteristics position meta-learning as a suitable choice for developing influential solutions in various healthcare contexts, where the available data is often insufficient, and the data collection methodologies are different. This survey discusses meta-learning broad applications in the healthcare domain to provide insight into how and where it can address critical healthcare challenges. We first describe the theoretical foundations and pivotal methods of meta-learning. We then divide the employed meta-learning approaches in the healthcare domain into two main categories of multi/single-task learning and many/few-shot learning and survey the studies. Finally, we highlight the current challenges in meta-learning research, discuss the potential solutions, and provide future perspectives on meta-learning in healthcare.
Authors: Feng Xie, Biwei Huang, Zhengming Chen, Ruichu Cai, Clark Glymour, Zhi Geng, Kun Zhang
Abstract: We investigate the task of learning causal structure in the presence of latent variables, including locating latent variables and determining their quantity, and identifying causal relationships among both latent and observed variables. To this end, we propose a Generalized Independent Noise (GIN) condition for linear non-Gaussian acyclic causal models that incorporate latent variables, which establishes the independence between a linear combination of certain measured variables and some other measured variables. Specifically, for two observed random vectors $\bf{Y}$ and $\bf{Z}$, GIN holds if and only if $\omega^{\intercal}\mathbf{Y}$ and $\mathbf{Z}$ are independent, where $\omega$ is a non-zero parameter vector determined by the cross-covariance between $\mathbf{Y}$ and $\mathbf{Z}$. We then give necessary and sufficient graphical criteria of the GIN condition in linear non-Gaussian acyclic models. Roughly speaking, GIN implies the existence of a set $\mathcal{S}$ such that $\mathcal{S}$ is causally earlier (w.r.t. the causal ordering) than $\mathbf{Y}$, and that every active (collider-free) path between $\mathbf{Y}$ and $\mathbf{Z}$ must contain a node from $\mathcal{S}$. Interestingly, we find that the independent noise condition (i.e., if there is no confounder, causes are independent of the residual derived from regressing the effect on the causes) can be seen as a special case of GIN. With such a connection between GIN and latent causal structures, we further leverage the proposed GIN condition, together with a well-designed search procedure, to efficiently estimate Linear, Non-Gaussian Latent Hierarchical Models (LiNGLaHs), where latent confounders may also be causally related and may even follow a hierarchical structure. We show that the causal structure of a LiNGLaH is identifiable in light of GIN conditions. Experimental results show the effectiveness of the proposed method.
Authors: Giovanni Minelli, Mirco Musolesi
Abstract: Robust coordination skills enable agents to operate cohesively in shared environments, together towards a common goal and, ideally, individually without hindering each other's progress. To this end, this paper presents Coordinated QMIX (CoMIX), a novel training framework for decentralized agents that enables emergent coordination through flexible policies, allowing at the same time independent decision-making at individual level. CoMIX models selfish and collaborative behavior as incremental steps in each agent's decision process. This allows agents to dynamically adapt their behavior to different situations balancing independence and collaboration. Experiments using a variety of simulation environments demonstrate that CoMIX outperforms baselines on collaborative tasks. The results validate our incremental approach as effective technique for improving coordination in multi-agent systems.
Authors: Alexander Bukharin, Yixiao Li, Pengcheng He, Tuo Zhao
Abstract: Reward design is a fundamental, yet challenging aspect of reinforcement learning (RL). Researchers typically utilize feedback signals from the environment to handcraft a reward function, but this process is not always effective due to the varying scale and intricate dependencies of the feedback signals. This paper shows by exploiting certain structures, one can ease the reward design process. Specifically, we propose a hierarchical reward modeling framework -- HERON for scenarios: (I) The feedback signals naturally present hierarchy; (II) The reward is sparse, but with less important surrogate feedback to help policy learning. Both scenarios allow us to design a hierarchical decision tree induced by the importance ranking of the feedback signals to compare RL trajectories. With such preference data, we can then train a reward model for policy learning. We apply HERON to several RL applications, and we find that our framework can not only train high performing agents on a variety of difficult tasks, but also provide additional benefits such as improved sample efficiency and robustness. Our code is available at \url{https://github.com/abukharin3/HERON}.
Authors: Xinyue Hu, Zenan Sun, Yi Nian, Yichen Wang, Yifang Dang, Fang Li, Jingna Feng, Evan Yu, Cui Tao
Abstract: Background: Alzheimer's disease and related dementias (ADRD) ranks as the sixth leading cause of death in the US, underlining the importance of accurate ADRD risk prediction. While recent advancement in ADRD risk prediction have primarily relied on imaging analysis, yet not all patients undergo medical imaging before an ADRD diagnosis. Merging machine learning with claims data can reveal additional risk factors and uncover interconnections among diverse medical codes. Objective: Our goal is to utilize Graph Neural Networks (GNNs) with claims data for ADRD risk prediction. Addressing the lack of human-interpretable reasons behind these predictions, we introduce an innovative method to evaluate relationship importance and its influence on ADRD risk prediction, ensuring comprehensive interpretation. Methods: We employed Variationally Regularized Encoder-decoder Graph Neural Network (VGNN) for estimating ADRD likelihood. We created three scenarios to assess the model's efficiency, using Random Forest and Light Gradient Boost Machine as baselines. We further used our relation importance method to clarify the key relationships for ADRD risk prediction. Results: VGNN surpassed other baseline models by 10% in the area under the receiver operating characteristic. The integration of the GNN model and relation importance interpretation could potentially play an essential role in providing valuable insight into factors that may contribute to or delay ADRD progression. Conclusions: Employing a GNN approach with claims data enhances ADRD risk prediction and provides insights into the impact of interconnected medical code relationships. This methodology not only enables ADRD risk modeling but also shows potential for other image analysis predictions using claims data.
Authors: Yash Sinha, Murari Mandal, Mohan Kankanhalli
Abstract: Graph unlearning has emerged as a pivotal method to delete information from a pre-trained graph neural network (GNN). One may delete nodes, a class of nodes, edges, or a class of edges. An unlearning method enables the GNN model to comply with data protection regulations (i.e., the right to be forgotten), adapt to evolving data distributions, and reduce the GPU-hours carbon footprint by avoiding repetitive retraining. Existing partitioning and aggregation-based methods have limitations due to their poor handling of local graph dependencies and additional overhead costs. More recently, GNNDelete offered a model-agnostic approach that alleviates some of these issues. Our work takes a novel approach to address these challenges in graph unlearning through knowledge distillation, as it distills to delete in GNN (D2DGN). It is a model-agnostic distillation framework where the complete graph knowledge is divided and marked for retention and deletion. It performs distillation with response-based soft targets and feature-based node embedding while minimizing KL divergence. The unlearned model effectively removes the influence of deleted graph elements while preserving knowledge about the retained graph elements. D2DGN surpasses the performance of existing methods when evaluated on various real-world graph datasets by up to $43.1\%$ (AUC) in edge and node unlearning tasks. Other notable advantages include better efficiency, better performance in removing target elements, preservation of performance for the retained elements, and zero overhead costs. Notably, our D2DGN surpasses the state-of-the-art GNNDelete in AUC by $2.4\%$, improves membership inference ratio by $+1.3$, requires $10.2\times10^6$ fewer FLOPs per forward pass and up to $\mathbf{3.2}\times$ faster.
Authors: Lele Cao, Gustaf Halvardsson, Andrew McCornack, Vilhelm von Ehrenheim, Pawel Herman
Abstract: This paper addresses the growing application of data-driven approaches within the Private Equity (PE) industry, particularly in sourcing investment targets (i.e., companies) for Venture Capital (VC) and Growth Capital (GC). We present a comprehensive review of the relevant approaches and propose a novel approach leveraging a Transformer-based Multivariate Time Series Classifier (TMTSC) for predicting the success likelihood of any candidate company. The objective of our research is to optimize sourcing performance for VC and GC investments by formally defining the sourcing problem as a multivariate time series classification task. We consecutively introduce the key components of our implementation which collectively contribute to the successful application of TMTSC in VC/GC sourcing: input features, model architecture, optimization target, and investor-centric data processing. Our extensive experiments on two real-world investment tasks, benchmarked towards three popular baselines, demonstrate the effectiveness of our approach in improving decision making within the VC and GC industry.
Authors: Ben Finkelshtein, Xingyue Huang, Michael Bronstein, \.Ismail \.Ilkan Ceylan
Abstract: Graph neural networks are popular architectures for graph machine learning, based on iterative computation of node representations of an input graph through a series of invariant transformations. A large class of graph neural networks follow a standard message-passing paradigm: at every layer, each node state is updated based on an aggregate of messages from its neighborhood. In this work, we propose a novel framework for training graph neural networks, where every node is viewed as a player that can choose to either 'listen', 'broadcast', 'listen and broadcast', or to 'isolate'. The standard message propagation scheme can then be viewed as a special case of this framework where every node 'listens and broadcasts' to all neighbors. Our approach offers a more flexible and dynamic message-passing paradigm, where each node can determine its own strategy based on their state, effectively exploring the graph topology while learning. We provide a theoretical analysis of the new message-passing scheme which is further supported by an extensive empirical analysis on a synthetic dataset and on real-world datasets.
Authors: Yinan Huang, William Lu, Joshua Robinson, Yu Yang, Muhan Zhang, Stefanie Jegelka, Pan Li
Abstract: Designing effective positional encodings for graphs is key to building powerful graph transformers and enhancing message-passing graph neural networks. Although widespread, using Laplacian eigenvectors as positional encodings faces two fundamental challenges: (1) \emph{Non-uniqueness}: there are many different eigendecompositions of the same Laplacian, and (2) \emph{Instability}: small perturbations to the Laplacian could result in completely different eigenspaces, leading to unpredictable changes in positional encoding. Despite many attempts to address non-uniqueness, most methods overlook stability, leading to poor generalization on unseen graph structures. We identify the cause of instability to be a ``hard partition'' of eigenspaces. Hence, we introduce Stable and Expressive Positional Encodings (SPE), an architecture for processing eigenvectors that uses eigenvalues to ``softly partition'' eigenspaces. SPE is the first architecture that is (1) provably stable, and (2) universally expressive for basis invariant functions whilst respecting all symmetries of eigenvectors. Besides guaranteed stability, we prove that SPE is at least as expressive as existing methods, and highly capable of counting graph structures. Finally, we evaluate the effectiveness of our method on molecular property prediction, and out-of-distribution generalization tasks, finding improved generalization compared to existing positional encoding methods. Our code is available at \url{https://github.com/Graph-COM/SPE}.
Authors: Sangjun Park, JinYeong Bak
Abstract: Making neural networks remember over the long term has been a longstanding issue. Although several external memory techniques have been introduced, most focus on retaining recent information in the short term. Regardless of its importance, information tends to be fatefully forgotten over time. We present Memoria, a memory system for artificial neural networks, drawing inspiration from humans and applying various neuroscientific and psychological theories. The experimental results prove the effectiveness of Memoria in the diverse tasks of sorting, language modeling, and classification, surpassing conventional techniques. Engram analysis reveals that Memoria exhibits the primacy, recency, and temporal contiguity effects which are characteristics of human memory.
Authors: Huijie Zhang, Jinfan Zhou, Yifu Lu, Minzhe Guo, Peng Wang, Liyue Shen, Qing Qu
Abstract: In this work, we investigate an intriguing and prevalent phenomenon of diffusion models which we term as "consistent model reproducibility": given the same starting noise input and a deterministic sampler, different diffusion models often yield remarkably similar outputs. We confirm this phenomenon through comprehensive experiments, implying that different diffusion models consistently reach the same data distribution and scoring function regardless of diffusion model frameworks, model architectures, or training procedures. More strikingly, our further investigation implies that diffusion models are learning distinct distributions affected by the training data size. This is supported by the fact that the model reproducibility manifests in two distinct training regimes: (i) "memorization regime", where the diffusion model overfits to the training data distribution, and (ii) "generalization regime", where the model learns the underlying data distribution. Our study also finds that this valuable property generalizes to many variants of diffusion models, including those for conditional use, solving inverse problems, and model fine-tuning. Finally, our work raises numerous intriguing theoretical questions for future investigation and highlights practical implications regarding training efficiency, model privacy, and the controlled generation of diffusion models.
Authors: Apoorva Nitsure, Youssef Mroueh, Mattia Rigotti, Kristjan Greenewald, Brian Belgodere, Mikhail Yurochkin, Jiri Navratil, Igor Melnyk, Jerret Ross
Abstract: We propose a distributional framework for benchmarking socio-technical risks of foundation models with quantified statistical significance. Our approach hinges on a new statistical relative testing based on first and second order stochastic dominance of real random variables. We show that the second order statistics in this test are linked to mean-risk models commonly used in econometrics and mathematical finance to balance risk and utility when choosing between alternatives. Using this framework, we formally develop a risk-aware approach for foundation model selection given guardrails quantified by specified metrics. Inspired by portfolio optimization and selection theory in mathematical finance, we define a metrics portfolio for each model as a means to aggregate a collection of metrics, and perform model selection based on the stochastic dominance of these portfolios. The statistical significance of our tests is backed theoretically by an asymptotic analysis via central limit theorems instantiated in practice via a bootstrap variance estimate. We use our framework to compare various large language models regarding risks related to drifting from instructions and outputting toxic content.
Authors: Ekin Ergen, Moritz Grillo
Abstract: We study the expressivity of ReLU neural networks in the setting of a binary classification problem from a topological perspective. Recently, empirical studies showed that neural networks operate by changing topology, transforming a topologically complicated data set into a topologically simpler one as it passes through the layers. This topological simplification has been measured by Betti numbers, which are algebraic invariants of a topological space. We use the same measure to establish lower and upper bounds on the topological simplification a ReLU neural network can achieve with a given architecture. We therefore contribute to a better understanding of the expressivity of ReLU neural networks in the context of binary classification problems by shedding light on their ability to capture the underlying topological structure of the data. In particular the results show that deep ReLU neural networks are exponentially more powerful than shallow ones in terms of topological simplification. This provides a mathematically rigorous explanation why deeper networks are better equipped to handle complex and topologically rich data sets.
Authors: Quentin Bouniot, Ievgen Redko, Anton Mallasto, Charlotte Laclau, Karol Arndt, Oliver Struckmeier, Markus Heinonen, Ville Kyrki, Samuel Kaski
Abstract: In the last decade, we have witnessed the introduction of several novel deep neural network (DNN) architectures exhibiting ever-increasing performance across diverse tasks. Explaining the upward trend of their performance, however, remains difficult as different DNN architectures of comparable depth and width -- common factors associated with their expressive power -- may exhibit a drastically different performance even when trained on the same dataset. In this paper, we introduce the concept of the non-linearity signature of DNN, the first theoretically sound solution for approximately measuring the non-linearity of deep neural networks. Built upon a score derived from closed-form optimal transport mappings, this signature provides a better understanding of the inner workings of a wide range of DNN architectures and learning paradigms, with a particular emphasis on the computer vision task. We provide extensive experimental results that highlight the practical usefulness of the proposed non-linearity signature and its potential for long-reaching implications. The code for our work is available at https://github.com/qbouniot/AffScoreDeep
Authors: Sabina J. Sloman, Ayush Bharti, Julien Martinelli, Samuel Kaski
Abstract: In many settings, such as scientific inference, optimization, and transfer learning, the learner has a well-defined objective, which can be treated as estimation of a target parameter, and no intrinsic interest in characterizing the entire data-generating process. Usually, the learner must also contend with additional sources of uncertainty or variables -- with nuisance parameters. Bayesian active learning, or sequential optimal experimental design, can straightforwardly accommodate the presence of nuisance parameters, and so is a natural active learning framework for such problems. However, the introduction of nuisance parameters can lead to bias in the Bayesian learner's estimate of the target parameters, a phenomenon we refer to as negative interference. We characterize the threat of negative interference and how it fundamentally changes the nature of the Bayesian active learner's task. We show that the extent of negative interference can be extremely large, and that accurate estimation of the nuisance parameters is critical to reducing it. The Bayesian active learner is confronted with a dilemma: whether to spend a finite acquisition budget in pursuit of estimation of the target or of the nuisance parameters. Our setting encompasses Bayesian transfer learning as a special case, and our results shed light on the phenomenon of negative transfer between learning environments.
Authors: Kai Yan, Alexander G. Schwing, Yu-xiong Wang
Abstract: In real-world scenarios, arbitrary interactions with the environment can often be costly, and actions of expert demonstrations are not always available. To reduce the need for both, offline Learning from Observations (LfO) is extensively studied: the agent learns to solve a task given only expert states and task-agnostic non-expert state-action pairs. The state-of-the-art DIstribution Correction Estimation (DICE) methods, as exemplified by SMODICE, minimize the state occupancy divergence between the learner's and empirical expert policies. However, such methods are limited to either $f$-divergences (KL and $chi^2$) or Wasserstein distance with Rubinstein duality, the latter of which constrains the underlying distance metric crucial to the performance of Wasserstein-based solutions. To enable more flexible distance metrics, we propose Primal Wasserstein DICE (PW-DICE). It minimizes the primal Wasserstein distance between the learner and expert state occupancies and leverages a contrastively learned distance metric. Theoretically, our framework is a generalization of SMODICE, and is the first work that unifies $f$-divergence and Wasserstein minimization. Empirically, we find that PW-DICE improves upon several state-of-the-art methods. The code is available at https://github.com/KaiYan289/PW-DICE.
Authors: Huayi Tang, Yong Liu
Abstract: We develop generalization bounds for transductive learning algorithms in the context of information theory and PAC-Bayesian theory, covering both the random sampling setting and the random splitting setting. We show that the transductive generalization gap can be bounded by the mutual information between training labels selection and the hypothesis. By introducing the concept of transductive supersamples, we translate results depicted by various information measures from the inductive learning setting to the transductive learning setting. We further establish PAC-Bayesian bounds with weaker assumptions on the loss function and numbers of training and test data points. Finally, we present the upper bounds for adaptive optimization algorithms and demonstrate the applications of results on semi-supervised learning and graph learning scenarios. Our theoretic results are validated on both synthetic and real-world datasets.
Authors: Shivam Gupta, Aditya Parulekar, Eric Price, Zhiyang Xun
Abstract: Diffusion models have become the most popular approach to deep generative modeling of images, largely due to their empirical performance and reliability. From a theoretical standpoint, a number of recent works~\cite{chen2022,chen2022improved,benton2023linear} have studied the iteration complexity of sampling, assuming access to an accurate diffusion model. In this work, we focus on understanding the \emph{sample complexity} of training such a model; how many samples are needed to learn an accurate diffusion model using a sufficiently expressive neural network? Prior work~\cite{BMR20} showed bounds polynomial in the dimension, desired Total Variation error, and Wasserstein error. We show an \emph{exponential improvement} in the dependence on Wasserstein error and depth, along with improved dependencies on other relevant parameters.
Authors: Anahita Baninajjar, Ahmed Rezine, Amir Aminifar
Abstract: Machine learning techniques often lack formal correctness guarantees, evidenced by the widespread adversarial examples that plague most deep-learning applications. This lack of formal guarantees resulted in several research efforts that aim at verifying Deep Neural Networks (DNNs), with a particular focus on safety-critical applications. However, formal verification techniques still face major scalability and precision challenges. The over-approximation introduced during the formal verification process to tackle the scalability challenge often results in inconclusive analysis. To address this challenge, we propose a novel framework to generate Verification-Friendly Neural Networks (VNNs). We present a post-training optimization framework to achieve a balance between preserving prediction performance and verification-friendliness. Our proposed framework results in VNNs that are comparable to the original DNNs in terms of prediction performance, while amenable to formal verification techniques. This essentially enables us to establish robustness for more VNNs than their DNN counterparts, in a time-efficient manner.
Authors: Weixi Song, Zuchao Li, Lefei Zhang, Hai Zhao, Bo Du
Abstract: With the prevalence of pre-training-fine-tuning paradigm, how to efficiently adapt the pre-trained model to the downstream tasks has been an intriguing issue. Parameter-Efficient Fine-Tuning (PEFT) methods have been proposed for low-cost adaptation. Although PEFT has demonstrated effectiveness and been widely applied, the underlying principles are still unclear. In this paper, we adopt the PAC-Bayesian generalization error bound, viewing pre-training as a shift of prior distribution which leads to a tighter bound for generalization error. We validate this shift from the perspectives of oscillations in the loss landscape and the quasi-sparsity in gradient distribution. Based on this, we propose a gradient-based sparse fine-tuning algorithm, named Sparse Increment Fine-Tuning (SIFT), and validate its effectiveness on a range of tasks including the GLUE Benchmark and Instruction-tuning. The code is accessible at https://github.com/song-wx/SIFT/.
Authors: Alexander Nikulin, Vladislav Kurenkov, Ilya Zisman, Artem Agarkov, Viacheslav Sinii, Sergey Kolesnikov
Abstract: Inspired by the diversity and depth of XLand and the simplicity and minimalism of MiniGrid, we present XLand-MiniGrid, a suite of tools and grid-world environments for meta-reinforcement learning research. Written in JAX, XLand-MiniGrid is designed to be highly scalable and can potentially run on GPU or TPU accelerators, democratizing large-scale experimentation with limited resources. Along with the environments, XLand-MiniGrid provides pre-sampled benchmarks with millions of unique tasks of varying difficulty and easy-to-use baselines that allow users to quickly start training adaptive agents. In addition, we have conducted a preliminary analysis of scaling and generalization, showing that our baselines are capable of reaching millions of steps per second during training and validating that the proposed benchmarks are challenging.
Authors: Satoki Ishikawa, Ryo Karakida
Abstract: Second-order optimization has been developed to accelerate the training of deep neural networks and it is being applied to increasingly larger-scale models. In this study, towards training on further larger scales, we identify a specific parameterization for second-order optimization that promotes feature learning in a stable manner even if the network width increases significantly. Inspired by a maximal update parameterization, we consider a one-step update of the gradient and reveal the appropriate scales of hyperparameters including random initialization, learning rates, and damping terms. Our approach covers two major second-order optimization algorithms, K-FAC and Shampoo, and we demonstrate that our parameterization achieves higher generalization performance in feature learning. In particular, it enables us to transfer the hyperparameters across models with different widths.
Authors: Junxiang Wang, Guangji Bai, Wei Cheng, Zhengzhang Chen, Liang Zhao, Haifeng Chen
Abstract: Time series domain adaptation stands as a pivotal and intricate challenge with diverse applications, including but not limited to human activity recognition, sleep stage classification, and machine fault diagnosis. Despite the numerous domain adaptation techniques proposed to tackle this complex problem, they primarily focus on domain adaptation from a single source domain. Yet, it is more crucial to investigate domain adaptation from multiple domains due to the potential for greater improvements. To address this, three important challenges need to be overcome: 1). The lack of exploration to utilize domain-specific information for domain adaptation, 2). The difficulty to learn domain-specific information that changes over time, and 3). The difficulty to evaluate learned domain-specific information. In order to tackle these challenges simultaneously, in this paper, we introduce PrOmpt-based domaiN Discrimination (POND), the first framework to utilize prompts for time series domain adaptation. Specifically, to address Challenge 1, we extend the idea of prompt tuning to time series analysis and learn prompts to capture common and domain-specific information from all source domains. To handle Challenge 2, we introduce a conditional module for each source domain to generate prompts from time series input data. For Challenge 3, we propose two criteria to select good prompts, which are used to choose the most suitable source domain for domain adaptation. The efficacy and robustness of our proposed POND model are extensively validated through experiments across 50 scenarios encompassing four datasets. Experimental results demonstrate that our proposed POND model outperforms all state-of-the-art comparison methods by up to $66\%$ on the F1-score.
Authors: Naiyu Yin, Tian Gao, Yue Yu, Qiang Ji
Abstract: Capturing the underlying structural causal relations represented by Directed Acyclic Graphs (DAGs) has been a fundamental task in various AI disciplines. Causal DAG learning via the continuous optimization framework has recently achieved promising performance in terms of both accuracy and efficiency. However, most methods make strong assumptions of homoscedastic noise, i.e., exogenous noises have equal variances across variables, observations, or even both. The noises in real data usually violate both assumptions due to the biases introduced by different data collection processes. To address the issue of heteroscedastic noise, we introduce relaxed and implementable sufficient conditions, proving the identifiability of a general class of SEM subject to these conditions. Based on the identifiable general SEM, we propose a novel formulation for DAG learning that accounts for the variation in noise variance across variables and observations. We then propose an effective two-phase iterative DAG learning algorithm to address the increasing optimization difficulties and to learn a causal DAG from data with heteroscedastic variable noise under varying variance. We show significant empirical gains of the proposed approaches over state-of-the-art methods on both synthetic data and real data.
Authors: Ram Dyuthi Sristi, Ofir Lindenbaum, Shira Lifshitz, Maria Lavzin, Jackie Schiller, Gal Mishne, Hadas Benisty
Abstract: Feature selection is a crucial tool in machine learning and is widely applied across various scientific disciplines. Traditional supervised methods generally identify a universal set of informative features for the entire population. However, feature relevance often varies with context, while the context itself may not directly affect the outcome variable. Here, we propose a novel architecture for contextual feature selection where the subset of selected features is conditioned on the value of context variables. Our new approach, Conditional Stochastic Gates (c-STG), models the importance of features using conditional Bernoulli variables whose parameters are predicted based on contextual variables. We introduce a hypernetwork that maps context variables to feature selection parameters to learn the context-dependent gates along with a prediction model. We further present a theoretical analysis of our model, indicating that it can improve performance and flexibility over population-level methods in complex feature selection settings. Finally, we conduct an extensive benchmark using simulated and real-world datasets across multiple domains demonstrating that c-STG can lead to improved feature selection capabilities while enhancing prediction accuracy and interpretability.
Authors: Vage Egiazarian, Andrei Panferov, Denis Kuznedelev, Elias Frantar, Artem Babenko, Dan Alistarh
Abstract: The emergence of accurate open large language models (LLMs) has led to a race towards performant quantization techniques which can enable their execution on end-user devices. In this paper, we revisit the problem of ``extreme'' LLM compression -- defined as targeting extremely low bit counts, such as 2 to 3 bits per parameter -- from the point of view of classic methods in Multi-Codebook Quantization (MCQ). Our algorithm, called AQLM, generalizes the classic Additive Quantization (AQ) approach for information retrieval to advance the state-of-the-art in LLM compression, via two innovations: 1) learned additive quantization of weight matrices in input-adaptive fashion, and 2) joint optimization of codebook parameters across each transformer blocks. Broadly, AQLM is the first scheme that is Pareto optimal in terms of accuracy-vs-model-size when compressing to less than 3 bits per parameter, and significantly improves upon all known schemes in the extreme compression (2bit) regime. In addition, AQLM is practical: we provide fast GPU and CPU implementations of AQLM for token generation, which enable us to match or outperform optimized FP16 implementations for speed, while executing in a much smaller memory footprint.
Authors: Jesse C. Cresswell, Yi Sui, Bhargava Kumar, No\"el Vouitsis
Abstract: In response to everyday queries, humans explicitly signal uncertainty and offer alternative answers when they are unsure. Machine learning models that output calibrated prediction sets through conformal prediction mimic this human behaviour; larger sets signal greater uncertainty while providing alternatives. In this work, we study the usefulness of conformal prediction sets as an aid for human decision making by conducting a pre-registered randomized controlled trial with conformal prediction sets provided to human subjects. With statistical significance, we find that when humans are given conformal prediction sets their accuracy on tasks improves compared to fixed-size prediction sets with the same coverage guarantee. The results show that quantifying model uncertainty with conformal prediction is helpful for human-in-the-loop decision making and human-AI teams.
Authors: Duy M. H. Nguyen, Nina Lukashina, Tai Nguyen, An T. Le, TrungTin Nguyen, Nhat Ho, Jan Peters, Daniel Sonntag, Viktor Zaverkin, Mathias Niepert
Abstract: A molecule's 2D representation consists of its atoms, their attributes, and the molecule's covalent bonds. A 3D (geometric) representation of a molecule is called a conformer and consists of its atom types and Cartesian coordinates. Every conformer has a potential energy, and the lower this energy, the more likely it occurs in nature. Most existing machine learning methods for molecular property prediction consider either 2D molecular graphs or 3D conformer structure representations in isolation. Inspired by recent work on using ensembles of conformers in conjunction with 2D graph representations, we propose $\mathrm{E}$(3)-invariant molecular conformer aggregation networks. The method integrates a molecule's 2D representation with that of multiple of its conformers. Contrary to prior work, we propose a novel 2D-3D aggregation mechanism based on a differentiable solver for the \emph{Fused Gromov-Wasserstein Barycenter} problem and the use of an efficient conformer generation method based on distance geometry. We show that the proposed aggregation mechanism is $\mathrm{E}$(3) invariant and propose an efficient GPU implementation. Moreover, we demonstrate that the aggregation mechanism helps to significantly outperform state-of-the-art molecule property prediction methods on established datasets.
Authors: Jingzhe Liu, Haitao Mao, Zhikai Chen, Tong Zhao, Neil Shah, Jiliang Tang
Abstract: Deep graph models (e.g., graph neural networks and graph transformers) have become important techniques for leveraging knowledge across various types of graphs. Yet, the scaling properties of deep graph models have not been systematically investigated, casting doubt on the feasibility of achieving large graph models through enlarging the model and dataset sizes. In this work, we delve into neural scaling laws on graphs from both model and data perspectives. We first verify the validity of such laws on graphs, establishing formulations to describe the scaling behaviors. For model scaling, we investigate the phenomenon of scaling law collapse and identify overfitting as the potential reason. Moreover, we reveal that the model depth of deep graph models can impact the model scaling behaviors, which differ from observations in other domains such as CV and NLP. For data scaling, we suggest that the number of graphs can not effectively metric the graph data volume in scaling law since the sizes of different graphs are highly irregular. Instead, we reform the data scaling law with the number of edges as the metric to address the irregular graph sizes. We further demonstrate the reformed law offers a unified view of the data scaling behaviors for various fundamental graph tasks including node classification, link prediction, and graph classification. This work provides valuable insights into neural scaling laws on graphs, which can serve as an essential step toward large graph models.
Authors: Huy Tran, Yikun Bai, Abihith Kothapalli, Ashkan Shahbazi, Xinran Liu, Rocio Diaz Martin, Soheil Kolouri
Abstract: Comparing spherical probability distributions is of great interest in various fields, including geology, medical domains, computer vision, and deep representation learning. The utility of optimal transport-based distances, such as the Wasserstein distance, for comparing probability measures has spurred active research in developing computationally efficient variations of these distances for spherical probability measures. This paper introduces a high-speed and highly parallelizable distance for comparing spherical measures using the stereographic projection and the generalized Radon transform, which we refer to as the Stereographic Spherical Sliced Wasserstein (S3W) distance. We carefully address the distance distortion caused by the stereographic projection and provide an extensive theoretical analysis of our proposed metric and its rotationally invariant variation. Finally, we evaluate the performance of the proposed metrics and compare them with recent baselines in terms of both speed and accuracy through a wide range of numerical studies, including gradient flows and self-supervised learning. Our code is available at https://github.com/mint-vu/s3wd.
Authors: Gaurav Pandey, Yatin Nandwani, Tahira Naseem, Mayank Mishra, Guangxuan Xu, Dinesh Raghu, Sachindra Joshi, Asim Munawar, Ram\'on Fernandez Astudillo
Abstract: Distribution matching methods for language model alignment such as Generation with Distributional Control (GDC) and Distributional Policy Gradient (DPG) have not received the same level of attention in reinforcement learning from human feedback (RLHF) as contrastive methods such as Sequence Likelihood Calibration (SLiC), Direct Preference Optimization (DPO) and its variants. We identify high variance of the gradient estimate as the primary reason for the lack of success of these methods and propose a self-normalized baseline to reduce the variance. We further generalize the target distribution in DPG, GDC and DPO by using Bayes' rule to define the reward-conditioned posterior. The resulting approach, referred to as BRAIn - Bayesian Reward-conditioned Amortized Inference acts as a bridge between distribution matching methods and DPO and significantly outperforms prior art in summarization and Antropic HH tasks.
Authors: Binghui Xie, Yatao Bian, Kaiwen zhou, Yongqiang Chen, Peilin Zhao, Bo Han, Wei Meng, James Cheng
Abstract: Learning neural subset selection tasks, such as compound selection in AI-aided drug discovery, have become increasingly pivotal across diverse applications. The existing methodologies in the field primarily concentrate on constructing models that capture the relationship between utility function values and subsets within their respective supersets. However, these approaches tend to overlook the valuable information contained within the superset when utilizing neural networks to model set functions. In this work, we address this oversight by adopting a probabilistic perspective. Our theoretical findings demonstrate that when the target value is conditioned on both the input set and subset, it is essential to incorporate an \textit{invariant sufficient statistic} of the superset into the subset of interest for effective learning. This ensures that the output value remains invariant to permutations of the subset and its corresponding superset, enabling identification of the specific superset from which the subset originated. Motivated by these insights, we propose a simple yet effective information aggregation module designed to merge the representations of subsets and supersets from a permutation invariance perspective. Comprehensive empirical evaluations across diverse tasks and datasets validate the enhanced efficacy of our approach over conventional methods, underscoring the practicality and potency of our proposed strategies in real-world contexts.
Authors: Michele Mastromattei, Fabio Massimo Zanzotto
Abstract: Neural network pruning has become increasingly crucial due to the complexity of these models and their widespread use in various fields. Existing pruning algorithms often suffer from limitations such as architecture specificity, excessive complexity and reliance on demanding calculations, rendering them impractical for real-world applications. This paper introduces KEN: a straightforward, universal and unstructured pruning algorithm based on Kernel Density Estimation (KDE). KEN aims to construct optimized transformers by selectively preserving the most significant parameters while restoring others to their pre-training state. This strategy preserves model performance while enabling storage of only the optimized subnetwork, leading to substantial memory savings. Extensive evaluations across seven different LLMs demonstrate that KEN achieves equal or better performance than their original unpruned versions, with a minimum parameter reduction of 25%. Furthermore, in-depth comparisons with established pruning and PEFT algorithms confirm KEN effectiveness. We further introduce KEN$_{viz}$, an explainable tool that visualizes the optimized model composition achieved by KEN from different points of view.
Authors: Md Shamim Hussain, Mohammed J. Zaki, Dharmashankar Subramanian
Abstract: Graph transformers typically lack third-order interactions, limiting their geometric understanding which is crucial for tasks like molecular geometry prediction. We propose the Triplet Graph Transformer (TGT) that enables direct communication between pairs within a 3-tuple of nodes via novel triplet attention and aggregation mechanisms. TGT is applied to molecular property prediction by first predicting interatomic distances from 2D graphs and then using these distances for downstream tasks. A novel three-stage training procedure and stochastic inference further improve training efficiency and model performance. Our model achieves new state-of-the-art (SOTA) results on open challenge benchmarks PCQM4Mv2 and OC20 IS2RE. We also obtain SOTA results on QM9, MOLPCBA, and LIT-PCBA molecular property prediction benchmarks via transfer learning. We also demonstrate the generality of TGT with SOTA results on the traveling salesman problem (TSP).
Authors: Kartik Ahuja, Amin Mansouri
Abstract: Out-of-distribution generalization capabilities of sequence-to-sequence models can be studied from the lens of two crucial forms of generalization: length generalization -- the ability to generalize to longer sequences than ones seen during training, and compositional generalization: the ability to generalize to token combinations not seen during training. In this work, we provide first provable guarantees on length and compositional generalization for common sequence-to-sequence models -- deep sets, transformers, state space models, and recurrent neural nets -- trained to minimize the prediction error. Taking a first principles perspective, we study the realizable case, i.e., the labeling function is realizable on the architecture. We show that limited capacity versions of these different architectures achieve both length and compositional generalization. Across different architectures, we also find that a linear relationship between the learned representation and the representation in the labeling function is necessary for length and compositional generalization.
Authors: Yichuan Mo, Yuji Wang, Zeming Wei, Yisen Wang
Abstract: While Large Language Models (LLMs) have achieved tremendous success in various applications, they are also susceptible to jailbreak attacks. Several primary defense strategies have been proposed to protect LLMs from producing harmful information, mostly with a particular focus on harmful content filtering or heuristical defensive prompt designs. However, how to achieve intrinsic robustness through the prompts remains an open problem. In this paper, motivated by adversarial training paradigms for achieving reliable robustness, we propose an approach named Prompt Adversarial Tuning (PAT) that trains a prompt control attached to the user prompt as a guard prefix. To achieve our defense goal whilst maintaining natural performance, we optimize the control prompt with both adversarial and benign prompts. Comprehensive experiments show that our method is effective against both black-box and white-box attacks, reducing the success rate of advanced attacks to nearly 0 while maintaining the model's utility on the benign task. The proposed defense strategy incurs only negligible computational overhead, charting a new perspective for future explorations in LLM security. Our code is available at https://github.com/rain152/PAT.
Authors: Gon Buzaglo, Itamar Harel, Mor Shpigel Nacson, Alon Brutzkus, Nathan Srebro, Daniel Soudry
Abstract: Background. A main theoretical puzzle is why over-parameterized Neural Networks (NNs) generalize well when trained to zero loss (i.e., so they interpolate the data). Usually, the NN is trained with Stochastic Gradient Descent (SGD) or one of its variants. However, recent empirical work examined the generalization of a random NN that interpolates the data: the NN was sampled from a seemingly uniform prior over the parameters, conditioned on that the NN perfectly classifies the training set. Interestingly, such a NN sample typically generalized as well as SGD-trained NNs. Contributions. We prove that such a random NN interpolator typically generalizes well if there exists an underlying narrow ``teacher NN'' that agrees with the labels. Specifically, we show that such a `flat' prior over the NN parameterization induces a rich prior over the NN functions, due to the redundancy in the NN structure. In particular, this creates a bias towards simpler functions, which require less relevant parameters to represent -- enabling learning with a sample complexity approximately proportional to the complexity of the teacher (roughly, the number of non-redundant parameters), rather than the student's.
Authors: Nate Gillman, Michael Freeman, Daksh Aggarwal, Chia-Hong Hsu, Calvin Luo, Yonglong Tian, Chen Sun
Abstract: As synthetic data becomes higher quality and proliferates on the internet, machine learning models are increasingly trained on a mix of human- and machine-generated data. Despite the successful stories of using synthetic data for representation learning, using synthetic data for generative model training creates "self-consuming loops" which may lead to training instability or even collapse, unless certain conditions are met. Our paper aims to stabilize self-consuming generative model training. Our theoretical results demonstrate that by introducing an idealized correction function, which maps a data point to be more likely under the true data distribution, self-consuming loops can be made exponentially more stable. We then propose self-correction functions, which rely on expert knowledge (e.g. the laws of physics programmed in a simulator), and aim to approximate the idealized corrector automatically and at scale. We empirically validate the effectiveness of self-correcting self-consuming loops on the challenging human motion synthesis task, and observe that it successfully avoids model collapse, even when the ratio of synthetic data to real data is as high as 100%.
Authors: Ziang Chen, Jialin Liu, Xiaohan Chen, Xinshang Wang, Wotao Yin
Abstract: Graph neural networks (GNNs) have been widely used to predict properties and heuristics of mixed-integer linear programs (MILPs) and hence accelerate MILP solvers. This paper investigates the capacity of GNNs to represent strong branching (SB), the most effective yet computationally expensive heuristic employed in the branch-and-bound algorithm. In the literature, message-passing GNN (MP-GNN), as the simplest GNN structure, is frequently used as a fast approximation of SB and we find that not all MILPs's SB can be represented with MP-GNN. We precisely define a class of "MP-tractable" MILPs for which MP-GNNs can accurately approximate SB scores. Particularly, we establish a universal approximation theorem: for any data distribution over the MP-tractable class, there always exists an MP-GNN that can approximate the SB score with arbitrarily high accuracy and arbitrarily high probability, which lays a theoretical foundation of the existing works on imitating SB with MP-GNN. For MILPs without the MP-tractability, unfortunately, a similar result is impossible, which can be illustrated by two MILP instances with different SB scores that cannot be distinguished by any MP-GNN, regardless of the number of parameters. Recognizing this, we explore another GNN structure called the second-order folklore GNN (2-FGNN) that overcomes this limitation, and the aforementioned universal approximation theorem can be extended to the entire MILP space using 2-FGNN, regardless of the MP-tractability. A small-scale numerical experiment is conducted to directly validate our theoretical findings.
Authors: Wei Xu, An Liu, Yiting Zhang, Vincent Lau
Abstract: Efficient learning and model compression algorithm for deep neural network (DNN) is a key workhorse behind the rise of deep learning (DL). In this work, we propose a message passing based Bayesian deep learning algorithm called EM-TDAMP to avoid the drawbacks of traditional stochastic gradient descent (SGD) based learning algorithms and regularization-based model compression methods. Specifically, we formulate the problem of DNN learning and compression as a sparse Bayesian inference problem, in which group sparse prior is employed to achieve structured model compression. Then, we propose an expectation maximization (EM) framework to estimate posterior distributions for parameters (E-step) and update hyperparameters (M-step), where the E-step is realized by a newly proposed turbo deep approximate message passing (TDAMP) algorithm. We further extend the EM-TDAMP and propose a novel Bayesian federated learning framework, in which and the clients perform TDAMP to efficiently calculate the local posterior distributions based on the local data, and the central server first aggregates the local posterior distributions to update the global posterior distributions and then update hyperparameters based on EM to accelerate convergence. We detail the application of EM-TDAMP to Boston housing price prediction and handwriting recognition, and present extensive numerical results to demonstrate the advantages of EM-TDAMP.
Authors: Parikshit Gopalan, Lunjia Hu, Guy N. Rothblum
Abstract: Consider a multi-class labelling problem, where the labels can take values in $[k]$, and a predictor predicts a distribution over the labels. In this work, we study the following foundational question: Are there notions of multi-class calibration that give strong guarantees of meaningful predictions and can be achieved in time and sample complexities polynomial in $k$? Prior notions of calibration exhibit a tradeoff between computational efficiency and expressivity: they either suffer from having sample complexity exponential in $k$, or needing to solve computationally intractable problems, or give rather weak guarantees. Our main contribution is a notion of calibration that achieves all these desiderata: we formulate a robust notion of projected smooth calibration for multi-class predictions, and give new recalibration algorithms for efficiently calibrating predictors under this definition with complexity polynomial in $k$. Projected smooth calibration gives strong guarantees for all downstream decision makers who want to use the predictor for binary classification problems of the form: does the label belong to a subset $T \subseteq [k]$: e.g. is this an image of an animal? It ensures that the probabilities predicted by summing the probabilities assigned to labels in $T$ are close to some perfectly calibrated binary predictor for that task. We also show that natural strengthenings of our definition are computationally hard to achieve: they run into information theoretic barriers or computational intractability. Underlying both our upper and lower bounds is a tight connection that we prove between multi-class calibration and the well-studied problem of agnostic learning in the (standard) binary prediction setting.
Authors: Emily Jin, Michael Bronstein, \.Ismail \.Ilkan Ceylan, Matthias Lanzinger
Abstract: A large body of work has investigated the properties of graph neural networks and identified several limitations, particularly pertaining to their expressive power. Their inability to count certain patterns (e.g., cycles) in a graph lies at the heart of such limitations, since many functions to be learned rely on the ability of counting such patterns. Two prominent paradigms aim to address this limitation by enriching the graph features with subgraph or homomorphism pattern counts. In this work, we show that both of these approaches are sub-optimal in a certain sense and argue for a more fine-grained approach, which incorporates the homomorphism counts of all structures in the ``basis'' of the target pattern. This yields strictly more expressive architectures without incurring any additional overhead in terms of computational complexity compared to existing approaches. We prove a series of theoretical results on node-level and graph-level motif parameters and empirically validate them on standard benchmark datasets.
Authors: Ziang Chen, Rong Ge
Abstract: In this work, we study the mean-field flow for learning subspace-sparse polynomials using stochastic gradient descent and two-layer neural networks, where the input distribution is standard Gaussian and the output only depends on the projection of the input onto a low-dimensional subspace. We propose a basis-free generalization of the merged-staircase property in Abbe et al. (2022) and establish a necessary condition for the SGD-learnability. In addition, we prove that the condition is almost sufficient, in the sense that a condition slightly stronger than the necessary condition can guarantee the exponential decay of the loss functional to zero.
Authors: Chin-Chia Michael Yeh, Yujie Fan, Xin Dai, Vivian Lai, Prince Osei Aboagye, Junpeng Wang, Huiyuan Chen, Yan Zheng, Zhongfang Zhuang, Liang Wang, Wei Zhang
Abstract: All-Multi-Layer Perceptron (all-MLP) mixer models have been shown to be effective for time series forecasting problems. However, when such a model is applied to high-dimensional time series (e.g., the time series in a spatial-temporal dataset), its performance is likely to degrade due to overfitting issues. In this paper, we propose an all-MLP time series forecasting architecture, referred to as RPMixer. Our method leverages the ensemble-like behavior of deep neural networks, where each individual block within the network acts like a base learner in an ensemble model, especially when identity mapping residual connections are incorporated. By integrating random projection layers into our model, we increase the diversity among the blocks' outputs, thereby enhancing the overall performance of RPMixer. Extensive experiments conducted on large-scale spatial-temporal forecasting benchmark datasets demonstrate that our proposed method outperforms alternative methods, including both spatial-temporal graph models and general forecasting models.
Authors: Ivan Marisca, Cesare Alippi, Filippo Maria Bianchi
Abstract: Given a set of synchronous time series, each associated with a sensor-point in space and characterized by inter-series relationships, the problem of spatiotemporal forecasting consists of predicting future observations for each point. Spatiotemporal graph neural networks achieve striking results by representing the relationships across time series as a graph. Nonetheless, most existing methods rely on the often unrealistic assumption that inputs are always available and fail to capture hidden spatiotemporal dynamics when part of the data is missing. In this work, we tackle this problem through hierarchical spatiotemporal downsampling. The input time series are progressively coarsened over time and space, obtaining a pool of representations that capture heterogeneous temporal and spatial dynamics. Conditioned on observations and missing data patterns, such representations are combined by an interpretable attention mechanism to generate the forecasts. Our approach outperforms state-of-the-art methods on synthetic and real-world benchmarks under different missing data distributions, particularly in the presence of contiguous blocks of missing values.
Authors: Alfredo Reichlin, Miguel Vasco, Hang Yin, Danica Kragic
Abstract: We address the problem of learning optimal behavior from sub-optimal datasets for goal-conditioned offline reinforcement learning. To do so, we propose the use of metric learning to approximate the optimal value function for goal-conditioned offline RL problems under sparse rewards, invertible actions and deterministic transitions. We introduce distance monotonicity, a property for representations to recover optimality and propose an optimization objective that leads to such property. We use the proposed value function to guide the learning of a policy in an actor-critic fashion, a method we name MetricRL. Experimentally, we show that our method estimates optimal behaviors from severely sub-optimal offline datasets without suffering from out-of-distribution estimation errors. We demonstrate that MetricRL consistently outperforms prior state-of-the-art goal-conditioned RL methods in learning optimal policies from sub-optimal offline datasets.
Authors: Zihao Luo, Xilie Xu, Feng Liu, Yun Sing Koh, Di Wang, Jingfeng Zhang
Abstract: Low-rank adaptation (LoRA) is an efficient strategy for adapting latent diffusion models (LDMs) on a private dataset to generate specific images by minimizing the adaptation loss. However, the LoRA-adapted LDMs are vulnerable to membership inference (MI) attacks that can judge whether a particular data point belongs to the private dataset, thus leading to the privacy leakage. To defend against MI attacks, we first propose a straightforward solution: Membership-Privacy-preserving LoRA (MP-LoRA). MP-LoRA is formulated as a min-max optimization problem where a proxy attack model is trained by maximizing its MI gain while the LDM is adapted by minimizing the sum of the adaptation loss and the MI gain of the proxy attack model. However, we empirically find that MP-LoRA has the issue of unstable optimization, and theoretically analyze that the potential reason is the unconstrained local smoothness, which impedes the privacy-preserving adaptation. To mitigate this issue, we further propose a Stable Membership-Privacy-preserving LoRA (SMP-LoRA) that adapts the LDM by minimizing the ratio of the adaptation loss to the MI gain. Besides, we theoretically prove that the local smoothness of SMP-LoRA can be constrained by the gradient norm, leading to improved convergence. Our experimental results corroborate that SMP-LoRA can indeed defend against MI attacks and generate high-quality images. Our code is available at https://github.com/WilliamLUO0/StablePrivateLoRA.
Authors: Michael Beukman, Samuel Coward, Michael Matthews, Mattie Fellows, Minqi Jiang, Michael Dennis, Jakob Foerster
Abstract: In unsupervised environment design, reinforcement learning agents are trained on environment configurations (levels) generated by an adversary that maximises some objective. Regret is a commonly used objective that theoretically results in a minimax regret (MMR) policy with desirable robustness guarantees; in particular, the agent's maximum regret is bounded. However, once the agent reaches this regret bound on all levels, the adversary will only sample levels where regret cannot be further reduced. Although there are possible performance improvements to be made outside of these regret-maximising levels, learning stagnates. In this work, we introduce Bayesian level-perfect MMR (BLP), a refinement of the minimax regret objective that overcomes this limitation. We formally show that solving for this objective results in a subset of MMR policies, and that BLP policies act consistently with a Perfect Bayesian policy over all levels. We further introduce an algorithm, ReMiDi, that results in a BLP policy at convergence. We empirically demonstrate that training on levels from a minimax regret adversary causes learning to prematurely stagnate, but that ReMiDi continues learning.
Authors: Youssef Allouah, Sadegh Farhadkhani, Rachid GuerraouI, Nirupam Gupta, Rafael Pinot, Geovani Rizk, Sasha Voitovych
Abstract: The possibility of adversarial (a.k.a., {\em Byzantine}) clients makes federated learning (FL) prone to arbitrary manipulation. The natural approach to robustify FL against adversarial clients is to replace the simple averaging operation at the server in the standard $\mathsf{FedAvg}$ algorithm by a \emph{robust averaging rule}. While a significant amount of work has been devoted to studying the convergence of federated {\em robust averaging} (which we denote by $\mathsf{FedRo}$), prior work has largely ignored the impact of {\em client subsampling} and {\em local steps}, two fundamental FL characteristics. While client subsampling increases the effective fraction of Byzantine clients, local steps increase the drift between the local updates computed by honest (i.e., non-Byzantine) clients. Consequently, a careless deployment of $\mathsf{FedRo}$ could yield poor performance. We validate this observation by presenting an in-depth analysis of $\mathsf{FedRo}$ tightly analyzing the impact of client subsampling and local steps. Specifically, we present a sufficient condition on client subsampling for nearly-optimal convergence of $\mathsf{FedRo}$ (for smooth non-convex loss). Also, we show that the rate of improvement in learning accuracy {\em diminishes} with respect to the number of clients subsampled, as soon as the sample size exceeds a threshold value. Interestingly, we also observe that under a careful choice of step-sizes, the learning error due to Byzantine clients decreases with the number of local steps. We validate our theory by experiments on the FEMNIST and CIFAR-$10$ image classification tasks.
Authors: Liam Hodgson, Danilo Bzdok
Abstract: The multivariate hypergeometric distribution describes sampling without replacement from a discrete population of elements divided into multiple categories. Addressing a gap in the literature, we tackle the challenge of estimating discrete distributions when both the total population size and the sizes of its constituent categories are unknown. Here, we propose a novel solution using the hypergeometric likelihood to solve this estimation challenge, even in the presence of severe under-sampling. We develop our approach to account for a data generating process where the ground-truth is a mixture of distributions conditional on a continuous latent variable, such as with collaborative filtering, using the variational autoencoder framework. Empirical data simulation demonstrates that our method outperforms other likelihood functions used to model count data, both in terms of accuracy of population size estimate and in its ability to learn an informative latent space. We demonstrate our method's versatility through applications in NLP, by inferring and estimating the complexity of latent vocabularies in text excerpts, and in biology, by accurately recovering the true number of gene transcripts from sparse single-cell genomics data.
Authors: Chen Jia
Abstract: Preference learning (PL) with large language models (LLMs) aims to align the LLMs' generations with human preferences. Previous work on reinforcement learning from human feedback (RLHF) has demonstrated promising results in in-distribution PL. However, due to the difficulty of obtaining human feedback, discretely training reward models for every encountered distribution is challenging. Thus, out-of-distribution (OOD) PL is practically useful for enhancing the generalization ability of LLMs with limited preference feedback. This work addresses OOD PL by optimizing a general reward model through a meta-learning approach. During meta-training, a bilevel optimization algorithm is utilized to learn a reward model capable of guiding policy learning to align with human preferences across various distributions. When encountering a test distribution, the meta-test procedure conducts regularized policy optimization using the learned reward model for PL. We theoretically demonstrate the convergence rate of the bilevel optimization algorithm under reasonable assumptions. Additionally, we conduct experiments on two text generation tasks across 20 held-out domains and outperform a variety of strong baselines across various evaluation metrics.
Authors: Jingfeng Wu, Peter L. Bartlett, Matus Telgarsky, Bin Yu
Abstract: We consider gradient descent (GD) with a constant stepsize applied to logistic regression with linearly separable data, where the constant stepsize $\eta$ is so large that the loss initially oscillates. We show that GD exits this initial oscillatory phase rapidly -- in $\mathcal{O}(\eta)$ steps -- and subsequently achieves an $\tilde{\mathcal{O}}(1 / (\eta t) )$ convergence rate after $t$ additional steps. Our results imply that, given a budget of $T$ steps, GD can achieve an accelerated loss of $\tilde{\mathcal{O}}(1/T^2)$ with an aggressive stepsize $\eta:= \Theta( T)$, without any use of momentum or variable stepsize schedulers. Our proof technique is versatile and also handles general classification loss functions (where exponential tails are needed for the $\tilde{\mathcal{O}}(1/T^2)$ acceleration), nonlinear predictors in the neural tangent kernel regime, and online stochastic gradient descent (SGD) with a large stepsize, under suitable separability conditions.
Authors: Miguel Liu-Schiaffini, Julius Berner, Boris Bonev, Thorsten Kurth, Kamyar Azizzadenesheli, Anima Anandkumar
Abstract: Neural operators learn mappings between function spaces, which is practical for learning solution operators of PDEs and other scientific modeling applications. Among them, the Fourier neural operator (FNO) is a popular architecture that performs global convolutions in the Fourier space. However, such global operations are often prone to over-smoothing and may fail to capture local details. In contrast, convolutional neural networks (CNN) can capture local features but are limited to training and inference at a single resolution. In this work, we present a principled approach to operator learning that can capture local features under two frameworks by learning differential operators and integral operators with locally supported kernels. Specifically, inspired by stencil methods, we prove that we obtain differential operators under an appropriate scaling of the kernel values of CNNs. To obtain local integral operators, we utilize suitable basis representations for the kernels based on discrete-continuous convolutions. Both these approaches preserve the properties of operator learning and, hence, the ability to predict at any resolution. Adding our layers to FNOs significantly improves their performance, reducing the relative L2-error by 34-72% in our experiments, which include a turbulent 2D Navier-Stokes and the spherical shallow water equations.
Authors: Leon Lang, Davis Foote, Stuart Russell, Anca Dragan, Erik Jenner, Scott Emmons
Abstract: Past analyses of reinforcement learning from human feedback (RLHF) assume that the human evaluators fully observe the environment. What happens when human feedback is based only on partial observations? We formally define two failure cases: deceptive inflation and overjustification. Modeling the human as Boltzmann-rational w.r.t. a belief over trajectories, we prove conditions under which RLHF is guaranteed to result in policies that deceptively inflate their performance, overjustify their behavior to make an impression, or both. Under the new assumption that the human's partial observability is known and accounted for, we then analyze how much information the feedback process provides about the return function. We show that sometimes, the human's feedback determines the return function uniquely up to an additive constant, but in other realistic cases, there is irreducible ambiguity. We propose exploratory research directions to help tackle these challenges and caution against blindly applying RLHF in partially observable settings.
Authors: Niclas G\"oring, Florian Hess, Manuel Brenner, Zahra Monfared, Daniel Durstewitz
Abstract: In science we are interested in finding the governing equations, the dynamical rules, underlying empirical phenomena. While traditionally scientific models are derived through cycles of human insight and experimentation, recently deep learning (DL) techniques have been advanced to reconstruct dynamical systems (DS) directly from time series data. State-of-the-art dynamical systems reconstruction (DSR) methods show promise in capturing invariant and long-term properties of observed DS, but their ability to generalize to unobserved domains remains an open challenge. Yet, this is a crucial property we would expect from any viable scientific theory. In this work, we provide a formal framework that addresses generalization in DSR. We explain why and how out-of-domain (OOD) generalization (OODG) in DSR profoundly differs from OODG considered elsewhere in machine learning. We introduce mathematical notions based on topological concepts and ergodic theory to formalize the idea of learnability of a DSR model. We formally prove that black-box DL techniques, without adequate structural priors, generally will not be able to learn a generalizing DSR model. We also show this empirically, considering major classes of DSR algorithms proposed so far, and illustrate where and why they fail to generalize across the whole phase space. Our study provides the first comprehensive mathematical treatment of OODG in DSR, and gives a deeper conceptual understanding of where the fundamental problems in OODG lie and how they could possibly be addressed in practice.
Authors: Xiaofan Cui, Muhammad Aadil Khan, Gabriele Pozzato, Surinder Singh, Ratnesh Sharma, Simona Onori
Abstract: The reuse of retired electric vehicle batteries in grid energy storage offers environmental and economic benefits. This study concentrates on health monitoring algorithms for retired batteries deployed in grid storage. Over 15 months of testing, we collect, analyze, and publicize a dataset of second-life batteries, implementing a cycling protocol simulating grid energy storage load profiles within a 3-4 V voltage window. Four machine-learning-based health estimation models, relying on online-accessible features and initial capacity, are compared, with the selected model achieving a mean absolute percentage error below 2.3% on test data. Additionally, an adaptive online health estimation algorithm is proposed by integrating a clustering-based method, thus limiting estimation errors during online deployment. These results showcase the feasibility of repurposing retired batteries for second-life applications. Based on obtained data and power demand, these second-life batteries exhibit potential for over a decade of grid energy storage use.
Authors: Siyu Chen, Heejune Sheen, Tianhao Wang, Zhuoran Yang
Abstract: We study the dynamics of gradient flow for training a multi-head softmax attention model for in-context learning of multi-task linear regression. We establish the global convergence of gradient flow under suitable choices of initialization. In addition, we prove that an interesting "task allocation" phenomenon emerges during the gradient flow dynamics, where each attention head focuses on solving a single task of the multi-task model. Specifically, we prove that the gradient flow dynamics can be split into three phases -- a warm-up phase where the loss decreases rather slowly and the attention heads gradually build up their inclination towards individual tasks, an emergence phase where each head selects a single task and the loss rapidly decreases, and a convergence phase where the attention parameters converge to a limit. Furthermore, we prove the optimality of gradient flow in the sense that the limiting model learned by gradient flow is on par with the best possible multi-head softmax attention model up to a constant factor. Our analysis also delineates a strict separation in terms of the prediction accuracy of ICL between single-head and multi-head attention models. The key technique for our convergence analysis is to map the gradient flow dynamics in the parameter space to a set of ordinary differential equations in the spectral domain, where the relative magnitudes of the semi-singular values of the attention weights determines task allocation. To our best knowledge, our work provides the first convergence result for the multi-head softmax attention model.
Authors: Cong Geng, Tian Han, Peng-Tao Jiang, Hao Zhang, Jinwei Chen, S{\o}ren Hauberg, Bo Li
Abstract: Generative models have shown strong generation ability while efficient likelihood estimation is less explored. Energy-based models~(EBMs) define a flexible energy function to parameterize unnormalized densities efficiently but are notorious for being difficult to train. Adversarial EBMs introduce a generator to form a minimax training game to avoid expensive MCMC sampling used in traditional EBMs, but a noticeable gap between adversarial EBMs and other strong generative models still exists. Inspired by diffusion-based models, we embedded EBMs into each denoising step to split a long-generated process into several smaller steps. Besides, we employ a symmetric Jeffrey divergence and introduce a variational posterior distribution for the generator's training to address the main challenges that exist in adversarial EBMs. Our experiments show significant improvement in generation compared to existing adversarial EBMs, while also providing a useful energy function for efficient density estimation.
Authors: Hongyu Zhang, Dongyi Zheng, Lin Zhong, Xu Yang, Jiyuan Feng, Yunqing Feng, Qing Liao
Abstract: In recent years, Cross-Domain Recommendation (CDR) has drawn significant attention, which utilizes user data from multiple domains to enhance the recommendation performance. However, current CDR methods require sharing user data across domains, thereby violating the General Data Protection Regulation (GDPR). Consequently, numerous approaches have been proposed for Federated Cross-Domain Recommendation (FedCDR). Nevertheless, the data heterogeneity across different domains inevitably influences the overall performance of federated learning. In this study, we propose FedHCDR, a novel Federated Cross-Domain Recommendation framework with Hypergraph signal decoupling. Specifically, to address the data heterogeneity across domains, we introduce an approach called hypergraph signal decoupling (HSD) to decouple the user features into domain-exclusive and domain-shared features. The approach employs high-pass and low-pass hypergraph filters to decouple domain-exclusive and domain-shared user representations, which are trained by the local-global bi-directional transfer algorithm. In addition, a hypergraph contrastive learning (HCL) module is devised to enhance the learning of domain-shared user relationship information by perturbing the user hypergraph. Extensive experiments conducted on three real-world scenarios demonstrate that FedHCDR outperforms existing baselines significantly.
Authors: Peimeng Guan, Naveed Iqbal, Mark A. Davenport, Mudassir Masood
Abstract: Model-based deep learning methods such as loop unrolling (LU) and deep equilibrium model}(DEQ) extensions offer outstanding performance in solving inverse problems (IP). These methods unroll the optimization iterations into a sequence of neural networks that in effect learn a regularization function from data. While these architectures are currently state-of-the-art in numerous applications, their success heavily relies on the accuracy of the forward model. This assumption can be limiting in many physical applications due to model simplifications or uncertainties in the apparatus. To address forward model mismatch, we introduce an untrained forward model residual block within the model-based architecture to match the data consistency in the measurement domain for each instance. We propose two variants in well-known model-based architectures (LU and DEQ) and prove convergence under mild conditions. Our approach offers a unified solution that is less parameter-sensitive, requires no additional data, and enables simultaneous fitting of the forward model and reconstruction in a single pass, benefiting both linear and nonlinear inverse problems. The experiments show significant quality improvement in removing artifacts and preserving details across three distinct applications, encompassing both linear and nonlinear inverse problems. Moreover, we highlight reconstruction effectiveness in intermediate steps and showcase robustness to random initialization of the residual block and a higher number of iterations during evaluation. Code is available at \texttt{https://github.com/InvProbs/A-adaptive-model-based-methods}.
URLs: https://github.com/InvProbs/A-adaptive-model-based-methods
Authors: Onur Celik, Aleksandar Taranovic, Gerhard Neumann
Abstract: Reinforcement learning (RL) is a powerful approach for acquiring a good-performing policy. However, learning diverse skills is challenging in RL due to the commonly used Gaussian policy parameterization. We propose \textbf{Di}verse \textbf{Skil}l \textbf{L}earning (Di-SkilL\footnote{Videos and code are available on the project webpage: \url{https://alrhub.github.io/di-skill-website/}}), an RL method for learning diverse skills using Mixture of Experts, where each expert formalizes a skill as a contextual motion primitive. Di-SkilL optimizes each expert and its associate context distribution to a maximum entropy objective that incentivizes learning diverse skills in similar contexts. The per-expert context distribution enables automatic curricula learning, allowing each expert to focus on its best-performing sub-region of the context space. To overcome hard discontinuities and multi-modalities without any prior knowledge of the environment's unknown context probability space, we leverage energy-based models to represent the per-expert context distributions and demonstrate how we can efficiently train them using the standard policy gradient objective. We show on challenging robot simulation tasks that Di-SkilL can learn diverse and performant skills.
Authors: Zhanyu Liu, Ke Hao, Guanjie Zheng, Yanwei Yu
Abstract: Time series data has been demonstrated to be crucial in various research fields. The management of large quantities of time series data presents challenges in terms of deep learning tasks, particularly for training a deep neural network. Recently, a technique named \textit{Dataset Condensation} has emerged as a solution to this problem. This technique generates a smaller synthetic dataset that has comparable performance to the full real dataset in downstream tasks such as classification. However, previous methods are primarily designed for image and graph datasets, and directly adapting them to the time series dataset leads to suboptimal performance due to their inability to effectively leverage the rich information inherent in time series data, particularly in the frequency domain. In this paper, we propose a novel framework named Dataset \textit{\textbf{Cond}}ensation for \textit{\textbf{T}}ime \textit{\textbf{S}}eries \textit{\textbf{C}}lassification via Dual Domain Matching (\textbf{CondTSC}) which focuses on the time series classification dataset condensation task. Different from previous methods, our proposed framework aims to generate a condensed dataset that matches the surrogate objectives in both the time and frequency domains. Specifically, CondTSC incorporates multi-view data augmentation, dual domain training, and dual surrogate objectives to enhance the dataset condensation process in the time and frequency domains. Through extensive experiments, we demonstrate the effectiveness of our proposed framework, which outperforms other baselines and learns a condensed synthetic dataset that exhibits desirable characteristics such as conforming to the distribution of the original data.
Authors: Ankit Sonthalia, Alexander Rubinstein, Ehsan Abbasnejad, Seong Joon Oh
Abstract: It has recently been conjectured that neural network solution sets reachable via stochastic gradient descent (SGD) are convex, considering permutation invariances (Entezari et al., 2022). This means that a linear path can connect two independent solutions with low loss, given the weights of one of the models are appropriately permuted. However, current methods to test this theory often require very wide networks to succeed. In this work, we conjecture that more generally, the SGD solution set is a "star domain" that contains a "star model" that is linearly connected to all the other solutions via paths with low loss values, modulo permutations. We propose the Starlight algorithm that finds a star model of a given learning task. We validate our claim by showing that this star model is linearly connected with other independently found solutions. As an additional benefit of our study, we demonstrate better uncertainty estimates on the Bayesian Model Averaging over the obtained star domain. Further, we demonstrate star models as potential substitutes for model ensembles. Our code is available at https://github.com/aktsonthalia/starlight.
Authors: Khaoula Chehbouni, Megha Roshan, Emmanuel Ma, Futian Andrew Wei, Afaf Taik, Jackie CK Cheung, Golnoosh Farnadi
Abstract: Recent progress in large language models (LLMs) has led to their widespread adoption in various domains. However, these advancements have also introduced additional safety risks and raised concerns regarding their detrimental impact on already marginalized populations. Despite growing mitigation efforts to develop safety safeguards, such as supervised safety-oriented fine-tuning and leveraging safe reinforcement learning from human feedback, multiple concerns regarding the safety and ingrained biases in these models remain. Furthermore, previous work has demonstrated that models optimized for safety often display exaggerated safety behaviors, such as a tendency to refrain from responding to certain requests as a precautionary measure. As such, a clear trade-off between the helpfulness and safety of these models has been documented in the literature. In this paper, we further investigate the effectiveness of safety measures by evaluating models on already mitigated biases. Using the case of Llama 2 as an example, we illustrate how LLMs' safety responses can still encode harmful assumptions. To do so, we create a set of non-toxic prompts, which we then use to evaluate Llama models. Through our new taxonomy of LLMs responses to users, we observe that the safety/helpfulness trade-offs are more pronounced for certain demographic groups which can lead to quality-of-service harms for marginalized populations.
Authors: Nathan Lambert, Valentina Pyatkin, Jacob Morrison, LJ Miranda, Bill Yuchen Lin, Khyathi Chandu, Nouha Dziri, Sachin Kumar, Tom Zick, Yejin Choi, Noah A. Smith, Hannaneh Hajishirzi
Abstract: Reward models (RMs) are at the crux of successfully using RLHF to align pretrained models to human preferences, yet there has been relatively little study that focuses on evaluation of those models. Evaluating reward models presents an opportunity to understand the opaque technologies used for alignment of language models and which values are embedded in them. Resources for reward model training and understanding are sparse in the nascent open-source community around them. To enhance scientific understanding of reward models, we present RewardBench, a benchmark dataset and code-base for evaluation. The RewardBench dataset is a collection of prompt-chosen-rejected trios spanning chat, reasoning, and safety, to benchmark how reward models perform on challenging, structured and out-of-distribution queries. We create specific comparison datasets for RMs that have subtle, but verifiable reasons (e.g. bugs, incorrect facts) why one answer should be preferred to another. On the RewardBench leaderboard, we evaluate reward models trained with a variety of methods, such as the direct MLE training of classifiers and the implicit reward modeling of Direct Preference Optimization (DPO). We present many findings on propensity for refusals, reasoning limitations, and instruction following shortcomings of various reward models towards a better understanding of the RLHF process.
Authors: Jeremy E. Cohen, Valentin Leplat
Abstract: Regularized nonnegative low-rank approximations such as sparse Nonnegative Matrix Factorization or sparse Nonnegative Tucker Decomposition are an important branch of dimensionality reduction models with enhanced interpretability. However, from a practical perspective, the choice of regularizers and regularization coefficients, as well as the design of efficient algorithms, is challenging because of the multifactor nature of these models and the lack of theory to back these choices. This paper aims at improving upon these issues. By studying a more general model called the Homogeneous Regularized Scale-Invariant, we prove that the scale-invariance inherent to low-rank approximation models causes an implicit regularization with both unexpected beneficial and detrimental effects. This observation allows to better understand the effect of regularization functions in low-rank approximation models, to guide the choice of the regularization hyperparameters, and to design balancing strategies to enhance the convergence speed of dedicated optimization algorithms. Some of these results were already known but restricted to specific instances of regularized low-rank approximations. We also derive a generic Majorization Minimization algorithm that handles many regularized nonnegative low-rank approximations, with convergence guarantees. We showcase our contributions on sparse Nonnegative Matrix Factorization, ridge-regularized Canonical Polyadic decomposition and sparse Nonnegative Tucker Decomposition.
Authors: Huiyi Wang, Haodong Lu, Lina Yao, Dong Gong
Abstract: Continual learning (CL) aims to continually accumulate knowledge from a non-stationary data stream without catastrophic forgetting of learned knowledge, requiring a balance between stability and adaptability. Relying on the generalizable representation in pre-trained models (PTMs), PTM-based CL methods perform effective continual adaptation on downstream tasks by adding learnable adapters or prompts upon the frozen PTMs. However, many existing PTM-based CL methods use restricted adaptation on a fixed set of these modules to avoid forgetting, suffering from limited CL ability. Periodically adding task-specific modules results in linear model growth rate and impaired knowledge reuse. We propose Self-Expansion of pre-trained models with Modularized Adaptation (SEMA), a novel approach to enhance the control of stability-plasticity balance in PTM-based CL. SEMA automatically decides to reuse or add adapter modules on demand in CL, depending on whether significant distribution shift that cannot be handled is detected at different representation levels. We design modular adapter consisting of a functional adapter and a representation descriptor. The representation descriptors are trained as a distribution shift indicator and used to trigger self-expansion signals. For better composing the adapters, an expandable weighting router is learned jointly for mixture of adapter outputs. SEMA enables better knowledge reuse and sub-linear expansion rate. Extensive experiments demonstrate the effectiveness of the proposed self-expansion method, achieving state-of-the-art performance compared to PTM-based CL methods without memory rehearsal.
Authors: David Valiente, Fernando Rodr\'iguez-Mas, Juan V. Alegre-Requena, David Dalmau, Juan C. Ferrer
Abstract: This work presents a set of optimal machine learning (ML) models to represent the temporal degradation suffered by the power conversion efficiency (PCE) of polymeric organic solar cells (OSCs) with a multilayer structure ITO/PEDOT:PSS/P3HT:PCBM/Al. To that aim, we generated a database with 996 entries, which includes up to 7 variables regarding both the manufacturing process and environmental conditions for more than 180 days. Then, we relied on a software framework that brings together a conglomeration of automated ML protocols that execute sequentially against our database by simply command-line interface. This easily permits hyper-optimizing and randomizing seeds of the ML models through exhaustive benchmarking so that optimal models are obtained. The accuracy achieved reaches values of the coefficient determination (R2) widely exceeding 0.90, whereas the root mean squared error (RMSE), sum of squared error (SSE), and mean absolute error (MAE)>1% of the target value, the PCE. Additionally, we contribute with validated models able to screen the behavior of OSCs never seen in the database. In that case, R2~0.96-0.97 and RMSE~1%, thus confirming the reliability of the proposal to predict. For comparative purposes, classical Bayesian regression fitting based on non-linear mean squares (LMS) are also presented, which only perform sufficiently for univariate cases of single OSCs. Hence they fail to outperform the breadth of the capabilities shown by the ML models. Finally, thanks to the standardized results offered by the ML framework, we study the dependencies between the variables of the dataset and their implications for the optimal performance and stability of the OSCs. Reproducibility is ensured by a standardized report altogether with the dataset, which are publicly available at Github.
Authors: Alessandro Brusaferri, Andrea Ballarino, Luigi Grossi, Fabrizio Laurini
Abstract: Probabilistic electricity price forecasting (PEPF) is subject of increasing interest, following the demand for proper quantification of prediction uncertainty, to support the operation in complex power markets with increasing share of renewable generation. Distributional neural networks ensembles have been recently shown to outperform state of the art PEPF benchmarks. Still, they require critical reliability enhancements, as fail to pass the coverage tests at various steps on the prediction horizon. In this work, we propose a novel approach to PEPF, extending the state of the art neural networks ensembles based methods through conformal inference based techniques, deployed within an on-line recalibration procedure. Experiments have been conducted on multiple market regions, achieving day-ahead forecasts with improved hourly coverage and stable probabilistic scores.
Authors: Xiwei Zhang, Tao Li
Abstract: We study the convergence of recursive regularized learning algorithms in the reproducing kernel Hilbert space (RKHS) with dependent and non-stationary online data streams. Firstly, we study the mean square asymptotic stability of a class of random difference equations in RKHS, whose non-homogeneous terms are martingale difference sequences dependent on the homogeneous ones. Secondly, we introduce the concept of random Tikhonov regularization path, and show that if the regularization path is slowly time-varying in some sense, then the output of the algorithm is consistent with the regularization path in mean square. Furthermore, if the data streams also satisfy the RKHS persistence of excitation condition, i.e. there exists a fixed length of time period, such that the conditional expectation of the operators induced by the input data accumulated over every time period has a uniformly strictly positive compact lower bound in the sense of the operator order with respect to time, then the output of the algorithm is consistent with the unknown function in mean square. Finally, for the case with independent and non-identically distributed data streams, the algorithm achieves the mean square consistency provided the marginal probability measures induced by the input data are slowly time-varying and the average measure over each fixed-length time period has a uniformly strictly positive lower bound.
Authors: Tianze Wang, Maryam Honari-Jahromi, Styliani Katsarou, Olga Mikheeva, Theodoros Panagiotakopoulos, Sahar Asadi, Oleg Smirnov
Abstract: Methods for learning latent user representations from historical behavior logs have gained traction for recommendation tasks in e-commerce, content streaming, and other settings. However, this area still remains relatively underexplored in video and mobile gaming contexts. In this work, we present a novel method for overcoming this limitation by extending a long-range Transformer model from the natural language processing domain to player behavior data. We discuss specifics of behavior tracking in games and propose preprocessing and tokenization approaches by viewing in-game events in an analogous way to words in sentences, thus enabling learning player representations in a self-supervised manner in the absence of ground-truth annotations. We experimentally demonstrate the efficacy of the proposed approach in fitting the distribution of behavior events by evaluating intrinsic language modeling metrics. Furthermore, we qualitatively analyze the emerging structure of the learned embedding space and show its value for generating insights into behavior patterns to inform downstream applications.
Authors: Lucas Lingle
Abstract: Large neural network models have become a mainstay of language, vision, and audio processing and synthesis, yet their initialization and learning rates are still set in a largely unsophisticated and potentially expensive fashion, with a potentially high cost incurred every time a new architecture or model size is to be trained. The $\mu$-Parameterization ($\mu$P) offers a potential solution to these challenges, yielding scaling rules for model initialization and learning rates, and reportedly enabling zero-shot hyperparameter transfer from small to large models in a variety of cases. Despite its evident promise, the $\mu$P method is not yet widely adopted, perhaps due to higher implementation complexity, many variations, or complex theoretical background. This work investigates $\mu$P empirically, focusing on the ubiquitous transformer architecture, and aims to answer a simple question: does $\mu$-Transfer yield optimal learning rates in practice? Studying models of up to 10B parameters and training budgets of up to 190B tokens, we find $\mu$-Transfer works as intended for the majority of important cases, yet also identify a few cases where it may not. Our experiment codebase is available at https://github.com/lucaslingle/mu_transformer/
Authors: Wenxuan Zhang, Youssef Mohamed, Bernard Ghanem, Philip H. S. Torr, Adel Bibi, Mohamed Elhoseiny
Abstract: We propose and study a realistic Continual Learning (CL) setting where learning algorithms are granted a restricted computational budget per time step while training. We apply this setting to large-scale semi-supervised Continual Learning scenarios with sparse label rates. Previous proficient CL methods perform very poorly in this challenging setting. Overfitting to the sparse labeled data and insufficient computational budget are the two main culprits for such a poor performance. Our new setting encourages learning methods to effectively and efficiently utilize the unlabeled data during training. To that end, we propose a simple but highly effective baseline, DietCL, which utilizes both unlabeled and labeled data jointly. DietCL meticulously allocates computational budget for both types of data. We validate our baseline, at scale, on several datasets, e.g., CLOC, ImageNet10K, and CGLM, under constraint budget setups. DietCL outperforms, by a large margin, all existing supervised CL algorithms as well as more recent continual semi-supervised methods. Our extensive analysis and ablations demonstrate that DietCL is stable under a full spectrum of label sparsity, computational budget, and various other ablations.
Authors: Herilalaina Rakotoarison, Steven Adriaensen, Neeratyoy Mallik, Samir Garibov, Edward Bergman, Frank Hutter
Abstract: With the increasing computational costs associated with deep learning, automated hyperparameter optimization methods, strongly relying on black-box Bayesian optimization (BO), face limitations. Freeze-thaw BO offers a promising grey-box alternative, strategically allocating scarce resources incrementally to different configurations. However, the frequent surrogate model updates inherent to this approach pose challenges for existing methods, requiring retraining or fine-tuning their neural network surrogates online, introducing overhead, instability, and hyper-hyperparameters. In this work, we propose FT-PFN, a novel surrogate for Freeze-thaw style BO. FT-PFN is a prior-data fitted network (PFN) that leverages the transformers' in-context learning ability to efficiently and reliably do Bayesian learning curve extrapolation in a single forward pass. Our empirical analysis across three benchmark suites shows that the predictions made by FT-PFN are more accurate and 10-100 times faster than those of the deep Gaussian process and deep ensemble surrogates used in previous work. Furthermore, we show that, when combined with our novel acquisition mechanism (MFPI-random), the resulting in-context freeze-thaw BO method (ifBO), yields new state-of-the-art performance in the same three families of deep learning HPO benchmarks considered in prior work.
Authors: Dimitris Bertsimas, Yu Ma
Abstract: Developing an integrated many-to-many framework leveraging multimodal data for multiple tasks is crucial to unifying healthcare applications ranging from diagnoses to operations. In resource-constrained hospital environments, a scalable and unified machine learning framework that improves previous forecast performances could improve hospital operations and save costs. We introduce M3H, an explainable Multimodal Multitask Machine Learning for Healthcare framework that consolidates learning from tabular, time-series, language, and vision data for supervised binary/multiclass classification, regression, and unsupervised clustering. It features a novel attention mechanism balancing self-exploitation (learning source-task), and cross-exploration (learning cross-tasks), and offers explainability through a proposed TIM score, shedding light on the dynamics of task learning interdependencies. M3H encompasses an unprecedented range of medical tasks and machine learning problem classes and consistently outperforms traditional single-task models by on average 11.6% across 40 disease diagnoses from 16 medical departments, three hospital operation forecasts, and one patient phenotyping task. The modular design of the framework ensures its generalizability in data processing, task definition, and rapid model prototyping, making it production ready for both clinical and operational healthcare settings, especially those in constrained environments.
Authors: Hinrikus Wolf, Luis B\"ottcher, Sarra Bouchkati, Philipp Lutat, Jens Breitung, Bastian Jung, Tina M\"ollemann, Viktor Todosijevi\'c, Jan Schiefelbein-Lach, Oliver Pohl, Andreas Ulbig, Martin Grohe
Abstract: In the course of the energy transition, the expansion of generation and consumption will change, and many of these technologies, such as PV systems, electric cars and heat pumps, will influence the power flow, especially in the distribution grids. Scalable methods that can make decisions for each grid connection are needed to enable congestion-free grid operation in the distribution grids. This paper presents a novel end-to-end approach to resolving congestion in distribution grids with deep reinforcement learning. Our architecture learns to curtail power and set appropriate reactive power to determine a non-congested and, thus, feasible grid state. State-of-the-art methods such as the optimal power flow (OPF) demand high computational costs and detailed measurements of every bus in a grid. In contrast, the presented method enables decisions under sparse information with just some buses observable in the grid. Distribution grids are generally not yet fully digitized and observable, so this method can be used for decision-making on the majority of low-voltage grids. On a real low-voltage grid the approach resolves 100\% of violations in the voltage band and 98.8\% of asset overloads. The results show that decisions can also be made on real grids that guarantee sufficient quality for congestion-free grid operation.
Authors: Joseph Janssen, Ardalan Tootchi, Ali A. Ameli
Abstract: Fine-resolution spatial patterns of water table depth (WTD) play a crucial role in shaping ecological resilience, hydrological connectivity, and anthropocentric objectives. Generally, a large-scale (e.g., continental or global) spatial map of static WTD can be simulated using either physically-based (PB) or machine learning-based (ML) models. We construct three fine-resolution (500 m) ML simulations of WTD, using the XGBoost algorithm and more than 20 million real and proxy observations of WTD, across the United States and Canada. The three ML models were constrained using known physical relations between WTD's drivers and WTD and were trained by sequentially adding real and proxy observations of WTD. We interpret the black box of our physically constrained ML models and compare it against available literature in groundwater hydrology. Through an extensive (pixel-by-pixel) evaluation, we demonstrate that our models can more accurately predict unseen real and proxy observations of WTD across most of North America's ecoregions compared to three available PB simulations of WTD. However, we still argue that large-scale WTD estimation is far from being a solved problem. We reason that due to biased observational data mainly collected from low-elevation floodplains, the misspecification of equations within physically-based models, and the over-flexibility of machine learning models, verifiably accurate simulations of WTD do not yet exist. Ultimately, we thoroughly discuss future directions that may help hydrogeologists decide how to proceed with WTD estimations, with a particular focus on the application of machine learning and the use of proxy satellite data.
Authors: Alice Cicirello
Abstract: This position paper takes a broad look at Physics-Enhanced Machine Learning (PEML) -- also known as Scientific Machine Learning -- with particular focus to those PEML strategies developed to tackle dynamical systems' challenges. The need to go beyond Machine Learning (ML) strategies is driven by: (i) limited volume of informative data, (ii) avoiding accurate-but-wrong predictions; (iii) dealing with uncertainties; (iv) providing Explainable and Interpretable inferences. A general definition of PEML is provided by considering four physics and domain knowledge biases, and three broad groups of PEML approaches are discussed: physics-guided, physics-encoded and physics-informed. The advantages and challenges in developing PEML strategies for guiding high-consequence decision making in engineering applications involving complex dynamical systems, are presented.
Authors: Vasilis Kontonis, Mingchen Ma, Christos Tzamos
Abstract: We consider an active learning setting where a learner is presented with a pool S of n unlabeled examples belonging to a domain X and asks queries to find the underlying labeling that agrees with a target concept h^* \in H. In contrast to traditional active learning that queries a single example for its label, we study more general region queries that allow the learner to pick a subset of the domain T \subset X and a target label y and ask a labeler whether h^*(x) = y for every example in the set T \cap S. Such more powerful queries allow us to bypass the limitations of traditional active learning and use significantly fewer rounds of interactions to learn but can potentially lead to a significantly more complex query language. Our main contribution is quantifying the trade-off between the number of queries and the complexity of the query language used by the learner. We measure the complexity of the region queries via the VC dimension of the family of regions. We show that given any hypothesis class H with VC dimension d, one can design a region query family Q with VC dimension O(d) such that for every set of n examples S \subset X and every h^* \in H, a learner can submit O(d log n) queries from Q to a labeler and perfectly label S. We show a matching lower bound by designing a hypothesis class H with VC dimension d and a dataset S \subset X of size n such that any learning algorithm using any query class with VC dimension less than O(d) must make poly(n) queries to label S perfectly. Finally, we focus on well-studied hypothesis classes including unions of intervals, high-dimensional boxes, and d-dimensional halfspaces, and obtain stronger results. In particular, we design learning algorithms that (i) are computationally efficient and (ii) work even when the queries are not answered based on the learner's pool of examples S but on some unknown superset L of S
Authors: Amutheezan Sivagnanam, Ava Pettet, Hunter Lee, Ayan Mukhopadhyay, Abhishek Dubey, Aron Laszka
Abstract: An emergency responder management (ERM) system dispatches responders, such as ambulances, when it receives requests for medical aid. ERM systems can also proactively reposition responders between predesignated waiting locations to cover any gaps that arise due to the prior dispatch of responders or significant changes in the distribution of anticipated requests. Optimal repositioning is computationally challenging due to the exponential number of ways to allocate responders between locations and the uncertainty in future requests. The state-of-the-art approach in proactive repositioning is a hierarchical approach based on spatial decomposition and online Monte Carlo tree search, which may require minutes of computation for each decision in a domain where seconds can save lives. We address the issue of long decision times by introducing a novel reinforcement learning (RL) approach, based on the same hierarchical decomposition, but replacing online search with learning. To address the computational challenges posed by large, variable-dimensional, and discrete state and action spaces, we propose: (1) actor-critic based agents that incorporate transformers to handle variable-dimensional states and actions, (2) projections to fixed-dimensional observations to handle complex states, and (3) combinatorial techniques to map continuous actions to discrete allocations. We evaluate our approach using real-world data from two U.S. cities, Nashville, TN and Seattle, WA. Our experiments show that compared to the state of the art, our approach reduces computation time per decision by three orders of magnitude, while also slightly reducing average ambulance response time by 5 seconds.
Authors: Chenhui Xu, Fuxun Yu, Maoliang Li, Zihao Zheng, Zirui Xu, Jinjun Xiong, Xiang Chen
Abstract: The past neural network design has largely focused on feature representation space dimension and its capacity scaling (e.g., width, depth), but overlooked the feature interaction space scaling. Recent advancements have shown shifted focus towards element-wise multiplication to facilitate higher-dimensional feature interaction space for better information transformation. Despite this progress, multiplications predominantly capture low-order interactions, thus remaining confined to a finite-dimensional interaction space. To transcend this limitation, classic kernel methods emerge as a promising solution to engage features in an infinite-dimensional space. We introduce InfiNet, a model architecture that enables feature interaction within an infinite-dimensional space created by RBF kernel. Our experiments reveal that InfiNet achieves new state-of-the-art, owing to its capability to leverage infinite-dimensional interactions, significantly enhancing model performance.
Authors: Kabir Dabholkar, Omri Barak
Abstract: Latent variable models serve as powerful tools to infer underlying dynamics from observed neural activity. However, due to the absence of ground truth data, prediction benchmarks are often employed as proxies. In this study, we reveal the limitations of the widely-used 'co-smoothing' prediction framework and propose an improved few-shot prediction approach that encourages more accurate latent dynamics. Utilizing a student-teacher setup with Hidden Markov Models, we demonstrate that the high co-smoothing model space can encompass models with arbitrary extraneous dynamics within their latent representations. To address this, we introduce a secondary metric -- a few-shot version of co-smoothing. This involves performing regression from the latent variables to held-out channels in the data using fewer trials. Our results indicate that among models with near-optimal co-smoothing, those with extraneous dynamics underperform in the few-shot co-smoothing compared to 'minimal' models devoid of such dynamics. We also provide analytical insights into the origin of this phenomenon. We further validate our findings on real neural data using two state-of-the-art methods: LFADS and STNDT. In the absence of ground truth, we suggest a proxy measure to quantify extraneous dynamics. By cross-decoding the latent variables of all model pairs with high co-smoothing, we identify models with minimal extraneous dynamics. We find a correlation between few-shot co-smoothing performance and this new measure. In summary, we present a novel prediction metric designed to yield latent variables that more accurately reflect the ground truth, offering a significant improvement for latent dynamics inference.
Authors: Indu Solomon, Aye Phyu Phyu Aung, Uttam Kumar, Senthilnath Jayavelu
Abstract: Continual learning (CL) models are designed to learn new tasks arriving sequentially without re-training the network. However, real-world ML applications have very limited label information and these models suffer from catastrophic forgetting. To address these issues, we propose an unsupervised CL model with task experts called Unsupervised Task Expert Lifelong Learning (U-TELL) to continually learn the data arriving in a sequence addressing catastrophic forgetting. During training of U-TELL, we introduce a new expert on arrival of a new task. Our proposed architecture has task experts, a structured data generator and a task assigner. Each task expert is composed of 3 blocks; i) a variational autoencoder to capture the task distribution and perform data abstraction, ii) a k-means clustering module, and iii) a structure extractor to preserve latent task data signature. During testing, task assigner selects a suitable expert to perform clustering. U-TELL does not store or replay task samples, instead, we use generated structured samples to train the task assigner. We compared U-TELL with five SOTA unsupervised CL methods. U-TELL outperformed all baselines on seven benchmarks and one industry dataset for various CL scenarios with a training time over 6 times faster than the best performing baseline.
Authors: Penghui Qi, Xinyi Wan, Nyamdavaa Amar, Min Lin
Abstract: Pipeline parallelism has been widely explored, but most existing schedules lack a systematic methodology. In this paper, we propose a framework to decompose pipeline schedules as repeating a building block and we show that the lifespan of the building block decides the peak activation memory of the pipeline schedule. Guided by the observations, we find that almost all existing pipeline schedules, to the best of our knowledge, are memory inefficient. To address this, we introduce a family of memory efficient building blocks with controllable activation memory, which can reduce the peak activation memory to 1/2 of 1F1B without sacrificing efficiency, and even to 1/3 with comparable throughput. We can also achieve almost zero pipeline bubbles while maintaining the same activation memory as 1F1B. Our evaluations demonstrate that in pure pipeline parallelism settings, our methods outperform 1F1B by from 7% to 55% in terms of throughput. When employing a grid search over hybrid parallelism hyperparameters in practical scenarios, our proposed methods demonstrate a 16% throughput improvement over the 1F1B baseline for large language models.
Authors: Noga Amit, Shafi Goldwasser, Orr Paradise, Guy Rothblum
Abstract: How can we trust the correctness of a learned model on a particular input of interest? Model accuracy is typically measured *on average* over a distribution of inputs, giving no guarantee for any fixed input. This paper proposes a theoretically-founded solution to this problem: to train *Self-Proving models* that prove the correctness of their output to a verification algorithm $V$ via an Interactive Proof. Self-Proving models satisfy that, with high probability over a random input, the model generates a correct output *and* successfully proves its correctness to $V\!$. The *soundness* property of $V$ guarantees that, for *every* input, no model can convince $V$ of the correctness of an incorrect output. Thus, a Self-Proving model proves correctness of most of its outputs, while *all* incorrect outputs (of any model) are detected by $V$. We devise a generic method for learning Self-Proving models, and we prove convergence bounds under certain assumptions. The theoretical framework and results are complemented by experiments on an arithmetic capability: computing the greatest common divisor (GCD) of two integers. Our learning method is used to train a Self-Proving transformer that computes the GCD *and* proves the correctness of its answer.
Authors: Biqing Qi, Junqi Gao, Kaiyan Zhang, Dong Li, Jianxing Liu, Ligang Wu, Bowen Zhou
Abstract: Despite the promising performance of state space models (SSMs) in long sequence modeling, limitations still exist. Advanced SSMs like S5 and S6 (Mamba) in addressing non-uniform sampling, their recursive structures impede efficient SSM computation via convolution. To overcome compatibility limitations in parallel convolutional computation, this paper proposes a novel non-recursive non-uniform sample processing strategy. Theoretical analysis of SSMs through the lens of Event-Triggered Control (ETC) theory reveals the Non-Stable State (NSS) problem, where deviations from sampling point requirements lead to error transmission and accumulation, causing the divergence of the SSM's hidden state. Our analysis further reveals that adjustments of input sequences with early memories can mitigate the NSS problem, achieving Sampling Step Adaptation (SSA). Building on this insight, we introduce a simple yet effective plug-and-play mechanism, State Memory Replay (SMR), which utilizes learnable memories to adjust the current state with multi-step information for generalization at sampling points different from those in the training data. This enables SSMs to stably model varying sampling points. Experiments on long-range modeling tasks in autoregressive language modeling and Long Range Arena demonstrate the general effectiveness of the SMR mechanism for a series of SSM models.
Authors: Hanye Zhao, Xiaoshen Han, Zhengbang Zhu, Minghuan Liu, Yong Yu, Weinan Zhang
Abstract: With the great success of diffusion models (DMs) in generating realistic synthetic vision data, many researchers have investigated their potential in decision-making and control. Most of these works utilized DMs to sample directly from the trajectory space, where DMs can be viewed as a combination of dynamics models and policies. In this work, we explore how to decouple DMs' ability as dynamics models in fully offline settings, allowing the learning policy to roll out trajectories. As DMs learn the data distribution from the dataset, their intrinsic policy is actually the behavior policy induced from the dataset, which results in a mismatch between the behavior policy and the learning policy. We propose Dynamics Diffusion, short as DyDiff, which can inject information from the learning policy to DMs iteratively. DyDiff ensures long-horizon rollout accuracy while maintaining policy consistency and can be easily deployed on model-free algorithms. We provide theoretical analysis to show the advantage of DMs on long-horizon rollout over models and demonstrate the effectiveness of DyDiff in the context of offline reinforcement learning, where the rollout dataset is provided but no online environment for interaction. Our code is at https://github.com/FineArtz/DyDiff.
Authors: Elliot Layne, Jason Hartford, S\'ebastien Lachapelle, Mathieu Blanchette, Dhanya Sridhar
Abstract: Many causal systems such as biological processes in cells can only be observed indirectly via measurements, such as gene expression. Causal representation learning -- the task of correctly mapping low-level observations to latent causal variables -- could advance scientific understanding by enabling inference of latent variables such as pathway activation. In this paper, we develop methods for inferring latent variables from multiple related datasets (environments) and tasks. As a running example, we consider the task of predicting a phenotype from gene expression, where we often collect data from multiple cell types or organisms that are related in known ways. The key insight is that the mapping from latent variables driven by gene expression to the phenotype of interest changes sparsely across closely related environments. To model sparse changes, we introduce Tree-Based Regularization (TBR), an objective that minimizes both prediction error and regularizes closely related environments to learn similar predictors. We prove that under assumptions about the degree of sparse changes, TBR identifies the true latent variables up to some simple transformations. We evaluate the theory empirically with both simulations and ground-truth gene expression data. We find that TBR recovers the latent causal variables better than related methods across these settings, even under settings that violate some assumptions of the theory.
Authors: Saurabh Bajaj, Hui Guan, Marco Serafini
Abstract: Graph Neural Networks (GNNs) have gained significant attention in recent years due to their ability to learn representations of graph structured data. Two common methods for training GNNs are mini-batch training and full-graph training. Since these two methods require different training pipelines and systems optimizations, two separate categories of GNN training systems emerged, each tailored for one method. Works that introduce systems belonging to a particular category predominantly compare them with other systems within the same category, offering limited or no comparison with systems from the other category. Some prior work also justifies its focus on one specific training method by arguing that it achieves higher accuracy than the alternative. The literature, however, has incomplete and contradictory evidence in this regard. In this paper, we provide a comprehensive empirical comparison of full-graph and mini-batch GNN training systems to get a clearer picture of the state of the art in the field. We find that the mini-batch training systems we consider consistently converge faster than the full-graph training ones across multiple datasets, GNN models, and system configurations, with speedups between 2.4x - 15.2x. We also find that both training techniques converge to similar accuracy values, so comparing systems across the two categories in terms of time-to-accuracy is a sound approach.
Authors: Naveen Raman, Zheyuan Ryan Shi, Fei Fang
Abstract: Restless multi-armed bandits (RMAB) extend multi-armed bandits so pulling an arm impacts future states. Despite the success of RMABs, a key limiting assumption is the separability of rewards into a sum across arms. We address this deficiency by proposing restless-multi-armed bandit with global rewards (RMAB-G), a generalization of RMABs to global non-separable rewards. To solve RMAB-G, we develop the Linear- and Shapley-Whittle indices, which extend Whittle indices from RMABs to RMAB-Gs. We prove approximation bounds but also point out how these indices could fail when reward functions are highly non-linear. To overcome this, we propose two sets of adaptive policies: the first computes indices iteratively, and the second combines indices with Monte-Carlo Tree Search (MCTS). Empirically, we demonstrate that our proposed policies outperform baselines and index-based policies with synthetic data and real-world data from food rescue.
Authors: Simone Brugiapaglia, Nick Dexter, Samir Karam, Weiqi Wang
Abstract: On the forefront of scientific computing, Deep Learning (DL), i.e., machine learning with Deep Neural Networks (DNNs), has emerged a powerful new tool for solving Partial Differential Equations (PDEs). It has been observed that DNNs are particularly well suited to weakening the effect of the curse of dimensionality, a term coined by Richard E. Bellman in the late `50s to describe challenges such as the exponential dependence of the sample complexity, i.e., the number of samples required to solve an approximation problem, on the dimension of the ambient space. However, although DNNs have been used to solve PDEs since the `90s, the literature underpinning their mathematical efficiency in terms of numerical analysis (i.e., stability, accuracy, and sample complexity), is only recently beginning to emerge. In this paper, we leverage recent advancements in function approximation using sparsity-based techniques and random sampling to develop and analyze an efficient high-dimensional PDE solver based on DL. We show, both theoretically and numerically, that it can compete with a novel stable and accurate compressive spectral collocation method. In particular, we demonstrate a new practical existence theorem, which establishes the existence of a class of trainable DNNs with suitable bounds on the network architecture and a sufficient condition on the sample complexity, with logarithmic or, at worst, linear scaling in dimension, such that the resulting networks stably and accurately approximate a diffusion-reaction PDE with high probability.
Authors: Yuxuan Chen, Rongpeng Li, Xiaoxue Yu, Zhifeng Zhao, Honggang Zhang
Abstract: Optimizing the deployment of large language models (LLMs) in edge computing environments is critical for enhancing privacy and computational efficiency. Toward efficient wireless LLM inference in edge computing, this study comprehensively analyzes the impact of different splitting points in mainstream open-source LLMs. On this basis, this study introduces a framework taking inspiration from model-based reinforcement learning (MBRL) to determine the optimal splitting point across the edge and user equipment (UE). By incorporating a reward surrogate model, our approach significantly reduces the computational cost of frequent performance evaluations. Extensive simulations demonstrate that this method effectively balances inference performance and computational load under varying network conditions, providing a robust solution for LLM deployment in decentralized settings.
Authors: Yuda Wang, Xuxin He, Shengxin Zhu
Abstract: Predicting user preferences and sequential dependencies based on historical behavior is the core goal of sequential recommendation. Although attention-based models have shown effectiveness in this field, they often struggle with inference inefficiency due to the quadratic computational complexity inherent in attention mechanisms, especially with long-range behavior sequences. Drawing inspiration from the recent advancements of state space models (SSMs) in control theory, which provide a robust framework for modeling and controlling dynamic systems, we introduce EchoMamba4Rec. Control theory emphasizes the use of SSMs for managing long-range dependencies and maintaining inferential efficiency through structured state matrices. EchoMamba4Rec leverages these control relationships in sequential recommendation and integrates bi-directional processing with frequency-domain filtering to capture complex patterns and dependencies in user interaction data more effectively. Our model benefits from the ability of state space models (SSMs) to learn and perform parallel computations, significantly enhancing computational efficiency and scalability. It features a bi-directional Mamba module that incorporates both forward and reverse Mamba components, leveraging information from both past and future interactions. Additionally, a filter layer operates in the frequency domain using learnable Fast Fourier Transform (FFT) and learnable filters, followed by an inverse FFT to refine item embeddings and reduce noise. We also integrate Gate Linear Units (GLU) to dynamically control information flow, enhancing the model's expressiveness and training stability. Experimental results demonstrate that EchoMamba significantly outperforms existing models, providing more accurate and personalized recommendations.
Authors: Yifei Wang, Sungyoon Kim, Paul Chu, Indu Subramaniam, Mert Pilanci
Abstract: We introduce randomized algorithms to Clifford's Geometric Algebra, generalizing randomized linear algebra to hypercomplex vector spaces. This novel approach has many implications in machine learning, including training neural networks to global optimality via convex optimization. Additionally, we consider fine-tuning large language model (LLM) embeddings as a key application area, exploring the intersection of geometric algebra and modern AI techniques. In particular, we conduct a comparative analysis of the robustness of transfer learning via embeddings, such as OpenAI GPT models and BERT, using traditional methods versus our novel approach based on convex optimization. We test our convex optimization transfer learning method across a variety of case studies, employing different embeddings (GPT-4 and BERT embeddings) and different text classification datasets (IMDb, Amazon Polarity Dataset, and GLUE) with a range of hyperparameter settings. Our results demonstrate that convex optimization and geometric algebra not only enhances the performance of LLMs but also offers a more stable and reliable method of transfer learning via embeddings.
Authors: Yang Liu, Peng Zhang, Yang Gao, Chuan Zhou, Zhao Li, Hongyang Chen
Abstract: In recent years, graph neural networks (GNNs) have become increasingly popular for solving NP-hard combinatorial optimization (CO) problems, such as maximum cut and maximum independent set. The core idea behind these methods is to represent a CO problem as a graph and then use GNNs to learn the node/graph embedding with combinatorial information. Although these methods have achieved promising results, given a specific CO problem, the design of GNN architectures still requires heavy manual work with domain knowledge. Existing automated GNNs are mostly focused on traditional graph learning problems, which is inapplicable to solving NP-hard CO problems. To this end, we present a new class of \textbf{AUTO}mated \textbf{G}NNs for solving \textbf{NP}-hard problems, namely \textbf{AutoGNP}. We represent CO problems by GNNs and focus on two specific problems, i.e., mixed integer linear programming and quadratic unconstrained binary optimization. The idea of AutoGNP is to use graph neural architecture search algorithms to automatically find the best GNNs for a given NP-hard combinatorial optimization problem. Compared with existing graph neural architecture search algorithms, AutoGNP utilizes two-hop operators in the architecture search space. Moreover, AutoGNP utilizes simulated annealing and a strict early stopping policy to avoid local optimal solutions. Empirical results on benchmark combinatorial problems demonstrate the superiority of our proposed model.
Authors: Gijs de Jong, Macha J. Meijer, Derck W. E. Prinzhorn, Harold Ruiter
Abstract: This work aims to reproduce the findings of the paper "Fair Attribute Completion on Graph with Missing Attributes" written by Guo, Chu, and Li arXiv:2302.12977 by investigating the claims made in the paper. This paper suggests that the results of the original paper are reproducible and thus, the claims hold. However, the claim that FairAC is a generic framework for many downstream tasks is very broad and could therefore only be partially tested. Moreover, we show that FairAC is generalizable to various datasets and sensitive attributes and show evidence that the improvement in group fairness of the FairAC framework does not come at the expense of individual fairness. Lastly, the codebase of FairAC has been refactored and is now easily applicable for various datasets and models.
Authors: Yang Liu, Chuan Zhou, Peng Zhang, Shirui Pan, Zhao Li, Hongyang Chen
Abstract: In recent years, there has been notable interest in investigating combinatorial optimization (CO) problems by neural-based framework. An emerging strategy to tackle these challenging problems involves the adoption of graph neural networks (GNNs) as an alternative to traditional algorithms, a subject that has attracted considerable attention. Despite the growing popularity of GNNs and traditional algorithm solvers in the realm of CO, there is limited research on their integrated use and the correlation between them within an end-to-end framework. The primary focus of our work is to formulate a more efficient and precise framework for CO by employing decision-focused learning on graphs. Additionally, we introduce a decision-focused framework that utilizes GNNs to address CO problems with auxiliary support. To realize an end-to-end approach, we have designed two cascaded modules: (a) an unsupervised trained graph predictive model, and (b) a solver for quadratic binary unconstrained optimization. Empirical evaluations are conducted on various classical tasks, including maximum cut, maximum independent set, and minimum vertex cover. The experimental results on classical CO problems (i.e. MaxCut, MIS, and MVC) demonstrate the superiority of our method over both the standalone GNN approach and classical methods.
Authors: Behrooz Tahmasebi, Ashkan Soleymani, Dara Bahri, Stefanie Jegelka, Patrick Jaillet
Abstract: Recently, there has been a surge in interest in developing optimization algorithms for overparameterized models as achieving generalization is believed to require algorithms with suitable biases. This interest centers on minimizing sharpness of the original loss function; the Sharpness-Aware Minimization (SAM) algorithm has proven effective. However, most literature only considers a few sharpness measures, such as the maximum eigenvalue or trace of the training loss Hessian, which may not yield meaningful insights for non-convex optimization scenarios like neural networks. Additionally, many sharpness measures are sensitive to parameter invariances in neural networks, magnifying significantly under rescaling parameters. Motivated by these challenges, we introduce a new class of sharpness measures in this paper, leading to new sharpness-aware objective functions. We prove that these measures are \textit{universally expressive}, allowing any function of the training loss Hessian matrix to be represented by appropriate hyperparameters. Furthermore, we show that the proposed objective functions explicitly bias towards minimizing their corresponding sharpness measures, and how they allow meaningful applications to models with parameter invariances (such as scale-invariances). Finally, as instances of our proposed general framework, we present \textit{Frob-SAM} and \textit{Det-SAM}, which are specifically designed to minimize the Frobenius norm and the determinant of the Hessian of the training loss, respectively. We also demonstrate the advantages of our general framework through extensive experiments.
Authors: Tian Wang, Chuang Wang
Abstract: Neural operators effectively solve PDE problems from data without knowing the explicit equations, which learn the map from the input sequences of observed samples to the predicted values. Most existed works build the model in the original geometric space, leading to high computational costs when the number of sample points is large. We present the Latent Neural Operator (LNO) solving PDEs in the latent space. In particular, we first propose Physics-Cross-Attention (PhCA) transforming representation from the geometric space to the latent space, then learn the operator in the latent space, and finally recover the real-world geometric space via the inverse PhCA map. Our model retains flexibility that can decode values in any position not limited to locations defined in training set, and therefore can naturally perform interpolation and extrapolation tasks particularly useful for inverse problems. Moreover, the proposed LNO improves in both prediction accuracy and computational efficiency. Experiments show that LNO reduces the GPU memory by 50%, speeds up training 1.8 times, and reaches state-of-the-art accuracy on four out of six benchmarks for forward problems and a benchmark for inverse problem.
Authors: Can Yaras, Peng Wang, Laura Balzano, Qing Qu
Abstract: While overparameterization in machine learning models offers great benefits in terms of optimization and generalization, it also leads to increased computational requirements as model sizes grow. In this work, we show that by leveraging the inherent low-dimensional structures of data and compressible dynamics within the model parameters, we can reap the benefits of overparameterization without the computational burdens. In practice, we demonstrate the effectiveness of this approach for deep low-rank matrix completion as well as fine-tuning language models. Our approach is grounded in theoretical findings for deep overparameterized low-rank matrix recovery, where we show that the learning dynamics of each weight matrix are confined to an invariant low-dimensional subspace. Consequently, we can construct and train compact, highly compressed factorizations possessing the same benefits as their overparameterized counterparts. In the context of deep matrix completion, our technique substantially improves training efficiency while retaining the advantages of overparameterization. For language model fine-tuning, we propose a method called "Deep LoRA", which improves the existing low-rank adaptation (LoRA) technique, leading to reduced overfitting and a simplified hyperparameter setup, while maintaining comparable efficiency. We validate the effectiveness of Deep LoRA on natural language tasks, particularly when fine-tuning with limited data. Our code is available at https://github.com/cjyaras/deep-lora-transformers.
Authors: Andy Zou, Long Phan, Justin Wang, Derek Duenas, Maxwell Lin, Maksym Andriushchenko, Rowan Wang, Zico Kolter, Matt Fredrikson, Dan Hendrycks
Abstract: AI systems can take harmful actions and are highly vulnerable to adversarial attacks. We present an approach, inspired by recent advances in representation engineering, that interrupts the models as they respond with harmful outputs with "circuit breakers." Existing techniques aimed at improving alignment, such as refusal training, are often bypassed. Techniques such as adversarial training try to plug these holes by countering specific attacks. As an alternative to refusal training and adversarial training, circuit-breaking directly controls the representations that are responsible for harmful outputs in the first place. Our technique can be applied to both text-only and multimodal language models to prevent the generation of harmful outputs without sacrificing utility -- even in the presence of powerful unseen attacks. Notably, while adversarial robustness in standalone image recognition remains an open challenge, circuit breakers allow the larger multimodal system to reliably withstand image "hijacks" that aim to produce harmful content. Finally, we extend our approach to AI agents, demonstrating considerable reductions in the rate of harmful actions when they are under attack. Our approach represents a significant step forward in the development of reliable safeguards to harmful behavior and adversarial attacks.
Authors: Eirini Katsidoniotaki, Biao Su, Eleni Kelasidi, Themistoklis P. Sapsis
Abstract: As the global population grows and climate change intensifies, sustainable food production is critical. Marine aquaculture offers a viable solution, providing a sustainable protein source. However, the industry's expansion requires novel technologies for remote management and autonomous operations. Digital twin technology can advance the aquaculture industry, but its adoption has been limited. Fish net cages, which are flexible floating structures, are critical yet vulnerable components of aquaculture farms. Exposed to harsh and dynamic marine environments, the cages experience significant loads and risk damage, leading to fish escapes, environmental impacts, and financial losses. We propose a multifidelity surrogate modeling framework for integration into a digital twin for real-time monitoring of aquaculture net cage structural dynamics under stochastic marine conditions. Central to this framework is the nonlinear autoregressive Gaussian process method, which learns complex, nonlinear cross-correlations between models of varying fidelity. It combines low-fidelity simulation data with a small set of high-fidelity field sensor measurements, which offer the real dynamics but are costly and spatially sparse. Validated at the SINTEF ACE fish farm in Norway, our digital twin receives online metocean data and accurately predicts net cage displacements and mooring line loads, aligning closely with field measurements. The proposed framework is beneficial where application-specific data are scarce, offering rapid predictions and real-time system representation. The developed digital twin prevents potential damages by assessing structural integrity and facilitates remote operations with unmanned underwater vehicles. Our work also compares GP and GCNs for predicting net cage deformation, highlighting the latter's effectiveness in complex structural applications.
Authors: Renpu Liu, Cong Shen, Jing Yang
Abstract: Federated representation learning (FRL) is a popular personalized federated learning (FL) framework where clients work together to train a common representation while retaining their personalized heads. Existing studies, however, largely focus on the over-parameterized regime. In this paper, we make the initial efforts to investigate FRL in the under-parameterized regime, where the FL model is insufficient to express the variations in all ground-truth models. We propose a novel FRL algorithm FLUTE, and theoretically characterize its sample complexity and convergence rate for linear models in the under-parameterized regime. To the best of our knowledge, this is the first FRL algorithm with provable performance guarantees in this regime. FLUTE features a data-independent random initialization and a carefully designed objective function that aids the distillation of subspace spanned by the global optimal representation from the misaligned local representations. On the technical side, we bridge low-rank matrix approximation techniques with the FL analysis, which may be of broad interest. We also extend FLUTE beyond linear representations. Experimental results demonstrate that FLUTE outperforms state-of-the-art FRL solutions in both synthetic and real-world tasks.
Authors: Zheng Huang, Qihui Yang, Dawei Zhou, Yujun Yan
Abstract: Although most graph neural networks (GNNs) can operate on graphs of any size, their classification performance often declines on graphs larger than those encountered during training. Existing methods insufficiently address the removal of size information from graph representations, resulting in sub-optimal performance and reliance on backbone models. In response, we propose DISGEN, a novel and model-agnostic framework designed to disentangle size factors from graph representations. DISGEN employs size- and task-invariant augmentations and introduces a decoupling loss that minimizes shared information in hidden representations, with theoretical guarantees for its effectiveness. Our empirical results show that DISGEN outperforms the state-of-the-art models by up to 6% on real-world datasets, underscoring its effectiveness in enhancing the size generalizability of GNNs. Our codes are available at: https://github.com/GraphmindDartmouth/DISGEN.
Authors: Sumanta Roy, Chandrasekhar Annavarapu, Pratanu Roy, Antareep Kumar Sarma
Abstract: We present an efficient physics-informed neural networks (PINNs) framework, termed Adaptive Interface-PINNs (AdaI-PINNs), to improve the modeling of interface problems with discontinuous coefficients and/or interfacial jumps. This framework is an enhanced version of its predecessor, Interface PINNs or I-PINNs (Sarma et al.; https://dx.doi.org/10.2139/ssrn.4766623), which involves domain decomposition and assignment of different predefined activation functions to the neural networks in each subdomain across a sharp interface, while keeping all other parameters of the neural networks identical. In AdaI-PINNs, the activation functions vary solely in their slopes, which are trained along with the other parameters of the neural networks. This makes the AdaI-PINNs framework fully automated without requiring preset activation functions. Comparative studies on one-dimensional, two-dimensional, and three-dimensional benchmark elliptic interface problems reveal that AdaI-PINNs outperform I-PINNs, reducing computational costs by 2-6 times while producing similar or better accuracy.
Authors: Zhigang Yao, Yuqing Xia
Abstract: There has been an emerging trend in non-Euclidean statistical analysis of aiming to recover a low dimensional structure, namely a manifold, underlying the high dimensional data. Recovering the manifold requires the noise to be of certain concentration. Existing methods address this problem by constructing an approximated manifold based on the tangent space estimation at each sample point. Although theoretical convergence for these methods is guaranteed, either the samples are noiseless or the noise is bounded. However, if the noise is unbounded, which is a common scenario, the tangent space estimation at the noisy samples will be blurred. Fitting a manifold from the blurred tangent space might increase the inaccuracy. In this paper, we introduce a new manifold-fitting method, by which the output manifold is constructed by directly estimating the tangent spaces at the projected points on the underlying manifold, rather than at the sample points, to decrease the error caused by the noise. Assuming the noise is unbounded, our new method provides theoretical convergence in high probability, in terms of the upper bound of the distance between the estimated and underlying manifold. The smoothness of the estimated manifold is also evaluated by bounding the supremum of twice difference above. Numerical simulations are provided to validate our theoretical findings and demonstrate the advantages of our method over other relevant manifold fitting methods. Finally, our method is applied to real data examples.
Authors: Alessandro Luongo, Changpeng Shao
Abstract: We propose new quantum algorithms for estimating spectral sums of positive semi-definite (PSD) matrices. The spectral sum of an PSD matrix $A$, for a function $f$, is defined as $ \text{Tr}[f(A)] = \sum_j f(\lambda_j)$, where $\lambda_j$ are the eigenvalues of $A$. Typical examples of spectral sums are the von Neumann entropy, the trace of $A^{-1}$, the log-determinant, and the Schatten $p$-norm, where the latter does not require the matrix to be PSD. The current best classical randomized algorithms estimating these quantities have a runtime that is at least linearly in the number of nonzero entries of the matrix and quadratic in the estimation error. Assuming access to a block-encoding of a matrix, our algorithms are sub-linear in the matrix size, and depend at most quadratically on other parameters, like the condition number and the approximation error, and thus can compete with most of the randomized and distributed classical algorithms proposed in the literature, and polynomially improve the runtime of other quantum algorithms proposed for the same problems. We show how the algorithms and techniques used in this work can be applied to three problems in spectral graph theory: approximating the number of triangles, the effective resistance, and the number of spanning trees within a graph.
Authors: Tobias Schlosser, Frederik Beuth, Danny Kowerko
Abstract: Whereas conventional state-of-the-art image processing systems of recording and output devices almost exclusively utilize square arranged methods, biological models, however, suggest an alternative, evolutionarily-based structure. Inspired by the human visual perception system, hexagonal image processing in the context of machine learning offers a number of key advantages that can benefit both researchers and users alike. The hexagonal deep learning framework Hexnet leveraged in this contribution serves therefore the generation of hexagonal images by utilizing hexagonal deep neural networks (H-DNN). As the results of our created test environment show, the proposed models can surpass current approaches of conventional image generation. While resulting in a reduction of the models' complexity in the form of trainable parameters, they furthermore allow an increase of test rates in comparison to their square counterparts.
Authors: Mohammad Emtiyaz Khan, H{\aa}vard Rue
Abstract: We show that many machine-learning algorithms are specific instances of a single algorithm called the \emph{Bayesian learning rule}. The rule, derived from Bayesian principles, yields a wide-range of algorithms from fields such as optimization, deep learning, and graphical models. This includes classical algorithms such as ridge regression, Newton's method, and Kalman filter, as well as modern deep-learning algorithms such as stochastic-gradient descent, RMSprop, and Dropout. The key idea in deriving such algorithms is to approximate the posterior using candidate distributions estimated by using natural gradients. Different candidate distributions result in different algorithms and further approximations to natural gradients give rise to variants of those algorithms. Our work not only unifies, generalizes, and improves existing algorithms, but also helps us design new ones.
Authors: Tobias Schlosser, Frederik Beuth, Trixy Meyer, Arunodhayan Sampath Kumar, Gabriel Stolze, Olga Furashova, Katrin Engelmann, Danny Kowerko
Abstract: In ophthalmology, intravitreal operative medication therapy (IVOM) is a widespread treatment for diseases related to the age-related macular degeneration (AMD), the diabetic macular edema (DME), as well as the retinal vein occlusion (RVO). However, in real-world settings, patients often suffer from loss of vision on time scales of years despite therapy, whereas the prediction of the visual acuity (VA) and the earliest possible detection of deterioration under real-life conditions is challenging due to heterogeneous and incomplete data. In this contribution, we present a workflow for the development of a research-compatible data corpus fusing different IT systems of the department of ophthalmology of a German maximum care hospital. The extensive data corpus allows predictive statements of the expected progression of a patient and his or her VA in each of the three diseases. For the disease AMD, we found out a significant deterioration of the visual acuity over time. Within our proposed multistage system, we subsequently classify the VA progression into the three groups of therapy "winners", "stabilizers", and "losers" (WSL classification scheme). Our OCT biomarker classification using an ensemble of deep neural networks results in a classification accuracy (F1-score) of over 98 %, enabling us to complete incomplete OCT documentations while allowing us to exploit them for a more precise VA modeling process. Our VA prediction requires at least four VA examinations and optionally OCT biomarkers from the same time period to predict the VA progression within a forecasted time frame, whereas our prediction is currently restricted to IVOM / no therapy. We achieve a final prediction accuracy of 69 % in macro average F1-score, while being in the same range as the ophthalmologists with 57.8 and 50 +- 10.7 % F1-score.
Authors: Lele Cao, Vilhelm von Ehrenheim, Sebastian Krakowski, Xiaoxue Li, Alexandra Lutz
Abstract: Startups often represent newly established business models associated with disruptive innovation and high scalability. They are commonly regarded as powerful engines for economic and social development. Meanwhile, startups are heavily constrained by many factors such as limited financial funding and human resources. Therefore, the chance for a startup to eventually succeed is as rare as "spotting a unicorn in the wild". Venture Capital (VC) strives to identify and invest in unicorn startups during their early stages, hoping to gain a high return. To avoid entirely relying on human domain expertise and intuition, investors usually employ data-driven approaches to forecast the success probability of startups. Over the past two decades, the industry has gone through a paradigm shift moving from conventional statistical approaches towards becoming machine-learning (ML) based. Notably, the rapid growth of data volume and variety is quickly ushering in deep learning (DL), a subset of ML, as a potentially superior approach in terms of capacity and expressivity. In this work, we carry out a literature review and synthesis on DL-based approaches, covering the entire DL life cycle. The objective is a) to obtain a thorough and in-depth understanding of the methodologies for startup evaluation using DL, and b) to distil valuable and actionable learning for practitioners. To the best of our knowledge, our work is the first of this kind.
Authors: Yang Zhang, Zhewei Wei, Ye Yuan, Chongxuan Li, Wenbing Huang
Abstract: Predicting the binding sites of target proteins plays a fundamental role in drug discovery. Most existing deep-learning methods consider a protein as a 3D image by spatially clustering its atoms into voxels and then feed the voxelized protein into a 3D CNN for prediction. However, the CNN-based methods encounter several critical issues: 1) defective in representing irregular protein structures; 2) sensitive to rotations; 3) insufficient to characterize the protein surface; 4) unaware of protein size shift. To address the above issues, this work proposes EquiPocket, an E(3)-equivariant Graph Neural Network (GNN) for binding site prediction, which comprises three modules: the first one to extract local geometric information for each surface atom, the second one to model both the chemical and spatial structure of protein and the last one to capture the geometry of the surface via equivariant message passing over the surface atoms. We further propose a dense attention output layer to alleviate the effect incurred by variable protein size. Extensive experiments on several representative benchmarks demonstrate the superiority of our framework to the state-of-the-art methods.
Authors: Francisco Plana, Andr\'es Abeliuk, Jorge P\'erez
Abstract: We present a simple and quick method to approximate network centrality indexes. Our approach, called QuickCent, is inspired by so-called fast and frugal heuristics, which are heuristics initially proposed to model some human decision and inference processes. The centrality index that we estimate is the harmonic centrality, which is a measure based on shortest-path distances, so infeasible to compute on large networks. We compare QuickCent with known machine learning algorithms on synthetic data generated with preferential attachment, and some empirical networks. Our experiments show that QuickCent is able to make estimates that are competitive in accuracy with the best alternative methods tested, either on synthetic scale-free networks or empirical networks. QuickCent has the feature of achieving low error variance estimates, even with a small training set. Moreover, QuickCent is comparable in efficiency -- accuracy and time cost -- to those produced by more complex methods. We discuss and provide some insight into how QuickCent exploits the fact that in some networks, such as those generated by preferential attachment, local density measures such as the in-degree, can be a proxy for the size of the network region to which a node has access, opening up the possibility of approximating centrality indices based on size such as the harmonic centrality. Our initial results show that simple heuristics and biologically inspired computational methods are a promising line of research in the context of network measure estimations.
Authors: Samuele Poppi, Sara Sarto, Marcella Cornia, Lorenzo Baraldi, Rita Cucchiara
Abstract: Machine Unlearning is an emerging paradigm for selectively removing the impact of training datapoints from a network. Unlike existing methods that target a limited subset or a single class, our framework unlearns all classes in a single round. We achieve this by modulating the network's components using memory matrices, enabling the network to demonstrate selective unlearning behavior for any class after training. By discovering weights that are specific to each class, our approach also recovers a representation of the classes which is explainable by design. We test the proposed framework on small- and medium-scale image classification datasets, with both convolution- and Transformer-based backbones, showcasing the potential for explainable solutions through unlearning.
Authors: Nikhil Singh, Chih-Wei Wu, Iroro Orife, Mahdi Kalayeh
Abstract: Audiovisual representation learning typically relies on the correspondence between sight and sound. However, there are often multiple audio tracks that can correspond with a visual scene. Consider, for example, different conversations on the same crowded street. The effect of such counterfactual pairs on audiovisual representation learning has not been previously explored. To investigate this, we use dubbed versions of movies and television shows to augment cross-modal contrastive learning. Our approach learns to represent alternate audio tracks, differing only in speech, similarly to the same video. Our results, from a comprehensive set of experiments investigating different training strategies, show this general approach improves performance on a range of downstream auditory and audiovisual tasks, without majorly affecting linguistic task performance overall. These findings highlight the importance of considering speech variation when learning scene-level audiovisual correspondences and suggest that dubbed audio can be a useful augmentation technique for training audiovisual models toward more robust performance on diverse downstream tasks.
Authors: Felix Benning, Leif D\"oring
Abstract: Classical worst-case optimization theory neither explains the success of optimization in machine learning, nor does it help with step size selection. We establish a connection between Bayesian Optimization (i.e. average case optimization theory) and classical optimization using a 'stochastic Taylor approximation' to rediscover gradient descent. This rediscovery yields a step size schedule we call Random Function Descent (RFD), which, in contrast to classical derivations, is scale invariant. Furthermore, our analysis of RFD step sizes yields a theoretical foundation for common step size heuristics such as gradient clipping and gradual learning rate warmup. We finally propose a statistical procedure for estimating the RFD step size schedule and validate this theory with a case study on the MNIST dataset.
Authors: Wan Tong Lou, Halvard Sutterud, Gino Cassella, W. M. C. Foulkes, Johannes Knolle, David Pfau, James S. Spencer
Abstract: Understanding superfluidity remains a major goal of condensed matter physics. Here we tackle this challenge utilizing the recently developed Fermionic neural network (FermiNet) wave function Ansatz [D. Pfau et al., Phys. Rev. Res. 2, 033429 (2020).] for variational Monte Carlo calculations. We study the unitary Fermi gas, a system with strong, short-range, two-body interactions known to possess a superfluid ground state but difficult to describe quantitatively. We demonstrate key limitations of the FermiNet Ansatz in studying the unitary Fermi gas and propose a simple modification based on the idea of an antisymmetric geminal power singlet (AGPs) wave function. The new AGPs FermiNet outperforms the original FermiNet significantly in paired systems, giving results which are more accurate than fixed-node diffusion Monte Carlo and are consistent with experiment. We prove mathematically that the new Ansatz, which only differs from the original Ansatz by the method of antisymmetrization, is a strict generalization of the original FermiNet architecture, despite the use of fewer parameters. Our approach shares several advantages with the original FermiNet: the use of a neural network removes the need for an underlying basis set; and the flexibility of the network yields extremely accurate results within a variational quantum Monte Carlo framework that provides access to unbiased estimates of arbitrary ground-state expectation values. We discuss how the method can be extended to study other superfluids.
Authors: Amir Ali Ahmadi, Abraar Chaudhry, Vikas Sindhwani, Stephen Tu
Abstract: A fundamental challenge in learning an unknown dynamical system is to reduce model uncertainty by making measurements while maintaining safety. We formulate a mathematical definition of what it means to safely learn a dynamical system by sequentially deciding where to initialize trajectories. The state of the system must stay within a safety region for a horizon of $T$ time steps under the action of all dynamical systems that (i) belong to a given initial uncertainty set, and (ii) are consistent with information gathered so far. First, we consider safely learning a linear dynamical system involving $n$ states. For the case $T=1$, we present an LP-based algorithm that either safely recovers the true dynamics from at most $n$ trajectories, or certifies that safe learning is impossible. For $T=2$, we give an SDP representation of the set of safe initial conditions and show that $\lceil n/2 \rceil$ trajectories generically suffice for safe learning. For $T = \infty$, we provide SDP-representable inner approximations of the set of safe initial conditions and show that one trajectory generically suffices for safe learning. We extend a number of our results to the cases where the initial uncertainty set contains sparse, low-rank, or permutation matrices, or when the system has a control input. Second, we consider safely learning a general class of nonlinear dynamical systems. For the case $T=1$, we give an SOCP-based representation of the set of safe initial conditions. For $T=\infty$, we provide semidefinite representable inner approximations to the set of safe initial conditions. We show how one can safely collect trajectories and fit a polynomial model of the nonlinear dynamics that is consistent with the initial uncertainty set and best agrees with the observations. We also present some extensions to cases where the measurements are noisy or the dynamical system involves disturbances.
Authors: Tarek Naous, Michael J. Ryan, Anton Lavrouk, Mohit Chandra, Wei Xu
Abstract: We present a comprehensive evaluation of large language models for multilingual readability assessment. Existing evaluation resources lack domain and language diversity, limiting the ability for cross-domain and cross-lingual analyses. This paper introduces ReadMe++, a multilingual multi-domain dataset with human annotations of 9757 sentences in Arabic, English, French, Hindi, and Russian, collected from 112 different data sources. This benchmark will encourage research on developing robust multilingual readability assessment methods. Using ReadMe++, we benchmark multilingual and monolingual language models in the supervised, unsupervised, and few-shot prompting settings. The domain and language diversity in ReadMe++ enable us to test more effective few-shot prompting, and identify shortcomings in state-of-the-art unsupervised methods. Our experiments also reveal exciting results of superior domain generalization and enhanced cross-lingual transfer capabilities by models trained on ReadMe++. We will make our data publicly available and release a python package tool for multilingual sentence readability prediction using our trained models at: https://github.com/tareknaous/readme
Authors: Kenshi Abe, Kaito Ariu, Mitsuki Sakamoto, Atsushi Iwasaki
Abstract: This paper proposes a payoff perturbation technique for the Mirror Descent (MD) algorithm in games where the gradient of the payoff functions is monotone in the strategy profile space, potentially containing additive noise. The optimistic family of learning algorithms, exemplified by optimistic MD, successfully achieves {\it last-iterate} convergence in scenarios devoid of noise, leading the dynamics to a Nash equilibrium. A recent re-emerging trend underscores the promise of the perturbation approach, where payoff functions are perturbed based on the distance from an anchoring, or {\it slingshot}, strategy. In response, we propose {\it Adaptively Perturbed MD} (APMD), which adjusts the magnitude of the perturbation by repeatedly updating the slingshot strategy at a predefined interval. This innovation empowers us to find a Nash equilibrium of the underlying game with guaranteed rates. Empirical demonstrations affirm that our algorithm exhibits significantly accelerated convergence.
Authors: Mohammed Rayyan Sheriff, Peyman Mohajerin Esfahani
Abstract: This article focuses on a class of distributionally robust optimization (DRO) problems where, unlike the growing body of the literature, the objective function is potentially nonlinear in the distribution. Existing methods to optimize nonlinear functions in probability space use the Frechet derivatives, which present both theoretical and computational challenges. Motivated by this, we propose an alternative notion for the derivative and corresponding smoothness based on Gateaux (G)-derivative for generic risk measures. These concepts are explained via three running risk measure examples of variance, entropic risk, and risk on finite support sets. We then propose a G-derivative based Frank-Wolfe (FW) algorithm for generic nonlinear optimization problems in probability spaces and establish its convergence under the proposed notion of smoothness in a completely norm-independent manner. We use the set-up of the FW algorithm to devise a methodology to compute a saddle point of the nonlinear DRO problem. Finally, we validate our theoretical results on two cases of the entropic and variance risk measures in the context of portfolio selection problems. In particular, we analyze their regularity conditions and "sufficient statistic", compute the respective FW-oracle in various settings, and confirm the theoretical outcomes through numerical validation.
Authors: Milad Sefidgaran, Romain Chor, Abdellatif Zaidi, Yijun Wan
Abstract: We investigate the generalization error of statistical learning models in a Federated Learning (FL) setting. Specifically, we study the evolution of the generalization error with the number of communication rounds $R$ between $K$ clients and a parameter server (PS), i.e., the effect on the generalization error of how often the clients' local models are aggregated at PS. In our setup, the more the clients communicate with PS the less data they use for local training in each round, such that the amount of training data per client is identical for distinct values of $R$. We establish PAC-Bayes and rate-distortion theoretic bounds on the generalization error that account explicitly for the effect of the number of rounds $R$, in addition to the number of participating devices $K$ and individual datasets size $n$. The bounds, which apply to a large class of loss functions and learning algorithms, appear to be the first of their kind for the FL setting. Furthermore, we apply our bounds to FL-type Support Vector Machines (FSVM); and derive (more) explicit bounds in this case. In particular, we show that the generalization bound of FSVM increases with $R$, suggesting that more frequent communication with PS diminishes the generalization power. This implies that the population risk decreases less fast with $R$ than does the empirical risk. Moreover, our bound suggests that the generalization error of FSVM decreases faster than that of centralized learning by a factor of $\mathcal{O}(\sqrt{\log(K)/K})$. Finally, we provide experimental results obtained using neural networks (ResNet-56) which show evidence that not only may our observations for FSVM hold more generally but also that the population risk may even start to increase beyond some value of $R$.
Authors: S. Rasoulzadeh, M. Wimmer, P. Stauss, I. Kovacic
Abstract: We present Strokes2Surface, an offline geometry reconstruction pipeline that recovers well-connected curve networks from imprecise 4D sketches to bridge concept design and digital modeling stages in architectural design. The input to our pipeline consists of 3D strokes' polyline vertices and their timestamps as the 4th dimension, along with additional metadata recorded throughout sketching. Inspired by architectural sketching practices, our pipeline combines a classifier and two clustering models to achieve its goal. First, with a set of extracted hand-engineered features from the sketch, the classifier recognizes the type of individual strokes between those depicting boundaries (Shape strokes) and those depicting enclosed areas (Scribble strokes). Next, the two clustering models parse strokes of each type into distinct groups, each representing an individual edge or face of the intended architectural object. Curve networks are then formed through topology recovery of consolidated Shape clusters and surfaced using Scribble clusters guiding the cycle discovery. Our evaluation is threefold: We confirm the usability of the Strokes2Surface pipeline in architectural design use cases via a user study, we validate our choice of features via statistical analysis and ablation studies on our collected dataset, and we compare our outputs against a range of reconstructions computed using alternative methods.
Authors: Lele Cao, Vilhelm von Ehrenheim, Mark Granroth-Wilding, Richard Anselmo Stahl, Andrew McCornack, Armin Catovic, Dhiana Deva Cavacanti Rocha
Abstract: In the investment industry, it is often essential to carry out fine-grained company similarity quantification for a range of purposes, including market mapping, competitor analysis, and mergers and acquisitions. We propose and publish a knowledge graph, named CompanyKG, to represent and learn diverse company features and relations. Specifically, 1.17 million companies are represented as nodes enriched with company description embeddings; and 15 different inter-company relations result in 51.06 million weighted edges. To enable a comprehensive assessment of methods for company similarity quantification, we have devised and compiled three evaluation tasks with annotated test sets: similarity prediction, competitor retrieval and similarity ranking. We present extensive benchmarking results for 11 reproducible predictive methods categorized into three groups: node-only, edge-only, and node+edge. To the best of our knowledge, CompanyKG is the first large-scale heterogeneous graph dataset originating from a real-world investment platform, tailored for quantifying inter-company similarity.
Authors: Yaswitha Gujju, Atsushi Matsuo, Rudy Raymond
Abstract: The past decade has witnessed significant advancements in quantum hardware, encompassing improvements in speed, qubit quantity, and quantum volume-a metric defining the maximum size of a quantum circuit effectively implementable on near-term quantum devices. This progress has led to a surge in Quantum Machine Learning (QML) applications on real hardware, aiming to achieve quantum advantage over classical approaches. This survey focuses on selected supervised and unsupervised learning applications executed on quantum hardware, specifically tailored for real-world scenarios. The exploration includes a thorough analysis of current QML implementation limitations on quantum hardware, covering techniques like encoding, ansatz structure, error mitigation, and gradient methods to address these challenges. Furthermore, the survey evaluates the performance of QML implementations in comparison to classical counterparts. In conclusion, we discuss existing bottlenecks related to applying QML on real quantum devices and propose potential solutions to overcome these challenges in the future.
Authors: Aryo Pradipta Gema, Pasquale Minervini, Luke Daines, Tom Hope, Beatrice Alex
Abstract: Adapting pretrained language models to novel domains, such as clinical applications, traditionally involves retraining their entire set of parameters. Parameter-Efficient Fine-Tuning (PEFT) techniques for fine-tuning language models significantly reduce computational requirements by selectively fine-tuning small subsets of parameters. In this study, we propose a two-step PEFT framework and evaluate it in the clinical domain. Our approach combines a specialised PEFT adapter layer designed for clinical domain adaptation with another adapter specialised for downstream tasks. We evaluate the framework on multiple clinical outcome prediction datasets, comparing it to clinically trained language models. Our framework achieves a better AUROC score averaged across all clinical downstream tasks compared to clinical language models. In particular, we observe large improvements of 4-5% AUROC in large-scale multilabel classification tasks, such as diagnoses and procedures classification. To our knowledge, this study is the first to provide an extensive empirical analysis of the interplay between PEFT techniques and domain adaptation in an important real-world domain of clinical applications.
Authors: Sabine Muzellec, Thomas Fel, Victor Boutin, L\'eo and\'eol, Rufin VanRullen, Thomas Serre
Abstract: Attribution methods correspond to a class of explainability methods (XAI) that aim to assess how individual inputs contribute to a model's decision-making process. We have identified a significant limitation in one type of attribution methods, known as ``white-box" methods. Although highly efficient, as we will show, these methods rely on a gradient signal that is often contaminated by high-frequency artifacts. To overcome this limitation, we introduce a new approach called "FORGrad". This simple method effectively filters out these high-frequency artifacts using optimal cut-off frequencies tailored to the unique characteristics of each model architecture. Our findings show that FORGrad consistently enhances the performance of already existing white-box methods, enabling them to compete effectively with more accurate yet computationally demanding "black-box" methods. We anticipate that our research will foster broader adoption of simpler and more efficient white-box methods for explainability, offering a better balance between faithfulness and computational efficiency.
Authors: Yanis Labrak, Mickael Rouvier, Richard Dufour
Abstract: We evaluate four state-of-the-art instruction-tuned large language models (LLMs) -- ChatGPT, Flan-T5 UL2, Tk-Instruct, and Alpaca -- on a set of 13 real-world clinical and biomedical natural language processing (NLP) tasks in English, such as named-entity recognition (NER), question-answering (QA), relation extraction (RE), etc. Our overall results demonstrate that the evaluated LLMs begin to approach performance of state-of-the-art models in zero- and few-shot scenarios for most tasks, and particularly well for the QA task, even though they have never seen examples from these tasks before. However, we observed that the classification and RE tasks perform below what can be achieved with a specifically trained model for the medical field, such as PubMedBERT. Finally, we noted that no LLM outperforms all the others on all the studied tasks, with some models being better suited for certain tasks than others.
Authors: Tae Yeob Kang, Haebom Lee, Sungho Suh
Abstract: In the field of optoelectronics, indium tin oxide (ITO) electrodes play a crucial role in various applications, such as displays, sensors, and solar cells. Effective fault diagnosis and root cause analysis of the ITO electrodes are essential to ensure the performance and reliability of the devices. However, traditional visual inspection is challenging with transparent ITO electrodes, and existing fault diagnosis methods have limitations in determining the root causes of the defects, often requiring destructive evaluations and secondary material characterization techniques. In this study, a fault diagnosis method with root cause analysis is proposed using scattering parameter (S-parameter) patterns, offering early detection, high diagnostic accuracy, and noise robustness. A comprehensive S-parameter pattern database is obtained according to various defect states of the ITO electrodes. Deep learning (DL) approaches, including multilayer perceptron (MLP), convolutional neural network (CNN), and transformer, are then used to simultaneously analyze the cause and severity of defects. Notably, it is demonstrated that the diagnostic performance under additive noise levels can be significantly enhanced by combining different channels of the S-parameters as input to the learning algorithms, as confirmed through the t-distributed stochastic neighbor embedding (t-SNE) dimension reduction visualization of the S-parameter patterns.
Authors: Hongzhi Qi, Qing Zhao, Jianqiang Li, Changwei Song, Wei Zhai, Dan Luo, Shuo Liu, Yi Jing Yu, Fan Wang, Huijing Zou, Bing Xiang Yang, Guanghui Fu
Abstract: On social media, users often express their personal feelings, which may exhibit cognitive distortions or even suicidal tendencies on certain specific topics. Early recognition of these signs is critical for effective psychological intervention. In this paper, we introduce two novel datasets from Chinese social media: SOS-HL-1K for suicidal risk classification and SocialCD-3K for cognitive distortions detection. The SOS-HL-1K dataset contained 1,249 posts and SocialCD-3K dataset was a multi-label classification dataset that containing 3,407 posts. We propose a comprehensive evaluation using two supervised learning methods and eight large language models (LLMs) on the proposed datasets. From the prompt engineering perspective, we experimented with two types of prompt strategies, including four zero-shot and five few-shot strategies. We also evaluated the performance of the LLMs after fine-tuning on the proposed tasks. The experimental results show that there is still a huge gap between LLMs relying only on prompt engineering and supervised learning. In the suicide classification task, this gap is 6.95% points in F1-score, while in the cognitive distortion task, the gap is even more pronounced, reaching 31.53% points in F1-score. However, after fine-tuning, this difference is significantly reduced. In the suicide and cognitive distortion classification tasks, the gap decreases to 4.31% and 3.14%, respectively. This research highlights the potential of LLMs in psychological contexts, but supervised learning remains necessary for more challenging tasks. All datasets and code are made available.
Authors: Muah Kim, Rick Fritschek, Rafael F. Schaefer
Abstract: Strong generative models can accurately learn channel distributions. This could save recurring costs for physical measurements of the channel. Moreover, the resulting differentiable channel model supports training neural encoders by enabling gradient-based optimization. The initial approach in the literature draws upon the modern advancements in image generation, utilizing generative adversarial networks (GANs) or their enhanced variants to generate channel distributions. In this paper, we address this channel approximation challenge with diffusion models (DMs), which have demonstrated high sample quality and mode coverage in image generation. In addition to testing the generative performance of the channel distributions, we use an end-to-end (E2E) coded-modulation framework underpinned by DMs and propose an efficient training algorithm. Our simulations with various channel models show that a DM can accurately learn channel distributions, enabling an E2E framework to achieve near-optimal symbol error rates (SERs). Furthermore, we examine the trade-off between mode coverage and sampling speed through skipped sampling using sliced Wasserstein distance (SWD) and the E2E SER. We investigate the effect of noise scheduling on this trade-off, demonstrating that with an appropriate choice of parameters and techniques, sampling time can be significantly reduced with a minor increase in SWD and SER. Finally, we show that the DM can generate a correlated fading channel, whereas a strong GAN variant fails to learn the covariance. This paper highlights the potential benefits of using DMs for learning channel distributions, which could be further investigated for various channels and advanced techniques of DMs.
Authors: Zhi-Yi Chin, Chieh-Ming Jiang, Ching-Chun Huang, Pin-Yu Chen, Wei-Chen Chiu
Abstract: While image data starts to enjoy the simple-but-effective self-supervised learning scheme built upon masking and self-reconstruction objective thanks to the introduction of tokenization procedure and vision transformer backbone, convolutional neural networks as another important and widely-adopted architecture for image data, though having contrastive-learning techniques to drive the self-supervised learning, still face the difficulty of leveraging such straightforward and general masking operation to benefit their learning process significantly. In this work, we aim to alleviate the burden of including masking operation into the contrastive-learning framework for convolutional neural networks as an extra augmentation method. In addition to the additive but unwanted edges (between masked and unmasked regions) as well as other adverse effects caused by the masking operations for ConvNets, which have been discussed by prior works, we particularly identify the potential problem where for one view in a contrastive sample-pair the randomly-sampled masking regions could be overly concentrated on important/salient objects thus resulting in misleading contrastiveness to the other view. To this end, we propose to explicitly take the saliency constraint into consideration in which the masked regions are more evenly distributed among the foreground and background for realizing the masking-based augmentation. Moreover, we introduce hard negative samples by masking larger regions of salient patches in an input image. Extensive experiments conducted on various datasets, contrastive learning mechanisms, and downstream tasks well verify the efficacy as well as the superior performance of our proposed method with respect to several state-of-the-art baselines.
Authors: Yingying Li, Jing Yu, Lauren Conger, Taylan Kargin, Adam Wierman
Abstract: This paper studies uncertainty set estimation for unknown linear systems. Uncertainty sets are crucial for the quality of robust control since they directly influence the conservativeness of the control design. Departing from the confidence region analysis of least squares estimation, this paper focuses on set membership estimation (SME). Though good numerical performances have attracted applications of SME in the control literature, the non-asymptotic convergence rate of SME for linear systems remains an open question. This paper provides the first convergence rate bounds for SME and discusses variations of SME under relaxed assumptions. We also provide numerical results demonstrating SME's practical promise.
Authors: Sahar Abdelnabi, Amr Gomaa, Sarath Sivaprasad, Lea Sch\"onherr, Mario Fritz
Abstract: There is an growing interest in using Large Language Models (LLMs) in multi-agent systems to tackle interactive real-world tasks that require effective collaboration and assessing complex situations. Yet, we still have a limited understanding of LLMs' communication and decision-making abilities in multi-agent setups. The fundamental task of negotiation spans many key features of communication, such as cooperation, competition, and manipulation potentials. Thus, we propose using scorable negotiation to evaluate LLMs. We create a testbed of complex multi-agent, multi-issue, and semantically rich negotiation games. To reach an agreement, agents must have strong arithmetic, inference, exploration, and planning capabilities while integrating them in a dynamic and multi-turn setup. We propose multiple metrics to rigorously quantify agents' performance and alignment with the assigned role. We provide procedures to create new games and increase games' difficulty to have an evolving benchmark. Importantly, we evaluate critical safety aspects such as the interaction dynamics between agents influenced by greedy and adversarial players. Our benchmark is highly challenging; GPT-3.5 and small models mostly fail, and GPT-4 and SoTA large models (e.g., Llama-3 70b) still underperform.
Authors: Xihuai Wang, Shao Zhang, Wenhao Zhang, Wentao Dong, Jingxiao Chen, Ying Wen, Weinan Zhang
Abstract: Zero-shot coordination (ZSC) is a new cooperative multi-agent reinforcement learning (MARL) challenge that aims to train an ego agent to work with diverse, unseen partners during deployment. The significant difference between the deployment-time partners' distribution and the training partners' distribution determined by the training algorithm makes ZSC a unique out-of-distribution (OOD) generalization challenge. The potential distribution gap between evaluation and deployment-time partners leads to inadequate evaluation, which is exacerbated by the lack of appropriate evaluation metrics. In this paper, we present ZSC-Eval, the first evaluation toolkit and benchmark for ZSC algorithms. ZSC-Eval consists of: 1) Generation of evaluation partner candidates through behavior-preferring rewards to approximate deployment-time partners' distribution; 2) Selection of evaluation partners by Best-Response Diversity (BR-Div); 3) Measurement of generalization performance with various evaluation partners via the Best-Response Proximity (BR-Prox) metric. We use ZSC-Eval to benchmark ZSC algorithms in Overcooked and Google Research Football environments and get novel empirical findings. We also conduct a human experiment of current ZSC algorithms to verify the ZSC-Eval's consistency with human evaluation. ZSC-Eval is now available at https://github.com/sjtu-marl/ZSC-Eval.
Authors: Xiaoxuan Liu, Lanxiang Hu, Peter Bailis, Alvin Cheung, Zhijie Deng, Ion Stoica, Hao Zhang
Abstract: Speculative decoding is a pivotal technique to accelerate the inference of large language models (LLMs) by employing a smaller draft model to predict the target model's outputs. However, its efficacy can be limited due to the low predictive accuracy of the draft model, particularly when faced with diverse text inputs and a significant capability gap between the draft and target models. We introduce online speculative decoding to address this challenge. The main idea is to continuously update the (multiple) draft model(s) on observed user query data. Adapting to query distribution mitigates the shifts between the training distribution of the draft model and the query distribution, enabling the draft model to more accurately predict the target model's outputs. We develop a prototype of online speculative decoding based on knowledge distillation and evaluate it using both synthetic and real query data. The results show a substantial increase in the token acceptance rate by 0.1 to 0.65, bringing 1.42x to 2.17x latency reduction. Our code is available at https://github.com/LiuXiaoxuanPKU/OSD.
Authors: Andrea Bernini, Fabrizio Silvestri, Gabriele Tolomei
Abstract: Community detection techniques are useful for social media platforms to discover tightly connected groups of users who share common interests. However, this functionality often comes at the expense of potentially exposing individuals to privacy breaches by inadvertently revealing their tastes or preferences. Therefore, some users may wish to preserve their anonymity and opt out of community detection for various reasons, such as affiliation with political or religious organizations, without leaving the platform. In this study, we address the challenge of community membership hiding, which involves strategically altering the structural properties of a network graph to prevent one or more nodes from being identified by a given community detection algorithm. We tackle this problem by formulating it as a constrained counterfactual graph objective, and we solve it via deep reinforcement learning. Extensive experiments demonstrate that our method outperforms existing baselines, striking the best balance between accuracy and cost.
Authors: Daniil Tiapkin, Denis Belomestny, Daniele Calandriello, Eric Moulines, Alexey Naumov, Pierre Perrault, Michal Valko, Pierre Menard
Abstract: Incorporating expert demonstrations has empirically helped to improve the sample efficiency of reinforcement learning (RL). This paper quantifies theoretically to what extent this extra information reduces RL's sample complexity. In particular, we study the demonstration-regularized reinforcement learning that leverages the expert demonstrations by KL-regularization for a policy learned by behavior cloning. Our findings reveal that using $N^{\mathrm{E}}$ expert demonstrations enables the identification of an optimal policy at a sample complexity of order $\widetilde{O}(\mathrm{Poly}(S,A,H)/(\varepsilon^2 N^{\mathrm{E}}))$ in finite and $\widetilde{O}(\mathrm{Poly}(d,H)/(\varepsilon^2 N^{\mathrm{E}}))$ in linear Markov decision processes, where $\varepsilon$ is the target precision, $H$ the horizon, $A$ the number of action, $S$ the number of states in the finite case and $d$ the dimension of the feature space in the linear case. As a by-product, we provide tight convergence guarantees for the behaviour cloning procedure under general assumptions on the policy classes. Additionally, we establish that demonstration-regularized methods are provably efficient for reinforcement learning from human feedback (RLHF). In this respect, we provide theoretical evidence showing the benefits of KL-regularization for RLHF in tabular and linear MDPs. Interestingly, we avoid pessimism injection by employing computationally feasible regularization to handle reward estimation uncertainty, thus setting our approach apart from the prior works.
Authors: Xiaolong Tu, Anik Mallik, Dawei Chen, Kyungtae Han, Onur Altintas, Haoxin Wang, Jiang Xie
Abstract: Today, deep learning optimization is primarily driven by research focused on achieving high inference accuracy and reducing latency. However, the energy efficiency aspect is often overlooked, possibly due to a lack of sustainability mindset in the field and the absence of a holistic energy dataset. In this paper, we conduct a threefold study, including energy measurement, prediction, and efficiency scoring, with an objective to foster transparency in power and energy consumption within deep learning across various edge devices. Firstly, we present a detailed, first-of-its-kind measurement study that uncovers the energy consumption characteristics of on-device deep learning. This study results in the creation of three extensive energy datasets for edge devices, covering a wide range of kernels, state-of-the-art DNN models, and popular AI applications. Secondly, we design and implement the first kernel-level energy predictors for edge devices based on our kernel-level energy dataset. Evaluation results demonstrate the ability of our predictors to provide consistent and accurate energy estimations on unseen DNN models. Lastly, we introduce two scoring metrics, PCS and IECS, developed to convert complex power and energy consumption data of an edge device into an easily understandable manner for edge device end-users. We hope our work can help shift the mindset of both end-users and the research community towards sustainability in edge computing, a principle that drives our research. Find data, code, and more up-to-date information at https://amai-gsu.github.io/DeepEn2023.
Authors: Guy Ohayon, Tomer Michaeli, Michael Elad
Abstract: We study the behavior of deterministic methods for solving inverse problems in imaging. These methods are commonly designed to achieve two goals: (1) attaining high perceptual quality, and (2) generating reconstructions that are consistent with the measurements. We provide a rigorous proof that the better a predictor satisfies these two requirements, the larger its Lipschitz constant must be, regardless of the nature of the degradation involved. In particular, to approach perfect perceptual quality and perfect consistency, the Lipschitz constant of the model must grow to infinity. This implies that such methods are necessarily more susceptible to adversarial attacks. We demonstrate our theory on single image super-resolution algorithms, addressing both noisy and noiseless settings. We also show how this undesired behavior can be leveraged to explore the posterior distribution, thereby allowing the deterministic model to imitate stochastic methods.
Authors: Nikolaos Ioannis Bountos, Maria Sdraka, Angelos Zavras, Ilektra Karasante, Andreas Karavias, Themistocles Herekakis, Angeliki Thanasou, Dimitrios Michail, Ioannis Papoutsis
Abstract: Global floods, exacerbated by climate change, pose severe threats to human life, infrastructure, and the environment. Recent catastrophic events in Pakistan and New Zealand underscore the urgent need for precise flood mapping to guide restoration efforts, understand vulnerabilities, and prepare for future occurrences. While Synthetic Aperture Radar (SAR) remote sensing offers day-and-night, all-weather imaging capabilities, its application in deep learning for flood segmentation is limited by the lack of large annotated datasets. To address this, we introduce Kuro Siwo, a manually annotated multi-temporal dataset, spanning 43 flood events globally. Our dataset maps more than 338 billion $m^2$ of land, with 33 billion designated as either flooded areas or permanent water bodies. Kuro Siwo includes a highly processed product optimized for flood mapping based on SAR Ground Range Detected, and a primal SAR Single Look Complex product with minimal preprocessing, designed to promote research on the exploitation of both the phase and amplitude information and to offer maximum flexibility for downstream task preprocessing. To leverage advances in large scale self-supervised pretraining methods for remote sensing data, we augment Kuro Siwo with a large unlabeled set of SAR samples. Finally, we provide an extensive benchmark, namely BlackBench, offering strong baselines for a diverse set of flood events from Europe, America, Africa, Asia and Australia.
Authors: Niv Cohen, Issar Tzachor, Yedid Hoshen
Abstract: This paper proposes to use set features for detecting anomalies in samples that consist of unusual combinations of normal elements. Many leading methods discover anomalies by detecting an unusual part of a sample. For example, state-of-the-art segmentation-based approaches, first classify each element of the sample (e.g., image patch) as normal or anomalous and then classify the entire sample as anomalous if it contains anomalous elements. However, such approaches do not extend well to scenarios where the anomalies are expressed by an unusual combination of normal elements. In this paper, we overcome this limitation by proposing set features that model each sample by the distribution of its elements. We compute the anomaly score of each sample using a simple density estimation method, using fixed features. Our approach outperforms the previous state-of-the-art in image-level logical anomaly detection and sequence-level time series anomaly detection.
Authors: Yu Yang, Qihong Yang, Yangtao Deng, Qiaolin He
Abstract: In this work, we propose an end-to-end adaptive sampling neural network (MMPDE-Net) based on the moving mesh method, which can adaptively generate new sampling points by solving the moving mesh PDE. This model focuses on improving the quality of sampling points generation. Moreover, we develop an iterative algorithm based on MMPDE-Net, which makes the sampling points more precise and controllable. Since MMPDE-Net is a framework independent of the deep learning solver, we combine it with physics-informed neural networks (PINN) to propose moving sampling PINN (MS-PINN) and demonstrate its effectiveness by error analysis under some assumptions. Finally, we demonstrate the performance improvement of MS-PINN compared to PINN through numerical experiments of four typical examples, which numerically verify the effectiveness of our method.
Authors: Weikang Qiu, Huangrui Chu, Selena Wang, Haolan Zuo, Xiaoxiao Li, Yize Zhao, Rex Ying
Abstract: Discovering reliable and informative relationships among brain regions from functional magnetic resonance imaging (fMRI) signals is essential in phenotypic predictions. Most of the current methods fail to accurately characterize those interactions because they only focus on pairwise connections and overlook the high-order relationships of brain regions. We propose that these high-order relationships should be maximally informative and minimally redundant (MIMR). However, identifying such high-order relationships is challenging and under-explored due to the exponential search space and the absence of a tractable objective. In response to this gap, we propose a novel method named HYBRID which aims to extract MIMR high-order relationships from fMRI data. HYBRID employs a CONSTRUCTOR to identify hyperedge structures, and a WEIGHTER to compute a weight for each hyperedge, which avoids searching in exponential space. HYBRID achieves the MIMR objective through an innovative information bottleneck framework named multi-head drop-bottleneck with theoretical guarantees. Our comprehensive experiments demonstrate the effectiveness of our model. Our model outperforms the state-of-the-art predictive model by an average of 11.2%, regarding the quality of hyperedges measured by CPM, a standard protocol for studying brain connections.
Authors: Yiheng Jiang, Sinho Chewi, Aram-Alexandre Pooladian
Abstract: We develop a theory of finite-dimensional polyhedral subsets over the Wasserstein space and optimization of functionals over them via first-order methods. Our main application is to the problem of mean-field variational inference, which seeks to approximate a distribution $\pi$ over $\mathbb{R}^d$ by a product measure $\pi^\star$. When $\pi$ is strongly log-concave and log-smooth, we provide (1) approximation rates certifying that $\pi^\star$ is close to the minimizer $\pi^\star_\diamond$ of the KL divergence over a \emph{polyhedral} set $\mathcal{P}_\diamond$, and (2) an algorithm for minimizing $\text{KL}(\cdot\|\pi)$ over $\mathcal{P}_\diamond$ with accelerated complexity $O(\sqrt \kappa \log(\kappa d/\varepsilon^2))$, where $\kappa$ is the condition number of $\pi$.
Authors: Paul Maria Scheikl, Nicolas Schreiber, Christoph Haas, Niklas Freymuth, Gerhard Neumann, Rudolf Lioutikov, Franziska Mathis-Ullrich
Abstract: Policy learning in robot-assisted surgery (RAS) lacks data efficient and versatile methods that exhibit the desired motion quality for delicate surgical interventions. To this end, we introduce Movement Primitive Diffusion (MPD), a novel method for imitation learning (IL) in RAS that focuses on gentle manipulation of deformable objects. The approach combines the versatility of diffusion-based imitation learning (DIL) with the high-quality motion generation capabilities of Probabilistic Dynamic Movement Primitives (ProDMPs). This combination enables MPD to achieve gentle manipulation of deformable objects, while maintaining data efficiency critical for RAS applications where demonstration data is scarce. We evaluate MPD across various simulated and real world robotic tasks on both state and image observations. MPD outperforms state-of-the-art DIL methods in success rate, motion quality, and data efficiency. Project page: https://scheiklp.github.io/movement-primitive-diffusion/
URLs: https://scheiklp.github.io/movement-primitive-diffusion/
Authors: Zeping Yu, Sophia Ananiadou
Abstract: Identifying important neurons for final predictions is essential for understanding the mechanisms of large language models. Due to computational constraints, current attribution techniques struggle to operate at neuron level. In this paper, we propose a static method for pinpointing significant neurons for different outputs. Compared to seven other methods, our approach demonstrates superior performance across three metrics. Additionally, since most static methods typically only identify "value neurons" directly contributing to the final prediction, we introduce a static method for identifying "query neurons" which activate these "value neurons". Finally, we apply our methods to analyze the localization of six distinct types of knowledge across both attention and feed-forward network (FFN) layers. Our method and analysis are helpful for understanding the mechanisms of knowledge storage and set the stage for future research in knowledge editing. We will release our data and code on github.
Authors: Yunye Gong, Robik Shrestha, Jared Claypoole, Michael Cogswell, Arijit Ray, Christopher Kanan, Ajay Divakaran
Abstract: We propose a novel VQA dataset, BloomVQA, to facilitate comprehensive evaluation of large vision-language models on comprehension tasks. Unlike current benchmarks that often focus on fact-based memorization and simple reasoning tasks without theoretical grounding, we collect multiple-choice samples based on picture stories that reflect different levels of comprehension, as laid out in Bloom's Taxonomy, a classic framework for learning assessment widely adopted in education research. Our data maps to a novel hierarchical graph representation which enables automatic data augmentation and novel measures characterizing model consistency. We perform graded evaluation and reliability analysis on recent multi-modal models. In comparison to low-level tasks, we observe decreased performance on tasks requiring advanced comprehension and cognitive skills with up to 38.0\% drop in VQA accuracy. In comparison to earlier models, GPT-4V demonstrates improved accuracy over all comprehension levels and shows a tendency of bypassing visual inputs especially for higher-level tasks. Current models also show consistency patterns misaligned with human comprehension in various scenarios, demonstrating the need for improvement based on theoretically-grounded criteria.
Authors: Cl\'ement Pierquin, Aur\'elien Bellet, Marc Tommasi, Matthieu Boussard
Abstract: Pufferfish privacy is a flexible generalization of differential privacy that allows to model arbitrary secrets and adversary's prior knowledge about the data. Unfortunately, designing general and tractable Pufferfish mechanisms that do not compromise utility is challenging. Furthermore, this framework does not provide the composition guarantees needed for a direct use in iterative machine learning algorithms. To mitigate these issues, we introduce a R\'enyi divergence-based variant of Pufferfish and show that it allows us to extend the applicability of the Pufferfish framework. We first generalize the Wasserstein mechanism to cover a wide range of noise distributions and introduce several ways to improve its utility. We also derive stronger guarantees against out-of-distribution adversaries. Finally, as an alternative to composition, we prove privacy amplification results for contractive noisy iterations and showcase the first use of Pufferfish in private convex optimization. A common ingredient underlying our results is the use and extension of shift reduction lemmas.
Authors: Lokesh Veeramacheneni (University of Bonn), Moritz Wolter (University of Bonn), Hildegard Kuehne (University of Bonn), Juergen Gall (University of Bonn)
Abstract: Modern metrics for generative learning like Fr\'echet Inception Distance (FID) demonstrate impressive performance. However, they suffer from various shortcomings, like a bias towards specific generators and datasets. To address this problem, we propose the Fr\'echet Wavelet Distance (FWD) as a domain-agnostic metric based on Wavelet Packet Transform ($W_p$). FWD provides a sight across a broad spectrum of frequencies in images with a high resolution, along with preserving both spatial and textural aspects. Specifically, we use Wp to project generated and dataset images to packet coefficient space. Further, we compute Fr\'echet distance with the resultant coefficients to evaluate the quality of a generator. This metric is general-purpose and dataset-domain agnostic, as it does not rely on any pre-trained network while being more interpretable because of frequency band transparency. We conclude with an extensive evaluation of a wide variety of generators across various datasets that the proposed FWD is able to generalize and improve robustness to domain shift and various corruptions compared to other metrics.
Authors: Sky CH-Wang, Benjamin Van Durme, Jason Eisner, Chris Kedzie
Abstract: We design probes trained on the internal representations of a transformer language model to predict its hallucinatory behavior on three grounded generation tasks. To train the probes, we annotate for span-level hallucination on both sampled (organic) and manually edited (synthetic) reference outputs. Our probes are narrowly trained and we find that they are sensitive to their training domain: they generalize poorly from one task to another or from synthetic to organic hallucinations. However, on in-domain data, they can reliably detect hallucinations at many transformer layers, achieving 95% of their peak performance as early as layer 4. Here, probing proves accurate for evaluating hallucination, outperforming several contemporary baselines and even surpassing an expert human annotator in response-level detection F1. Similarly, on span-level labeling, probes are on par or better than the expert annotator on two out of three generation tasks. Overall, we find that probing is a feasible and efficient alternative to language model hallucination evaluation when model states are available.
Authors: Muhammad Wasim Nawaz, Muhammad Ahmad Tahir, Ahsan Mehmood, Muhammad Mahboob Ur Rahman, Kashif Riaz, Qammer H. Abbasi
Abstract: We develop and evaluate two novel purpose-built deep learning (DL) models for synthesis of the arterial blood pressure (ABP) waveform in a cuff-less manner, using a single-site photoplethysmography (PPG) signal. We train and evaluate our DL models on the data of 209 subjects from the public UCI dataset on cuff-less blood pressure (CLBP) estimation. Our transformer model consists of an encoder-decoder pair that incorporates positional encoding, multi-head attention, layer normalization, and dropout techniques for ABP waveform synthesis. Secondly, under our frequency-domain (FD) learning approach, we first obtain the discrete cosine transform (DCT) coefficients of the PPG and ABP signals, and then learn a linear/non-linear (L/NL) regression between them. The transformer model (FD L/NL model) synthesizes the ABP waveform with a mean absolute error (MAE) of 3.01 (4.23). Further, the synthesis of ABP waveform also allows us to estimate the systolic blood pressure (SBP) and diastolic blood pressure (DBP) values. To this end, the transformer model reports an MAE of 3.77 mmHg and 2.69 mmHg, for SBP and DBP, respectively. On the other hand, the FD L/NL method reports an MAE of 4.37 mmHg and 3.91 mmHg, for SBP and DBP, respectively. Both methods fulfill the AAMI criterion. As for the BHS criterion, our transformer model (FD L/NL regression model) achieves grade A (grade B).
Authors: Yujun Mao, Yoon Kim, Yilun Zhou
Abstract: Recent large language models (LLMs) have shown indications of mathematical reasoning ability on challenging competition-level problems, especially with self-generated verbalizations of intermediate reasoning steps (i.e., chain-of-thought prompting). However, current evaluations mainly focus on the end-to-end final answer correctness, and it is unclear whether LLMs can make use of helpful side information such as problem-specific hints. In this paper, we propose a challenging benchmark dataset for enabling such analyses. The Concept and Hint-Annotated Math Problems (CHAMP) consists of high school math competition problems, annotated with concepts, or general math facts, and hints, or problem-specific tricks. These annotations allow us to explore the effects of additional information, such as relevant hints, misleading concepts, or related problems. This benchmark is difficult, with the best model only scoring 58.1% in standard settings. With concepts and hints, performance sometimes improves, indicating that some models can make use of such side information. Furthermore, we annotate model-generated solutions for their correctness. Using this corpus, we find that models often arrive at the correct final answer through wrong reasoning steps. In addition, we test whether models are able to verify these solutions, and find that most models struggle.
Authors: Weize Kong, Spurthi Amba Hombaiah, Mingyang Zhang, Qiaozhu Mei, Michael Bendersky
Abstract: Prompt engineering is critical for the development of LLM-based applications. However, it is usually done manually in a "trial and error" fashion that can be time consuming, ineffective, and sub-optimal. Even for the prompts which seemingly work well, there is always a lingering question: can the prompts be made better with further modifications? To address these problems, we investigate automated prompt engineering in this paper. Specifically, we propose PRewrite, an automated method to rewrite an under-optimized prompt to a more effective prompt. We instantiate the prompt rewriter using a LLM. The rewriter LLM is trained using reinforcement learning to optimize the performance on a given downstream task. We conduct experiments on diverse benchmark datasets, which demonstrates the effectiveness of PRewrite.
Authors: Luwei Bai
Abstract: We introduce a strict saddle property for $\ell_p$ regularized functions, and propose an iterative reweighted $\ell_1$ algorithm to solve the $\ell_p$ regularized problems. The algorithm is guaranteed to converge only to local minimizers when randomly initialized. The strict saddle property is shown generic on these sparse optimization problems. Those analyses as well as the proposed algorithm can be easily extended to general nonconvex regularized problems.
Authors: Yanyan Yang, Lili Wang, Xiaoya Zhai, Kai Chen, Wenming Wu, Yunkai Zhao, Ligang Liu, Xiao-Ming Fu
Abstract: Mechanical metamaterial is a synthetic material that can possess extraordinary physical characteristics, such as abnormal elasticity, stiffness, and stability, by carefully designing its internal structure. To make metamaterials contain delicate local structures with unique mechanical properties, it is a potential method to represent them through high-resolution voxels. However, it brings a substantial computational burden. To this end, this paper proposes a fast inverse design method, whose core is an advanced deep generative AI algorithm, to generate voxel-based mechanical metamaterials. Specifically, we use the self-conditioned diffusion model, capable of generating a microstructure with a resolution of $128^3$ to approach the specified homogenized tensor matrix in just 3 seconds. Accordingly, this rapid reverse design tool facilitates the exploration of extreme metamaterials, the sequence interpolation in metamaterials, and the generation of diverse microstructures for multi-scale design. This flexible and adaptive generative tool is of great value in structural engineering or other mechanical systems and can stimulate more subsequent research.
Authors: Talip Ucar, Aubin Ramon, Dino Oglic, Rebecca Croasdale-Wood, Tom Diethe, Pietro Sormanni
Abstract: We investigate the potential of patent data for improving the antibody humanness prediction using a multi-stage, multi-loss training process. Humanness serves as a proxy for the immunogenic response to antibody therapeutics, one of the major causes of attrition in drug discovery and a challenging obstacle for their use in clinical settings. We pose the initial learning stage as a weakly-supervised contrastive-learning problem, where each antibody sequence is associated with possibly multiple identifiers of function and the objective is to learn an encoder that groups them according to their patented properties. We then freeze a part of the contrastive encoder and continue training it on the patent data using the cross-entropy loss to predict the humanness score of a given antibody sequence. We illustrate the utility of the patent data and our approach by performing inference on three different immunogenicity datasets, unseen during training. Our empirical results demonstrate that the learned model consistently outperforms the alternative baselines and establishes new state-of-the-art on five out of six inference tasks, irrespective of the used metric.
Authors: Jacob Fein-Ashley, Tian Ye, Sachini Wickramasinghe, Bingyi Zhang, Rajgopal Kannan, Viktor Prasanna
Abstract: Image classifiers often rely on convolutional neural networks (CNN) for their tasks, which are inherently more heavyweight than multilayer perceptrons (MLPs), which can be problematic in real-time applications. Additionally, many image classification models work on both RGB and grayscale datasets. Classifiers that operate solely on grayscale images are much less common. Grayscale image classification has diverse applications, including but not limited to medical image classification and synthetic aperture radar (SAR) automatic target recognition (ATR). Thus, we present a novel grayscale (single channel) image classification approach using a vectorized view of images. We exploit the lightweightness of MLPs by viewing images as a vector and reducing our problem setting to the grayscale image classification setting. We find that using a single graph convolutional layer batch-wise increases accuracy and reduces variance in the performance of our model. Moreover, we develop a customized accelerator on FPGA for the proposed model with several optimizations to improve its performance. Our experimental results on benchmark grayscale image datasets demonstrate the effectiveness of the proposed model, achieving vastly lower latency (up to 16$\times$ less) and competitive or leading performance compared to other state-of-the-art image classification models on various domain-specific grayscale image classification datasets.
Authors: Ollie Liu, Deqing Fu, Dani Yogatama, Willie Neiswanger
Abstract: The potential of large language models (LLMs) as decision support tools is increasingly being explored in fields such as business, engineering, and medicine, which often face challenging tasks of decision-making under uncertainty. In this paper, we show that directly prompting LLMs on these types of decision-making problems can yield poor results, especially as the problem complexity increases. To aid in these tasks, we propose DeLLMa (Decision-making Large Language Model assistant), a framework designed to enhance decision-making accuracy in uncertain environments. DeLLMa involves a multi-step scaffolding procedure, drawing upon principles from decision theory and utility theory, to provide a rational and human-auditable decision-making process. We validate our framework on multiple realistic decision-making environments, demonstrating that DeLLMa can consistently enhance the decision-making performance of leading language models, and achieve up to a 40% increase in accuracy over competing methods.
Authors: Armand Foucault (IMT), Franck Mamalet (UT), Fran\c{c}ois Malgouyres (IMT)
Abstract: In recent years, Orthogonal Recurrent Neural Networks (ORNNs) have gained popularity due to their ability to manage tasks involving long-term dependencies, such as the copy-task, and their linear complexity. However, existing ORNNs utilize full precision weights and activations, which prevents their deployment on compact devices.In this paper, we explore the quantization of the weight matrices in ORNNs, leading to Quantized approximately Orthogonal RNNs (QORNNs). The construction of such networks remained an open problem, acknowledged for its inherent instability. We propose and investigate two strategies to learn QORNN by combining quantization-aware training (QAT) and orthogonal projections. We also study post-training quantization of the activations for pure integer computation of the recurrent loop. The most efficient models achieve results similar to state-of-the-art full-precision ORNN, LSTM and FastRNN on a variety of standard benchmarks, even with 4-bits quantization.
Authors: Reduan Achtibat, Sayed Mohammad Vakilzadeh Hatefi, Maximilian Dreyer, Aakriti Jain, Thomas Wiegand, Sebastian Lapuschkin, Wojciech Samek
Abstract: Large Language Models are prone to biased predictions and hallucinations, underlining the paramount importance of understanding their model-internal reasoning process. However, achieving faithful attributions for the entirety of a black-box transformer model and maintaining computational efficiency is an unsolved challenge. By extending the Layer-wise Relevance Propagation attribution method to handle attention layers, we address these challenges effectively. While partial solutions exist, our method is the first to faithfully and holistically attribute not only input but also latent representations of transformer models with the computational efficiency similar to a single backward pass. Through extensive evaluations against existing methods on LLaMa 2, Mixtral 8x7b, Flan-T5 and vision transformer architectures, we demonstrate that our proposed approach surpasses alternative methods in terms of faithfulness and enables the understanding of latent representations, opening up the door for concept-based explanations. We provide an LRP library at https://github.com/rachtibat/LRP-eXplains-Transformers.
URLs: https://github.com/rachtibat/LRP-eXplains-Transformers.
Authors: Kasimir Tanner, Matteo Vilucchio, Bruno Loureiro, Florent Krzakala
Abstract: This work investigates adversarial training in the context of margin-based linear classifiers in the high-dimensional regime where the dimension $d$ and the number of data points $n$ diverge with a fixed ratio $\alpha = n / d$. We introduce a tractable mathematical model where the interplay between the data and adversarial attacker geometries can be studied, while capturing the core phenomenology observed in the adversarial robustness literature. Our main theoretical contribution is an exact asymptotic description of the sufficient statistics for the adversarial empirical risk minimiser, under generic convex and non-increasing losses. Our result allow us to precisely characterise which directions in the data are associated with a higher generalisation/robustness trade-off, as defined by a robustness and a usefulness metric. In particular, we unveil the existence of directions which can be defended without penalising accuracy. Finally, we show the advantage of defending non-robust features during training, identifying a uniform protection as an inherently effective defence mechanism.
Authors: Tor Lattimore
Abstract: Bandit convex optimisation is a fundamental framework for studying zeroth-order convex optimisation. These notes cover the many tools used for this problem, including cutting plane methods, interior point methods, continuous exponential weights, gradient descent and online Newton step. The nuances between the many assumptions and setups are explained. Although there is not much truly new here, some existing tools are applied in novel ways to obtain new algorithms. A few bounds are improved in minor ways.
Authors: Kwang-Sung Jun, Jungtaek Kim
Abstract: Adapting to a priori unknown noise level is a very important but challenging problem in sequential decision-making as efficient exploration typically requires knowledge of the noise level, which is often loosely specified. We report significant progress in addressing this issue for linear bandits in two respects. First, we propose a novel confidence set that is `semi-adaptive' to the unknown sub-Gaussian parameter $\sigma_*^2$ in the sense that the (normalized) confidence width scales with $\sqrt{d\sigma_*^2 + \sigma_0^2}$ where $d$ is the dimension and $\sigma_0^2$ is the specified sub-Gaussian parameter (known) that can be much larger than $\sigma_*^2$. This is a significant improvement over $\sqrt{d\sigma_0^2}$ of the standard confidence set of Abbasi-Yadkori et al. (2011), especially when $d$ is large or $\sigma_*^2=0$. We show that this leads to an improved regret bound in linear bandits. Second, for bounded rewards, we propose a novel variance-adaptive confidence set that has much improved numerical performance upon prior art. We then apply this confidence set to develop, as we claim, the first practical variance-adaptive linear bandit algorithm via an optimistic approach, which is enabled by our novel regret analysis technique. Both of our confidence sets rely critically on `regret equality' from online learning. Our empirical evaluation in diverse Bayesian optimization tasks shows that our proposed algorithms demonstrate better or comparable performance compared to existing methods.
Authors: Rodrigo Veiga, Anastasia Remizova, Nicolas Macris
Abstract: We investigate the test risk of continuous-time stochastic gradient flow dynamics in learning theory. Using a path integral formulation we provide, in the regime of a small learning rate, a general formula for computing the difference between test risk curves of pure gradient and stochastic gradient flows. We apply the general theory to a simple model of weak features, which displays the double descent phenomenon, and explicitly compute the corrections brought about by the added stochastic term in the dynamics, as a function of time and model parameters. The analytical results are compared to simulations of discrete-time stochastic gradient descent and show good agreement.
Authors: Ming Li, Lichang Chen, Jiuhai Chen, Shwai He, Jiuxiang Gu, Tianyi Zhou
Abstract: Instruction tuning is critical to large language models (LLMs) for achieving better instruction following and task adaptation capabilities but its success heavily relies on the training data quality. Many recent methods focus on improving the data quality but often overlook the compatibility of the data with the student model being finetuned. This paper introduces Selective Reflection-Tuning, a novel paradigm that synergizes a teacher LLM's reflection and introspection for improving existing data quality with the data selection capability of the student LLM, to automatically refine existing instruction-tuning data. This teacher-student collaboration produces high-quality and student-compatible instruction-response pairs, resulting in sample-efficient instruction tuning and LLMs of superior performance. Selective Reflection-Tuning is a data augmentation and synthesis that generally improves LLM finetuning and self-improvement without collecting brand-new data. We apply our method to Alpaca and WizardLM data and achieve much stronger and top-tier 7B and 13B LLMs.
Authors: Yanis Labrak, Adrien Bazoge, Emmanuel Morin, Pierre-Antoine Gourraud, Mickael Rouvier, Richard Dufour
Abstract: Large Language Models (LLMs) have demonstrated remarkable versatility in recent years, offering potential applications across specialized domains such as healthcare and medicine. Despite the availability of various open-source LLMs tailored for health contexts, adapting general-purpose LLMs to the medical domain presents significant challenges. In this paper, we introduce BioMistral, an open-source LLM tailored for the biomedical domain, utilizing Mistral as its foundation model and further pre-trained on PubMed Central. We conduct a comprehensive evaluation of BioMistral on a benchmark comprising 10 established medical question-answering (QA) tasks in English. We also explore lightweight models obtained through quantization and model merging approaches. Our results demonstrate BioMistral's superior performance compared to existing open-source medical models and its competitive edge against proprietary counterparts. Finally, to address the limited availability of data beyond English and to assess the multilingual generalization of medical LLMs, we automatically translated and evaluated this benchmark into 7 other languages. This marks the first large-scale multilingual evaluation of LLMs in the medical domain. Datasets, multilingual evaluation benchmarks, scripts, and all the models obtained during our experiments are freely released.
Authors: Ming Li, Jiuhai Chen, Lichang Chen, Tianyi Zhou
Abstract: Making LLMs speak for different, especially minority groups of people, and generate statements supporting their diverse or even controversial perspectives is critical to creating an inclusive environment. However, existing LLMs lack sufficient controllability to the stance of their generated content, which often contains inconsistent, neutral, or biased statements. In this paper, we improve the controllability of LLMs in generating statements supporting an argument the user defined in the prompt. We find that multi-round debates between two LLMs with opposite stances generate higher-quality and more salient statements for each, which are important training data to improve the controllability of LLMs. Motivated by this, we develop a novel debate & tuning (DEBATUNE) pipeline finetuning LLMs to generate the statements obtained via debate. To examine DEBATUNE, we curate the largest dataset of debate topics so far, which covers 710 controversial topics and corresponding arguments for each topic. Evaluations by the GPT-4 judge with a novel controversy controllability metric show that LLMs' capability of generating diverse perspectives is significantly improved by DEBATUNE. Moreover, such controllability can be generalized to unseen topics, generating high-quality statements supporting controversial arguments.
Authors: Kyoungseok Jang, Chicheng Zhang, Kwang-Sung Jun
Abstract: We study low-rank matrix trace regression and the related problem of low-rank matrix bandits. Assuming access to the distribution of the covariates, we propose a novel low-rank matrix estimation method called LowPopArt and provide its recovery guarantee that depends on a novel quantity denoted by B(Q) that characterizes the hardness of the problem, where Q is the covariance matrix of the measurement distribution. We show that our method can provide tighter recovery guarantees than classical nuclear norm penalized least squares (Koltchinskii et al., 2011) in several problems. To perform efficient estimation with a limited number of measurements from an arbitrarily given measurement set A, we also propose a novel experimental design criterion that minimizes B(Q) with computational efficiency. We leverage our novel estimator and design of experiments to derive two low-rank linear bandit algorithms for general arm sets that enjoy improved regret upper bounds. This improves over previous works on low-rank bandits, which make somewhat restrictive assumptions that the arm set is the unit ball or that an efficient exploration distribution is given. To our knowledge, our experimental design criterion is the first one tailored to low-rank matrix estimation beyond the naive reduction to linear regression, which can be of independent interest.
Authors: Abe Bohan Hou, Jingyu Zhang, Yichen Wang, Daniel Khashabi, Tianxing He
Abstract: Recent watermarked generation algorithms inject detectable signatures during language generation to facilitate post-hoc detection. While token-level watermarks are vulnerable to paraphrase attacks, SemStamp (Hou et al., 2023) applies watermark on the semantic representation of sentences and demonstrates promising robustness. SemStamp employs locality-sensitive hashing (LSH) to partition the semantic space with arbitrary hyperplanes, which results in a suboptimal tradeoff between robustness and speed. We propose k-SemStamp, a simple yet effective enhancement of SemStamp, utilizing k-means clustering as an alternative of LSH to partition the embedding space with awareness of inherent semantic structure. Experimental results indicate that k-SemStamp saliently improves its robustness and sampling efficiency while preserving the generation quality, advancing a more effective tool for machine-generated text detection.
Authors: Yavuz Faruk Bakman, Duygu Nur Yaldiz, Baturalp Buyukates, Chenyang Tao, Dimitrios Dimitriadis, Salman Avestimehr
Abstract: Generative Large Language Models (LLMs) are widely utilized for their excellence in various tasks. However, their tendency to produce inaccurate or misleading outputs poses a potential risk, particularly in high-stakes environments. Therefore, estimating the correctness of generative LLM outputs is an important task for enhanced reliability. Uncertainty Estimation (UE) in generative LLMs is an evolving domain, where SOTA probability-based methods commonly employ length-normalized scoring. In this work, we propose Meaning-Aware Response Scoring (MARS) as an alternative to length-normalized scoring for UE methods. MARS is a novel scoring function that considers the semantic contribution of each token in the generated sequence in the context of the question. We demonstrate that integrating MARS into UE methods results in a universal and significant improvement in UE performance. We conduct experiments using three distinct closed-book question-answering datasets across five popular pre-trained LLMs. Lastly, we validate the efficacy of MARS on a Medical QA dataset. Code can be found https://github.com/Ybakman/LLM_Uncertainity.
Authors: Jinhao Duan, Renming Zhang, James Diffenderfer, Bhavya Kailkhura, Lichao Sun, Elias Stengel-Eskin, Mohit Bansal, Tianlong Chen, Kaidi Xu
Abstract: As Large Language Models (LLMs) are integrated into critical real-world applications, their strategic and logical reasoning abilities are increasingly crucial. This paper evaluates LLMs' reasoning abilities in competitive environments through game-theoretic tasks, e.g., board and card games that require pure logic and strategic reasoning to compete with opponents. We first propose GTBench, a language-driven environment composing 10 widely recognized tasks, across a comprehensive game taxonomy: complete versus incomplete information, dynamic versus static, and probabilistic versus deterministic scenarios. Then, we (1) Characterize the game-theoretic reasoning of LLMs; and (2) Perform LLM-vs.-LLM competitions as reasoning evaluation. We observe that (1) LLMs have distinct behaviors regarding various gaming scenarios; for example, LLMs fail in complete and deterministic games yet they are competitive in probabilistic gaming scenarios; (2) Most open-source LLMs, e.g., CodeLlama-34b-Instruct and Llama-2-70b-chat, are less competitive than commercial LLMs, e.g., GPT-4, in complex games, yet the recently released Llama-3-70b-Instruct makes up for this shortcoming. In addition, code-pretraining greatly benefits strategic reasoning, while advanced reasoning methods such as Chain-of-Thought (CoT) and Tree-of-Thought (ToT) do not always help. We further characterize the game-theoretic properties of LLMs, such as equilibrium and Pareto Efficiency in repeated games. Detailed error profiles are provided for a better understanding of LLMs' behavior. We hope our research provides standardized protocols and serves as a foundation to spur further explorations in the strategic reasoning of LLMs.
Authors: Jianrui Zhang, Mu Cai, Tengyang Xie, Yong Jae Lee
Abstract: We propose CounterCurate, a framework to comprehensively improve the visio-linguistic compositional reasoning capability for both contrastive and generative multimodal models. In particular, we identify two critical under-explored problems: the neglect of the physically grounded reasoning (counting and position understanding) and the potential of using highly capable text and image generation models for semantic counterfactual fine-tuning. Our work pioneers an approach that addresses these gaps. We first spotlight the near-chance performance of multimodal models like CLIP and LLaVA in physically grounded compositional reasoning. We then apply simple data augmentation using grounded image generation model GLIGEN to generate fine-tuning data, resulting in significant performance improvements: +33% and +37% for CLIP and LLaVA, respectively, on our newly curated Flickr30k-Positions benchmark. Moreover, we exploit the capabilities of high-performing text generation and image generation models, specifically GPT-4V and DALLE-3, to curate challenging semantic counterfactuals, thereby further enhancing compositional reasoning capabilities on benchmarks such as SugarCrepe, where CounterCurate outperforms GPT-4V. To facilitate future research, we release our code, dataset, benchmark, and checkpoints at https://countercurate.github.io.
Authors: Dominik Schr\"oder, Daniil Dmitriev, Hugo Cui, Bruno Loureiro
Abstract: For a large class of feature maps we provide a tight asymptotic characterisation of the test error associated with learning the readout layer, in the high-dimensional limit where the input dimension, hidden layer widths, and number of training samples are proportionally large. This characterization is formulated in terms of the population covariance of the features. Our work is partially motivated by the problem of learning with Gaussian rainbow neural networks, namely deep non-linear fully-connected networks with random but structured weights, whose row-wise covariances are further allowed to depend on the weights of previous layers. For such networks we also derive a closed-form formula for the feature covariance in terms of the weight matrices. We further find that in some cases our results can capture feature maps learned by deep, finite-width neural networks trained under gradient descent.
Authors: Amit Dhurandhar, Rahul Nair, Moninder Singh, Elizabeth Daly, Karthikeyan Natesan Ramamurthy
Abstract: Evaluation and ranking of large language models (LLMs) has become an important problem with the proliferation of these models and their impact. Evaluation methods either require human responses which are expensive to acquire or use pairs of LLMs to evaluate each other which can be unreliable. In this paper, we provide a novel perspective where, given a dataset of prompts (viz. questions, instructions, etc.) and a set of LLMs, we rank them without access to any ground truth or reference responses. Inspired by real life where both an expert and a knowledgeable person can identify a novice our main idea is to consider triplets of models, where each one of them evaluates the other two, correctly identifying the worst model in the triplet with high probability. We also analyze our idea and provide sufficient conditions for it to succeed. Applying this idea repeatedly, we propose two methods to rank LLMs. In experiments on different generative tasks (summarization, multiple-choice, and dialog), our methods reliably recover close to true rankings without reference data. This points to a viable low-resource mechanism for practical use.
Authors: Lele Cao, Valentin Buchner, Zineb Senane, Fangkai Yang
Abstract: Multimodal Large Language Models (MLLMs) are commonly evaluated using costly annotated multimodal benchmarks. However, these benchmarks often struggle to keep pace with the rapidly advancing requirements of MLLM evaluation. We propose GenCeption, a novel and annotation-free MLLM evaluation framework that merely requires unimodal data to assess inter-modality semantic coherence and inversely reflects the models' inclination to hallucinate. Analogous to the popular DrawCeption game, GenCeption initiates with a non-textual sample and undergoes a series of iterative description and generation steps. Semantic drift across iterations is quantified using the GC@T metric. Our empirical findings validate GenCeption's efficacy, showing strong correlations with popular MLLM benchmarking results. GenCeption may be extended to mitigate training data contamination by utilizing ubiquitous, previously unseen unimodal data.
Authors: Yanis Labrak, Adrien Bazoge, Beatrice Daille, Mickael Rouvier, Richard Dufour
Abstract: Subword tokenization has become the prevailing standard in the field of natural language processing (NLP) over recent years, primarily due to the widespread utilization of pre-trained language models. This shift began with Byte-Pair Encoding (BPE) and was later followed by the adoption of SentencePiece and WordPiece. While subword tokenization consistently outperforms character and word-level tokenization, the precise factors contributing to its success remain unclear. Key aspects such as the optimal segmentation granularity for diverse tasks and languages, the influence of data sources on tokenizers, and the role of morphological information in Indo-European languages remain insufficiently explored. This is particularly pertinent for biomedical terminology, characterized by specific rules governing morpheme combinations. Despite the agglutinative nature of biomedical terminology, existing language models do not explicitly incorporate this knowledge, leading to inconsistent tokenization strategies for common terms. In this paper, we seek to delve into the complexities of subword tokenization in French biomedical domain across a variety of NLP tasks and pinpoint areas where further enhancements can be made. We analyze classical tokenization algorithms, including BPE and SentencePiece, and introduce an original tokenization strategy that integrates morpheme-enriched word segmentation into existing tokenization methods.
Authors: Anja Deli\'c, Matej Grci\'c, Sini\v{s}a \v{S}egvi\'c
Abstract: Outlier detection is an essential capability in safety-critical applications of supervised visual recognition. Most of the existing methods deliver best results by encouraging standard closed-set models to produce low-confidence predictions in negative training data. However, that approach conflates prediction uncertainty with recognition of the negative class. We therefore reconsider direct prediction of K+1 logits that correspond to K groundtruth classes and one outlier class. This setup allows us to formulate a novel anomaly score as an ensemble of in-distribution uncertainty and the posterior of the outlier class which we term negative objectness. Now outliers can be independently detected due to i) high prediction uncertainty or ii) similarity with negative data. We embed our method into a dense prediction architecture with mask-level recognition over K+2 classes. The training procedure encourages the novel K+2-th class to learn negative objectness at pasted negative instances. Our models outperform the current state-of-the art on standard benchmarks for image-wide and pixel-level outlier detection with and without training on real negative data.
Authors: Zhaorun Chen, Zhuokai Zhao, Hongyin Luo, Huaxiu Yao, Bo Li, Jiawei Zhou
Abstract: While large vision-language models (LVLMs) have demonstrated impressive capabilities in interpreting multi-modal contexts, they invariably suffer from object hallucinations (OH). We introduce HALC, a novel decoding algorithm designed to mitigate OH in LVLMs. HALC leverages distinct fine-grained optimal visual information in vision-language tasks and operates on both local and global contexts simultaneously. Specifically, HALC integrates a robust auto-focal grounding mechanism (locally) to correct hallucinated tokens on the fly, and a specialized beam search algorithm (globally) to significantly reduce OH while preserving text generation quality. Additionally, HALC can be integrated into any LVLMs as a plug-and-play module without extra training. Extensive experimental studies demonstrate the effectiveness of HALC in reducing OH, outperforming state-of-the-arts across four benchmarks.
Authors: Anudeex Shetty, Yue Teng, Ke He, Qiongkai Xu
Abstract: Embedding as a Service (EaaS) has become a widely adopted solution, which offers feature extraction capabilities for addressing various downstream tasks in Natural Language Processing (NLP). Prior studies have shown that EaaS can be prone to model extraction attacks; nevertheless, this concern could be mitigated by adding backdoor watermarks to the text embeddings and subsequently verifying the attack models post-publication. Through the analysis of the recent watermarking strategy for EaaS, EmbMarker, we design a novel CSE (Clustering, Selection, Elimination) attack that removes the backdoor watermark while maintaining the high utility of embeddings, indicating that the previous watermarking approach can be breached. In response to this new threat, we propose a new protocol to make the removal of watermarks more challenging by incorporating multiple possible watermark directions. Our defense approach, WARDEN, notably increases the stealthiness of watermarks and has been empirically shown to be effective against CSE attack.
Authors: Fu Chen, Qinglin Zhao, Li Feng, Chuangtao Chen, Yangbin Lin, Jianhong Lin
Abstract: The rapid advancement of quantum computing has increasingly highlighted its potential in the realm of machine learning, particularly in the context of natural language processing (NLP) tasks. Quantum machine learning (QML) leverages the unique capabilities of quantum computing to offer novel perspectives and methodologies for complex data processing and pattern recognition challenges. This paper introduces a novel Quantum Mixed-State Attention Network (QMSAN), which integrates the principles of quantum computing with classical machine learning algorithms, especially self-attention networks, to enhance the efficiency and effectiveness in handling NLP tasks. QMSAN model employs a quantum attention mechanism based on mixed states, enabling efficient direct estimation of similarity between queries and keys within the quantum domain, leading to more effective attention weight acquisition. Additionally, we propose an innovative quantum positional encoding scheme, implemented through fixed quantum gates within the quantum circuit, to enhance the model's accuracy. Experimental validation on various datasets demonstrates that QMSAN model outperforms existing quantum and classical models in text classification, achieving significant performance improvements. QMSAN model not only significantly reduces the number of parameters but also exceeds classical self-attention networks in performance, showcasing its strong capability in data representation and information extraction. Furthermore, our study investigates the model's robustness in different quantum noise environments, showing that QMSAN possesses commendable robustness to low noise.
Authors: Yubiao Yue, Zhenzhang Li
Abstract: Since the era of deep learning, convolutional neural networks (CNNs) and vision transformers (ViTs) have been extensively studied and widely used in medical image classification tasks. Unfortunately, CNN's limitations in modeling long-range dependencies result in poor classification performances. In contrast, ViTs are hampered by the quadratic computational complexity of their self-attention mechanism, making them difficult to deploy in real-world settings with limited computational resources. Recent studies have shown that state space models (SSMs) represented by Mamba can effectively model long-range dependencies while maintaining linear computational complexity. Inspired by it, we proposed MedMamba, the first vision Mamba for generalized medical image classification. Concretely, we introduced a novel hybrid basic block named SS-Conv-SSM, which integrates the convolutional layers for extracting local features with the abilities of SSM to capture long-range dependencies, aiming to model medical images from different image modalities efficiently. By employing the grouped convolution strategy and channel-shuffle operation, MedMamba successfully provides fewer model parameters and a lower computational burden for efficient applications. To demonstrate the potential of MedMamba, we conducted extensive experiments using 16 datasets containing ten imaging modalities and 411,007 images. Experimental results show that the proposed MedMamba demonstrates competitive performance in classifying various medical images compared with the state-of-the-art methods. Our work is aims to establish a new baseline for medical image classification and provide valuable insights for developing more powerful SSM-based artificial intelligence algorithms and application systems in the medical field. The source codes and all pre-trained weights of MedMamba are available at https://github.com/YubiaoYue/MedMamba.
Authors: Qibing Ren, Chang Gao, Jing Shao, Junchi Yan, Xin Tan, Wai Lam, Lizhuang Ma
Abstract: The rapid advancement of Large Language Models (LLMs) has brought about remarkable generative capabilities but also raised concerns about their potential misuse. While strategies like supervised fine-tuning and reinforcement learning from human feedback have enhanced their safety, these methods primarily focus on natural languages, which may not generalize to other domains. This paper introduces CodeAttack, a framework that transforms natural language inputs into code inputs, presenting a novel environment for testing the safety generalization of LLMs. Our comprehensive studies on state-of-the-art LLMs including GPT-4, Claude-2, and Llama-2 series reveal a new and universal safety vulnerability of these models against code input: CodeAttack bypasses the safety guardrails of all models more than 80\% of the time. We find that a larger distribution gap between CodeAttack and natural language leads to weaker safety generalization, such as encoding natural language input with data structures. Furthermore, we give our hypotheses about the success of CodeAttack: the misaligned bias acquired by LLMs during code training, prioritizing code completion over avoiding the potential safety risk. Finally, we analyze potential mitigation measures. These findings highlight new safety risks in the code domain and the need for more robust safety alignment algorithms to match the code capabilities of LLMs.
Authors: Johannes Schneider, Rene Abraham, Christian Meske
Abstract: Generative Artificial Intelligence (GenAI), specifically large language models like ChatGPT, has swiftly entered organizations without adequate governance, posing both opportunities and risks. Despite extensive debates on GenAI's transformative nature and regulatory measures, limited research addresses organizational governance, encompassing technical and business perspectives. Our review paper fills this gap by surveying recent works with the purpose of developing a framework for GenAI governance within companies. This framework outlines the scope, objectives, and governance mechanisms tailored to harness business opportunities as well as mitigate risks associated with GenAI integration. Our research contributes a focused approach to GenAI governance, offering practical insights for companies navigating the challenges of GenAI adoption and highlighting research gaps.
Authors: Yuzhao Heng, Chunyuan Deng, Yitong Li, Yue Yu, Yinghao Li, Rongzhi Zhang, Chao Zhang
Abstract: Although Large Language Models (LLMs) exhibit remarkable adaptability across domains, these models often fall short in structured knowledge extraction tasks such as named entity recognition (NER). This paper explores an innovative, cost-efficient strategy to harness LLMs with modest NER capabilities for producing superior NER datasets. Our approach diverges from the basic class-conditional prompts by instructing LLMs to self-reflect on the specific domain, thereby generating domain-relevant attributes (such as category and emotions for movie reviews), which are utilized for creating attribute-rich training data. Furthermore, we preemptively generate entity terms and then develop NER context data around these entities, effectively bypassing the LLMs' challenges with complex structures. Our experiments across both general and niche domains reveal significant performance enhancements over conventional data generation methods while being more cost-effective than existing alternatives.
Authors: R. Kenny Jones, Siddhartha Chaudhuri, Daniel Ritchie
Abstract: People grasp flexible visual concepts from a few examples. We explore a neurosymbolic system that learns how to infer programs that capture visual concepts in a domain-general fashion. We introduce Template Programs: programmatic expressions from a domain-specific language that specify structural and parametric patterns common to an input concept. Our framework supports multiple concept-related tasks, including few-shot generation and co-segmentation through parsing. We develop a learning paradigm that allows us to train networks that infer Template Programs directly from visual datasets that contain concept groupings. We run experiments across multiple visual domains: 2D layouts, Omniglot characters, and 3D shapes. We find that our method outperforms task-specific alternatives, and performs competitively against domain-specific approaches for the limited domains where they exist.
Authors: Nishant Kumar, Ziyan Tao, Jaikirat Singh, Yang Li, Peiwen Sun, Binghui Zhao, Stefan Gumhold
Abstract: Image fusion typically employs non-invertible neural networks to merge multiple source images into a single fused image. However, for clinical experts, solely relying on fused images may be insufficient for making diagnostic decisions, as the fusion mechanism blends features from source images, thereby making it difficult to interpret the underlying tumor pathology. We introduce FusionINN, a novel decomposable image fusion framework, capable of efficiently generating fused images and also decomposing them back to the source images. FusionINN is designed to be bijective by including a latent image alongside the fused image, while ensuring minimal transfer of information from the source images to the latent representation. To the best of our knowledge, we are the first to investigate the decomposability of fused images, which is particularly crucial for life-sensitive applications such as medical image fusion compared to other tasks like multi-focus or multi-exposure image fusion. Our extensive experimentation validates FusionINN over existing discriminative and generative fusion methods, both subjectively and objectively. Moreover, compared to a recent denoising diffusion-based fusion model, our approach offers faster and qualitatively better fusion results.
Authors: Brian B. Moser, Federico Raue, Andreas Dengel
Abstract: In image Super-Resolution (SR), relying on large datasets for training is a double-edged sword. While offering rich training material, they also demand substantial computational and storage resources. In this work, we analyze dataset pruning to solve these challenges. We introduce a novel approach that reduces a dataset to a core-set of training samples, selected based on their loss values as determined by a simple pre-trained SR model. By focusing the training on just 50\% of the original dataset, specifically on the samples characterized by the highest loss values, we achieve results comparable to or surpassing those obtained from training on the entire dataset. Interestingly, our analysis reveals that the top 5\% of samples with the highest loss values negatively affect the training process. Excluding these samples and adjusting the selection to favor easier samples further enhances training outcomes. Our work opens new perspectives to the untapped potential of dataset pruning in image SR. It suggests that careful selection of training data based on loss-value metrics can lead to better SR models, challenging the conventional wisdom that more data inevitably leads to better performance.
Authors: Thomas Melistas, Nikos Spyrou, Nefeli Gkouti, Pedro Sanchez, Athanasios Vlontzos, Yannis Panagakis, Giorgos Papanastasiou, Sotirios A. Tsaftaris
Abstract: Generative AI has revolutionised visual content editing, empowering users to effortlessly modify images and videos. However, not all edits are equal. To perform realistic edits in domains such as natural image or medical imaging, modifications must respect causal relationships inherent to the data generation process. Such image editing falls into the counterfactual image generation regime. Evaluating counterfactual image generation is substantially complex: not only it lacks observable ground truths, but also requires adherence to causal constraints. Although several counterfactual image generation methods and evaluation metrics exist, a comprehensive comparison within a unified setting is lacking. We present a comparison framework to thoroughly benchmark counterfactual image generation methods. We integrate all models that have been used for the task at hand and expand them to novel datasets and causal graphs, demonstrating the superiority of Hierarchical VAEs across most datasets and metrics. Our framework is implemented in a user-friendly Python package that can be extended to incorporate additional SCMs, causal methods, generative models, and datasets for the community to build on.
Authors: Lei Ding, Jeshwanth Bheemanpally, Yi Zhang
Abstract: Many people use search engines to find online guidance to solve computer or mobile device problems. Users frequently encounter challenges in identifying effective solutions from search results, often wasting time trying ineffective solutions that seem relevant yet fail to solve real problems. This paper introduces a novel approach to improving the accuracy and relevance of online technical support search results through automated search results verification and reranking. Taking "How-to" queries specific to on-device execution as a starting point, we developed the first solution that allows an AI agent to interpret and execute step-by-step instructions in the search results in a controlled Android environment. We further integrated the agent's findings into a reranking mechanism that orders search results based on the success indicators of the tested solutions. The paper details the architecture of our solution and a comprehensive evaluation of the system through a series of tests across various application domains. The results demonstrate a significant improvement in the quality and reliability of the top-ranked results. Our findings suggest a paradigm shift in how search engine ranking for online technical support help can be optimized, offering a scalable and automated solution to the pervasive challenge of finding effective and reliable online help.
Authors: Ahmed Maged, Salah Haridy, Herman Shen
Abstract: As the manufacturing industry advances with sensor integration and automation, the opaque nature of deep learning models in machine learning poses a significant challenge for fault detection and diagnosis. And despite the related predictive insights Artificial Intelligence (AI) can deliver, advanced machine learning engines often remain a black box. This paper reviews the eXplainable AI (XAI) tools and techniques in this context. We explore various XAI methodologies, focusing on their role in making AI decision-making transparent, particularly in critical scenarios where humans are involved. We also discuss current limitations and potential future research that aims to balance explainability with model performance while improving trustworthiness in the context of AI applications for critical industrial use cases.
Authors: Tommie Kerssies, Daan de Geus, Gijs Dubbelman
Abstract: Recent vision foundation models (VFMs) have demonstrated proficiency in various tasks but require supervised fine-tuning to perform the task of semantic segmentation effectively. Benchmarking their performance is essential for selecting current models and guiding future model developments for this task. The lack of a standardized benchmark complicates comparisons. Therefore, the primary objective of this paper is to study how VFMs should be benchmarked for semantic segmentation. To do so, various VFMs are fine-tuned under various settings, and the impact of individual settings on the performance ranking and training time is assessed. Based on the results, the recommendation is to fine-tune the ViT-B variants of VFMs with a 16x16 patch size and a linear decoder, as these settings are representative of using a larger model, more advanced decoder and smaller patch size, while reducing training time by more than 13 times. Using multiple datasets for training and evaluation is also recommended, as the performance ranking across datasets and domain shifts varies. Linear probing, a common practice for some VFMs, is not recommended, as it is not representative of end-to-end fine-tuning. The benchmarking setup recommended in this paper enables a performance analysis of VFMs for semantic segmentation. The findings of such an analysis reveal that pretraining with promptable segmentation is not beneficial, whereas masked image modeling (MIM) with abstract representations is crucial, even more important than the type of supervision used. The code for efficiently fine-tuning VFMs for semantic segmentation can be accessed through the project page at: https://tue-mps.github.io/benchmark-vfm-ss/.
Authors: Andrew Zammit-Mangion, Matthew Sainsbury-Dale, Rapha\"el Huser
Abstract: Simulation-based methods for making statistical inference have evolved dramatically over the past 50 years, keeping pace with technological advancements. The field is undergoing a new revolution as it embraces the representational capacity of neural networks, optimisation libraries and graphics processing units for learning complex mappings between data and inferential targets. The resulting tools are amortised, in the sense that they allow inference to be made quickly through fast feedforward operations. In this article we review recent progress made in the context of point estimation, approximate Bayesian inference, summary-statistic construction, and likelihood approximation. The review also covers available software, and includes a simple illustration to showcase the wide array of tools available for amortised inference and the benefits they offer over state-of-the-art Markov chain Monte Carlo methods. The article concludes with an overview of relevant topics and an outlook on future research directions.
Authors: Ge Gao, Alexey Taymanov, Eduardo Salinas, Paul Mineiro, Dipendra Misra
Abstract: We study interactive learning of LLM-based language agents based on user edits made to the agent's output. In a typical setting such as writing assistants, the user interacts with a language agent to generate a response given a context, and may optionally edit the agent response to personalize it based on their latent preference, in addition to improving the correctness. The edit feedback is naturally generated, making it a suitable candidate for improving the agent's alignment with the user's preference, and for reducing the cost of user edits over time. We propose a learning framework, PRELUDE that infers a description of the user's latent preference based on historic edit data. The inferred user preference descriptions are used to define prompts for generating responses in the future. This avoids fine-tuning the agent, which is costly, challenging to scale with the number of users, and may even degrade its performance on other tasks. Furthermore, learning descriptive preference improves interpretability, allowing the user to view and modify the learned preference. However, user preference can be complex, subtle, and vary based on context, making it challenging to learn. To address this, we propose a simple yet effective algorithm named CIPHER that leverages the LLM to infer the user preference for a given context based on user edits. In the future, CIPHER retrieves inferred preferences from the k-closest contexts in the history, and forms an aggregate preference for response generation. We introduce two interactive environments -- summarization and email writing, and use a GPT-4 simulated user for evaluation. On both tasks, CIPHER outperforms several baselines by achieving the lowest edit distance cost while only having a small overhead in LLM query cost. Our analysis reports that user preferences learned by CIPHER show significant similarity to the ground truth latent preferences.
Authors: Letitia Parcalabescu, Anette Frank
Abstract: Vision and language model (VLM) decoders are currently the best-performing architectures on multimodal tasks. Next to predictions, they can also produce explanations, either in post-hoc or CoT settings. However, it is not clear how much they use the vision and text modalities when generating predictions or explanations. In this work, we investigate if VLMs rely on modalities differently when they produce explanations as opposed to providing answers. We also evaluate the self-consistency of VLM decoders in both post-hoc and CoT explanation settings, by extending existing unimodal tests and measures to VLM decoders. We find that VLMs are less self-consistent than LLMs. Text contributions in VL decoders are more important than image contributions in all examined tasks. Moreover, the contributions of images are significantly stronger for explanation generation compared to answer generation. This difference is even larger in CoT compared to post-hoc explanations. Lastly, we provide an up-to-date benchmarking of state-of-the-art VL decoders on the VALSE benchmark, which before only covered VL encoders. We find that VL decoders still struggle with most phenomena tested by VALSE.
Authors: Hanyin Wang, Chufan Gao, Bolun Liu, Qiping Xu, Guleid Hussein, Mohamad El Labban, Kingsley Iheasirim, Hariprasad Korsapati, Chuck Outcalt, Jimeng Sun
Abstract: Proprietary Large Language Models (LLMs) such as GPT-4 and Gemini have demonstrated promising capabilities in clinical text summarization tasks. However, due to patient data privacy concerns and computational costs, many healthcare providers prefer using small, locally-hosted models over external generic LLMs. This study presents a comprehensive domain- and task-specific adaptation process for the open-source LLaMA-2 13 billion parameter model, enabling it to generate high-quality clinical notes from outpatient patient-doctor dialogues. Our process incorporates continued pre-training, supervised fine-tuning, and reinforcement learning from both AI and human feedback. We introduced a new approach, DistillDirect, for performing on-policy reinforcement learning with Gemini 1.0 Pro as the teacher model. Our resulting model, LLaMA-Clinic, can generate clinical notes comparable in quality to those authored by physicians. In a blinded physician reader study, the majority (90.4%) of individual evaluations rated the notes generated by LLaMA-Clinic as "acceptable" or higher across all three criteria: real-world readiness, completeness, and accuracy. In the more challenging "Assessment and Plan" section, LLaMA-Clinic scored higher (4.2/5) in real-world readiness than physician-authored notes (4.1/5). Our cost analysis for inference shows that our LLaMA-Clinic model achieves a 3.75-fold cost reduction compared to an external generic LLM service. Additionally, we highlight key considerations for future clinical note-generation tasks, emphasizing the importance of pre-defining a best-practice note format, rather than relying on LLMs to determine this for clinical practice. We have made our newly created synthetic clinic dialogue-note dataset and the physician feedback dataset publicly available to foster future research.
Authors: Giorgio Tosti Balducci, Boyang Chen, Matthias M\"oller, Marc Gerritsma, Roeland De Breuker
Abstract: Modeling open hole failure of composites is a complex task, consisting in a highly nonlinear response with interacting failure modes. Numerical modeling of this phenomenon has traditionally been based on the finite element method, but requires to tradeoff between high fidelity and computational cost. To mitigate this shortcoming, recent work has leveraged machine learning to predict the strength of open hole composite specimens. Here, we also propose using data-based models but to tackle open hole composite failure from a classification point of view. More specifically, we show how to train surrogate models to learn the ultimate failure envelope of an open hole composite plate under in-plane loading. To achieve this, we solve the classification problem via support vector machine (SVM) and test different classifiers by changing the SVM kernel function. The flexibility of kernel-based SVM also allows us to integrate the recently developed quantum kernels in our algorithm and compare them with the standard radial basis function (RBF) kernel. Finally, thanks to kernel-target alignment optimization, we tune the free parameters of all kernels to best separate safe and failure-inducing loading states. The results show classification accuracies higher than 90% for RBF, especially after alignment, followed closely by the quantum kernel classifiers.
Authors: Jakob Roche
Abstract: This paper presents the detection of a periodic dimming event in the lightcurve of the G1.5IV-V type star KIC 1718360. This is based on visible-light observations conducted by both the TESS and Kepler space telescopes. Analysis of the data seems to point toward a high rotation rate in the star, with a rotational period of 2.938 days. The high variability seen within the star's lightcurve points toward classification as a rotating variable. The initial observation was made in Kepler Quarter 16 data using the One-Class SVM machine learning method. Subsequent observations by the TESS space telescope corroborated these findings. It appears that KIC 1718360 is a nearby rotating variable that appears in little to no major catalogs as such. A secondary, additional periodic dip is also present, indicating a possible exoplanetary companion.
Authors: Yifeng Tian, Nishant Panda, Yen Ting Lin
Abstract: We present the Liouville Flow Importance Sampler (LFIS), an innovative flow-based model for generating samples from unnormalized density functions. LFIS learns a time-dependent velocity field that deterministically transports samples from a simple initial distribution to a complex target distribution, guided by a prescribed path of annealed distributions. The training of LFIS utilizes a unique method that enforces the structure of a derived partial differential equation to neural networks modeling velocity fields. By considering the neural velocity field as an importance sampler, sample weights can be computed through accumulating errors along the sample trajectories driven by neural velocity fields, ensuring unbiased and consistent estimation of statistical quantities. We demonstrate the effectiveness of LFIS through its application to a range of benchmark problems, on many of which LFIS achieved state-of-the-art performance.
Authors: Matteo Zecchin, Osvaldo Simeone
Abstract: Adaptive Risk Control (ARC) is an online calibration strategy based on set prediction that offers worst-case deterministic long-term risk control, as well as statistical marginal coverage guarantees. ARC adjusts the size of the prediction set by varying a single scalar threshold based on feedback from past decisions. In this work, we introduce Localized Adaptive Risk Control (L-ARC), an online calibration scheme that targets statistical localized risk guarantees ranging from conditional risk to marginal risk, while preserving the worst-case performance of ARC. L-ARC updates a threshold function within a reproducing kernel Hilbert space (RKHS), with the kernel determining the level of localization of the statistical risk guarantee. The theoretical results highlight a trade-off between localization of the statistical risk and convergence speed to the long-term risk target. Thanks to localization, L-ARC is demonstrated via experiments to produce prediction sets with risk guarantees across different data subpopulations, significantly improving the fairness of the calibrated model for tasks such as image segmentation and beam selection in wireless networks.
Authors: Oswaldo Ludwig
Abstract: This paper introduces a new biologically-inspired training method named Continual Learning through Adjustment Suppression and Sparsity Promotion (CLASSP). CLASSP is based on two main principles observed in neuroscience, particularly in the context of synaptic transmission and Long-Term Potentiation (LTP). The first principle is a decay rate over the weight adjustment, which is implemented as a generalization of the AdaGrad optimization algorithm. This means that weights that have received many updates should have lower learning rates as they likely encode important information about previously seen data. However, this principle results in a diffuse distribution of updates throughout the model, as it promotes updates for weights that haven't been previously updated, while a sparse update distribution is preferred to leave weights unassigned for future tasks. Therefore, the second principle introduces a threshold on the loss gradient. This promotes sparse learning by updating a weight only if the loss gradient with respect to that weight is above a certain threshold, i.e. only updating weights with a significant impact on the current loss. Both principles reflect phenomena observed in LTP, where a threshold effect and a gradual saturation of potentiation have been observed. CLASSP is implemented in a Python/PyTorch class, making it applicable to any model. When compared with Elastic Weight Consolidation (EWC) using Computer Vision and sentiment analysis datasets, CLASSP demonstrates superior performance in terms of accuracy and memory footprint.
Authors: Sihan Zeng, Thinh T. Doan
Abstract: Two-time-scale optimization is a framework introduced in Zeng et al. (2024) that abstracts a range of policy evaluation and policy optimization problems in reinforcement learning (RL). Akin to bi-level optimization under a particular type of stochastic oracle, the two-time-scale optimization framework has an upper level objective whose gradient evaluation depends on the solution of a lower level problem, which is to find the root of a strongly monotone operator. In this work, we propose a new method for solving two-time-scale optimization that achieves significantly faster convergence than the prior arts. The key idea of our approach is to leverage an averaging step to improve the estimates of the operators in both lower and upper levels before using them to update the decision variables. These additional averaging steps eliminate the direct coupling between the main variables, enabling the accelerated performance of our algorithm. We characterize the finite-time convergence rates of the proposed algorithm under various conditions of the underlying objective function, including strong convexity, convexity, Polyak-Lojasiewicz condition, and general non-convexity. These rates significantly improve over the best-known complexity of the standard two-time-scale stochastic approximation algorithm. When applied to RL, we show how the proposed algorithm specializes to novel online sample-based methods that surpass or match the performance of the existing state of the art. Finally, we support our theoretical results with numerical simulations in RL.
Authors: Joongkyu Lee, Min-hwan Oh
Abstract: In this paper, we study the contextual multinomial logit (MNL) bandit problem in which a learning agent sequentially selects an assortment based on contextual information, and user feedback follows an MNL choice model. There has been a significant discrepancy between lower and upper regret bounds, particularly regarding the maximum assortment size $K$. Additionally, the variation in reward structures between these bounds complicates the quest for optimality. Under uniform rewards, where all items have the same expected reward, we establish a regret lower bound of $\Omega(d\sqrt{\smash[b]{T/K}})$ and propose a constant-time algorithm, OFU-MNL+, that achieves a matching upper bound of $\tilde{O}(d\sqrt{\smash[b]{T/K}})$. Under non-uniform rewards, we prove a lower bound of $\Omega(d\sqrt{T})$ and an upper bound of $\tilde{O}(d\sqrt{T})$, also achievable by OFU-MNL+. Our empirical studies support these theoretical findings. To the best of our knowledge, this is the first work in the contextual MNL bandit literature to prove minimax optimality -- for either uniform or non-uniform reward setting -- and to propose a computationally efficient algorithm that achieves this optimality up to logarithmic factors.
Authors: D. Aristoff, M. Johnson, G. Simpson, R. J. Webber
Abstract: In the study of stochastic systems, the committor function describes the probability that a system starting from an initial configuration $x$ will reach a set $B$ before a set $A$. This paper introduces an efficient and interpretable algorithm for approximating the committor, called the "fast committor machine" (FCM). The FCM uses simulated trajectory data to build a kernel-based model of the committor. The kernel function is constructed to emphasize low-dimensional subspaces which optimally describe the $A$ to $B$ transitions. The coefficients in the kernel model are determined using randomized linear algebra, leading to a runtime that scales linearly in the number of data points. In numerical experiments involving a triple-well potential and alanine dipeptide, the FCM yields higher accuracy and trains more quickly than a neural network with the same number of parameters. The FCM is also more interpretable than the neural net.
Authors: Gaoxiang Li, Yu Zhuang
Abstract: Physical Unclonable Functions (PUFs) are emerging as promising security primitives for IoT devices, providing device fingerprints based on physical characteristics. Despite their strengths, PUFs are vulnerable to machine learning (ML) attacks, including conventional and reliability-based attacks. Conventional ML attacks have been effective in revealing vulnerabilities of many PUFs, and reliability-based ML attacks are more powerful tools that have detected vulnerabilities of some PUFs that are resistant to conventional ML attacks. Since reliability-based ML attacks leverage information of PUFs' unreliability, we were tempted to examine the feasibility of building defense using reliability enhancing techniques, and have discovered that majority voting with reasonably high repeats provides effective defense against existing reliability-based ML attack methods. It is known that majority voting reduces but does not eliminate unreliability, we are motivated to investigate if new attack methods exist that can capture the low unreliability of highly but not-perfectly reliable PUFs, which led to the development of a new reliability representation and the new representation-enabled attack method that has experimentally cracked PUFs enhanced with majority voting of high repetitions.
Authors: Christopher Rawles, Sarah Clinckemaillie, Yifan Chang, Jonathan Waltz, Gabrielle Lau, Marybeth Fair, Alice Li, William Bishop, Wei Li, Folawiyo Campbell-Ajala, Daniel Toyama, Robert Berry, Divya Tyamagundlu, Timothy Lillicrap, Oriana Riva
Abstract: Autonomous agents that execute human tasks by controlling computers can enhance human productivity and application accessibility. However, progress in this field will be driven by realistic and reproducible benchmarks. We present AndroidWorld, a fully functional Android environment that provides reward signals for 116 programmatic tasks across 20 real-world Android apps. Unlike existing interactive environments, which provide a static test set, AndroidWorld dynamically constructs tasks that are parameterized and expressed in natural language in unlimited ways, thus enabling testing on a much larger and more realistic suite of tasks. Reward signals are derived from the computer's system state, making them durable across task variations and extensible across different apps. To demonstrate AndroidWorld's benefits and mode of operation, we introduce a new computer control agent, M3A. M3A can complete 30.6% of the AndroidWorld's tasks, leaving ample room for future work. Furthermore, we adapt a popular desktop web agent to work on Android, which we find to be less effective on mobile, suggesting future research is needed to achieve universal, cross-domain agents. Finally, we conduct a robustness analysis by testing M3A against a range of task variations on a representative subset of tasks, demonstrating that variations in task parameters can significantly alter a task's complexity and, consequently, an agent's performance, highlighting the importance of testing agents under diverse conditions. AndroidWorld and the experiments in this paper are available at https://github.com/google-research/android_world.
Authors: Manoj Kumar, Neil Houlsby, Emiel Hoogeboom
Abstract: We investigate the task of adapting image generative models to different datasets without finetuneing. To this end, we introduce Semantica, an image-conditioned diffusion model capable of generating images based on the semantics of a conditioning image. Semantica is trained exclusively on web-scale image pairs, that is it receives a random image from a webpage as conditional input and models another random image from the same webpage. Our experiments highlight the expressivity of pretrained image encoders and necessity of semantic-based data filtering in achieving high-quality image generation. Once trained, it can adaptively generate new images from a dataset by simply using images from that dataset as input. We study the transfer properties of Semantica on ImageNet, LSUN Churches, LSUN Bedroom and SUN397.
Authors: Xiyao Wang, Jiuhai Chen, Zhaoyang Wang, Yuhang Zhou, Yiyang Zhou, Huaxiu Yao, Tianyi Zhou, Tom Goldstein, Parminder Bhatia, Furong Huang, Cao Xiao
Abstract: Large vision-language models (LVLMs) have achieved impressive results in various visual question-answering and reasoning tasks through vision instruction tuning on specific datasets. However, there is still significant room for improvement in the alignment between visual and language modalities. Previous methods to enhance this alignment typically require external models or data, heavily depending on their capabilities and quality, which inevitably sets an upper bound on performance. In this paper, we propose SIMA, a framework that enhances visual and language modality alignment through self-improvement, eliminating the needs for external models or data. SIMA leverages prompts from existing vision instruction tuning datasets to self-generate responses and employs an in-context self-critic mechanism to select response pairs for preference tuning. The key innovation is the introduction of three vision metrics during the in-context self-critic process, which can guide the LVLM in selecting responses that enhance image comprehension. Through experiments across 14 hallucination and comprehensive benchmarks, we demonstrate that SIMA not only improves model performance across all benchmarks but also achieves superior modality alignment, outperforming previous approaches.
Authors: Abhishek Kumar, Robert Morabito, Sanzhar Umbet, Jad Kabbara, Ali Emami
Abstract: As the use of Large Language Models (LLMs) becomes more widespread, understanding their self-evaluation of confidence in generated responses becomes increasingly important as it is integral to the reliability of the output of these models. We introduce the concept of Confidence-Probability Alignment, that connects an LLM's internal confidence, quantified by token probabilities, to the confidence conveyed in the model's response when explicitly asked about its certainty. Using various datasets and prompting techniques that encourage model introspection, we probe the alignment between models' internal and expressed confidence. These techniques encompass using structured evaluation scales to rate confidence, including answer options when prompting, and eliciting the model's confidence level for outputs it does not recognize as its own. Notably, among the models analyzed, OpenAI's GPT-4 showed the strongest confidence-probability alignment, with an average Spearman's $\hat{\rho}$ of 0.42, across a wide range of tasks. Our work contributes to the ongoing efforts to facilitate risk assessment in the application of LLMs and to further our understanding of model trustworthiness.
Authors: Felipe Maia Polo, Ronald Xu, Lucas Weber, M\'irian Silva, Onkar Bhardwaj, Leshem Choshen, Allysson Flavio Melo de Oliveira, Yuekai Sun, Mikhail Yurochkin
Abstract: Most popular benchmarks for comparing LLMs rely on a limited set of prompt templates, which may not fully capture the LLMs' abilities and can affect the reproducibility of results on leaderboards. Many recent works empirically verify prompt sensitivity and advocate for changes in LLM evaluation. In this paper, we consider the problem of estimating the performance distribution across many prompt variants instead of finding a single prompt to evaluate with. We introduce PromptEval, a method for estimating performance across a large set of prompts borrowing strength across prompts and examples to produce accurate estimates under practical evaluation budgets. The resulting distribution can be used to obtain performance quantiles to construct various robust performance metrics (e.g., top 95% quantile or median). We prove that PromptEval consistently estimates the performance distribution and demonstrate its efficacy empirically on three prominent LLM benchmarks: MMLU, BIG-bench Hard, and LMentry. For example, PromptEval can accurately estimate performance quantiles across 100 prompt templates on MMLU with a budget equivalent to two single-prompt evaluations. Our code and data can be found at https://github.com/felipemaiapolo/prompt-eval.
Authors: Tijana Zrnic
Abstract: We introduce PPBoot: a bootstrap-based method for prediction-powered inference. PPBoot is applicable to arbitrary estimation problems and is very simple to implement, essentially only requiring one application of the bootstrap. Through a series of examples, we demonstrate that PPBoot often performs nearly identically to (and sometimes better than) the earlier PPI(++) method based on asymptotic normality$\unicode{x2013}$when the latter is applicable$\unicode{x2013}$without requiring any asymptotic characterizations. Given its versatility, PPBoot could simplify and expand the scope of application of prediction-powered inference to problems where central limit theorems are hard to prove.
Authors: Divya Nori, Wengong Jin
Abstract: The growing significance of RNA engineering in diverse biological applications has spurred interest in developing AI methods for structure-based RNA design. While diffusion models have excelled in protein design, adapting them for RNA presents new challenges due to RNA's conformational flexibility and the computational cost of fine-tuning large structure prediction models. To this end, we propose RNAFlow, a flow matching model for protein-conditioned RNA sequence-structure design. Its denoising network integrates an RNA inverse folding model and a pre-trained RosettaFold2NA network for generation of RNA sequences and structures. The integration of inverse folding in the structure denoising process allows us to simplify training by fixing the structure prediction network. We further enhance the inverse folding model by conditioning it on inferred conformational ensembles to model dynamic RNA conformations. Evaluation on protein-conditioned RNA structure and sequence generation tasks demonstrates RNAFlow's advantage over existing RNA design methods.
Authors: Zekai Liu, Xiaoqi Li, Hongli Peng, Wenkai Li
Abstract: The openness and transparency of Ethereum transaction data make it easy to be exploited by any entities, executing malicious attacks. The sandwich attack manipulates the Automated Market Maker (AMM) mechanism, profiting from manipulating the market price through front or after-running transactions. To identify and prevent sandwich attacks, we propose a cascade classification framework GasTrace. GasTrace analyzes various transaction features to detect malicious accounts, notably through the analysis and modeling of Gas features. In the initial classification, we utilize the Support Vector Machine (SVM) with the Radial Basis Function (RBF) kernel to generate the predicted probabilities of accounts, further constructing a detailed transaction network. Subsequently, the behavior features are captured by the Graph Attention Network (GAT) technique in the second classification. Through cascade classification, GasTrace can analyze and classify the sandwich attacks. Our experimental results demonstrate that GasTrace achieves a remarkable detection and generation capability, performing an accuracy of 96.73% and an F1 score of 95.71% for identifying sandwich attack accounts.
Authors: Yueming Lyu, Kim Yong Tan, Yew Soon Ong, Ivor W. Tsang
Abstract: Diffusion models have demonstrated great potential in generating high-quality content for images, natural language, protein domains, etc. However, how to perform user-preferred targeted generation via diffusion models with only black-box target scores of users remains challenging. To address this issue, we first formulate the fine-tuning of the targeted reserve-time stochastic differential equation (SDE) associated with a pre-trained diffusion model as a sequential black-box optimization problem. Furthermore, we propose a novel covariance-adaptive sequential optimization algorithm to optimize cumulative black-box scores under unknown transition dynamics. Theoretically, we prove a $O(\frac{d^2}{\sqrt{T}})$ convergence rate for cumulative convex functions without smooth and strongly convex assumptions. Empirically, experiments on both numerical test problems and target-guided 3D-molecule generation tasks show the superior performance of our method in achieving better target scores.
Authors: Jean-Marie Lemercier, Simon Rouard, Jade Copet, Yossi Adi, Alexandre D\'efossez
Abstract: Music generation schemes using language modeling rely on a vocabulary of audio tokens, generally provided as codes in a discrete latent space learnt by an auto-encoder. Multi-stage quantizers are often employed to produce these tokens, therefore the decoding strategy used for token prediction must be adapted to account for multiple codebooks: either it should model the joint distribution over all codebooks, or fit the product of the codebook marginal distributions. Modelling the joint distribution requires a costly increase in the number of auto-regressive steps, while fitting the product of the marginals yields an inexact model unless the codebooks are mutually independent. In this work, we introduce an independence-promoting loss to regularize the auto-encoder used as the tokenizer in language models for music generation. The proposed loss is a proxy for mutual information based on the maximum mean discrepancy principle, applied in reproducible kernel Hilbert spaces. Our criterion is simple to implement and train, and it is generalizable to other multi-stream codecs. We show that it reduces the statistical dependence between codebooks during auto-encoding. This leads to an increase in the generated music quality when modelling the product of the marginal distributions, while generating audio much faster than the joint distribution model.
Authors: Amelia Kawasaki, Andrew Davis, Houssam Abbas
Abstract: The widespread adoption of Large Language Models (LLMs), exemplified by OpenAI's ChatGPT, brings to the forefront the imperative to defend against adversarial threats on these models. These attacks, which manipulate an LLM's output by introducing malicious inputs, undermine the model's integrity and the trust users place in its outputs. In response to this challenge, our paper presents an innovative defensive strategy, given white box access to an LLM, that harnesses residual activation analysis between transformer layers of the LLM. We apply a novel methodology for analyzing distinctive activation patterns in the residual streams for attack prompt classification. We curate multiple datasets to demonstrate how this method of classification has high accuracy across multiple types of attack scenarios, including our newly-created attack dataset. Furthermore, we enhance the model's resilience by integrating safety fine-tuning techniques for LLMs in order to measure its effect on our capability to detect attacks. The results underscore the effectiveness of our approach in enhancing the detection and mitigation of adversarial inputs, advancing the security framework within which LLMs operate.
Authors: Harshvardhan Mestha, Tejas Agrawal, Karan Bania, Shreyas V, Yash Bhisikar
Abstract: Large vision-language models (VLMs) are shown to learn rich joint image-text representations enabling high performances in relevant downstream tasks. However, they fail to showcase their quantitative understanding of objects, and they lack good counting-aware representation. This paper conducts a reproducibility study of 'Teaching CLIP to Count to Ten' (Paiss et al., 2023), which presents a method to finetune a CLIP model (Radford et al., 2021) to improve zero-shot counting accuracy in an image while maintaining the performance for zero-shot classification by introducing a counting-contrastive loss term. We improve the model's performance on a smaller subset of their training data with lower computational resources. We verify these claims by reproducing their study with our own code. The implementation can be found at https://github.com/SforAiDl/CountCLIP.
Authors: Tom Huix, Anna Korba, Alain Durmus, Eric Moulines
Abstract: Variational inference (VI) is a popular approach in Bayesian inference, that looks for the best approximation of the posterior distribution within a parametric family, minimizing a loss that is typically the (reverse) Kullback-Leibler (KL) divergence. Despite its empirical success, the theoretical properties of VI have only received attention recently, and mostly when the parametric family is the one of Gaussians. This work aims to contribute to the theoretical study of VI in the non-Gaussian case by investigating the setting of Mixture of Gaussians with fixed covariance and constant weights. In this view, VI over this specific family can be casted as the minimization of a Mollified relative entropy, i.e. the KL between the convolution (with respect to a Gaussian kernel) of an atomic measure supported on Diracs, and the target distribution. The support of the atomic measure corresponds to the localization of the Gaussian components. Hence, solving variational inference becomes equivalent to optimizing the positions of the Diracs (the particles), which can be done through gradient descent and takes the form of an interacting particle system. We study two sources of error of variational inference in this context when optimizing the mollified relative entropy. The first one is an optimization result, that is a descent lemma establishing that the algorithm decreases the objective at each iteration. The second one is an approximation error, that upper bounds the objective between an optimal finite mixture and the target distribution.
Authors: Jiaoda Li, Yifan Hou, Mrinmaya Sachan, Ryan Cotterell
Abstract: Large language models (LLMs) exhibit an intriguing ability to learn a novel task from in-context examples presented in a demonstration, termed in-context learning (ICL). Understandably, a swath of research has been dedicated to uncovering the theories underpinning ICL. One popular hypothesis explains ICL by task selection. LLMs identify the task based on the demonstration and generalize it to the prompt. Another popular hypothesis is that ICL is a form of meta-learning, i.e., the models learn a learning algorithm at pre-training time and apply it to the demonstration. Finally, a third hypothesis argues that LLMs use the demonstration to select a composition of tasks learned during pre-training to perform ICL. In this paper, we empirically explore these three hypotheses that explain LLMs' ability to learn in context with a suite of experiments derived from common text classification tasks. We invalidate the first two hypotheses with counterexamples and provide evidence in support of the last hypothesis. Our results suggest an LLM could learn a novel task in context via composing tasks learned during pre-training.