Authors: Weilin Cai, Juyong Jiang, Fan Wang, Jing Tang, Sunghun Kim, Jiayi Huang
Abstract: Large language models (LLMs) have garnered unprecedented advancements across diverse fields, ranging from natural language processing to computer vision and beyond. The prowess of LLMs is underpinned by their substantial model size, extensive and diverse datasets, and the vast computational power harnessed during training, all of which contribute to the emergent abilities of LLMs (e.g., in-context learning) that are not present in small models. Within this context, the mixture of experts (MoE) has emerged as an effective method for substantially scaling up model capacity with minimal computation overhead, gaining significant attention from academia and industry. Despite its growing prevalence, there lacks a systematic and comprehensive review of the literature on MoE. This survey seeks to bridge that gap, serving as an essential resource for researchers delving into the intricacies of MoE. We first briefly introduce the structure of the MoE layer, followed by proposing a new taxonomy of MoE. Next, we overview the core designs for various MoE models including both algorithmic and systemic aspects, alongside collections of available open-source implementations, hyperparameter configurations and empirical evaluations. Furthermore, we delineate the multifaceted applications of MoE in practice, and outline some potential directions for future research. To facilitate ongoing updates and the sharing of cutting-edge developments in MoE research, we have established a resource repository accessible at https://github.com/withinmiaov/A-Survey-on-Mixture-of-Experts.
URLs: https://github.com/withinmiaov/A-Survey-on-Mixture-of-Experts.
Authors: Ximing Wen, Rosina O. Weber, Anik Sen, Darryl Hannan, Steven C. Nesbit, Vincent Chan, Alberto Goffi, Michael Morris, John C. Hunninghake, Nicholas E. Villalobos, Edward Kim, Christopher J. MacLellan
Abstract: Point-of-Care Ultrasound (POCUS) is the practice of clinicians conducting and interpreting ultrasound scans right at the patient's bedside. However, the expertise needed to interpret these images is considerable and may not always be present in emergency situations. This reality makes algorithms such as machine learning classifiers extremely valuable to augment human decisions. POCUS devices are becoming available at a reasonable cost in the size of a mobile phone. The challenge of turning POCUS devices into life-saving tools is that interpretation of ultrasound images requires specialist training and experience. Unfortunately, the difficulty to obtain positive training images represents an important obstacle to building efficient and accurate classifiers. Hence, the problem we try to investigate is how to explore strategies to increase accuracy of classifiers trained with scarce data. We hypothesize that training with a few data instances may not suffice for classifiers to generalize causing them to overfit. Our approach uses an Explainable AI-Augmented approach to help the algorithm learn more from less and potentially help the classifier better generalize.
Authors: Varun Madhavan, Amal S Sebastian, Bharath Ramsundar, Venkatasubramanian Viswanathan
Abstract: In this work, we describe a novel approach to building a neural PDE solver leveraging recent advances in transformer based neural network architectures. Our model can provide solutions for different values of PDE parameters without any need for retraining the network. The training is carried out in a self-supervised manner, similar to pretraining approaches applied in language and vision tasks. We hypothesize that the model is in effect learning a family of operators (for multiple parameters) mapping the initial condition to the solution of the PDE at any future time step t. We compare this approach with the Fourier Neural Operator (FNO), and demonstrate that it can generalize over the space of PDE parameters, despite having a higher prediction error for individual parameter values compared to the FNO. We show that performance on a specific parameter can be improved by finetuning the model with very small amounts of data. We also demonstrate that the model scales with data as well as model size.
Authors: Luuk Gubbels, Marco Puts, Piet Daas
Abstract: Online platform businesses can be identified by using web-scraped texts. This is a classification problem that combines elements of natural language processing and rare event detection. Because online platforms are rare, accurately identifying them with Machine Learning algorithms is challenging. Here, we describe the development of a Machine Learning-based text classification approach that reduces the number of false positives as much as possible. It greatly reduces the bias in the estimates obtained by using calibrated probabilities and ensembles.
Authors: Pawe{\l} \'Swider
Abstract: One of the most crucial tasks in the future will be to understand what is going on in neural networks, as they will become even more powerful and widely deployed. This work aims to use TDA methods to analyze neural representations. We develop methods for analyzing representations from different architectures and check how one should use them to obtain valid results. Our findings indicate that removing outliers does not have much impact on the results and that we should compare representations with the same number of elements. We applied these methods for ResNet, VGG19, and ViT architectures and found substantial differences along with some similarities. Additionally, we determined that models with similar architecture tend to have a similar topology of representations and models with a larger number of layers change their topology more smoothly. Furthermore, we found that the topology of pre-trained and finetuned models starts to differ in the middle and final layers while remaining quite similar in the initial layers. These findings demonstrate the efficacy of TDA in the analysis of neural network behavior.
Authors: Daniel Marczak, Bart{\l}omiej Twardowski, Tomasz Trzci\'nski, Sebastian Cygert
Abstract: This paper introduces a continual learning approach named MagMax, which utilizes model merging to enable large pre-trained models to continuously learn from new data without forgetting previously acquired knowledge. Distinct from traditional continual learning methods that aim to reduce forgetting during task training, MagMax combines sequential fine-tuning with a maximum magnitude weight selection for effective knowledge integration across tasks. Our initial contribution is an extensive examination of model merging techniques, revealing that simple approaches like weight averaging and random weight selection surprisingly hold up well in various continual learning contexts. More importantly, we present MagMax, a novel model-merging strategy that enables continual learning of large pre-trained models for successive tasks. Our thorough evaluation demonstrates the superiority of MagMax in various scenarios, including class- and domain-incremental learning settings.
Authors: Luca Zancato, Arjun Seshadri, Yonatan Dukler, Aditya Golatkar, Yantao Shen, Benjamin Bowman, Matthew Trager, Alessandro Achille, Stefano Soatto
Abstract: We describe a family of architectures to support transductive inference by allowing memory to grow to a finite but a-priori unknown bound while making efficient use of finite resources for inference. Current architectures use such resources to represent data either eidetically over a finite span ("context" in Transformers), or fading over an infinite span (in State Space Models, or SSMs). Recent hybrid architectures have combined eidetic and fading memory, but with limitations that do not allow the designer or the learning process to seamlessly modulate the two, nor to extend the eidetic memory span. We leverage ideas from Stochastic Realization Theory to develop a class of models called B'MOJO to seamlessly combine eidetic and fading memory within an elementary composable module. The overall architecture can be used to implement models that can access short-term eidetic memory "in-context," permanent structural memory "in-weights," fading memory "in-state," and long-term eidetic memory "in-storage" by natively incorporating retrieval from an asynchronously updated memory. We show that Transformers, existing SSMs such as Mamba, and hybrid architectures such as Jamba are special cases of B'MOJO and describe a basic implementation, to be open sourced, that can be stacked and scaled efficiently in hardware. We test B'MOJO on transductive inference tasks, such as associative recall, where it outperforms existing SSMs and Hybrid models; as a baseline, we test ordinary language modeling where B'MOJO achieves perplexity comparable to similarly-sized Transformers and SSMs up to 1.4B parameters, while being up to 10% faster to train. Finally, we show that B'MOJO's ability to modulate eidetic and fading memory results in better inference on longer sequences tested up to 32K tokens, four-fold the length of the longest sequences seen during training.
Authors: Jeremy Carleton, Prathik Vijaykumar, Divyanshu Saxena, Dheeraj Narasimha, Srinivas Shakkottai, Aditya Akella
Abstract: We address the challenge of online convex optimization where the objective function's gradient exhibits sparsity, indicating that only a small number of dimensions possess non-zero gradients. Our aim is to leverage this sparsity to obtain useful estimates of the objective function's gradient even when the only information available is a limited number of function samples. Our motivation stems from distributed queueing systems like microservices-based applications, characterized by request-response workloads. Here, each request type proceeds through a sequence of microservices to produce a response, and the resource allocation across the collection of microservices is controlled to balance end-to-end latency with resource costs. While the number of microservices is substantial, the latency function primarily reacts to resource changes in a few, rendering the gradient sparse. Our proposed method, CONGO (Compressive Online Gradient Optimization), combines simultaneous perturbation with compressive sensing to estimate gradients. We establish analytical bounds on the requisite number of compressive sensing samples per iteration to maintain bounded bias of gradient estimates, ensuring sub-linear regret. By exploiting sparsity, we reduce the samples required per iteration to match the gradient's sparsity, rather than the problem's original dimensionality. Numerical experiments and real-world microservices benchmarks demonstrate CONGO's superiority over multiple stochastic gradient descent approaches, as it quickly converges to performance comparable to policies pre-trained with workload awareness.
Authors: Xihong Su, Marek Petrik
Abstract: Multi-model Markov decision process (MMDP) is a promising framework for computing policies that are robust to parameter uncertainty in MDPs. MMDPs aim to find a policy that maximizes the expected return over a distribution of MDP models. Because MMDPs are NP-hard to solve, most methods resort to approximations. In this paper, we derive the policy gradient of MMDPs and propose CADP, which combines a coordinate ascent method and a dynamic programming algorithm for solving MMDPs. The main innovation of CADP compared with earlier algorithms is to take the coordinate ascent perspective to adjust model weights iteratively to guarantee monotone policy improvements to a local maximum. A theoretical analysis of CADP proves that it never performs worse than previous dynamic programming algorithms like WSU. Our numerical results indicate that CADP substantially outperforms existing methods on several benchmark problems.
Authors: Kwanghyuk Park, Xinjuan Chen, Dongjin Lee, Jiaxi Gu, Jae-Hun Jung
Abstract: In this paper, we introduce the finite difference weighted essentially non-oscillatory (WENO) scheme based on the neural network for hyperbolic conservation laws. We employ the supervised learning and design two loss functions, one with the mean squared error and the other with the mean squared logarithmic error, where the WENO3-JS weights are computed as the labels. Each loss function consists of two components where the first component compares the difference between the weights from the neural network and WENO3-JS weights, while the second component matches the output weights of the neural network and the linear weights. The former of the loss function enforces the neural network to follow the WENO properties, implying that there is no need for the post-processing layer. Additionally the latter leads to better performance around discontinuities. As a neural network structure, we choose the shallow neural network (SNN) for computational efficiency with the Delta layer consisting of the normalized undivided differences. These constructed WENO3-SNN schemes show the outperformed results in one-dimensional examples and improved behavior in two-dimensional examples, compared with the simulations from WENO3-JS and WENO3-Z.
Authors: Fred Lu, Ryan R. Curtin, Edward Raff, Francis Ferraro, James Holt
Abstract: As the size of datasets used in statistical learning continues to grow, distributed training of models has attracted increasing attention. These methods partition the data and exploit parallelism to reduce memory and runtime, but suffer increasingly from communication costs as the data size or the number of iterations grows. Recent work on linear models has shown that a surrogate likelihood can be optimized locally to iteratively improve on an initial solution in a communication-efficient manner. However, existing versions of these methods experience multiple shortcomings as the data size becomes massive, including diverging updates and efficiently handling sparsity. In this work we develop solutions to these problems which enable us to learn a communication-efficient distributed logistic regression model even beyond millions of features. In our experiments we demonstrate a large improvement in accuracy over distributed algorithms with only a few distributed update steps needed, and similar or faster runtimes. Our code is available at \url{https://github.com/FutureComputing4AI/ProxCSL}.
Authors: Matthew Lau, Haoran Wang, Alec Helbling, Matthew Hul, ShengYun Peng, Martin Andreoni, Willian T. Lunardi, Wenke Lee
Abstract: The robustness of machine learning models has been questioned by the existence of adversarial examples. We examine the threat of adversarial examples in practical applications that require lightweight models for one-class classification. Building on Ilyas et al. (2019), we investigate the vulnerability of lightweight one-class classifiers to adversarial attacks and possible reasons for it. Our results show that lightweight one-class classifiers learn features that are not robust (e.g. texture) under stronger attacks. However, unlike in multi-class classification (Ilyas et al., 2019), these non-robust features are not always useful for the one-class task, suggesting that learning these unpredictive and non-robust features is an unwanted consequence of training.
Authors: Adriano Hernandez
Abstract: Large language models (LLMs) sometimes exhibit dangerous unintended behaviors. Finding and fixing these is challenging because the attack surface is massive -- it is not tractable to exhaustively search for all possible inputs that may elicit such behavior. One specific and particularly challenging case is that if data-poisoning-injected trojans, since there is no way to know what they are to search for them. To our knowledge, there is no generally applicable method to unlearn unknown trojans injected during pre-training. This work seeks to provide a general purpose recipe (filters) and a specific implementation (LoRA) filters that work in practice on small to medium sized models. The focus is primarily empirical, though some perplexing behavior opens the door to the fundamental question of how LLMs store and process information. Not unexpectedly, we find that our filters work best on the residual stream and the latest layers.
Authors: Arinbjorn Kolbeinsson, Kyle O'Brien, Tianjin Huang, Shanghua Gao, Shiwei Liu, Jonathan Richard Schwarz, Anurag Vaidya, Faisal Mahmood, Marinka Zitnik, Tianlong Chen, Thomas Hartvigsen
Abstract: Test-time interventions for language models can enhance factual accuracy, mitigate harmful outputs, and improve model efficiency without costly retraining. But despite a flood of new methods, different types of interventions are largely developing independently. In practice, multiple interventions must be applied sequentially to the same model, yet we lack standardized ways to study how interventions interact. We fill this gap by introducing composable interventions, a framework to study the effects of using multiple interventions on the same language models, featuring new metrics and a unified codebase. Using our framework, we conduct extensive experiments and compose popular methods from three emerging intervention categories -- Knowledge Editing, Model Compression, and Machine Unlearning. Our results from 310 different compositions uncover meaningful interactions: compression hinders editing and unlearning, composing interventions hinges on their order of application, and popular general-purpose metrics are inadequate for assessing composability. Taken together, our findings showcase clear gaps in composability, suggesting a need for new multi-objective interventions. All of our code is public: https://github.com/hartvigsen-group/composable-interventions.
URLs: https://github.com/hartvigsen-group/composable-interventions.
Authors: Yan Liu, Bin Guo, Nuo Li, Yasan Ding, Zhouyangzi Zhang, Zhiwen Yu
Abstract: Artificial Intelligence of Things (AIoT) is an emerging frontier based on the deep fusion of Internet of Things (IoT) and Artificial Intelligence (AI) technologies. Although advanced deep learning techniques enhance the efficient data processing and intelligent analysis of complex IoT data, they still suffer from notable challenges when deployed to practical AIoT applications, such as constrained resources, and diverse task requirements. Knowledge transfer is an effective method to enhance learning performance by avoiding the exorbitant costs associated with data recollection and model retraining. Notably, although there are already some valuable and impressive surveys on transfer learning, these surveys introduce approaches in a relatively isolated way and lack the recent advances of various knowledge transfer techniques for AIoT field. This survey endeavors to introduce a new concept of knowledge transfer, referred to as Crowd Knowledge Transfer (CrowdTransfer), which aims to transfer prior knowledge learned from a crowd of agents to reduce the training cost and as well as improve the performance of the model in real-world complicated scenarios. Particularly, we present four transfer modes from the perspective of crowd intelligence, including derivation, sharing, evolution and fusion modes. Building upon conventional transfer learning methods, we further delve into advanced crowd knowledge transfer models from three perspectives for various AIoT applications. Furthermore, we explore some applications of AIoT areas, such as human activity recognition, urban computing, multi-robot system, and smart factory. Finally, we discuss the open issues and outline future research directions of knowledge transfer in AIoT community.
Authors: Long Wei, Peiyan Hu, Ruiqi Feng, Haodong Feng, Yixuan Du, Tao Zhang, Rui Wang, Yue Wang, Zhi-Ming Ma, Tailin Wu
Abstract: Controlling the evolution of complex physical systems is a fundamental task across science and engineering. Classical techniques suffer from limited applicability or huge computational costs. On the other hand, recent deep learning and reinforcement learning-based approaches often struggle to optimize long-term control sequences under the constraints of system dynamics. In this work, we introduce Diffusion Physical systems Control (DiffPhyCon), a new class of method to address the physical systems control problem. DiffPhyCon excels by simultaneously minimizing both the learned generative energy function and the predefined control objectives across the entire trajectory and control sequence. Thus, it can explore globally and identify near-optimal control sequences. Moreover, we enhance DiffPhyCon with prior reweighting, enabling the discovery of control sequences that significantly deviate from the training distribution. We test our method in 1D Burgers' equation and 2D jellyfish movement control in a fluid environment. Our method outperforms widely applied classical approaches and state-of-the-art deep learning and reinforcement learning methods. Notably, DiffPhyCon unveils an intriguing fast-close-slow-open pattern observed in the jellyfish, aligning with established findings in the field of fluid dynamics.
Authors: Meenatchi Sundaram Muthu Selva Annamalai
Abstract: Differentially Private Stochastic Gradient Descent (DP-SGD) is a popular iterative algorithm used to train machine learning models while formally guaranteeing the privacy of users. However the privacy analysis of DP-SGD makes the unrealistic assumption that all intermediate iterates (aka internal state) of the algorithm are released since in practice, only the final trained model, i.e., the final iterate of the algorithm is released. In this hidden state setting, prior work has provided tighter analyses, albeit only when the loss function is constrained, e.g., strongly convex and smooth or linear. On the other hand, the privacy leakage observed empirically from hidden state DP-SGD, even when using non-convex loss functions suggest that there is in fact a gap between the theoretical privacy analysis and the privacy guarantees achieved in practice. Therefore, it remains an open question whether privacy amplification for DP-SGD is possible in the hidden state setting for general loss functions. Unfortunately, this work answers the aforementioned research question negatively. By carefully constructing a loss function for DP-SGD, we show that for specific loss functions, the final iterate of DP-SGD alone leaks as much information as the sequence of all iterates combined. Furthermore, we empirically verify this result by evaluating the privacy leakage from the final iterate of DP-SGD with our loss function and show that this matches the theoretical upper bound guaranteed by DP exactly. Therefore, we show that the current privacy analysis fo DP-SGD is tight for general loss functions and conclude that no privacy amplification is possible for DP-SGD in general for all (possibly non-convex) loss functions.
Authors: Guojian Wang, Faguo Wu, Xiao Zhang, Tianyuan Chen, Xuyang Chen, Lin Zhao
Abstract: In this paper, we investigate preference-based reinforcement learning (PbRL) that allows reinforcement learning (RL) agents to learn from human feedback. This is particularly valuable when defining a fine-grain reward function is not feasible. However, this approach is inefficient and impractical for promoting deep exploration in hard-exploration tasks with long horizons and sparse rewards. To tackle this issue, we introduce LOPE: Learning Online with trajectory Preference guidancE, an end-to-end preference-guided RL framework that enhances exploration efficiency in hard-exploration tasks. Our intuition is that LOPE directly adjusts the focus of online exploration by considering human feedback as guidance, avoiding learning a separate reward model from preferences. Specifically, LOPE includes a two-step sequential policy optimization process consisting of trust-region-based policy improvement and preference guidance steps. We reformulate preference guidance as a novel trajectory-wise state marginal matching problem that minimizes the maximum mean discrepancy distance between the preferred trajectories and the learned policy. Furthermore, we provide a theoretical analysis to characterize the performance improvement bound and evaluate the LOPE's effectiveness. When assessed in various challenging hard-exploration environments, LOPE outperforms several state-of-the-art methods regarding convergence rate and overall performance. The code used in this study is available at \url{https://github.com/buaawgj/LOPE}.
Authors: Leye Zhang, Xiangxiang Tian, Chengli Zhang, Hongjun Zhang
Abstract: Deep Q-network algorithm is used to select economic span of bridge. Selection of bridge span has a significant impact on the total cost of bridge, and a reasonable selection of span can reduce engineering cost. Economic span of bridge is theoretically analyzed, and the theoretical solution formula of economic span is deduced. Construction process of bridge simulation environment is described in detail, including observation space, action space and reward function of the environment. Agent is constructed, convolutional neural network is used to approximate Q function,{\epsilon} greedy policy is used for action selection, and experience replay is used for training. The test verifies that the agent can successfully learn optimal policy and realize economic span selection of bridge. This study provides a potential decision-making tool for bridge design.
Authors: Maoxin Ji, Qiong Wu, Pingyi Fan, Nan Cheng, Wen Chen, Jiangzhou Wang, Khaled B. Letaief
Abstract: In the rapidly evolving landscape of Internet of Vehicles (IoV) technology, Cellular Vehicle-to-Everything (C-V2X) communication has attracted much attention due to its superior performance in coverage, latency, and throughput. Resource allocation within C-V2X is crucial for ensuring the transmission of safety information and meeting the stringent requirements for ultra-low latency and high reliability in Vehicle-to-Vehicle (V2V) communication. This paper proposes a method that integrates Graph Neural Networks (GNN) with Deep Reinforcement Learning (DRL) to address this challenge. By constructing a dynamic graph with communication links as nodes and employing the Graph Sample and Aggregation (GraphSAGE) model to adapt to changes in graph structure, the model aims to ensure a high success rate for V2V communication while minimizing interference on Vehicle-to-Infrastructure (V2I) links, thereby ensuring the successful transmission of V2V link information and maintaining high transmission rates for V2I links. The proposed method retains the global feature learning capabilities of GNN and supports distributed network deployment, allowing vehicles to extract low-dimensional features that include structural information from the graph network based on local observations and to make independent resource allocation decisions. Simulation results indicate that the introduction of GNN, with a modest increase in computational load, effectively enhances the decision-making quality of agents, demonstrating superiority to other methods. This study not only provides a theoretically efficient resource allocation strategy for V2V and V2I communications but also paves a new technical path for resource management in practical IoV environments.
Authors: Yu Cheng, Junjie Guo, Shiqing Long, You Wu, Mengfang Sun, Rong Zhang
Abstract: The innovative GNN-CL model proposed in this paper marks a breakthrough in the field of financial fraud detection by synergistically combining the advantages of graph neural networks (gnn), convolutional neural networks (cnn) and long short-term memory (LSTM) networks. This convergence enables multifaceted analysis of complex transaction patterns, improving detection accuracy and resilience against complex fraudulent activities. A key novelty of this paper is the use of multilayer perceptrons (MLPS) to estimate node similarity, effectively filtering out neighborhood noise that can lead to false positives. This intelligent purification mechanism ensures that only the most relevant information is considered, thereby improving the model's understanding of the network structure. Feature weakening often plagues graph-based models due to the dilution of key signals. In order to further address the challenge of feature weakening, GNN-CL adopts reinforcement learning strategies. By dynamically adjusting the weights assigned to central nodes, it reinforces the importance of these influential entities to retain important clues of fraud even in less informative data. Experimental evaluations on Yelp datasets show that the results highlight the superior performance of GNN-CL compared to existing methods.
Authors: Rachneet Kaur, Zhen Zeng, Tucker Balch, Manuela Veloso
Abstract: Recent advancements in language modeling have shown promising results when applied to time series data. In particular, fine-tuning pre-trained large language models (LLMs) for time series classification tasks has achieved state-of-the-art (SOTA) performance on standard benchmarks. However, these LLM-based models have a significant drawback due to the large model size, with the number of trainable parameters in the millions. In this paper, we propose an alternative approach to leveraging the success of language modeling in the time series domain. Instead of fine-tuning LLMs, we utilize a language embedding model to embed time series and then pair the embeddings with a simple classification head composed of convolutional neural networks (CNN) and multilayer perceptron (MLP). We conducted extensive experiments on well-established time series classification benchmark datasets. We demonstrated LETS-C not only outperforms the current SOTA in classification accuracy but also offers a lightweight solution, using only 14.5% of the trainable parameters on average compared to the SOTA model. Our findings suggest that leveraging language encoders to embed time series data, combined with a simple yet effective classification head, offers a promising direction for achieving high-performance time series classification while maintaining a lightweight model architecture.
Authors: Christofer Fellicious, Sahib Julka, Lorenz Wendlinger, Michael Granitzer
Abstract: In real-world applications, input data distributions are rarely static over a period of time, a phenomenon known as concept drift. Such concept drifts degrade the model's prediction performance, and therefore we require methods to overcome these issues. The initial step is to identify concept drifts and have a training method in place to recover the model's performance. Most concept drift detection methods work on detecting concept drifts and signalling the requirement to retrain the model. However, in real-world cases, there could be concept drifts that recur over a period of time. In this paper, we present an unsupervised method based on Generative Adversarial Networks(GAN) to detect concept drifts and identify whether a specific concept drift occurred in the past. Our method reduces the time and data the model requires to get up to speed for recurring drifts. Our key results indicate that our proposed model can outperform the current state-of-the-art models in most datasets. We also test our method on a real-world use case from astrophysics, where we detect the bow shock and magnetopause crossings with better results than the existing methods in the domain.
Authors: Xin Xu, Eibe Frank, Geoffrey Holmes
Abstract: We explore multiple-instance verification, a problem setting where a query instance is verified against a bag of target instances with heterogeneous, unknown relevancy. We show that naive adaptations of attention-based multiple instance learning (MIL) methods and standard verification methods like Siamese neural networks are unsuitable for this setting: directly combining state-of-the-art (SOTA) MIL methods and Siamese networks is shown to be no better, and sometimes significantly worse, than a simple baseline model. Postulating that this may be caused by the failure of the representation of the target bag to incorporate the query instance, we introduce a new pooling approach named ``cross-attention pooling'' (CAP). Under the CAP framework, we propose two novel attention functions to address the challenge of distinguishing between highly similar instances in a target bag. Through empirical studies on three different verification tasks, we demonstrate that CAP outperforms adaptations of SOTA MIL methods and the baseline by substantial margins, in terms of both classification accuracy and quality of the explanations provided for the classifications. Ablation studies confirm the superior ability of the new attention functions to identify key instances.
Authors: Mingjia Yin, Chuhan Wu, Yufei Wang, Hao Wang, Wei Guo, Yasheng Wang, Yong Liu, Ruiming Tang, Defu Lian, Enhong Chen
Abstract: Data is the cornerstone of large language models (LLMs), but not all data is useful for model learning. Carefully selected data can better elicit the capabilities of LLMs with much less computational overhead. Most methods concentrate on evaluating the quality of individual samples in data selection, while the combinatorial effects among samples are neglected. Even if each sample is of perfect quality, their combinations may be suboptimal in teaching LLMs due to their intrinsic homogeneity or contradiction. In this paper, we aim to uncover the underlying relationships between LLM performance and data selection. Inspired by the information compression nature of LLMs, we uncover an ``entropy law'' that connects LLM performance with data compression ratio and first-epoch training loss, which reflect the information redundancy of a dataset and the mastery of inherent knowledge encoded in this dataset, respectively. Through both theoretical deduction and empirical evaluation, we find that model performance is negatively correlated to the compression ratio of training data, which usually yields a lower training loss. Based on the findings of the entropy law, we propose a quite efficient and universal data selection method named \textbf{ZIP} for training LLMs, which aim to prioritize data subsets exhibiting a low compression ratio. Based on a multi-stage algorithm that selects diverse data in a greedy manner, we can obtain a good data subset with satisfactory diversity. Extensive experiments have been conducted to validate the entropy law and the superiority of ZIP across different LLM backbones and alignment stages. We also present an interesting application of entropy law that can detect potential performance risks at the beginning of model training.
Authors: Fabio Valerio Massoli, Christos Louizos, Arash Behboodi
Abstract: Compressed sensing combines the power of convex optimization techniques with a sparsity-inducing prior on the signal space to solve an underdetermined system of equations. For many problems, the sparsifying dictionary is not directly given, nor its existence can be assumed. Besides, the sensing matrix can change across different scenarios. Addressing these issues requires solving a sparse representation learning problem, namely dictionary learning, taking into account the epistemic uncertainty of the learned dictionaries and, finally, jointly learning sparse representations and reconstructions under varying sensing matrix conditions. We address both concerns by proposing a variant of the LISTA architecture. First, we introduce Augmented Dictionary Learning ISTA (A-DLISTA), which incorporates an augmentation module to adapt parameters to the current measurement setup. Then, we propose to learn a distribution over dictionaries via a variational approach, dubbed Variational Learning ISTA (VLISTA). VLISTA exploits A-DLISTA as the likelihood model and approximates a posterior distribution over the dictionaries as part of an unfolded LISTA-based recovery algorithm. As a result, VLISTA provides a probabilistic way to jointly learn the dictionary distribution and the reconstruction algorithm with varying sensing matrices. We provide theoretical and experimental support for our architecture and show that our model learns calibrated uncertainties.
Authors: Gyeong Taek Lee, Oh-Ran Kwon
Abstract: In the manufacturing process, sensor data collected from equipment is crucial for building predictive models to manage processes and improve productivity. However, in the field, it is challenging to gather sufficient data to build robust models. This study proposes a novel predictive model based on the Transformer, utilizing statistical feature embedding and window positional encoding. Statistical features provide an effective representation of sensor data, and the embedding enables the Transformer to learn both time- and sensor-related information. Window positional encoding captures precise time details from the feature embedding. The model's performance is evaluated in two problems: fault detection and virtual metrology, showing superior results compared to baseline models. This improvement is attributed to the efficient use of parameters, which is particularly beneficial for sensor data that often has limited sample sizes. The results support the model's applicability across various manufacturing industries, demonstrating its potential for enhancing process management and yield.
Authors: Guillermo Infante, Anders Jonsson, Vicen\c{c} G\'omez
Abstract: We introduce a novel approach to hierarchical reinforcement learning for Linearly-solvable Markov Decision Processes (LMDPs) in the infinite-horizon average-reward setting. Unlike previous work, our approach allows learning low-level and high-level tasks simultaneously, without imposing limiting restrictions on the low-level tasks. Our method relies on partitions of the state space that create smaller subtasks that are easier to solve, and the equivalence between such partitions to learn more efficiently. We then exploit the compositionality of low-level tasks to exactly represent the value function of the high-level task. Experiments show that our approach can outperform flat average-reward reinforcement learning by one or several orders of magnitude.
Authors: Long H. Pham, Jun Sun
Abstract: On the one hand, there has been considerable progress on neural network verification in recent years, which makes certifying neural networks a possibility. On the other hand, neural networks in practice are often re-trained over time to cope with new data distribution or for solving different tasks (a.k.a. continual learning). Once re-trained, the verified correctness of the neural network is likely broken, particularly in the presence of the phenomenon known as catastrophic forgetting. In this work, we propose an approach called certified continual learning which improves existing continual learning methods by preserving, as long as possible, the established correctness properties of a verified network. Our approach is evaluated with multiple neural networks and on two different continual learning methods. The results show that our approach is efficient and the trained models preserve their certified correctness and often maintain high utility.
Authors: Zitai Wang, Qianqian Xu, Zhiyong Yang, Peisong Wen, Yuan He, Xiaochun Cao, Qingming Huang
Abstract: Multi-label ranking, which returns multiple top-ranked labels for each instance, has a wide range of applications for visual tasks. Due to its complicated setting, prior arts have proposed various measures to evaluate model performances. However, both theoretical analysis and empirical observations show that a model might perform inconsistently on different measures. To bridge this gap, this paper proposes a novel measure named Top-K Pairwise Ranking (TKPR), and a series of analyses show that TKPR is compatible with existing ranking-based measures. In light of this, we further establish an empirical surrogate risk minimization framework for TKPR. On one hand, the proposed framework enjoys convex surrogate losses with the theoretical support of Fisher consistency. On the other hand, we establish a sharp generalization bound for the proposed framework based on a novel technique named data-dependent contraction. Finally, empirical results on benchmark datasets validate the effectiveness of the proposed framework.
Authors: Arsenii Mustafin, Aleksei Pakharev, Alex Olshevsky, Ioannis Ch. Paschalidis
Abstract: Markov Decision Process (MDP) is a common mathematical model for sequential decision-making problems. In this paper, we present a new geometric interpretation of MDP, which is useful for analyzing the dynamics of main MDP algorithms. Based on this interpretation, we demonstrate that MDPs can be split into equivalence classes with indistinguishable algorithm dynamics. The related normalization procedure allows for the design of a new class of MDP-solving algorithms that find optimal policies without computing policy values.
Authors: \'Alvaro Fern\'andez-Campa-Gonz\'alez, Jorge Paz-Ruza, Amparo Alonso-Betanzos, Bertha Guijarro-Berdi\~nas
Abstract: Among the existing approaches for visual-based Recommender System (RS) explainability, utilizing user-uploaded item images as efficient, trustable explanations is a promising option. However, current models following this paradigm assume that, for any user, all images uploaded by other users can be considered negative training examples (i.e. bad explanatory images), an inadvertedly naive labelling assumption that contradicts the rationale of the approach. This work proposes a new explainer training pipeline by leveraging Positive-Unlabelled (PU) Learning techniques to train image-based explainer with refined subsets of reliable negative examples for each user selected through a novel user-personalized, two-step, similarity-based PU Learning algorithm. Computational experiments show this PU-based approach outperforms the state-of-the-art non-PU method in six popular real-world datasets, proving that an improvement of visual-based RS explainability can be achieved by maximizing training data quality rather than increasing model complexity.
Authors: Augustine N. Mavor-Parker, Matthew J. Sargent, Caswell Barry, Lewis Griffin, Clare Lyle
Abstract: Periodic activation functions, often referred to as learned Fourier features have been widely demonstrated to improve sample efficiency and stability in a variety of deep RL algorithms. Potentially incompatible hypotheses have been made about the source of these improvements. One is that periodic activations learn low frequency representations and as a result avoid overfitting to bootstrapped targets. Another is that periodic activations learn high frequency representations that are more expressive, allowing networks to quickly fit complex value functions. We analyse these claims empirically, finding that periodic representations consistently converge to high frequencies regardless of their initialisation frequency. We also find that while periodic activation functions improve sample efficiency, they exhibit worse generalization on states with added observation noise -- especially when compared to otherwise equivalent networks with ReLU activation functions. Finally, we show that weight decay regularization is able to partially offset the overfitting of periodic activation functions, delivering value functions that learn quickly while also generalizing.
Authors: Eugene Golikov
Abstract: We consider nonlinear networks as perturbations of linear ones. Based on this approach, we present novel generalization bounds that become non-vacuous for networks that are close to being linear. The main advantage over the previous works which propose non-vacuous generalization bounds is that our bounds are a-priori: performing the actual training is not required for evaluating the bounds. To the best of our knowledge, they are the first non-vacuous generalization bounds for neural nets possessing this property.
Authors: Johannes Viehweg, Dominik Walther, Prof. Dr. -Ing. Patrick M\"ader
Abstract: The prediction of time series is a challenging task relevant in such diverse applications as analyzing financial data, forecasting flow dynamics or understanding biological processes. Especially chaotic time series that depend on a long history pose an exceptionally difficult problem. While machine learning has shown to be a promising approach for predicting such time series, it either demands long training time and much training data when using deep recurrent neural networks. Alternative, when using a reservoir computing approach it comes with high uncertainty and typically a high number of random initializations and extensive hyper-parameter tuning when using a reservoir computing approach. In this paper, we focus on the reservoir computing approach and propose a new mapping of input data into the reservoir's state space. Furthermore, we incorporate this method in two novel network architectures increasing parallelizability, depth and predictive capabilities of the neural network while reducing the dependence on randomness. For the evaluation, we approximate a set of time series from the Mackey-Glass equation, inhabiting non-chaotic as well as chaotic behavior and compare our approaches in regard to their predictive capabilities to echo state networks and gated recurrent units. For the chaotic time series, we observe an error reduction of up to $85.45\%$ and up to $87.90\%$ in contrast to echo state networks and gated recurrent units respectively. Furthermore, we also observe tremendous improvements for non-chaotic time series of up to $99.99\%$ in contrast to existing approaches.
Authors: Fotios Lygerakis, Elmar Rueckert
Abstract: Traditional Variational Autoencoders (VAEs) are constrained by the limitations of the Evidence Lower Bound (ELBO) formulation, particularly when utilizing simplistic, non-analytic, or unknown prior distributions. These limitations inhibit the VAE's ability to generate high-quality samples and provide clear, interpretable latent representations. This work introduces the Entropy Decomposed Variational Autoencoder (ED-VAE), a novel re-formulation of the ELBO that explicitly includes entropy and cross-entropy components. This reformulation significantly enhances model flexibility, allowing for the integration of complex and non-standard priors. By providing more detailed control over the encoding and regularization of latent spaces, ED-VAE not only improves interpretability but also effectively captures the complex interactions between latent variables and observed data, thus leading to better generative performance.
Authors: Lucas Correia, Jan-Christoph Goos, Philipp Klein, Thomas B\"ack, Anna V. Kononova
Abstract: As attention to recorded data grows in the realm of automotive testing and manual evaluation reaches its limits, there is a growing need for automatic online anomaly detection. This real-world data is complex in many ways and requires the modelling of testee behaviour. To address this, we propose a temporal variational autoencoder (TeVAE) that can detect anomalies with minimal false positives when trained on unlabelled data. Our approach also avoids the bypass phenomenon and introduces a new method to remap individual windows to a continuous time series. Furthermore, we propose metrics to evaluate the detection delay and root-cause capability of our approach and present results from experiments on a real-world industrial data set. When properly configured, TeVAE flags anomalies only 6% of the time wrongly and detects 65% of anomalies present. It also has the potential to perform well with a smaller training and validation subset but requires a more sophisticated threshold estimation method.
Authors: Sourabh Kapoor, Arnab Sharma, Michael R\"oder, Caglar Demir, Axel-Cyrille Ngonga Ngomo
Abstract: Knowledge Graph Embedding (KGE) transforms a discrete Knowledge Graph (KG) into a continuous vector space facilitating its use in various AI-driven applications like Semantic Search, Question Answering, or Recommenders. While KGE approaches are effective in these applications, most existing approaches assume that all information in the given KG is correct. This enables attackers to influence the output of these approaches, e.g., by perturbing the input. Consequently, the robustness of such KGE approaches has to be addressed. Recent work focused on adversarial attacks. However, non-adversarial attacks on all attack surfaces of these approaches have not been thoroughly examined. We close this gap by evaluating the impact of non-adversarial attacks on the performance of 5 state-of-the-art KGE algorithms on 5 datasets with respect to attacks on 3 attack surfaces-graph, parameter, and label perturbation. Our evaluation results suggest that label perturbation has a strong effect on the KGE performance, followed by parameter perturbation with a moderate and graph with a low effect.
Authors: Sahel Vahedi Noori, Bin Hu, Geir Dullerud, Peter Seiler
Abstract: This paper derives a complete set of quadratic constraints (QCs) for the repeated ReLU. The complete set of QCs is described by a collection of $2^{n_v}$ matrix copositivity conditions where $n_v$ is the dimension of the repeated ReLU. We also show that only two functions satisfy all QCs in our complete set: the repeated ReLU and a repeated "flipped" ReLU. Thus our complete set of QCs bounds the repeated ReLU as tight as possible up to the sign invariance inherent in quadratic forms. We derive a similar complete set of incremental QCs for repeated ReLU, which can potentially lead to less conservative Lipschitz bounds for ReLU networks than the standard LipSDP approach. Finally, we illustrate the use of the complete set of QCs to assess stability and performance for recurrent neural networks with ReLU activation functions. The stability/performance condition combines Lyapunov/dissipativity theory with the QCs for repeated ReLU. A numerical implementation is given and demonstrated via a simple example.
Authors: Jiajun Liang, Qian Zhang, Wei Deng, Qifan Song, Guang Lin
Abstract: This work introduces a novel and efficient Bayesian federated learning algorithm, namely, the Federated Averaging stochastic Hamiltonian Monte Carlo (FA-HMC), for parameter estimation and uncertainty quantification. We establish rigorous convergence guarantees of FA-HMC on non-iid distributed data sets, under the strong convexity and Hessian smoothness assumptions. Our analysis investigates the effects of parameter space dimension, noise on gradients and momentum, and the frequency of communication (between the central node and local nodes) on the convergence and communication costs of FA-HMC. Beyond that, we establish the tightness of our analysis by showing that the convergence rate cannot be improved even for continuous FA-HMC process. Moreover, extensive empirical studies demonstrate that FA-HMC outperforms the existing Federated Averaging-Langevin Monte Carlo (FA-LD) algorithm.
Authors: Samuel Tonks, Cuong Nguyer, Steve Hood, Ryan Musso, Ceridwen Hopely, Steve Titus, Minh Doan, Iain Styles, Alexander Krull
Abstract: The large volume and variety of imaging data from high-throughput screening (HTS) in the pharmaceutical industry present an excellent resource for training virtual staining models. However, the potential of models trained under one set of experimental conditions to generalize to other conditions remains underexplored. This study systematically investigates whether data from three cell types (lung, ovarian, and breast) and two phenotypes (toxic and non-toxic conditions) commonly found in HTS can effectively train virtual staining models to generalize across three typical HTS distribution shifts: unseen phenotypes, unseen cell types, and the combination of both. Utilizing a dataset of 772,416 paired bright-field, cytoplasm, nuclei, and DNA-damage stain images, we evaluate the generalization capabilities of models across pixel-based, instance-wise, and biological-feature-based levels. Our findings indicate that training virtual nuclei and cytoplasm models on non-toxic condition samples not only generalizes to toxic condition samples but leads to improved performance across all evaluation levels compared to training on toxic condition samples. Generalization to unseen cell types shows variability depending on the cell type; models trained on ovarian or lung cell samples often perform well under other conditions, while those trained on breast cell samples consistently show poor generalization. Generalization to unseen cell types and phenotypes shows good generalization across all levels of evaluation compared to addressing unseen cell types alone. This study represents the first large-scale, data-centric analysis of the generalization capability of virtual staining models trained on diverse HTS datasets, providing valuable strategies for experimental training data generation.
Authors: Amey Agrawal, Anmol Agarwal, Nitin Kedia, Jayashree Mohan, Souvik Kundu, Nipun Kwatra, Ramachandran Ramjee, Alexey Tumanov
Abstract: Serving large language models (LLMs) in production can incur substantial costs, which has prompted recent advances in inference system optimizations. Today, these systems are evaluated against conventional latency and throughput metrics (eg. TTFT, TBT, Normalised Latency and TPOT). However, these metrics fail to fully capture the nuances of LLM inference, leading to an incomplete assessment of user-facing performance crucial for real-time applications such as chat and translation. In this paper, we first identify the pitfalls of current performance metrics in evaluating LLM inference systems. We then propose Metron, a comprehensive performance evaluation framework that includes fluidity-index -- a novel metric designed to reflect the intricacies of the LLM inference process and its impact on real-time user experience. Finally, we evaluate various existing open-source platforms and model-as-a-service offerings using Metron, discussing their strengths and weaknesses. Metron is available at https://github.com/project-metron/metron.
Authors: Nikita Dhawan, Leonardo Cotta, Karen Ullrich, Rahul G. Krishnan, Chris J. Maddison
Abstract: Knowing the effect of an intervention is critical for human decision-making, but current approaches for causal effect estimation rely on manual data collection and structuring, regardless of the causal assumptions. This increases both the cost and time-to-completion for studies. We show how large, diverse observational text data can be mined with large language models (LLMs) to produce inexpensive causal effect estimates under appropriate causal assumptions. We introduce NATURAL, a novel family of causal effect estimators built with LLMs that operate over datasets of unstructured text. Our estimators use LLM conditional distributions (over variables of interest, given the text data) to assist in the computation of classical estimators of causal effect. We overcome a number of technical challenges to realize this idea, such as automating data curation and using LLMs to impute missing information. We prepare six (two synthetic and four real) observational datasets, paired with corresponding ground truth in the form of randomized trials, which we used to systematically evaluate each step of our pipeline. NATURAL estimators demonstrate remarkable performance, yielding causal effect estimates that fall within 3 percentage points of their ground truth counterparts, including on real-world Phase 3/4 clinical trials. Our results suggest that unstructured text data is a rich source of causal effect information, and NATURAL is a first step towards an automated pipeline to tap this resource.
Authors: Fardin Jalil Piran, Prathyush P. Poduval, Hamza Errahmouni Barkam, Mohsen Imani, Farhad Imani
Abstract: In-situ sensing, in conjunction with learning models, presents a unique opportunity to address persistent defect issues in Additive Manufacturing (AM) processes. However, this integration introduces significant data privacy concerns, such as data leakage, sensor data compromise, and model inversion attacks, revealing critical details about part design, material composition, and machine parameters. Differential Privacy (DP) models, which inject noise into data under mathematical guarantees, offer a nuanced balance between data utility and privacy by obscuring traces of sensing data. However, the introduction of noise into learning models, often functioning as black boxes, complicates the prediction of how specific noise levels impact model accuracy. This study introduces the Differential Privacy-HyperDimensional computing (DP-HD) framework, leveraging the explainability of the vector symbolic paradigm to predict the noise impact on the accuracy of in-situ monitoring, safeguarding sensitive data while maintaining operational efficiency. Experimental results on real-world high-speed melt pool data of AM for detecting overhang anomalies demonstrate that DP-HD achieves superior operational efficiency, prediction accuracy, and robust privacy protection, outperforming state-of-the-art Machine Learning (ML) models. For example, when implementing the same level of privacy protection (with a privacy budget set at 1), our model achieved an accuracy of 94.43\%, surpassing the performance of traditional models such as ResNet50 (52.30\%), GoogLeNet (23.85\%), AlexNet (55.78\%), DenseNet201 (69.13\%), and EfficientNet B2 (40.81\%). Notably, DP-HD maintains high performance under substantial noise additions designed to enhance privacy, unlike current models that suffer significant accuracy declines under high privacy constraints.
Authors: Alexander David Goldie, Chris Lu, Matthew Thomas Jackson, Shimon Whiteson, Jakob Nicolaus Foerster
Abstract: While reinforcement learning (RL) holds great potential for decision making in the real world, it suffers from a number of unique difficulties which often need specific consideration. In particular: it is highly non-stationary; suffers from high degrees of plasticity loss; and requires exploration to prevent premature convergence to local optima and maximize return. In this paper, we consider whether learned optimization can help overcome these problems. Our method, Learned Optimization for Plasticity, Exploration and Non-stationarity (OPEN), meta-learns an update rule whose input features and output structure are informed by previously proposed solutions to these difficulties. We show that our parameterization is flexible enough to enable meta-learning in diverse learning contexts, including the ability to use stochasticity for exploration. Our experiments demonstrate that when meta-trained on single and small sets of environments, OPEN outperforms or equals traditionally used optimizers. Furthermore, OPEN shows strong generalization across a distribution of environments and a range of agent architectures.
Authors: Xiaowen Jiang, Anton Rodomanov, Sebastian U. Stich
Abstract: In developing efficient optimization algorithms, it is crucial to account for communication constraints -- a significant challenge in modern federated learning settings. The best-known communication complexity among non-accelerated algorithms is achieved by DANE, a distributed proximal-point algorithm that solves local subproblems in each iteration and that can exploit second-order similarity among individual functions. However, to achieve such communication efficiency, the accuracy requirement for solving the local subproblems is slightly sub-optimal. Inspired by the hybrid projection-proximal point method, in this work, we i) propose a novel distributed algorithm S-DANE. This method adopts a more stabilized prox-center in the proximal step compared with DANE, and matches its deterministic communication complexity. Moreover, the accuracy condition of the subproblem is milder, leading to enhanced local computation efficiency. Furthermore, it supports partial client participation and arbitrary stochastic local solvers, making it more attractive in practice. We further ii) accelerate S-DANE, and show that the resulting algorithm achieves the best-known communication complexity among all existing methods for distributed convex optimization, with the same improved local computation efficiency as S-DANE.
Authors: Ruochen Jin, Bojian Hou, Jiancong Xiao, Weijie Su, Li Shen
Abstract: Task arithmetic has recently emerged as a cost-effective and scalable approach to edit pre-trained models directly in weight space, by adding the fine-tuned weights of different tasks. The performance has been further improved by a linear property which is illustrated by weight disentanglement. Yet, conventional linearization methods (e.g., NTK linearization) not only double the time and training cost but also have a disadvantage on single-task performance. We propose a simple yet effective and efficient method that only fine-tunes linear layers, which improves weight disentanglement and efficiency simultaneously. Specifically, our study reveals that only fine-tuning the linear layers in the attention modules makes the whole model occur in a linear regime, significantly improving weight disentanglement. To further understand how our method improves the disentanglement of task arithmetic, we present a comprehensive study of task arithmetic by differentiating the role of representation model and task-specific model. In particular, we find that the representation model plays an important role in improving weight disentanglement whereas the task-specific models such as the classification heads can degenerate the weight disentanglement performance. Overall, our work uncovers novel insights into the fundamental mechanisms of task arithmetic and offers a more reliable and effective approach to editing pre-trained models.
Authors: Jiaang Li, Yova Kementchedjhieva, Constanza Fierro, Anders S{\o}gaard
Abstract: Large-scale pretrained language models (LMs) are said to ``lack the ability to connect utterances to the world'' (Bender and Koller, 2020), because they do not have ``mental models of the world' '(Mitchell and Krakauer, 2023). If so, one would expect LM representations to be unrelated to representations induced by vision models. We present an empirical evaluation across four families of LMs (BERT, GPT-2, OPT and LLaMA-2) and three vision model architectures (ResNet, SegFormer, and MAE). Our experiments show that LMs partially converge towards representations isomorphic to those of vision models, subject to dispersion, polysemy and frequency. This has important implications for both multi-modal processing and the LM understanding debate (Mitchell and Krakauer, 2023).
Authors: Jared Deighton, Wyatt Mackey, Ioannis Schizas, David L. Boothe Jr., Vasileios Maroulas
Abstract: Mammals navigate novel environments and exhibit resilience to sparse environmental sensory cues via place and grid cells, which encode position in space. While the efficiency of grid cell coding has been extensively studied, the computational role of place cells is less well understood. This gap arises partially because spatial information measures have, until now, been limited to single place cells. We derive and implement a higher-order spatial information measure, allowing for the study of the emergence of multiple place cells in a self-supervised manner. We show that emergent place cells have many desirable features, including high-accuracy spatial decoding. This is the first work in which higher-order spatial information measures that depend solely on place cells' firing rates have been derived and which focuses on the emergence of multiple place cells via self-supervised learning. By quantifying the spatial information of multiple place cells, we enhance our understanding of place cell formation and capabilities in recurrent neural networks, thereby improving the potential navigation capabilities of artificial systems in novel environments without objective location information.
Authors: Styliani-Christina Fragkouli, Dhwani Solanki, Leyla J Castro, Fotis E Psomopoulos, N\'uria Queralt-Rosinach, Davide Cirillo, Lisa C Crossman
Abstract: Over the last three to five years, it has become possible to generate machine learning synthetic data for healthcare-related uses. However, concerns have been raised about potential negative factors associated with the possibilities of artificial dataset generation. These include the potential misuse of generative artificial intelligence (AI) in fields such as cybercrime, the use of deepfakes and fake news to deceive or manipulate, and displacement of human jobs across various market sectors. Here, we consider both current and future positive advances and possibilities with synthetic datasets. Synthetic data offers significant benefits, particularly in data privacy, research, in balancing datasets and reducing bias in machine learning models. Generative AI is an artificial intelligence genre capable of creating text, images, video or other data using generative models. The recent explosion of interest in GenAI was heralded by the invention and speedy move to use of large language models (LLM). These computational models are able to achieve general-purpose language generation and other natural language processing tasks and are based on transformer architectures, which made an evolutionary leap from previous neural network architectures. Fuelled by the advent of improved GenAI techniques and wide scale usage, this is surely the time to consider how synthetic data can be used to advance infectious disease research. In this commentary we aim to create an overview of the current and future position of synthetic data in infectious disease research.
Authors: Daniel Navia, Rodrigo Bruna, Francisco Fern\'andez, Cristobal Mancilla, Mat\'ias Rojas, Mauricio Estrada, Paulina Quintanilla
Abstract: This work presents the development and validation of a digital twin for a semi-autogenous grinding (SAG) mill controlled by an expert control system. The digital twin consists of three interconnected modules that emulate the behavior of a closed-loop system: (1) fuzzy logic for the expert control system, (2) a state-space model for the regulatory control, and (3) a recurrent neural network (RNN) for the SAG mill process. The model was trained with data corresponding to 68 hours of operation and validated with 8 hours of test data. The digital twin predicts the dynamic behavior of the mill's bearing pressure, motor power, tonnage, solids percentage, and rotational speed within a 2.5-minute horizon with a 30-second sampling time. The RNN comprises two serial modules for detection and training. The disturbance detection evaluates the need for training by comparing the recent prediction error with the expected error using hypothesis tests for mean, variance, and probability distribution. If the detection module is activated, the parameters of the neural model are re-estimated with recent data. The detection module was configured with test data to eliminate false positives. Results indicate that the digital twin can satisfactorily supervise the SAG mill, which is operated with the expert control system. Future work will focus on integrating this digital twin into real-time optimization strategies with industrial validation.
Authors: Swetha T, Seshaiah M, Hemalatha KL, ManjunathaKumar BH, Murthy SVN
Abstract: The proliferation of malicious URLs has become a significant threat to internet security, encompassing SPAM, phishing, malware, and defacement attacks. Traditional detection methods struggle to keep pace with the evolving nature of these threats. Detecting malicious URLs in real-time requires advanced techniques capable of handling large datasets and identifying novel attack patterns. The challenge lies in developing a robust model that combines efficient feature extraction with accurate classification. We propose a hybrid machine learning approach combining Self-Organizing Map based Radial Movement Optimization (SOM-RMO) for feature extraction and Radial Basis Function Network (RBFN) based Tabu Search for classification. SOM-RMO effectively reduces dimensionality and highlights significant features, while RBFN, optimized with Tabu Search, classifies URLs with high precision. The proposed model demonstrates superior performance in detecting various malicious URL attacks. On a benchmark dataset, our approach achieved an accuracy of 96.5%, precision of 95.2%, recall of 94.8%, and an F1-score of 95.0%, outperforming traditional methods significantly.
Authors: Papri Saha
Abstract: Machine learning and quantum computing are being progressively explored to shed light on possible computational approaches to deal with hitherto unsolvable problems. Classical methods for machine learning are ubiquitous in pattern recognition, with support vector machines (SVMs) being a prominent technique for network classification. However, there are limitations to the successful resolution of such classification instances when the input feature space becomes large, and the successive evaluation of so-called kernel functions becomes computationally exorbitant. The use of principal component analysis (PCA) substantially minimizes the dimensionality of feature space thereby enabling computational speed-ups of supervised learning: the creation of a classifier. Further, the application of quantum-based learning to the PCA reduced input feature space might offer an exponential speedup with fewer parameters. The present learning model is evaluated on a real clinical application: the diagnosis of Progressive Supranuclear Palsy (PSP) disorder. The results suggest that quantum machine learning has led to noticeable advancement and outperforms classical frameworks. The optimized variational quantum classifier classifies the PSP dataset with 86% accuracy as compared to conventional SVM. The other technique, a quantum kernel estimator, approximates the kernel function on the quantum machine and optimizes a classical SVM. In particular, we have demonstrated the successful application of the present model on both a quantum simulator and real chips of the IBM quantum platform.
Authors: Krunal Kishor Patel
Abstract: In this article, we introduce the concept of discounted pseudocosts, inspired by discounted total reward in reinforcement learning, and explore their application in mixed-integer linear programming (MILP). Traditional pseudocosts estimate changes in the objective function due to variable bound changes during the branch-and-bound process. By integrating reinforcement learning concepts, we propose a novel approach incorporating a forward-looking perspective into pseudocost estimation. We present the motivation behind discounted pseudocosts and discuss how they represent the anticipated reward for branching after one level of exploration in the MILP problem space. Initial experiments on MIPLIB 2017 benchmark instances demonstrate the potential of discounted pseudocosts to enhance branching strategies and accelerate the solution process for challenging MILP problems.
Authors: Pranshav Gajjar, Vijay K. Shah
Abstract: Large Language Models (LLMs) can revolutionize how we deploy and operate Open Radio Access Networks (O-RAN) by enhancing network analytics, anomaly detection, and code generation and significantly increasing the efficiency and reliability of a plethora of O-RAN tasks. In this paper, we present ORAN-Bench-13K, the first comprehensive benchmark designed to evaluate the performance of Large Language Models (LLMs) within the context of O-RAN. Our benchmark consists of 13,952 meticulously curated multiple-choice questions generated from 116 O-RAN specification documents. We leverage a novel three-stage LLM framework, and the questions are categorized into three distinct difficulties to cover a wide spectrum of ORAN-related knowledge. We thoroughly evaluate the performance of several state-of-the-art LLMs, including Gemini, Chat-GPT, and Mistral. Additionally, we propose ORANSight, a Retrieval-Augmented Generation (RAG)-based pipeline that demonstrates superior performance on ORAN-Bench-13K compared to other tested closed-source models. Our findings indicate that current popular LLM models are not proficient in O-RAN, highlighting the need for specialized models. We observed a noticeable performance improvement when incorporating the RAG-based ORANSight pipeline, with a Macro Accuracy of 0.784 and a Weighted Accuracy of 0.776, which was on average 21.55% and 22.59% better than the other tested LLMs.
Authors: Ramya Deshpande, Francesco Mottes, Ariana-Dalia Vlad, Michael P. Brenner, Alma dal Co
Abstract: Understanding the rules underlying organismal development is a major unsolved problem in biology. Each cell in a developing organism responds to signals in its local environment by dividing, excreting, consuming, or reorganizing, yet how these individual actions coordinate over a macroscopic number of cells to grow complex structures with exquisite functionality is unknown. Here we use recent advances in automatic differentiation to discover local interaction rules and genetic networks that yield emergent, systems-level characteristics in a model of development. We consider a growing tissue with cellular interactions are mediated by morphogen diffusion, differential cell adhesion and mechanical stress. Each cell has an internal genetic network that it uses to make decisions based on its local environment. We show that one can simultaneously learn parameters governing the cell interactions and the genetic network for complex developmental scenarios, including the symmetry breaking of an embryo from an initial cell, the creation of emergent chemical gradients,homogenization of growth via mechanical stress, programmed growth into a prespecified shape, and the ability to repair from damage. When combined with recent experimental advances measuring spatio-temporal dynamics and gene expression of cells in a growing tissue, the methodology outlined here offers a promising path to unravelling the cellular basis of development.
Authors: Murilo Gustineli, Anthony Miyaguchi, Ian Stalter
Abstract: We present a transfer learning approach using a self-supervised Vision Transformer (DINOv2) for the PlantCLEF 2024 competition, focusing on the multi-label plant species classification. Our method leverages both base and fine-tuned DINOv2 models to extract generalized feature embeddings. We train classifiers to predict multiple plant species within a single image using these rich embeddings. To address the computational challenges of the large-scale dataset, we employ Spark for distributed data processing, ensuring efficient memory management and processing across a cluster of workers. Our data processing pipeline transforms images into grids of tiles, classifying each tile, and aggregating these predictions into a consolidated set of probabilities. Our results demonstrate the efficacy of combining transfer learning with advanced data processing techniques for multi-label image classification tasks. Our code is available at https://github.com/dsgt-kaggle-clef/plantclef-2024.
Authors: Ahmed Maged, Herman Shen
Abstract: Automated f ault detection and monitoring in engineering are critical but frequently difficult owing to the necessity for collecting and labeling large amounts of defective samples . We present an unsupervised method that uses the high end Segment Anything Model (SAM) and a moving window approach. SAM has gained recognition in AI image segmentation communities for its accuracy and versatility. However, its performance can be inconsistent when dealing with certain unexpected shapes , such as shadows and subtle surface irregularities. This limitation raise s concerns about its applicability for fault detection in real world scenarios We aim to overcome these challenges without requiring fine tun ing or labeled data. Our technique divides pictures into smaller windows, which are subsequently processed using SAM. This increases the accuracy of fault identification by focusing on localized details. We compute the sizes of the segmented sections and then us e a clustering technique to discover consistent fault areas while filtering out noise. To further improve the method's robustness , we propose adding the Exponentially Weighted Moving Average (EWMA) technique for continuous monitoring in industrial settings, which would improve the method's capacity to trace faults over time. We compare our method to various well established methods u sing a real case study where our model achieve s 0.96 accuracy compared to 0. 8 5 for the second best method. W e also compare our method us ing two open source datasets where our model attains a consistent 0. 86 accuracy across the datasets compared to 0.53 and 0.54 for second best model s.
Authors: Mengzhe Geng, Xurong Xie, Jiajun Deng, Zengrui Jin, Guinan Li, Tianzi Wang, Shujie Hu, Zhaoqing Li, Helen Meng, Xunying Liu
Abstract: The application of data-intensive automatic speech recognition (ASR) technologies to dysarthric and elderly adult speech is confronted by their mismatch against healthy and nonaged voices, data scarcity and large speaker-level variability. To this end, this paper proposes two novel data-efficient methods to learn homogeneous dysarthric and elderly speaker-level features for rapid, on-the-fly test-time adaptation of DNN/TDNN and Conformer ASR models. These include: 1) speaker-level variance-regularized spectral basis embedding (VR-SBE) features that exploit a special regularization term to enforce homogeneity of speaker features in adaptation; and 2) feature-based learning hidden unit contributions (f-LHUC) transforms that are conditioned on VR-SBE features. Experiments are conducted on four tasks across two languages: the English UASpeech and TORGO dysarthric speech datasets, the English DementiaBank Pitt and Cantonese JCCOCC MoCA elderly speech corpora. The proposed on-the-fly speaker adaptation techniques consistently outperform baseline iVector and xVector adaptation by statistically significant word or character error rate reductions up to 5.32% absolute (18.57% relative) and batch-mode LHUC speaker adaptation by 2.24% absolute (9.20% relative), while operating with real-time factors speeding up to 33.6 times against xVectors during adaptation. The efficacy of the proposed adaptation techniques is demonstrated in a comparison against current ASR technologies including SSL pre-trained systems on UASpeech, where our best system produces a state-of-the-art WER of 23.33%. Analyses show VR-SBE features and f-LHUC transforms are insensitive to speaker-level data quantity in testtime adaptation. T-SNE visualization reveals they have stronger speaker-level homogeneity than baseline iVectors, xVectors and batch-mode LHUC transforms.
Authors: Matthew J. Colbrook, Igor Mezi\'c, Alexei Stepanenko
Abstract: Dynamical systems provide a comprehensive way to study complex and changing behaviors across various sciences. Many modern systems are too complicated to analyze directly or we do not have access to models, driving significant interest in learning methods. Koopman operators have emerged as a dominant approach because they allow the study of nonlinear dynamics using linear techniques by solving an infinite-dimensional spectral problem. However, current algorithms face challenges such as lack of convergence, hindering practical progress. This paper addresses a fundamental open question: \textit{When can we robustly learn the spectral properties of Koopman operators from trajectory data of dynamical systems, and when can we not?} Understanding these boundaries is crucial for analysis, applications, and designing algorithms. We establish a foundational approach that combines computational analysis and ergodic theory, revealing the first fundamental barriers -- universal for any algorithm -- associated with system geometry and complexity, regardless of data quality and quantity. For instance, we demonstrate well-behaved smooth dynamical systems on tori where non-trivial eigenfunctions of the Koopman operator cannot be determined by any sequence of (even randomized) algorithms, even with unlimited training data. Additionally, we identify when learning is possible and introduce optimal algorithms with verification that overcome issues in standard methods. These results pave the way for a sharp classification theory of data-driven dynamical systems based on how many limits are needed to solve a problem. These limits characterize all previous methods, presenting a unified view. Our framework systematically determines when and how Koopman spectral properties can be learned.
Authors: Mujtaba Hussain Mirza, Maria Rosaria Briglia, Senad Beadini, Iacopo Masi
Abstract: By reinterpreting a robust discriminative classifier as Energy-based Model (EBM), we offer a new take on the dynamics of adversarial training (AT). Our analysis of the energy landscape during AT reveals that untargeted attacks generate adversarial images much more in-distribution (lower energy) than the original data from the point of view of the model. Conversely, we observe the opposite for targeted attacks. On the ground of our thorough analysis, we present new theoretical and practical results that show how interpreting AT energy dynamics unlocks a better understanding: (1) AT dynamic is governed by three phases and robust overfitting occurs in the third phase with a drastic divergence between natural and adversarial energies (2) by rewriting the loss of TRadeoff-inspired Adversarial DEfense via Surrogate-loss minimization (TRADES) in terms of energies, we show that TRADES implicitly alleviates overfitting by means of aligning the natural energy with the adversarial one (3) we empirically show that all recent state-of-the-art robust classifiers are smoothing the energy landscape and we reconcile a variety of studies about understanding AT and weighting the loss function under the umbrella of EBMs. Motivated by rigorous evidence, we propose Weighted Energy Adversarial Training (WEAT), a novel sample weighting scheme that yields robust accuracy matching the state-of-the-art on multiple benchmarks such as CIFAR-10 and SVHN and going beyond in CIFAR-100 and Tiny-ImageNet. We further show that robust classifiers vary in the intensity and quality of their generative capabilities, and offer a simple method to push this capability, reaching a remarkable Inception Score (IS) and FID using a robust classifier without training for generative modeling. The code to reproduce our results is available at http://github.com/OmnAI-Lab/Robust-Classifiers-under-the-lens-of-EBM/ .
URLs: http://github.com/OmnAI-Lab/Robust-Classifiers-under-the-lens-of-EBM/
Authors: Marco Mussi, Simone Drago, Alberto Maria Metelli
Abstract: We consider Kernelized Bandits (KBs) to optimize a function $f : \mathcal{X} \rightarrow [0,1]$ belonging to the Reproducing Kernel Hilbert Space (RKHS) $\mathcal{H}_k$. Mainstream works on kernelized bandits focus on a subgaussian noise model in which observations of the form $f(\mathbf{x}_t)+\epsilon_t$, being $\epsilon_t$ a subgaussian noise, are available (Chowdhury and Gopalan, 2017). Differently, we focus on the case in which we observe realizations $y_t \sim \text{Ber}(f(\mathbf{x}_t))$ sampled from a Bernoulli distribution with parameter $f(\mathbf{x}_t)$. While the Bernoulli model has been investigated successfully in multi-armed bandits (Garivier and Capp\'e, 2011), logistic bandits (Faury et al., 2022), bandits in metric spaces (Magureanu et al., 2014), it remains an open question whether tight results can be obtained for KBs. This paper aims to draw the attention of the online learning community to this open problem.
Authors: Shouchang Guo
Abstract: The goals of functional Magnetic Resonance Imaging (fMRI) include high spatial and temporal resolutions with a high signal-to-noise ratio (SNR). To simultaneously improve spatial and temporal resolutions and maintain the high SNR advantage of OSSI, we present novel pipelines for fast acquisition and high-resolution fMRI reconstruction and physics parameter quantification. We propose a patch-tensor low-rank model, a physics-based manifold model, and a voxel-wise attention network. With novel models for acquisition and reconstruction, we demonstrate that we can improve SNR and resolution simultaneously without compromising scan time. All the proposed models outperform other comparison approaches with higher resolution and more functional information.
Authors: Eshant English, Eliot Wong-Toi, Matteo Fontana, Stephan Mandt, Padhraic Smyth, Christoph Lippert
Abstract: Conformal prediction provides machine learning models with prediction sets that offer theoretical guarantees, but the underlying assumption of exchangeability limits its applicability to time series data. Furthermore, existing approaches struggle to handle multi-step ahead prediction tasks, where uncertainty estimates across multiple future time points are crucial. We propose JANET (Joint Adaptive predictioN-region Estimation for Time-series), a novel framework for constructing conformal prediction regions that are valid for both univariate and multivariate time series. JANET generalises the inductive conformal framework and efficiently produces joint prediction regions with controlled K-familywise error rates, enabling flexible adaptation to specific application needs. Our empirical evaluation demonstrates JANET's superior performance in multi-step prediction tasks across diverse time series datasets, highlighting its potential for reliable and interpretable uncertainty quantification in sequential data.
Authors: Steffen W. R. Werner, Benjamin Peherstorfer
Abstract: Stability is a basic requirement when studying the behavior of dynamical systems. However, stabilizing dynamical systems via reinforcement learning is challenging because only little data can be collected over short time horizons before instabilities are triggered and data become meaningless. This work introduces a reinforcement learning approach that is formulated over latent manifolds of unstable dynamics so that stabilizing policies can be trained from few data samples. The unstable manifolds are minimal in the sense that they contain the lowest dimensional dynamics that are necessary for learning policies that guarantee stabilization. This is in stark contrast to generic latent manifolds that aim to approximate all -- stable and unstable -- system dynamics and thus are higher dimensional and often require higher amounts of data. Experiments demonstrate that the proposed approach stabilizes even complex physical systems from few data samples for which other methods that operate either directly in the system state space or on generic latent manifolds fail.
Authors: Yangyi Chen, Xingyao Wang, Hao Peng, Heng Ji
Abstract: We present SOLO, a single transformer for Scalable visiOn-Language mOdeling. Current large vision-language models (LVLMs) such as LLaVA mostly employ heterogeneous architectures that connect pre-trained visual encoders with large language models (LLMs) to facilitate visual recognition and complex reasoning. Although achieving remarkable performance with relatively lightweight training, we identify four primary scalability limitations: (1) The visual capacity is constrained by pre-trained visual encoders, which are typically an order of magnitude smaller than LLMs. (2) The heterogeneous architecture complicates the use of established hardware and software infrastructure. (3) Study of scaling laws on such architecture must consider three separate components - visual encoder, connector, and LLMs, which complicates the analysis. (4) The use of existing visual encoders typically requires following a pre-defined specification of image inputs pre-processing, for example, by reshaping inputs to fixed-resolution square images, which presents difficulties in processing and training on high-resolution images or those with unusual aspect ratio. A unified single Transformer architecture, like SOLO, effectively addresses these scalability concerns in LVLMs; however, its limited adoption in the modern context likely stems from the absence of reliable training recipes that balance both modalities and ensure stable training for billion-scale models. In this paper, we introduce the first open-source training recipe for developing SOLO, an open-source 7B LVLM using moderate academic resources. The training recipe involves initializing from LLMs, sequential pre-training on ImageNet and web-scale data, and instruction fine-tuning on our curated high-quality datasets. On extensive evaluation, SOLO demonstrates performance comparable to LLaVA-v1.5-7B, particularly excelling in visual mathematical reasoning.
Authors: Divyagna Bavikadi, Dyuman Aditya, Devendra Parkar, Paulo Shakarian, Graham Mueller, Chad Parvis, Gerardo I. Simari
Abstract: The ability to generate artificial human movement patterns while meeting location and time constraints is an important problem in the security community, particularly as it enables the study of the analog problem of detecting such patterns while maintaining privacy. We frame this problem as an instance of abduction guided by a novel parsimony function represented as an aggregate truth value over an annotated logic program. This approach has the added benefit of affording explainability to an analyst user. By showing that any subset of such a program can provide a lower bound on this parsimony requirement, we are able to abduce movement trajectories efficiently through an informed (i.e., A*) search. We describe how our implementation was enhanced with the application of multiple techniques in order to be scaled and integrated with a cloud-based software stack that included bottom-up rule learning, geolocated knowledge graph retrieval/management, and interfaces with government systems for independently conducted government-run tests for which we provide results. We also report on our own experiments showing that we not only provide exact results but also scale to very large scenarios and provide realistic agent trajectories that can go undetected by machine learning anomaly detectors.
Authors: Hang Yu, Qidi Fang, Shijie Fang, Reuben M. Aronson, Elaine Schaertl Short
Abstract: Enhancing the expressiveness of human teaching is vital for both improving robots' learning from humans and the human-teaching-robot experience. In this work, we characterize and test a little-used teaching signal: \textit{progress}, designed to represent the completion percentage of a task. We conducted two online studies with 76 crowd-sourced participants and one public space study with 40 non-expert participants to validate the capability of this progress signal. We find that progress indicates whether the task is successfully performed, reflects the degree of task completion, identifies unproductive but harmless behaviors, and is likely to be more consistent across participants. Furthermore, our results show that giving progress does not require extra workload and time. An additional contribution of our work is a dataset of 40 non-expert demonstrations from the public space study through an ice cream topping-adding task, which we observe to be multi-policy and sub-optimal, with sub-optimality not only from teleoperation errors but also from exploratory actions and attempts. The dataset is available at \url{https://github.com/TeachingwithProgress/Non-Expert\_Demonstrations}.
URLs: https://github.com/TeachingwithProgress/Non-Expert\_Demonstrations
Authors: Yikun Bai
Abstract: In this note, we generalize the classical optimal partial transport (OPT) problem by modifying the mass destruction/creation term to function-based terms, introducing what we term ``generalized optimal partial transport'' problems. We then discuss the dual formulation of these problems and the associated Sinkhorn solver. Finally, we explore how these new OPT problems relate to classical optimal transport (OT) problems and introduce a linear programming solver tailored for these generalized scenarios.
Authors: Yongqi Leng, Deyi Xiong
Abstract: While large language models (LLMs) have demonstrated superior multi-task capabilities, understanding the learning mechanisms behind this is still a challenging problem. In this paper, we attempt to understand such mechanisms from the perspective of neurons. Specifically, we detect task-sensitive neurons in LLMs via gradient attribution on task-specific data. Through extensive deactivation and fine-tuning experiments, we demonstrate that the detected neurons are highly correlated with the given task, which we term as task-specific neurons. With these identified task-specific neurons, we delve into two common problems in multi-task learning and continuous learning: Generalization and Catastrophic Forgetting. We find that the overlap of task-specific neurons is strongly associated with generalization and specialization across tasks. Interestingly, at certain layers of LLMs, there is a high similarity in the parameters of different task-specific neurons, and such similarity is highly correlated with the generalization performance. Inspired by these findings, we propose a neuron-level continuous fine-tuning method that only fine-tunes the current task-specific neurons during continuous learning, and extensive experiments demonstrate the effectiveness of the proposed method. Our study provides insights into the interpretability of LLMs in multi-task learning.
Authors: Shouchang Guo, Sonam Damani, Keng-hao Chang
Abstract: Ads relevance models are crucial in determining the relevance between user search queries and ad offers, often framed as a classification problem. The complexity of modeling increases significantly with multiple ad types and varying scenarios that exhibit both similarities and differences. In this work, we introduce a novel multi-faceted attention model that performs task aware feature combination and cross task interaction modeling. Our technique formulates the feature combination problem as "language" modeling with auto-regressive attentions across both feature and task dimensions. Specifically, we introduce a new dimension of task ID encoding for task representations, thereby enabling precise relevance modeling across diverse ad scenarios with substantial improvement in generality capability for unseen tasks. We demonstrate that our model not only effectively handles the increased computational and maintenance demands as scenarios proliferate, but also outperforms generalized DNN models and even task-specific models across a spectrum of ad applications using a single unified model.
Authors: Mehrsa Pourya, Sebastian Neumayer, Michael Unser
Abstract: We propose a regularization scheme for image reconstruction that leverages the power of deep learning while hinging on classic sparsity-promoting models. Many deep-learning-based models are hard to interpret and cumbersome to analyze theoretically. In contrast, our scheme is interpretable because it corresponds to the minimization of a series of convex problems. For each problem in the series, a mask is generated based on the previous solution to refine the regularization strength spatially. In this way, the model becomes progressively attentive to the image structure. For the underlying update operator, we prove the existence of a fixed point. As a special case, we investigate a mask generator for which the fixed-point iterations converge to a critical point of an explicit energy functional. In our experiments, we match the performance of state-of-the-art learned variational models for the solution of inverse problems. Additionally, we offer a promising balance between interpretability, theoretical guarantees, reliability, and performance.
Authors: Rui Jin, Derun Li, Dehui Xiang, Lei Zhang, Hailing Zhou, Fei Shi, Weifang Zhu, Jing Cai, Tao Peng, Xinjian Chen
Abstract: Prostate cancer represents a major threat to health. Early detection is vital in reducing the mortality rate among prostate cancer patients. One approach involves using multi-modality (CT, MRI, US, etc.) computer-aided diagnosis (CAD) systems for the prostate region. However, prostate segmentation is challenging due to imperfections in the images and the prostate's complex tissue structure. The advent of precision medicine and a significant increase in clinical capacity have spurred the need for various data-driven tasks in the field of medical imaging. Recently, numerous machine learning and data mining tools have been integrated into various medical areas, including image segmentation. This article proposes a new classification method that differentiates supervision types, either in number or kind, during the training phase. Subsequently, we conducted a survey on artificial intelligence (AI)-based automatic prostate segmentation methods, examining the advantages and limitations of each. Additionally, we introduce variants of evaluation metrics for the verification and performance assessment of the segmentation method and summarize the current challenges. Finally, future research directions and development trends are discussed, reflecting the outcomes of our literature survey, suggesting high-precision detection and treatment of prostate cancer as a promising avenue.
Authors: Balint Bicski, Adrian Pekar
Abstract: This research presents a novel method for predicting service degradation (SD) in computer networks by leveraging early flow features. Our approach focuses on the observable (O) segments of network flows, particularly analyzing Packet Inter-Arrival Time (PIAT) values and other derived metrics, to infer the behavior of non-observable (NO) segments. Through a comprehensive evaluation, we identify an optimal O/NO split threshold of 10 observed delay samples, balancing prediction accuracy and resource utilization. Evaluating models including Logistic Regression, XGBoost, and Multi-Layer Perceptron, we find XGBoost outperforms others, achieving an F1-score of 0.74, balanced accuracy of 0.84, and AUROC of 0.97. Our findings highlight the effectiveness of incorporating comprehensive early flow features and the potential of our method to offer a practical solution for monitoring network traffic in resource-constrained environments. This approach ensures enhanced user experience and network performance by preemptively addressing potential SD, providing the basis for a robust framework for maintaining high-quality network services.
Authors: Xunjiang Gu, Guanyu Song, Igor Gilitschenski, Marco Pavone, Boris Ivanovic
Abstract: Understanding road geometry is a critical component of the autonomous vehicle (AV) stack. While high-definition (HD) maps can readily provide such information, they suffer from high labeling and maintenance costs. Accordingly, many recent works have proposed methods for estimating HD maps online from sensor data. The vast majority of recent approaches encode multi-camera observations into an intermediate representation, e.g., a bird's eye view (BEV) grid, and produce vector map elements via a decoder. While this architecture is performant, it decimates much of the information encoded in the intermediate representation, preventing downstream tasks (e.g., behavior prediction) from leveraging them. In this work, we propose exposing the rich internal features of online map estimation methods and show how they enable more tightly integrating online mapping with trajectory forecasting. In doing so, we find that directly accessing internal BEV features yields up to 73% faster inference speeds and up to 29% more accurate predictions on the real-world nuScenes dataset.
Authors: Chengjie Wang, Chengming Xu, Zhenye Gan, Jianlong Hu, Wenbing Zhu, Lizhuag Ma
Abstract: Positive and Unlabeled (PU) learning, a binary classification model trained with only positive and unlabeled data, generally suffers from overfitted risk estimation due to inconsistent data distributions. To address this, we introduce a pseudo-supervised PU learning framework (PSPU), in which we train the PU model first, use it to gather confident samples for the pseudo supervision, and then apply these supervision to correct the PU model's weights by leveraging non-PU objectives. We also incorporate an additional consistency loss to mitigate noisy sample effects. Our PSPU outperforms recent PU learning methods significantly on MNIST, CIFAR-10, CIFAR-100 in both balanced and imbalanced settings, and enjoys competitive performance on MVTecAD for industrial anomaly detection.
Authors: Gian Marco Visani, Michael N. Pun, William Galvin, Eric Daniel, Kevin Borisiak, Utheri Wagura, Armita Nourmohammad
Abstract: Predicting the stability and fitness effects of amino acid mutations in proteins is a cornerstone of biological discovery and engineering. Various experimental techniques have been developed to measure mutational effects, providing us with extensive datasets across a diverse range of proteins. By training on these data, traditional computational modeling and more recent machine learning approaches have advanced significantly in predicting mutational effects. Here, we introduce HERMES, a 3D rotationally equivariant structure-based neural network model for mutational effect and stability prediction. Pre-trained to predict amino acid propensity from its surrounding 3D structure, HERMES can be fine-tuned for mutational effects using our open-source code. We present a suite of HERMES models, pre-trained with different strategies, and fine-tuned to predict the stability effect of mutations. Benchmarking against other models shows that HERMES often outperforms or matches their performance in predicting mutational effect on stability, binding, and fitness. HERMES offers versatile tools for evaluating mutational effects and can be fine-tuned for specific predictive objectives.
Authors: Arthur Aubret, C\'eline Teuli\`ere, Jochen Triesch
Abstract: Self-supervised learning (SSL) has revolutionized visual representation learning, but has not achieved the robustness of human vision. A reason for this could be that SSL does not leverage all the data available to humans during learning. When learning about an object, humans often purposefully turn or move around objects and research suggests that these interactions can substantially enhance their learning. Here we explore whether such object-related actions can boost SSL. For this, we extract the actions performed to change from one ego-centric view of an object to another in four video datasets. We then introduce a new loss function to learn visual and action embeddings by aligning the performed action with the representations of two images extracted from the same clip. This permits the performed actions to structure the latent visual representation. Our experiments show that our method consistently outperforms previous methods on downstream category recognition. In our analysis, we find that the observed improvement is associated with a better viewpoint-wise alignment of different objects from the same category. Overall, our work demonstrates that embodied interactions with objects can improve SSL of object categories.
Authors: Yu-Guan Hsieh, Cheng-Yu Hsieh, Shih-Ying Yeh, Louis B\'ethune, Hadi Pour Ansari, Pavan Kumar Anasosalu Vasu, Chun-Liang Li, Ranjay Krishna, Oncel Tuzel, Marco Cuturi
Abstract: Humans describe complex scenes with compositionality, using simple text descriptions enriched with links and relationships. While vision-language research has aimed to develop models with compositional understanding capabilities, this is not reflected yet in existing datasets which, for the most part, still use plain text to describe images. In this work, we propose a new annotation strategy, graph-based captioning (GBC) that describes an image using a labelled graph structure, with nodes of various types. The nodes in GBC are created using, in a first stage, object detection and dense captioning tools nested recursively to uncover and describe entity nodes, further linked together in a second stage by highlighting, using new types of nodes, compositions and relations among entities. Since all GBC nodes hold plain text descriptions, GBC retains the flexibility found in natural language, but can also encode hierarchical information in its edges. We demonstrate that GBC can be produced automatically, using off-the-shelf multimodal LLMs and open-vocabulary detection models, by building a new dataset, GBC10M, gathering GBC annotations for about 10M images of the CC12M dataset. We use GBC10M to showcase the wealth of node captions uncovered by GBC, as measured with CLIP training. We show that using GBC nodes' annotations -- notably those stored in composition and relation nodes -- results in significant performance boost on downstream models when compared to other dataset formats. To further explore the opportunities provided by GBC, we also propose a new attention mechanism that can leverage the entire GBC graph, with encouraging experimental results that show the extra benefits of incorporating the graph structure. Our datasets are released at \url{https://huggingface.co/graph-based-captions}.
Authors: Leon Bungert, Jeff Calder, Max Mihailescu, Kodjo Houssou, Amber Yuan
Abstract: In this paper we prove discrete to continuum convergence rates for Poisson Learning, a graph-based semi-supervised learning algorithm that is based on solving the graph Poisson equation with a source term consisting of a linear combination of Dirac deltas located at labeled points and carrying label information. The corresponding continuum equation is a Poisson equation with measure data in a Euclidean domain $\Omega \subset \mathbb{R}^d$. The singular nature of these equations is challenging and requires an approach with several distinct parts: (1) We prove quantitative error estimates when convolving the measure data of a Poisson equation with (approximately) radial function supported on balls. (2) We use quantitative variational techniques to prove discrete to continuum convergence rates on random geometric graphs with bandwidth $\varepsilon>0$ for bounded source terms. (3) We show how to regularize the graph Poisson equation via mollification with the graph heat kernel, and we study fine asymptotics of the heat kernel on random geometric graphs. Combining these three pillars we obtain $L^1$ convergence rates that scale, up to logarithmic factors, like $O(\varepsilon^{\frac{1}{d+2}})$ for general data distributions, and $O(\varepsilon^{\frac{2-\sigma}{d+4}})$ for uniformly distributed data, where $\sigma>0$. These rates are valid with high probability if $\varepsilon\gg\left({\log n}/{n}\right)^q$ where $n$ denotes the number of vertices of the graph and $q \approx \frac{1}{3d}$.
Authors: Ashok Dahal, Luigi Lombardo
Abstract: For decades, solutions to regional scale landslide prediction have mostly relied on data-driven models, by definition, disconnected from the physics of the failure mechanism. The success and spread of such tools came from the ability to exploit proxy variables rather than explicit geotechnical ones, as the latter are prohibitive to acquire over broad landscapes. Our work implements a Physics Informed Neural Network (PINN) approach, thereby adding to a standard data-driven architecture, an intermediate constraint to solve for the permanent deformation typical of Newmark slope stability methods. This translates into a neural network tasked with explicitly retrieving geotechnical parameters from common proxy variables and then minimize a loss function with respect to the available coseismic landside inventory. The results are very promising, because our model not only produces excellent predictive performance in the form of standard susceptibility output, but in the process, also generates maps of the expected geotechnical properties at a regional scale. Such architecture is therefore framed to tackle coseismic landslide prediction, something that, if confirmed in other studies, could open up towards PINN-based near-real-time predictions.
Authors: Mengjie Qian, Siyuan Tang, Rao Ma, Kate M. Knill, Mark J. F. Gales
Abstract: Foundation ASR models often support many languages, e.g. 100 languages in Whisper. However, there has been limited work on integrating an additional, typically low-resource, language, while maintaining performance on the original language set. Fine-tuning, while simple, may degrade the accuracy of the original set. We compare three approaches that exploit adaptation parameters: soft language code tuning, train only the language code; soft prompt tuning, train prepended tokens; and LoRA where a small set of additional parameters are optimised. Elastic Weight Consolidation (EWC) offers an alternative compromise with the potential to maintain performance in specific target languages. Results show that direct fine-tuning yields the best performance for the new language but degrades existing language capabilities. EWC can address this issue for specific languages. If only adaptation parameters are used, the language capabilities are maintained but at the cost of performance in the new language.
Authors: Lorenzo Cassano, Jacopo D'Abramo, Siraj Munir, Stefano Ferretti
Abstract: In this paper, we present a study of a Federated Learning (FL) system, based on the use of decentralized architectures to ensure trust and increase reliability. The system is based on the idea that the FL collaborators upload the (ciphered) model parameters on the Inter-Planetary File System (IPFS) and interact with a dedicated smart contract to track their behavior. Thank to this smart contract, the phases of parameter updates are managed efficiently, thereby strengthening data security. We have carried out an experimental study that exploits two different methods of weight aggregation, i.e., a classic averaging scheme and a federated proximal aggregation. The results confirm the feasibility of the proposal.
Authors: Ashok S. Kumar, Nancy Nayak, Sheetal Kalyani, Himal A. Suraweera
Abstract: In this work, we propose a method to improve the energy efficiency and fairness of simultaneously transmitting and reflecting reconfigurable intelligent surfaces (STAR-RIS) for mobile users, ensuring reduced power consumption while maintaining reliable communication. To achieve this, we introduce a new parameter known as the subsurface assignment variable, which determines the number of STAR-RIS elements allocated to each user. We then formulate a novel optimization problem by concurrently optimizing the phase shifts of the STAR-RIS and subsurface assignment variable. We leverage the deep reinforcement learning (DRL) technique to address this optimization problem. The DRL model predicts the phase shifts of the STAR-RIS and efficiently allocates elements of STAR-RIS to the users. Additionally, we incorporate a penalty term in the DRL model to facilitate intelligent deactivation of STAR-RIS elements when not in use to enhance energy efficiency. Through extensive experiments, we show that the proposed method can achieve fairly high and nearly equal data rates for all users in both the transmission and reflection spaces in an energy-efficient manner.
Authors: Yang Liu, Weixing Chen, Yongjie Bai, Jingzhou Luo, Xinshuai Song, Kaixuan Jiang, Zhida Li, Ganlong Zhao, Junyi Lin, Guanbin Li, Wen Gao, Liang Lin
Abstract: Embodied Artificial Intelligence (Embodied AI) is crucial for achieving Artificial General Intelligence (AGI) and serves as a foundation for various applications that bridge cyberspace and the physical world. Recently, the emergence of Multi-modal Large Models (MLMs) and World Models (WMs) have attracted significant attention due to their remarkable perception, interaction, and reasoning capabilities, making them a promising architecture for the brain of embodied agents. However, there is no comprehensive survey for Embodied AI in the era of MLMs. In this survey, we give a comprehensive exploration of the latest advancements in Embodied AI. Our analysis firstly navigates through the forefront of representative works of embodied robots and simulators, to fully understand the research focuses and their limitations. Then, we analyze four main research targets: 1) embodied perception, 2) embodied interaction, 3) embodied agent, and 4) sim-to-real adaptation, covering the state-of-the-art methods, essential paradigms, and comprehensive datasets. Additionally, we explore the complexities of MLMs in virtual and real embodied agents, highlighting their significance in facilitating interactions in dynamic digital and physical environments. Finally, we summarize the challenges and limitations of embodied AI and discuss their potential future directions. We hope this survey will serve as a foundational reference for the research community and inspire continued innovation. The associated project can be found at https://github.com/HCPLab-SYSU/Embodied_AI_Paper_List.
URLs: https://github.com/HCPLab-SYSU/Embodied_AI_Paper_List.
Authors: Shahana Ibrahim, Panagiotis A. Traganitis, Xiao Fu, Georgios B. Giannakis
Abstract: One of the primary catalysts fueling advances in artificial intelligence (AI) and machine learning (ML) is the availability of massive, curated datasets. A commonly used technique to curate such massive datasets is crowdsourcing, where data are dispatched to multiple annotators. The annotator-produced labels are then fused to serve downstream learning and inference tasks. This annotation process often creates noisy labels due to various reasons, such as the limited expertise, or unreliability of annotators, among others. Therefore, a core objective in crowdsourcing is to develop methods that effectively mitigate the negative impact of such label noise on learning tasks. This feature article introduces advances in learning from noisy crowdsourced labels. The focus is on key crowdsourcing models and their methodological treatments, from classical statistical models to recent deep learning-based approaches, emphasizing analytical insights and algorithmic developments. In particular, this article reviews the connections between signal processing (SP) theory and methods, such as identifiability of tensor and nonnegative matrix factorization, and novel, principled solutions of longstanding challenges in crowdsourcing -- showing how SP perspectives drive the advancements of this field. Furthermore, this article touches upon emerging topics that are critical for developing cutting-edge AI/ML systems, such as crowdsourcing in reinforcement learning with human feedback (RLHF) and direct preference optimization (DPO) that are key techniques for fine-tuning large language models (LLMs).
Authors: Francisco Giral, Ignacio G\'omez, Soledad Le Clainche
Abstract: The proliferation of unmanned aerial vehicles (UAVs) in controlled airspace presents significant risks, including potential collisions, disruptions to air traffic, and security threats. Ensuring the safe and efficient operation of airspace, particularly in urban environments and near critical infrastructure, necessitates effective methods to intercept unauthorized or non-cooperative UAVs. This work addresses the critical need for robust, adaptive systems capable of managing such threats through the use of Reinforcement Learning (RL). We present a novel approach utilizing RL to train fixed-wing UAV pursuer agents for intercepting dynamic evader targets. Our methodology explores both model-based and model-free RL algorithms, specifically DreamerV3, Truncated Quantile Critics (TQC), and Soft Actor-Critic (SAC). The training and evaluation of these algorithms were conducted under diverse scenarios, including unseen evasion strategies and environmental perturbations. Our approach leverages high-fidelity flight dynamics simulations to create realistic training environments. This research underscores the importance of developing intelligent, adaptive control systems for UAV interception, significantly contributing to the advancement of secure and efficient airspace management. It demonstrates the potential of RL to train systems capable of autonomously achieving these critical tasks.
Authors: Manpreet Singh, Ravdeep Pasricha, Ravi Prasad Kondapalli, Kiran R, Nitish Singh, Akshita Agarwalla, Manoj R, Manish Prabhakar, Laurent Bou\'e
Abstract: One of the most critical tasks of Microsoft sellers is to meticulously track and nurture potential business opportunities through proactive engagement and tailored solutions. Recommender systems play a central role to help sellers achieve their goals. In this paper, we present a content recommendation model which surfaces various types of content (technical documentation, comparison with competitor products, customer success stories etc.) that sellers can share with their customers or use for their own self-learning. The model operates at the opportunity level which is the lowest possible granularity and the most relevant one for sellers. It is based on semantic matching between metadata from the contents and carefully selected attributes of the opportunities. Considering the volume of seller-managed opportunities in organizations such as Microsoft, we show how to perform efficient semantic matching over a very large number of opportunity-content combinations. The main challenge is to ensure that the top-5 relevant contents for each opportunity are recommended out of a total of $\approx 40,000$ published contents. We achieve this target through an extensive comparison of different model architectures and feature selection. Finally, we further examine the quality of the recommendations in a quantitative manner using a combination of human domain experts as well as by using the recently proposed "LLM as a judge" framework.
Authors: Tim R. Davidson, Viacheslav Surkov, Veniamin Veselovsky, Giuseppe Russo, Robert West, Caglar Gulcehre
Abstract: A rapidly growing number of applications rely on a small set of closed-source language models (LMs). This dependency might introduce novel security risks if LMs develop self-recognition capabilities. Inspired by human identity verification methods, we propose a novel approach for assessing self-recognition in LMs using model-generated "security questions". Our test can be externally administered to keep track of frontier models as it does not require access to internal model parameters or output probabilities. We use our test to examine self-recognition in ten of the most capable open- and closed-source LMs currently publicly available. Our extensive experiments found no empirical evidence of general or consistent self-recognition in any examined LM. Instead, our results suggest that given a set of alternatives, LMs seek to pick the "best" answer, regardless of its origin. Moreover, we find indications that preferences about which models produce the best answers are consistent across LMs. We additionally uncover novel insights on position bias considerations for LMs in multiple-choice settings.
Authors: Yu-An Liu, Ruqing Zhang, Jiafeng Guo, Maarten de Rijke, Yixing Fan, Xueqi Cheng
Abstract: Recent advances in neural information retrieval (IR) models have significantly enhanced their effectiveness over various IR tasks. The robustness of these models, essential for ensuring their reliability in practice, has also garnered significant attention. With a wide array of research on robust IR being proposed, we believe it is the opportune moment to consolidate the current status, glean insights from existing methodologies, and lay the groundwork for future development. We view the robustness of IR to be a multifaceted concept, emphasizing its necessity against adversarial attacks, out-of-distribution (OOD) scenarios and performance variance. With a focus on adversarial and OOD robustness, we dissect robustness solutions for dense retrieval models (DRMs) and neural ranking models (NRMs), respectively, recognizing them as pivotal components of the neural IR pipeline. We provide an in-depth discussion of existing methods, datasets, and evaluation metrics, shedding light on challenges and future directions in the era of large language models. To the best of our knowledge, this is the first comprehensive survey on the robustness of neural IR models, and we will also be giving our first tutorial presentation at SIGIR 2024 \url{https://sigir2024-robust-information-retrieval.github.io}. Along with the organization of existing work, we introduce a Benchmark for robust IR (BestIR), a heterogeneous evaluation benchmark for robust neural information retrieval, which is publicly available at \url{https://github.com/Davion-Liu/BestIR}. We hope that this study provides useful clues for future research on the robustness of IR models and helps to develop trustworthy search engines \url{https://github.com/Davion-Liu/Awesome-Robustness-in-Information-Retrieval}.
URLs: https://sigir2024-robust-information-retrieval.github.io, https://github.com/Davion-Liu/BestIR, https://github.com/Davion-Liu/Awesome-Robustness-in-Information-Retrieval
Authors: Emiliano Penaloza, Nathaniel Stevens
Abstract: Detecting anomalous behavior in dynamic networks remains a constant challenge. This problem is further exacerbated when the underlying topology of these networks is affected by individual highly-dimensional node attributes. We address this issue by tracking a network's modularity as a proxy of its community structure. We leverage Graph Neural Networks (GNNs) to estimate each snapshot's modularity. GNNs can account for both network structure and high-dimensional node attributes, providing a comprehensive approach for estimating network statistics. Our method is validated through simulations that demonstrate its ability to detect changes in highly-attributed networks by analyzing shifts in modularity. Moreover, we find our method is able to detect a real-world event within the \#Iran Twitter reply network, where each node has high-dimensional textual attributes.
Authors: Rapha\"el Tinarrage, Henrique Ennes, Lucas E. Resck, Lucas T. Gomes, Jean R. Ponciano, Jorge Poco
Abstract: Binding precedents (S\'umulas Vinculantes) constitute a juridical instrument unique to the Brazilian legal system and whose objectives include the protection of the Federal Supreme Court against repetitive demands. Studies of the effectiveness of these instruments in decreasing the Court's exposure to similar cases, however, indicate that they tend to fail in such a direction, with some of the binding precedents seemingly creating new demands. We empirically assess the legal impact of five binding precedents, 11, 14, 17, 26 and 37, at the highest court level through their effects on the legal subjects they address. This analysis is only possible through the comparison of the Court's ruling about the precedents' themes before they are created, which means that these decisions should be detected through techniques of Similar Case Retrieval. The contributions of this article are therefore twofold: on the mathematical side, we compare the uses of different methods of Natural Language Processing -- TF-IDF, LSTM, BERT, and regex -- for Similar Case Retrieval, whereas on the legal side, we contrast the inefficiency of these binding precedents with a set of hypotheses that may justify their repeated usage. We observe that the deep learning models performed significantly worse in the specific Similar Case Retrieval task and that the reasons for binding precedents to fail in responding to repetitive demand are heterogeneous and case-dependent, making it impossible to single out a specific cause.
Authors: Linh Tran, Sanjay Chari, Md. Saikat Islam Khan, Aaron Zachariah, Stacy Patterson, Oshani Seneviratne
Abstract: We present the Differentially Private Blockchain-Based Vertical Federal Learning (DP-BBVFL) algorithm that provides verifiability and privacy guarantees for decentralized applications. DP-BBVFL uses a smart contract to aggregate the feature representations, i.e., the embeddings, from clients transparently. We apply local differential privacy to provide privacy for embeddings stored on a blockchain, hence protecting the original data. We provide the first prototype application of differential privacy with blockchain for vertical federated learning. Our experiments with medical data show that DP-BBVFL achieves high accuracy with a tradeoff in training time due to on-chain aggregation. This innovative fusion of differential privacy and blockchain technology in DP-BBVFL could herald a new era of collaborative and trustworthy machine learning applications across several decentralized application domains.
Authors: Haiqian Yang, Anh Q. Nguyen, Dapeng Bi, Markus J. Buehler, Ming Guo
Abstract: During developmental processes such as embryogenesis, how a group of cells fold into specific structures, is a central question in biology that defines how living organisms form. Establishing tissue-level morphology critically relies on how every single cell decides to position itself relative to its neighboring cells. Despite its importance, it remains a major challenge to understand and predict the behavior of every cell within the living tissue over time during such intricate processes. To tackle this question, we propose a geometric deep learning model that can predict multicellular folding and embryogenesis, accurately capturing the highly convoluted spatial interactions among cells. We demonstrate that multicellular data can be represented with both granular and foam-like physical pictures through a unified graph data structure, considering both cellular interactions and cell junction networks. We successfully use our model to achieve two important tasks, interpretable 4-D morphological sequence alignment, and predicting local cell rearrangements before they occur at single-cell resolution. Furthermore, using an activation map and ablation studies, we demonstrate that cell geometries and cell junction networks together regulate local cell rearrangement which is critical for embryo morphogenesis. This approach provides a novel paradigm to study morphogenesis, highlighting a unified data structure and harnessing the power of geometric deep learning to accurately model the mechanisms and behaviors of cells during development. It offers a pathway toward creating a unified dynamic morphological atlas for a variety of developmental processes such as embryogenesis.
Authors: Nathan Cloos, Moufan Li, Markus Siegel, Scott L. Brincat, Earl K. Miller, Guangyu Robert Yang, Christopher J. Cueva
Abstract: What metrics should guide the development of more realistic models of the brain? One proposal is to quantify the similarity between models and brains using methods such as linear regression, Centered Kernel Alignment (CKA), and angular Procrustes distance. To better understand the limitations of these similarity measures we analyze neural activity recorded in five experiments on nonhuman primates, and optimize synthetic datasets to become more similar to these neural recordings. How similar can these synthetic datasets be to neural activity while failing to encode task relevant variables? We find that some measures like linear regression and CKA, differ from angular Procrustes, and yield high similarity scores even when task relevant variables cannot be linearly decoded from the synthetic datasets. Synthetic datasets optimized to maximize similarity scores initially learn the first principal component of the target dataset, but angular Procrustes captures higher variance dimensions much earlier than methods like linear regression and CKA. We show in both theory and simulations how these scores change when different principal components are perturbed. And finally, we jointly optimize multiple similarity scores to find their allowed ranges, and show that a high angular Procrustes similarity, for example, implies a high CKA score, but not the converse.
Authors: Catherine Tony, Nicol\'as E. D\'iaz Ferreyra, Markus Mutas, Salem Dhiff, Riccardo Scandariato
Abstract: Large Language Models (LLMs) are gaining momentum in software development with prompt-driven programming enabling developers to create code from natural language (NL) instructions. However, studies have questioned their ability to produce secure code and, thereby, the quality of prompt-generated software. Alongside, various prompting techniques that carefully tailor prompts have emerged to elicit optimal responses from LLMs. Still, the interplay between such prompting strategies and secure code generation remains under-explored and calls for further investigations. OBJECTIVE: In this study, we investigate the impact of different prompting techniques on the security of code generated from NL instructions by LLMs. METHOD: First we perform a systematic literature review to identify the existing prompting techniques that can be used for code generation tasks. A subset of these techniques are evaluated on GPT-3, GPT-3.5, and GPT-4 models for secure code generation. For this, we used an existing dataset consisting of 150 NL security-relevant code-generation prompts. RESULTS: Our work (i) classifies potential prompting techniques for code generation (ii) adapts and evaluates a subset of the identified techniques for secure code generation tasks and (iii) observes a reduction in security weaknesses across the tested LLMs, especially after using an existing technique called Recursive Criticism and Improvement (RCI), contributing valuable insights to the ongoing discourse on LLM-generated code security.
Authors: Yung-Sung Chuang, Linlu Qiu, Cheng-Yu Hsieh, Ranjay Krishna, Yoon Kim, James Glass
Abstract: When asked to summarize articles or answer questions given a passage, large language models (LLMs) can hallucinate details and respond with unsubstantiated answers that are inaccurate with respect to the input context. This paper describes a simple approach for detecting such contextual hallucinations. We hypothesize that contextual hallucinations are related to the extent to which an LLM attends to information in the provided context versus its own generations. Based on this intuition, we propose a simple hallucination detection model whose input features are given by the ratio of attention weights on the context versus newly generated tokens (for each attention head). We find that a linear classifier based on these lookback ratio features is as effective as a richer detector that utilizes the entire hidden states of an LLM or a text-based entailment model. The lookback ratio-based detector -- Lookback Lens -- is found to transfer across tasks and even models, allowing a detector that is trained on a 7B model to be applied (without retraining) to a larger 13B model. We further apply this detector to mitigate contextual hallucinations, and find that a simple classifier-guided decoding approach is able to reduce the amount of hallucination, for example by 9.6% in the XSum summarization task.
Authors: Tong Chen, Akari Asai, Niloofar Mireshghallah, Sewon Min, James Grimmelmann, Yejin Choi, Hannaneh Hajishirzi, Luke Zettlemoyer, Pang Wei Koh
Abstract: Evaluating the degree of reproduction of copyright-protected content by language models (LMs) is of significant interest to the AI and legal communities. Although both literal and non-literal similarities are considered by courts when assessing the degree of reproduction, prior research has focused only on literal similarities. To bridge this gap, we introduce CopyBench, a benchmark designed to measure both literal and non-literal copying in LM generations. Using copyrighted fiction books as text sources, we provide automatic evaluation protocols to assess literal and non-literal copying, balanced against the model utility in terms of the ability to recall facts from the copyrighted works and generate fluent completions. We find that, although literal copying is relatively rare, two types of non-literal copying -- event copying and character copying -- occur even in models as small as 7B parameters. Larger models demonstrate significantly more copying, with literal copying rates increasing from 0.2% to 10.5% and non-literal copying from 2.3% to 6.9% when comparing Llama3-8B and 70B models, respectively. We further evaluate the effectiveness of current strategies for mitigating copying and show that (1) training-time alignment can reduce literal copying but may increase non-literal copying, and (2) current inference-time mitigation methods primarily reduce literal but not non-literal copying.
Authors: Liqun Ma, Mingjie Sun, Zhiqiang Shen
Abstract: This work presents a Fully BInarized Large Language Model (FBI-LLM), demonstrating for the first time how to train a large-scale binary language model from scratch (not the partial binary or ternary LLM like BitNet b1.58) to match the performance of its full-precision counterparts (e.g., FP16 or BF16) in transformer-based LLMs. It achieves this by employing an autoregressive distillation (AD) loss with maintaining equivalent model dimensions (130M, 1.3B, 7B) and training data volume as regular LLM pretraining, while delivering competitive results in terms of perplexity and task-specific effectiveness. Intriguingly, by analyzing the training trajectory, we find that the pretrained weight is not necessary for training binarized LLMs from scratch. This research encourages a new computational framework and may facilitate the future design of specialized hardware tailored for fully 1-bit LLMs. We make all models, code, and training dataset fully accessible and transparent to support further research (Code: https://github.com/LiqunMa/FBI-LLM. Model: https://huggingface.co/LiqunMa/).
URLs: https://github.com/LiqunMa/FBI-LLM., https://huggingface.co/LiqunMa/).
Authors: Michael Klamkin, Mathieu Tanneau, Terrence W. K. Mak, Pascal Van Hentenryck
Abstract: This paper considers optimization proxies for Optimal Power Flow (OPF), i.e., machine-learning models that approximate the input/output relationship of OPF. Recent work has focused on showing that such proxies can be of high fidelity. However, their training requires significant data, each instance necessitating the (offline) solving of an OPF. To meet the requirements of market-clearing applications, this paper proposes Bucketized Active Sampling (BAS), a novel active learning framework that aims at training the best possible OPF proxy within a time limit. BAS partitions the input domain into buckets and uses an acquisition function to determine where to sample next. By applying the same partitioning to the validation set, BAS leverages labeled validation samples in the selection of unlabeled samples. BAS also relies on an adaptive learning rate that increases and decreases over time. Experimental results demonstrate the benefits of BAS.
Authors: Dominik Hintersdorf, Lukas Struppek, Manuel Brack, Felix Friedrich, Patrick Schramowski, Kristian Kersting
Abstract: With the rise of deep learning in various applications, privacy concerns around the protection of training data have become a critical area of research. Whereas prior studies have focused on privacy risks in single-modal models, we introduce a novel method to assess privacy for multi-modal models, specifically vision-language models like CLIP. The proposed Identity Inference Attack (IDIA) reveals whether an individual was included in the training data by querying the model with images of the same person. Letting the model choose from a wide variety of possible text labels, the model reveals whether it recognizes the person and, therefore, was used for training. Our large-scale experiments on CLIP demonstrate that individuals used for training can be identified with very high accuracy. We confirm that the model has learned to associate names with depicted individuals, implying the existence of sensitive information that can be extracted by adversaries. Our results highlight the need for stronger privacy protection in large-scale models and suggest that IDIAs can be used to prove the unauthorized use of data for training and to enforce privacy laws.
Authors: Futoon M. Abushaqra, Hao Xue, Yongli Ren, Flora D. Salim
Abstract: Ordinary Differential Equations (ODE) based models have become popular as foundation models for solving many time series problems. Combining neural ODEs with traditional RNN models has provided the best representation for irregular time series. However, ODE-based models typically require the trajectory of hidden states to be defined based on either the initial observed value or the most recent observation, raising questions about their effectiveness when dealing with longer sequences and extended time intervals. In this article, we explore the behaviour of the ODE models in the context of time series data with varying degrees of sparsity. We introduce SeqLink, an innovative neural architecture designed to enhance the robustness of sequence representation. Unlike traditional approaches that solely rely on the hidden state generated from the last observed value, SeqLink leverages ODE latent representations derived from multiple data samples, enabling it to generate robust data representations regardless of sequence length or data sparsity level. The core concept behind our model is the definition of hidden states for the unobserved values based on the relationships between samples (links between sequences). Through extensive experiments on partially observed synthetic and real-world datasets, we demonstrate that SeqLink improves the modelling of intermittent time series, consistently outperforming state-of-the-art approaches.
Authors: Abdullah Caglar Oksuz, Anisa Halimi, Erman Ayday
Abstract: Explainable Artificial Intelligence (XAI) aims to uncover the decision-making processes of AI models. However, the data used for such explanations can pose security and privacy risks. Existing literature identifies attacks on machine learning models, including membership inference, model inversion, and model extraction attacks. These attacks target either the model or the training data, depending on the settings and parties involved. XAI tools can increase the vulnerability of model extraction attacks, which is a concern when model owners prefer black-box access, thereby keeping model parameters and architecture private. To exploit this risk, we propose AUTOLYCUS, a novel retraining (learning) based model extraction attack framework against interpretable models under black-box settings. As XAI tools, we exploit Local Interpretable Model-Agnostic Explanations (LIME) and Shapley values (SHAP) to infer decision boundaries and create surrogate models that replicate the functionality of the target model. LIME and SHAP are mainly chosen for their realistic yet information-rich explanations, coupled with their extensive adoption, simplicity, and usability. We evaluate AUTOLYCUS on six machine learning datasets, measuring the accuracy and similarity of the surrogate model to the target model. The results show that AUTOLYCUS is highly effective, requiring significantly fewer queries compared to state-of-the-art attacks, while maintaining comparable accuracy and similarity. We validate its performance and transferability on multiple interpretable ML models, including decision trees, logistic regression, naive bayes, and k-nearest neighbor. Additionally, we show the resilience of AUTOLYCUS against proposed countermeasures.
Authors: Junwei Su, Difan Zou, Chuan Wu
Abstract: Continual learning seeks to empower models to progressively acquire information from a sequence of tasks. This approach is crucial for many real-world systems, which are dynamic and evolve over time. Recent research has witnessed a surge in the exploration of Graph Neural Networks (GNN) in Node-wise Graph Continual Learning (NGCL), a practical yet challenging paradigm involving the continual training of a GNN on node-related tasks. Despite recent advancements in continual learning strategies for GNNs in NGCL, a thorough theoretical understanding, especially regarding its learnability, is lacking. Learnability concerns the existence of a learning algorithm that can produce a good candidate model from the hypothesis/weight space, which is crucial for model selection in NGCL development. This paper introduces the first theoretical exploration of the learnability of GNN in NGCL, revealing that learnability is heavily influenced by structural shifts due to the interconnected nature of graph data. Specifically, GNNs may not be viable for NGCL under significant structural changes, emphasizing the need to manage structural shifts. To mitigate the impact of structural shifts, we propose a novel experience replay method termed Structure-Evolution-Aware Experience Replay (SEA-ER). SEA-ER features an innovative experience selection strategy that capitalizes on the topological awareness of GNNs, alongside a unique replay strategy that employs structural alignment to effectively counter catastrophic forgetting and diminish the impact of structural shifts on GNNs in NGCL. Our extensive experiments validate our theoretical insights and the effectiveness of SEA-ER.
Authors: Kazuki Irie, R\'obert Csord\'as, J\"urgen Schmidhuber
Abstract: Unsupervised learning of discrete representations in neural networks (NNs) from continuous ones is essential for many modern applications. Vector Quantisation (VQ) has become popular for this, in particular in the context of generative models, such as Variational Auto-Encoders (VAEs), where the exponential moving average-based VQ (EMA-VQ) algorithm is often used. Here, we study an alternative VQ algorithm based on Kohonen's learning rule for the Self-Organising Map (KSOM; 1982). EMA-VQ is a special case of KSOM. KSOM is known to offer two potential benefits: empirically, it converges faster than EMA-VQ, and KSOM-generated discrete representations form a topological structure on the grid whose nodes are the discrete symbols, resulting in an artificial version of the brain's topographic map. We revisit these properties by using KSOM in VQ-VAEs for image processing. In our experiments, the speed-up compared to well-configured EMA-VQ is only observable at the beginning of training, but KSOM is generally much more robust, e.g., w.r.t. the choice of initialisation schemes.
Authors: Wenlong Chen, Yingzhen Li
Abstract: Transformer models have achieved profound success in prediction tasks in a wide range of applications in natural language processing, speech recognition and computer vision. Extending Transformer's success to safety-critical domains requires calibrated uncertainty estimation which remains under-explored. To address this, we propose Sparse Gaussian Process attention (SGPA), which performs Bayesian inference directly in the output space of multi-head attention blocks (MHAs) in transformer to calibrate its uncertainty. It replaces the scaled dot-product operation with a valid symmetric kernel and uses sparse Gaussian processes (SGP) techniques to approximate the posterior processes of MHA outputs. Empirically, on a suite of prediction tasks on text, images and graphs, SGPA-based Transformers achieve competitive predictive accuracy, while noticeably improving both in-distribution calibration and out-of-distribution robustness and detection.
Authors: Xu Tan, Jiawei Yang, Junqi Chen, Sylwan Rahardja, Susanto Rahardja
Abstract: AutoEncoders (AEs) are commonly used for machine learning tasks due to their intrinsic learning ability. This unique characteristic can be capitalized for Outlier Detection (OD). However conventional AE-based methods face the issue of overconfident decisions and unexpected reconstruction results of outliers, limiting their performance in OD. To mitigate these issues, the Mean Squared Error (MSE) and Negative Logarithmic Likelihood (NLL) were firstly analyzed, and the importance of incorporating aleatoric uncertainty to AE-based OD was elucidated. Then the Weighted Negative Logarithmic Likelihood (WNLL) was proposed to adjust for the effect of uncertainty for different OD scenarios. Moreover, the Mean-Shift Scoring (MSS) method was proposed to utilize the local relationship of data to reduce the issue of false inliers caused by AE. Experiments on 32 real-world OD datasets proved the effectiveness of the proposed methods. The combination of WNLL and MSS achieved 41% relative performance improvement compared to the best baseline. In addition, MSS improved the detection performance of multiple AE-based outlier detectors by an average of 20%. The proposed methods have the potential to advance AE's development in OD.
Authors: Alexander Nikitin, Letizia Iannucci, Samuel Kaski
Abstract: Temporally indexed data are essential in a wide range of fields and of interest to machine learning researchers. Time series data, however, are often scarce or highly sensitive, which precludes the sharing of data between researchers and industrial organizations and the application of existing and new data-intensive ML methods. A possible solution to this bottleneck is to generate synthetic data. In this work, we introduce Time Series Generative Modeling (TSGM), an open-source framework for the generative modeling of synthetic time series. TSGM includes a broad repertoire of machine learning methods: generative models, probabilistic, and simulator-based approaches. The framework enables users to evaluate the quality of the produced data from different angles: similarity, downstream effectiveness, predictive consistency, diversity, and privacy. The framework is extensible, which allows researchers to rapidly implement their own methods and compare them in a shareable environment. TSGM was tested on open datasets and in production and proved to be beneficial in both cases. Additionally to the library, the project allows users to employ command line interfaces for synthetic data generation which lowers the entry threshold for those without a programming background.
Authors: Chanyoung Chung, Jaejun Lee, Joyce Jiyoung Whang
Abstract: A hyper-relational knowledge graph has been recently studied where a triplet is associated with a set of qualifiers; a qualifier is composed of a relation and an entity, providing auxiliary information for a triplet. While existing hyper-relational knowledge graph embedding methods assume that the entities are discrete objects, some information should be represented using numeric values, e.g., (J.R.R., was born in, 1892). Also, a triplet (J.R.R., educated at, Oxford Univ.) can be associated with a qualifier such as (start time, 1911). In this paper, we propose a unified framework named HyNT that learns representations of a hyper-relational knowledge graph containing numeric literals in either triplets or qualifiers. We define a context transformer and a prediction transformer to learn the representations based not only on the correlations between a triplet and its qualifiers but also on the numeric information. By learning compact representations of triplets and qualifiers and feeding them into the transformers, we reduce the computation cost of using transformers. Using HyNT, we can predict missing numeric values in addition to missing entities or relations in a hyper-relational knowledge graph. Experimental results show that HyNT significantly outperforms state-of-the-art methods on real-world datasets.
Authors: Guojun Xiong, Gang Yan, Shiqiang Wang, Jian Li
Abstract: With the increasing demand for large-scale training of machine learning models, fully decentralized optimization methods have recently been advocated as alternatives to the popular parameter server framework. In this paradigm, each worker maintains a local estimate of the optimal parameter vector, and iteratively updates it by waiting and averaging all estimates obtained from its neighbors, and then corrects it on the basis of its local dataset. However, the synchronization phase is sensitive to stragglers. An efficient way to mitigate this effect is to consider asynchronous updates, where each worker computes stochastic gradients and communicates with other workers at its own pace. Unfortunately, fully asynchronous updates suffer from staleness of stragglers' parameters. To address these limitations, we propose a fully decentralized algorithm DSGD-AAU with adaptive asynchronous updates via adaptively determining the number of neighbor workers for each worker to communicate with. We show that DSGD-AAU achieves a linear speedup for convergence and demonstrate its effectiveness via extensive experiments.
Authors: Ping Hu, Virginia Bordignon, Mert Kayaalp, Ali H. Sayed
Abstract: This paper studies the probability of error associated with the social machine learning framework, which involves an independent training phase followed by a cooperative decision-making phase over a graph. This framework addresses the problem of classifying a stream of unlabeled data in a distributed manner. In this work, we examine the classification task with limited observations during the decision-making phase, which requires a non-asymptotic performance analysis. We establish a condition for consistent training and derive an upper bound on the probability of error for classification. The results clarify the dependence on the statistical properties of the data and the combination policy used over the graph. They also establish the exponential decay of the probability of error with respect to the number of unlabeled samples.
Authors: Ngo Nghi Truyen Huynh (INRAE), Pierre-Andr\'e Garambois (INRAE), Fran\c{c}ois Colleoni (INRAE), Benjamin Renard (INRAE), H\'el\`ene Roux (IMFT), Julie Demargne (HYDRIS), Maxime Jay-Allemand (HYDRIS), Pierre Javelle (INRAE)
Abstract: Estimating spatially distributed hydrological parameters in ungauged catchments poses a challenging regionalization problem and requires imposing spatial constraints given the sparsity of discharge data. A possible approach is to search for a transfer function that quantitatively relates physical descriptors to conceptual model parameters. This paper introduces a Hybrid Data Assimilation and Parameter Regionalization (HDA-PR) approach incorporating learnable regionalization mappings, based on either multi-linear regressions or artificial neural networks (ANNs), into a differentiable hydrological model. This approach demonstrates how two differentiable codes can be linked and their gradients chained, enabling the exploitation of heterogeneous datasets across extensive spatio-temporal computational domains within a high-dimensional regionalization context, using accurate adjoint-based gradients. The inverse problem is tackled with a multi-gauge calibration cost function accounting for information from multiple observation sites. HDA-PR was tested on high-resolution, hourly and kilometric regional modeling of 126 flash-flood-prone catchments in the French Mediterranean region. The results highlight a strong regionalization performance of HDA-PR especially in the most challenging upstream-to-downstream extrapolation scenario with ANN, achieving median Nash-Sutcliffe efficiency (NSE) scores from 0.6 to 0.71 for spatial, temporal, spatio-temporal validations, and improving NSE by up to 30% on average compared to the baseline model calibrated with lumped parameters. ANN enables to learn a non-linear descriptors-to-parameters mapping which provides better model controllability than a linear mapping for complex calibration cases.
Authors: Liyuan Wang, Jingyi Xie, Xingxing Zhang, Hang Su, Jun Zhu
Abstract: In this work, we present a general framework for continual learning of sequentially arrived tasks with the use of pre-training, which has emerged as a promising direction for artificial intelligence systems to accommodate real-world dynamics. From a theoretical perspective, we decompose its objective into three hierarchical components, including within-task prediction, task-identity inference, and task-adaptive prediction. Then we propose an innovative approach to explicitly optimize these components with parameter-efficient fine-tuning (PEFT) techniques and representation statistics. We empirically demonstrate the superiority and generality of our approach in downstream continual learning, and further explore the applicability of PEFT techniques in upstream continual learning. We also discuss the biological basis of the proposed framework with recent advances in neuroscience.
Authors: Amir Noorizadegan, D. L. Young, Y. C. Hon, C. S. Chen
Abstract: In this study, we investigate how the updating of weights during forward operation and the computation of gradients during backpropagation impact the optimization process, training procedure, and overall performance of the neural network, particularly the multi-layer perceptrons (MLPs). This paper introduces a novel neural network structure called the Power-Enhancing residual network, inspired by highway network and residual network, designed to improve the network's capabilities for both smooth and non-smooth functions approximation in 2D and 3D settings. By incorporating power terms into residual elements, the architecture enhances the stability of weight updating, thereby facilitating better convergence and accuracy. The study explores network depth, width, and optimization methods, showing the architecture's adaptability and performance advantages. Consistently, the results emphasize the exceptional accuracy of the proposed Power-Enhancing residual network, particularly for non-smooth functions. Real-world examples also confirm its superiority over plain neural network in terms of accuracy, convergence, and efficiency. Moreover, the proposed architecture is also applied to solving the inverse Burgers' equation, demonstrating superior performance. In conclusion, the Power-Enhancing residual network offers a versatile solution that significantly enhances neural network capabilities by emphasizing the importance of stable weight updates for effective training in deep neural networks. The codes implemented are available at: \url{https://github.com/CMMAi/ResNet_for_PINN}.
Authors: Haoyu He, Zizheng Pan, Jing Liu, Jianfei Cai, Bohan Zhuang
Abstract: The paradigm of pre-training and fine-tuning has laid the foundation for deploying deep learning models. However, most fine-tuning methods are designed to meet a specific resource budget. Recently, considering diverse deployment scenarios with various resource budgets, SN-Net is introduced to quickly obtain numerous new networks (stitches) from the pre-trained models (anchors) in a model family via model stitching. Although promising, SN-Net confronts new challenges when adapting it to new target domains, including huge memory and storage requirements and a long and sub-optimal multistage adaptation process. In this work, we present a novel framework, Efficient Stitchable Task Adaptation (ESTA), to efficiently produce a palette of fine-tuned models that adhere to diverse resource constraints. Specifically, we first tailor parameter-efficient fine-tuning to share low-rank updates among the stitches while maintaining independent bias terms. In this way, we largely reduce fine-tuning memory burdens and mitigate the interference among stitches that arises in task adaptation. Furthermore, we streamline a simple yet effective one-stage deployment pipeline, which estimates the important stitches to deploy with training-time gradient statistics. By assigning higher sampling probabilities to important stitches, we also get a boosted Pareto frontier. Extensive experiments on 25 downstream visual recognition tasks demonstrate that our ESTA is capable of generating stitches with smooth accuracy-efficiency trade-offs and surpasses the direct SN-Net adaptation by remarkable margins with significantly lower training time and fewer trainable parameters. Furthermore, we demonstrate the flexibility and scalability of our ESTA framework by stitching LLMs from LLaMA family, obtaining chatbot stitches of assorted sizes. Source code is available at https://github.com/ziplab/Stitched_LLaMA
Authors: Charika De Alvis, Dishanika Denipitiyage, Suranga Seneviratne
Abstract: Integrating supervised contrastive loss to cross entropy-based communication has recently been proposed as a solution to address the long-tail learning problem. However, when the class imbalance ratio is high, it requires adjusting the supervised contrastive loss to support the tail classes, as the conventional contrastive learning is biased towards head classes by default. To this end, we present Rebalanced Contrastive Learning (RCL), an efficient means to increase the long tail classification accuracy by addressing three main aspects: 1. Feature space balancedness - Equal division of the feature space among all the classes, 2. Intra-Class compactness - Reducing the distance between same-class embeddings, 3. Regularization - Enforcing larger margins for tail classes to reduce overfitting. RCL adopts class frequency-based SoftMax loss balancing to supervised contrastive learning loss and exploits scalar multiplied features fed to the contrastive learning loss to enforce compactness. We implement RCL on the Balanced Contrastive Learning (BCL) Framework, which has the SOTA performance. Our experiments on three benchmark datasets demonstrate the richness of the learnt embeddings and increased top-1 balanced accuracy RCL provides to the BCL framework. We further demonstrate that the performance of RCL as a standalone loss also achieves state-of-the-art level accuracy.
Authors: Anran Xu, Lindsey J. Heagy
Abstract: Regularization is critical for solving ill-posed geophysical inverse problems. Explicit regularization is often used, but there are opportunities to explore the implicit regularization effects that are inherent in a Neural Network structure. Researchers have discovered that the Convolutional Neural Network (CNN) architecture inherently enforces a regularization that is advantageous for addressing diverse inverse problems in computer vision, including de-noising and in-painting. In this study, we examine the applicability of this implicit regularization to geophysical inversions. The CNN maps an arbitrary vector to the model space. The predicted subsurface model is then fed into a forward numerical simulation to generate corresponding predicted measurements. Subsequently, the objective function value is computed by comparing these predicted measurements with the observed measurements. The backpropagation algorithm is employed to update the trainable parameters of the CNN during the inversion. Note that the CNN in our proposed method does not require training before the inversion, rather, the CNN weights are estimated in the inversion process, hence this is a test-time learning (TTL) approach. In this study, we choose to focus on the Direct Current (DC) resistivity inverse problem, which is representative of typical Tikhonov-style geophysical inversions (e.g. gravity, electromagnetic, etc.), to test our hypothesis. The experimental results demonstrate that the implicit regularization can be useful in some DC resistivity inversions. We also provide a discussion of the potential sources of this implicit regularization introduced from the CNN architecture and discuss some practical guides for applying the proposed method to other geophysical methods.
Authors: Canlin Zhang, Xiuwen Liu
Abstract: Link prediction is a crucial research area in knowledge graphs, with many downstream applications. In many real-world scenarios, inductive link prediction is required, where predictions have to be made among unseen entities. Embedding-based models usually need fine-tuning on new entity embeddings, and hence are difficult to be directly applied to inductive link prediction tasks. Logical rules captured by rule-based models can be directly applied to new entities with the same graph typologies, but the captured rules are discrete and usually lack generosity. Graph neural networks (GNNs) can generalize topological information to new graphs taking advantage of deep neural networks, which however may still need fine-tuning on new entity embeddings. In this paper, we propose SiaILP, a path-based model for inductive link prediction using siamese neural networks. Our model only depends on relation and path embeddings, which can be generalized to new entities without fine-tuning. Experiments show that our model achieves several new state-of-the-art performances in link prediction tasks using inductive versions of WN18RR, FB15k-237, and Nell995. Our code is available at \url{https://github.com/canlinzhang/SiaILP}.
Authors: Daniel Coquelin, Katharina Fl\"ugel, Marie Weiel, Nicholas Kiefer, Charlotte Debus, Achim Streit, Markus G\"otz
Abstract: This study explores the learning dynamics of neural networks by analyzing the singular value decomposition (SVD) of their weights throughout training. Our investigation reveals that an orthogonal basis within each multidimensional weight's SVD representation stabilizes during training. Building upon this, we introduce Orthogonality-Informed Adaptive Low-Rank (OIALR) training, a novel training method exploiting the intrinsic orthogonality of neural networks. OIALR seamlessly integrates into existing training workflows with minimal accuracy loss, as demonstrated by benchmarking on various datasets and well-established network architectures. With appropriate hyperparameter tuning, OIALR can surpass conventional training setups, including those of state-of-the-art models.
Authors: Patrick Cook, Danny Jammooa, Morten Hjorth-Jensen, Daniel D. Lee, Dean Lee
Abstract: We present a general class of machine learning algorithms called parametric matrix models. In contrast with most existing machine learning models that imitate the biology of neurons, parametric matrix models use matrix equations that emulate the physics of quantum systems. Similar to how physics problems are usually solved, parametric matrix models learn the governing equations that lead to the desired outputs. Parametric matrix models can be efficiently trained from empirical data, and the equations may use algebraic, differential, or integral relations. While originally designed for scientific computing, we prove that parametric matrix models are universal function approximators that can be applied to general machine learning problems. After introducing the underlying theory, we apply parametric matrix models to a series of different challenges that show their performance for a wide range of problems. For all the challenges tested here, parametric matrix models produce accurate results within an efficient and interpretable computational framework that allows for input feature extrapolation.
Authors: Andy Zhou, Bo Li, Haohan Wang
Abstract: Despite advances in AI alignment, large language models (LLMs) remain vulnerable to adversarial attacks or jailbreaking, in which adversaries can modify prompts to induce unwanted behavior. While some defenses have been proposed, they have not been adapted to newly proposed attacks and more challenging threat models. To address this, we propose an optimization-based objective for defending LLMs against jailbreaking attacks and an algorithm, Robust Prompt Optimization (RPO) to create robust system-level defenses. Our approach directly incorporates the adversary into the defensive objective and optimizes a lightweight and transferable suffix, enabling RPO to adapt to worst-case adaptive attacks. Our theoretical and experimental results show improved robustness to both jailbreaks seen during optimization and unknown jailbreaks, reducing the attack success rate (ASR) on GPT-4 to 6% and Llama-2 to 0% on JailbreakBench, setting the state-of-the-art. Code can be found at https://github.com/lapisrocks/rpo
Authors: Xingyu Qu, Samuel Horvath
Abstract: Recent studies suggest that with sufficiently wide models, most SGD solutions can, up to permutation, converge into the same basin. This phenomenon, known as the model re-basin regime, has significant implications for model averaging by ensuring the linear mode connectivity. However, current re-basin strategies are ineffective in many scenarios due to a lack of comprehensive understanding of underlying mechanisms. Addressing this gap, this paper provides novel insights into understanding and improving the standard practice. Firstly, we decompose re-normalization into rescaling and reshift, uncovering that rescaling plays a crucial role in re-normalization while re-basin performance is sensitive to shifts in model activation. The finding calls for a more nuanced handling of the activation shift. Secondly, we identify that the merged model suffers from the issue of activation collapse and magnitude collapse. Varying the learning rate, weight decay, and initialization method can mitigate the issues and improve model performance. Lastly, we propose a new perspective to unify the re-basin and pruning, under which a lightweight yet effective post-pruning technique is derived, which can significantly improve the model performance after pruning. Our implementation is available at https://github.com/XingyuQu/rethink-re-basin.
Authors: Ayush K Tarun, Vikram S Chundawat, Murari Mandal, Hong Ming Tan, Bowei Chen, Mohan Kankanhalli
Abstract: Quantifying the value of data within a machine learning workflow can play a pivotal role in making more strategic decisions in machine learning initiatives. The existing Shapley value based frameworks for data valuation in machine learning are computationally expensive as they require considerable amount of repeated training of the model to obtain the Shapley value. In this paper, we introduce an efficient data valuation framework EcoVal, to estimate the value of data for machine learning models in a fast and practical manner. Instead of directly working with individual data sample, we determine the value of a cluster of similar data points. This value is further propagated amongst all the member cluster points. We show that the overall value of the data can be determined by estimating the intrinsic and extrinsic value of each data. This is enabled by formulating the performance of a model as a \textit{production function}, a concept which is popularly used to estimate the amount of output based on factors like labor and capital in a traditional free economic market. We provide a formal proof of our valuation technique and elucidate the principles and mechanisms that enable its accelerated performance. We demonstrate the real-world applicability of our method by showcasing its effectiveness for both in-distribution and out-of-sample data. This work addresses one of the core challenges of efficient data valuation at scale in machine learning models. The code is available at \underline{https://github.com/respai-lab/ecoval}.
Authors: Taisuke Kobayashi
Abstract: Experience replay (ER) used in (deep) reinforcement learning is considered to be applicable only to off-policy algorithms. However, there have been some cases in which ER has been applied for on-policy algorithms, suggesting that off-policyness might be a sufficient condition for applying ER. This paper reconsiders more strict "experience replayable conditions" (ERC) and proposes the way of modifying the existing algorithms to satisfy ERC. In light of this, it is postulated that the instability of policy improvements represents a pivotal factor in ERC. The instability factors are revealed from the viewpoint of metric learning as i) repulsive forces from negative samples and ii) replays of inappropriate experiences. Accordingly, the corresponding stabilization tricks are derived. As a result, it is confirmed through numerical simulations that the proposed stabilization tricks make ER applicable to an advantage actor-critic, an on-policy algorithm. Moreover, its learning performance is comparable to that of a soft actor-critic, a state-of-the-art off-policy algorithm.
Authors: Creston Brooks, Robert Calef, Charlie Cowen-Breen, Anna Sappington
Abstract: Beam search with masked language models (MLMs) is challenging in part because joint probability distributions over sequences are not readily available, unlike for autoregressive models. However, estimating such distributions has important domain-specific applications such as ancient text restoration and protein engineering. Here we present probabilistically-sound methods for beam search with MLMs. First, we clarify the conditions under which it is theoretically sound to perform text infilling with MLMs using standard beam search. When these conditions fail, we provide a probabilistically-sound modification with no additional computational complexity and demonstrate that it is superior to the aforementioned beam search in the expected conditions. We then present empirical results comparing several infilling approaches with MLMs across several domains.
Authors: Xiaoying Zhang, Jean-Francois Ton, Wei Shen, Hongning Wang, Yang Liu
Abstract: We introduce Adversarial Policy Optimization (AdvPO), a novel solution to the pervasive issue of reward over-optimization in Reinforcement Learning from Human Feedback (RLHF) for Large Language Models (LLMs). Over-optimization occurs when a reward model serves as an imperfect proxy for human preference, and RL-driven policy optimization erroneously exploits reward inaccuracies. In this paper, we begin by introducing a lightweight way to quantify uncertainties in rewards, relying solely on the last layer embeddings of the reward model, without the need for computationally expensive reward ensembles. AdvPO then addresses a distributionally robust optimization problem centred around the confidence interval of the reward model's predictions for policy improvement. Through comprehensive experiments on the Anthropic HH and TL;DR summarization datasets, we illustrate the efficacy of AdvPO in mitigating the overoptimization issue, consequently resulting in enhanced performance as evaluated through human-assisted evaluation.
Authors: Chongyu Fan, Jiancheng Liu, Alfred Hero, Sijia Liu
Abstract: The trustworthy machine learning (ML) community is increasingly recognizing the crucial need for models capable of selectively 'unlearning' data points after training. This leads to the problem of machine unlearning (MU), aiming to eliminate the influence of chosen data points on model performance, while still maintaining the model's utility post-unlearning. Despite various MU methods for data influence erasure, evaluations have largely focused on random data forgetting, ignoring the vital inquiry into which subset should be chosen to truly gauge the authenticity of unlearning performance. To tackle this issue, we introduce a new evaluative angle for MU from an adversarial viewpoint. We propose identifying the data subset that presents the most significant challenge for influence erasure, i.e., pinpointing the worst-case forget set. Utilizing a bi-level optimization principle, we amplify unlearning challenges at the upper optimization level to emulate worst-case scenarios, while simultaneously engaging in standard training and unlearning at the lower level, achieving a balance between data influence erasure and model utility. Our proposal offers a worst-case evaluation of MU's resilience and effectiveness. Through extensive experiments across different datasets (including CIFAR-10, 100, CelebA, Tiny ImageNet, and ImageNet) and models (including both image classifiers and generative models), we expose critical pros and cons in existing (approximate) unlearning strategies. Our results illuminate the complex challenges of MU in practice, guiding the future development of more accurate and robust unlearning algorithms. The code is available at https://github.com/OPTML-Group/Unlearn-WorstCase.
Authors: Yushan Huang, Ranya Aloufi, Xavier Cadet, Yuchen Zhao, Payam Barnaghi, Hamed Haddadi
Abstract: We propose MicroT, a low-energy, multi-task adaptive model framework for resource-constrained MCUs. We divide the original model into a feature extractor and a classifier. The feature extractor is obtained through self-supervised knowledge distillation and further optimized into part and full models through model splitting and joint training. These models are then deployed on MCUs, with classifiers added and trained on local tasks, ultimately performing stage-decision for joint inference. In this process, the part model initially processes the sample, and if the confidence score falls below the set threshold, the full model will resume and continue the inference. We evaluate MicroT on two models, three datasets, and two MCU boards. Our experimental evaluation shows that MicroT effectively improves model performance and reduces energy consumption when dealing with multiple local tasks. Compared to the unoptimized feature extractor, MicroT can improve accuracy by up to 9.87%. On MCUs, compared to the standard full model inference, MicroT can save up to about 29.13% in energy consumption. MicroT also allows users to adaptively adjust the stage-decision ratio as needed, better balancing model performance and energy consumption. Under the standard stage-decision ratio configuration, MicroT can increase accuracy by 5.91% and save about 14.47% of energy consumption.
Authors: Yu Cai, Hao Chen, Kwang-Ting Cheng
Abstract: Medical anomaly detection aims to identify abnormal findings using only normal training data, playing a crucial role in health screening and recognizing rare diseases. Reconstruction-based methods, particularly those utilizing autoencoders (AEs), are dominant in this field. They work under the assumption that AEs trained on only normal data cannot reconstruct unseen abnormal regions well, thereby enabling the anomaly detection based on reconstruction errors. However, this assumption does not always hold due to the mismatch between the reconstruction training objective and the anomaly detection task objective, rendering these methods theoretically unsound. This study focuses on providing a theoretical foundation for AE-based reconstruction methods in anomaly detection. By leveraging information theory, we elucidate the principles of these methods and reveal that the key to improving AE in anomaly detection lies in minimizing the information entropy of latent vectors. Experiments on four datasets with two image modalities validate the effectiveness of our theory. To the best of our knowledge, this is the first effort to theoretically clarify the principles and design philosophy of AE for anomaly detection. The code is available at \url{https://github.com/caiyu6666/AE4AD}.
Authors: Daniele Mari, Andr\'e F. R. Guarda, Nuno M. M. Rodrigues, Simone Milani, Fernando Pereira
Abstract: The widespread usage of point clouds (PC) for immersive visual applications has resulted in the use of very heterogeneous receiving conditions and devices, notably in terms of network, hardware, and display capabilities. In this scenario, quality scalability, i.e., the ability to reconstruct a signal at different qualities by progressively decoding a single bitstream, is a major requirement that has yet to be conveniently addressed, notably in most learning-based PC coding solutions. This paper proposes a quality scalability scheme, named Scalable Quality Hyperprior (SQH), adaptable to learning-based static point cloud geometry codecs, which uses a Quality-conditioned Latents Probability Estimator (QuLPE) to decode a high-quality version of a PC learning-based representation, based on an available lower quality base layer. SQH is integrated in the future JPEG PC coding standard, allowing to create a layered bitstream that can be used to progressively decode the PC geometry with increasing quality and fidelity. Experimental results show that SQH offers the quality scalability feature with very limited or no compression performance penalty at all when compared with the corresponding non-scalable solution, thus preserving the significant compression gains over other state-of-the-art PC codecs.
Authors: Lars Ullrich, Alex McMaster, Knut Graichen
Abstract: Trajectory planning in autonomous driving is highly dependent on predicting the emergent behavior of other road users. Learning-based methods are currently showing impressive results in simulation-based challenges, with transformer-based architectures technologically leading the way. Ultimately, however, predictions are needed in the real world. In addition to the shifts from simulation to the real world, many vehicle- and country-specific shifts, i.e. differences in sensor systems, fusion and perception algorithms as well as traffic rules and laws, are on the agenda. Since models that can cover all system setups and design domains at once are not yet foreseeable, model adaptation plays a central role. Therefore, a simulation-based study on transfer learning techniques is conducted on basis of a transformer-based model. Furthermore, the study aims to provide insights into possible trade-offs between computational time and performance to support effective transfers into the real world.
Authors: Roussel Desmond Nzoyem, David A. W. Barton, Tom Deakin
Abstract: Neural Ordinary Differential Equations typically struggle to generalize to new dynamical behaviors created by parameter changes in the underlying system, even when the dynamics are close to previously seen behaviors. The issue gets worse when the changing parameters are unobserved, i.e., their value or influence is not directly measurable when collecting data. We introduce Neural Context Flow (NCF), a framework that encodes said unobserved parameters in a latent context vector as input to a vector field. NCFs leverage differentiability of the vector field with respect to the parameters, along with first-order Taylor expansion to allow any context vector to influence trajectories from other parameters. We validate our method and compare it to established Multi-Task and Meta-Learning alternatives, showing competitive performance in mean squared error for in-domain and out-of-distribution evaluation on the Lotka-Volterra, Glycolytic Oscillator, and Gray-Scott problems. This study holds practical implications for foundational models in science and related areas that benefit from conditional neural ODEs. Our code is openly available at https://github.com/ddrous/ncflow.
Authors: Ruikai Yang, Fan He, Mingzhen He, Kaijie Wang, Xiaolin Huang
Abstract: Data imputation, the process of filling in missing feature elements for incomplete data sets, plays a crucial role in data-driven learning. A fundamental belief is that data imputation is helpful for learning performance, and it follows that the pursuit of better classification can guide the data imputation process. While some works consider using label information to assist in this task, their simplistic utilization of labels lacks flexibility and may rely on strict assumptions. In this paper, we propose a new framework that effectively leverages supervision information to complete missing data in a manner conducive to classification. Specifically, this framework operates in two stages. Firstly, it leverages labels to supervise the optimization of similarity relationships among data, represented by the kernel matrix, with the goal of enhancing classification accuracy. To mitigate overfitting that may occur during this process, a perturbation variable is introduced to improve the robustness of the framework. Secondly, the learned kernel matrix serves as additional supervision information to guide data imputation through regression, utilizing the block coordinate descent method. The superiority of the proposed method is evaluated on four real-world data sets by comparing it with state-of-the-art imputation methods. Remarkably, our algorithm significantly outperforms other methods when the data is missing more than 60\% of the features
Authors: Jun Xia, Yi Zhang, Yiyu Shi
Abstract: Although Federated Learning (FL) is promising in knowledge sharing for heterogeneous Artificial Intelligence of Thing (AIoT) devices, their training performance and energy efficacy are severely restricted in practical battery-driven scenarios due to the ``wooden barrel effect'' caused by the mismatch between homogeneous model paradigms and heterogeneous device capability. As a result, due to various kinds of differences among devices, it is hard for existing FL methods to conduct training effectively in energy-constrained scenarios, such as battery constraints of devices. To tackle the above issues, we propose an energy-aware FL framework named DR-FL, which considers the energy constraints in both clients and heterogeneous deep learning models to enable energy-efficient FL. Unlike Vanilla FL, DR-FL adopts our proposed Muti-Agents Reinforcement Learning (MARL)-based dual-selection method, which allows participated devices to make contributions to the global model effectively and adaptively based on their computing capabilities and energy capacities in a MARL-based manner. Experiments conducted with various widely recognized datasets demonstrate that DR-FL has the capability to optimize the exchange of knowledge among diverse models in large-scale AIoT systems while adhering to energy limitations. Additionally, it improves the performance of each individual heterogeneous device's model.
Authors: Michael Lu, Matin Aghaei, Anant Raj, Sharan Vaswani
Abstract: We consider (stochastic) softmax policy gradient (PG) methods for bandits and tabular Markov decision processes (MDPs). While the PG objective is non-concave, recent research has used the objective's smoothness and gradient domination properties to achieve convergence to an optimal policy. However, these theoretical results require setting the algorithm parameters according to unknown problem-dependent quantities (e.g. the optimal action or the true reward vector in a bandit problem). To address this issue, we borrow ideas from the optimization literature to design practical, principled PG methods in both the exact and stochastic settings. In the exact setting, we employ an Armijo line-search to set the step-size for softmax PG and demonstrate a linear convergence rate. In the stochastic setting, we utilize exponentially decreasing step-sizes, and characterize the convergence rate of the resulting algorithm. We show that the proposed algorithm offers similar theoretical guarantees as the state-of-the art results, but does not require the knowledge of oracle-like quantities. For the multi-armed bandit setting, our techniques result in a theoretically-principled PG algorithm that does not require explicit exploration, the knowledge of the reward gap, the reward distributions, or the noise. Finally, we empirically compare the proposed methods to PG approaches that require oracle knowledge, and demonstrate competitive performance.
Authors: Fran\c{c}ois Rozet, G\'er\^ome Andry, Fran\c{c}ois Lanusse, Gilles Louppe
Abstract: Diffusion models recently proved to be remarkable priors for Bayesian inverse problems. However, training these models typically requires access to large amounts of clean data, which could prove difficult in some settings. In this work, we present a novel method based on the expectation-maximization algorithm for training diffusion models from incomplete and noisy observations only. Unlike previous works, our method leads to proper diffusion models, which is crucial for downstream tasks. As part of our method, we propose and motivate a new posterior sampling scheme for unconditional diffusion models. We present empirical evidence supporting the effectiveness of our method.
Authors: Shi Mu, Shunpan Liang, Xiang Li
Abstract: Medication recommendation combines patient medical history with biomedical knowledge to assist doctors in determining medication combinations more accurately and safely. Existing works based on molecular knowledge neglect the 3D geometric structure of molecules and fail to learn the high-dimensional information of medications, leading to structural confusion. Additionally, it does not extract key substructures from a single patient visit, resulting in the failure to identify medication molecules suitable for the current patient visit. To address the above limitations, we propose a bimodal molecular recommendation framework named BiMoRec, which introduces 3D molecular structures to obtain atomic 3D coordinates and edge indices, overcoming the inherent lack of high-dimensional molecular information in 2D molecular structures. To retain the fast training and prediction efficiency of the recommendation system, we use bimodal graph contrastive pretraining to maximize the mutual information between the two molecular modalities, achieving the fusion of 2D and 3D molecular graphs and re-evaluating substructures at the visit level. Specifically, we use deep learning networks to construct a pretraining method that acquires 2D and 3D molecular structure representations and substructure representations, and obtain mutual information through contrastive learning. We then generate fused molecular representations using the trained GNN module and re-determine the relevance of substructure representations in combination with the patient's clinical history. Finally, we generate the final medication combination based on the extracted substructure sequences. Our implementation on the MIMIC-III and MIMIC-IV datasets demonstrates that our method achieves state-of-the-art performance. Compared to the second-best baseline, our model improves accuracy by 2.07%, with DDI at the same level as the baseline.
Authors: Alexander Bukharin, Ilgee Hong, Haoming Jiang, Zichong Li, Qingru Zhang, Zixuan Zhang, Tuo Zhao
Abstract: Reinforcement learning from human feedback (RLHF) provides a principled framework for aligning AI systems with human preference data. For various reasons, e.g., personal bias, context ambiguity, lack of training, etc, human annotators may give incorrect or inconsistent preference labels. To tackle this challenge, we propose a robust RLHF approach -- $R^3M$, which models the potentially corrupted preference label as sparse outliers. Accordingly, we formulate the robust reward learning as an $\ell_1$-regularized maximum likelihood estimation problem. Computationally, we develop an efficient alternating optimization algorithm, which only incurs negligible computational overhead compared with the standard RLHF approach. Theoretically, we prove that under proper regularity conditions, $R^3M$ can consistently learn the underlying reward and identify outliers, provided that the number of outlier labels scales sublinearly with the preference sample size. Furthermore, we remark that $R^3M$ is versatile and can be extended to various preference optimization methods, including direct preference optimization (DPO). Our experiments on robotic control and natural language generation with large language models (LLMs) show that $R^3M$ improves robustness of the reward against several types of perturbations to the preference data.
Authors: Zhiyu Wu, Jinshi Cui
Abstract: Existing semi-supervised learning algorithms adopt pseudo-labeling and consistency regulation techniques to introduce supervision signals for unlabeled samples. To overcome the inherent limitation of threshold-based pseudo-labeling, prior studies have attempted to align the confidence threshold with the evolving learning status of the model, which is estimated through the predictions made on the unlabeled data. In this paper, we further reveal that classifier weights can reflect the differentiated learning status across categories and consequently propose a class-specific adaptive threshold mechanism. Additionally, considering that even the optimal threshold scheme cannot resolve the problem of discarding unlabeled samples, a binary classification consistency regulation approach is designed to distinguish candidate classes from negative options for all unlabeled samples. By combining the above strategies, we present a novel SSL algorithm named AllMatch, which achieves improved pseudo-label accuracy and a 100% utilization ratio for the unlabeled data. We extensively evaluate our approach on multiple benchmarks, encompassing both balanced and imbalanced settings. The results demonstrate that AllMatch consistently outperforms existing state-of-the-art methods.
Authors: Yixin Jin, Wenjing Zhou, Meiqi Wang, Meng Li, Xintao Li, Tianyu Hu
Abstract: This paper examines an online multi-task learning (OMTL) method, which processes data sequentially to predict labels across related tasks. The framework learns task weights and their relatedness concurrently. Unlike previous models that assumed static task relatedness, our approach treats tasks as initially independent, updating their relatedness iteratively using newly calculated weight vectors. We introduced three rules to update the task relatedness matrix: OMTLCOV, OMTLLOG, and OMTLVON, and compared them against a conventional method (CMTL) that uses a fixed relatedness value. Performance evaluations on three datasets a spam dataset and two EEG datasets from construction workers under varying conditions demonstrated that our OMTL methods outperform CMTL, improving accuracy by 1% to 3% on EEG data, and maintaining low error rates around 12% on the spam dataset.
Authors: Yifang Chen, Shuohang Wang, Ziyi Yang, Hiteshi Sharma, Nikos Karampatziakis, Donghan Yu, Kevin Jamieson, Simon Shaolei Du, Yelong Shen
Abstract: Reinforcement learning with human feedback (RLHF), as a widely adopted approach in current large language model pipelines, is \textit{bottlenecked by the size of human preference data}. While traditional methods rely on offline preference dataset constructions, recent approaches have shifted towards online settings, where a learner uses a small amount of labeled seed data and a large pool of unlabeled prompts to iteratively construct new preference data through self-generated responses and high-quality reward/preference feedback. However, most current online algorithms still focus on preference labeling during policy model updating with given feedback oracles, which incurs significant expert query costs. \textit{We are the first to explore cost-effective proxy reward oracles construction strategies for further labeling preferences or rewards with extremely limited labeled data and expert query budgets}. Our approach introduces two key innovations: (1) on-policy query to avoid OOD and imbalance issues in seed data, and (2) active learning to select the most informative data for preference queries. Using these methods, we train a evaluation model with minimal expert-labeled data, which then effectively labels nine times more preference pairs for further RLHF training. For instance, our model using Direct Preference Optimization (DPO) gains around over 1% average improvement on AlpacaEval2, MMLU-5shot and MMLU-0shot, with only 1.7K query cost. Our methodology is orthogonal to other direct expert query-based strategies and therefore might be integrated with them to further reduce query costs.
Authors: Yuxuan Zhu, Michael Mandulak, Kerui Wu, George Slota, Yuseok Jeon, Ka-Ho Chow, Lei Yu
Abstract: Graph Neural Networks (GNNs) are gaining popularity across various domains due to their effectiveness in learning graph-structured data. Nevertheless, they have been shown to be susceptible to backdoor poisoning attacks, which pose serious threats to real-world applications. Meanwhile, graph reduction techniques, including coarsening and sparsification, which have long been employed to improve the scalability of large graph computational tasks, have recently emerged as effective methods for accelerating GNN training on large-scale graphs. However, the current development and deployment of graph reduction techniques for large graphs overlook the potential risks of data poisoning attacks against GNNs. It is not yet clear how graph reduction interacts with existing backdoor attacks. This paper conducts a thorough examination of the robustness of graph reduction methods in scalable GNN training in the presence of state-of-the-art backdoor attacks. We performed a comprehensive robustness analysis across six coarsening methods and six sparsification methods for graph reduction, under three GNN backdoor attacks against three GNN architectures. Our findings indicate that the effectiveness of graph reduction methods in mitigating attack success rates varies significantly, with some methods even exacerbating the attacks. Through detailed analyses of triggers and poisoned nodes, we interpret our findings and enhance our understanding of how graph reduction influences robustness against backdoor attacks. These results highlight the critical need for incorporating robustness considerations in graph reduction for GNN training, ensuring that enhancements in computational efficiency do not compromise the security of GNN systems.
Authors: Xiucheng Wang, Qi Qiu, Nan Cheng
Abstract: In this paper, we investigate a special class of quadratic-constrained quadratic programming (QCQP) with semi-definite constraints. Traditionally, since such a problem is non-convex and N-hard, the neural network (NN) is regarded as a promising method to obtain a high-performing solution. However, due to the inherent prediction error, it is challenging to ensure all solution output by the NN is feasible. Although some existing methods propose some naive methods, they only focus on reducing the constraint violation probability, where not all solutions are feasibly guaranteed. To deal with the above challenge, in this paper a computing efficient and reliable projection is proposed, where all solution output by the NN are ensured to be feasible. Moreover, unsupervised learning is used, so the NN can be trained effectively and efficiently without labels. Theoretically, the solution of the NN after projection is proven to be feasible, and we also prove the projection method can enhance the convergence performance and speed of the NN. To evaluate our proposed method, the quality of service (QoS)-contained beamforming scenario is studied, where the simulation results show the proposed method can achieve high-performance which is competitive with the lower bound.
Authors: Peiyu Yang, Naveed Akhtar, Mubarak Shah, Ajmal Mian
Abstract: Trustworthy machine learning necessitates meticulous regulation of model reliance on non-robust features. We propose a framework to delineate and regulate such features by attributing model predictions to the input. Within our approach, robust feature attributions exhibit a certain consistency, while non-robust feature attributions are susceptible to fluctuations. This behavior allows identification of correlation between model reliance on non-robust features and smoothness of marginal density of the input samples. Hence, we uniquely regularize the gradients of the marginal density w.r.t. the input features for robustness. We also devise an efficient implementation of our regularization to address the potential numerical instability of the underlying optimization process. Moreover, we analytically reveal that, as opposed to our marginal density smoothing, the prevalent input gradient regularization smoothens conditional or joint density of the input, which can cause limited robustness. Our experiments validate the effectiveness of the proposed method, providing clear evidence of its capability to address the feature leakage problem and mitigate spurious correlations. Extensive results further establish that our technique enables the model to exhibit robustness against perturbations in pixel values, input gradients, and density.
Authors: Xiaojin Zhang, Mingcong Xu, Wei Chen
Abstract: In this paper, we first give an introduction to the theoretical basis of the privacy-utility equilibrium in federated learning based on Bayesian privacy definitions and total variation distance privacy definitions. We then present the \textit{Learn-to-Distort-Data} framework, which provides a principled approach to navigate the privacy-utility equilibrium by explicitly modeling the distortion introduced by the privacy-preserving mechanism as a learnable variable and optimizing it jointly with the model parameters. We demonstrate the applicability of our framework to a variety of privacy-preserving mechanisms on the basis of data distortion and highlight its connections to related areas such as adversarial training, input robustness, and unlearnable examples. These connections enable leveraging techniques from these areas to design effective algorithms for privacy-utility equilibrium in federated learning under the \textit{Learn-to-Distort-Data} framework.
Authors: Ryan Fayyazi, Christian Weilbach, Frank Wood
Abstract: Inter-neuron communication delays are ubiquitous in physically realized neural networks such as biological neural circuits and neuromorphic hardware. These delays have significant and often disruptive consequences on network dynamics during training and inference. It is therefore essential that communication delays be accounted for, both in computational models of biological neural networks and in large-scale neuromorphic systems. Nonetheless, communication delays have yet to be comprehensively addressed in either domain. In this paper, we first show that delays prevent state-of-the-art continuous-time neural networks called Latent Equilibrium (LE) networks from learning even simple tasks despite significant overparameterization. We then propose to compensate for communication delays by predicting future signals based on currently available ones. This conceptually straightforward approach, which we call prospective messaging (PM), uses only neuron-local information, and is flexible in terms of memory and computation requirements. We demonstrate that incorporating PM into delayed LE networks prevents reaction lags, and facilitates successful learning on Fourier synthesis and autoregressive video prediction tasks.
Authors: Rui Zhu, Shuvrarghya Ghosh, Subhashis Ghosal
Abstract: Semi-supervised learning is a model training method that uses both labeled and unlabeled data. This paper proposes a fully Bayes semi-supervised learning algorithm that can be applied to any binary classification problem. We assume the labels are missing at random when using unlabeled data in a semi-supervised setting. We assume that the observations follow two multivariate normal distributions depending on their true class labels after some common unknown transformation is applied to each component of the observation vector. The function is expanded in a B-splines series and a prior is put on the coefficients. We consider a normal prior on the coefficients and constrain the values to meet the requirement for normality and identifiability constraints. The precision matrices of the two Gaussian distributions have a conjugate Wishart prior, while the means have improper uniform priors. The resulting posterior is still conditionally conjugate, and the Gibbs sampler aided by a data augmentation technique can thus be adopted. An extensive simulation study compares the proposed method with several other available methods. The proposed method is also applied to real datasets on diagnosing breast cancer and classification of signals. We conclude that the proposed method has a better prediction accuracy in various cases.
Authors: Tao Ni, Qing Wang, Gabriela Ferraro
Abstract: Extracting multiple relations from text sentences is still a challenge for current Open Relation Extraction (Open RE) tasks. In this paper, we develop several Open RE models based on the bidirectional LSTM-CRF (BiLSTM-CRF) neural network and different contextualized word embedding methods. We also propose a new tagging scheme to solve overlapping problems and enhance models' performance. From the evaluation results and comparisons between models, we select the best combination of tagging scheme, word embedder, and BiLSTM-CRF network to achieve an Open RE model with a remarkable extracting ability on multiple-relation sentences.
Authors: Dawei Chen, Samuel Yang-Zhao, John Lloyd, Kee Siong Ng
Abstract: This paper introduces factored conditional filters, new filtering algorithms for simultaneously tracking states and estimating parameters in high-dimensional state spaces. The conditional nature of the algorithms is used to estimate parameters and the factored nature is used to decompose the state space into low-dimensional subspaces in such a way that filtering on these subspaces gives distributions whose product is a good approximation to the distribution on the entire state space. The conditions for successful application of the algorithms are that observations be available at the subspace level and that the transition model can be factored into local transition models that are approximately confined to the subspaces; these conditions are widely satisfied in computer science, engineering, and geophysical filtering applications. We give experimental results on tracking epidemics and estimating parameters in large contact networks that show the effectiveness of our approach.
Authors: Gideon Kowadlo, Abdelrahman Ahmed, Amir Mayan, David Rawlinson
Abstract: Continual learning and few-shot learning are important frontiers in progress toward broader Machine Learning (ML) capabilities. Recently, there has been intense interest in combining both. One of the first examples to do so was the Continual few-shot Learning (CFSL) framework of Antoniou et al. arXiv:2004.11967. In this study, we extend CFSL in two ways that capture a broader range of challenges, important for intelligent agent behaviour in real-world conditions. First, we increased the number of classes by an order of magnitude, making the results more comparable to standard continual learning experiments. Second, we introduced an 'instance test' which requires recognition of specific instances of classes -- a capability of animal cognition that is usually neglected in ML. For an initial exploration of ML model performance under these conditions, we selected representative baseline models from the original CFSL work and added a model variant with replay. As expected, learning more classes is more difficult than the original CFSL experiments, and interestingly, the way in which image instances and classes are presented affects classification performance. Surprisingly, accuracy in the baseline instance test is comparable to other classification tasks, but poor given significant occlusion and noise. The use of replay for consolidation substantially improves performance for both types of tasks, but particularly for the instance test.
Authors: Aurelian Gheondea, Cankat Tilki
Abstract: We prove a few representer theorems for a localised version of the regularised and multiview support vector machine learning problem introduced by H.Q. Minh, L. Bazzani, and V. Murino, Journal of Machine Learning Research, 17(2016) 1-72, that involves operator valued positive semidefinite kernels and their reproducing kernel Hilbert spaces. The results concern general cases when convex or nonconvex loss functions and finite or infinite dimensional input spaces are considered. We show that the general framework allows infinite dimensional input spaces and nonconvex loss functions for some special cases, in particular in case the loss functions are Gateaux differentiable. Detailed calculations are provided for the exponential least square loss function that lead to partially nonlinear equations for which a particular unconstrained potential reduction Newton's approximation method can be used.
Authors: Daniel Golovin, Gabor Bartok, Eric Chen, Emily Donahue, Tzu-Kuo Huang, Efi Kokiopoulou, Ruoyan Qin, Nikhil Sarda, Justin Sybrandt, Vincent Tjeng
Abstract: In many software systems, heuristics are used to make decisions - such as cache eviction, task scheduling, and information presentation - that have a significant impact on overall system behavior. While machine learning may outperform these heuristics, replacing existing heuristics in a production system safely and reliably can be prohibitively costly. We present SmartChoices, a novel approach that reduces the cost to deploy production-ready ML solutions for contextual bandits problems. SmartChoices' interface cleanly separates problem formulation from implementation details: engineers describe their use case by defining datatypes for the context, arms, and feedback that are passed to SmartChoices APIs, while SmartChoices manages encoding & logging data and training, evaluating & deploying policies. Our implementation codifies best practices, is efficient enough for use in low-level applications, and provides valuable production features off the shelf via a shared library. Overall, SmartChoices enables non-experts to rapidly deploy production-ready ML solutions by eliminating many sources of technical debt common to ML systems. Engineers have independently used SmartChoices to improve a wide range of software including caches, batch processing workloads, and UI layouts, resulting in better latency, throughput, and click-through rates.
Authors: Hongyang Chen, Yuhong Yang, Zhongyuan Wang, Weiping Tu, Haojun Ai, Song Lin
Abstract: This study explores how sentence types affect the Lombard effect and intelligibility enhancement, focusing on comparisons between natural and grid sentences. Using the Lombard Chinese-TIMIT (LCT) corpus and the Enhanced MAndarin Lombard Grid (EMALG) corpus, we analyze changes in phonetic and acoustic features across different noise levels. Our results show that grid sentences produce more pronounced Lombard effects than natural sentences. Then, we develop and test a normal-to-Lombard conversion model, trained separately on LCT and EMALG corpora. Through subjective and objective evaluations, natural sentences are superior in maintaining speech quality in intelligibility enhancement. In contrast, grid sentences could provide superior intelligibility due to the more pronounced Lombard effect. This study provides a valuable perspective on enhancing speech communication in noisy environments.
Authors: Utsab Saha, Sawradip Saha, Tahmid Kabir, Shaikh Anowarul Fattah, Mohammad Saquib
Abstract: A person's movement or relative positioning can be effectively captured by different types of sensors and corresponding sensor output can be utilized in various manipulative techniques for the classification of different human activities. This letter proposes an effective scheme for human activity recognition, which introduces two unique approaches within a multi-structural architecture, named FusionActNet. The first approach aims to capture the static and dynamic behavior of a particular action by using two dedicated residual networks and the second approach facilitates the final decision-making process by introducing a guidance module. A two-stage training process is designed where at the first stage, residual networks are pre-trained separately by using static (where the human body is immobile) and dynamic (involving movement of the human body) data. In the next stage, the guidance module along with the pre-trained static or dynamic models are used to train the given sensor data. Here the guidance module learns to emphasize the most relevant prediction vector obtained from the static or dynamic models, which helps to effectively classify different human activities. The proposed scheme is evaluated using two benchmark datasets and compared with state-of-the-art methods. The results clearly demonstrate that our method outperforms existing approaches in terms of accuracy, precision, recall, and F1 score, achieving 97.35% and 95.35% accuracy on the UCI HAR and Motion-Sense datasets, respectively which highlights both the effectiveness and stability of the proposed scheme.
Authors: Davide Piras, Lucas Lombriser
Abstract: We present DE-VAE, a variational autoencoder (VAE) architecture to search for a compressed representation of dynamical dark energy (DE) models in observational studies of the cosmic large-scale structure. DE-VAE is trained on matter power spectra boosts generated at wavenumbers $k\in(0.01-2.5) \ h/\rm{Mpc}$ and at four redshift values $z\in(0.1,0.48,0.78,1.5)$ for the most typical dynamical DE parametrization with two extra parameters describing an evolving DE equation of state. The boosts are compressed to a lower-dimensional representation, which is concatenated with standard cold dark matter (CDM) parameters and then mapped back to reconstructed boosts; both the compression and the reconstruction components are parametrized as neural networks. Remarkably, we find that a single latent parameter is sufficient to predict 95% (99%) of DE power spectra generated over a broad range of cosmological parameters within $1\sigma$ ($2\sigma$) of a Gaussian error which includes cosmic variance, shot noise and systematic effects for a Stage IV-like survey. This single parameter shows a high mutual information with the two DE parameters, and these three variables can be linked together with an explicit equation through symbolic regression. Considering a model with two latent variables only marginally improves the accuracy of the predictions, and adding a third latent variable has no significant impact on the model's performance. We discuss how the DE-VAE architecture can be extended from a proof of concept to a general framework to be employed in the search for a common lower-dimensional parametrization of a wide range of beyond-$\Lambda$CDM models and for different cosmological datasets. Such a framework could then both inform the development of cosmological surveys by targeting optimal probes, and provide theoretical insight into the common phenomenological aspects of beyond-$\Lambda$CDM models.
Authors: Yuan Gao, Jian Huang, Yuling Jiao
Abstract: Gaussian denoising has emerged as a powerful method for constructing simulation-free continuous normalizing flows for generative modeling. Despite their empirical successes, theoretical properties of these flows and the regularizing effect of Gaussian denoising have remained largely unexplored. In this work, we aim to address this gap by investigating the well-posedness of simulation-free continuous normalizing flows built on Gaussian denoising. Through a unified framework termed Gaussian interpolation flow, we establish the Lipschitz regularity of the flow velocity field, the existence and uniqueness of the flow, and the Lipschitz continuity of the flow map and the time-reversed flow map for several rich classes of target distributions. This analysis also sheds light on the auto-encoding and cycle consistency properties of Gaussian interpolation flows. Additionally, we study the stability of these flows in source distributions and perturbations of the velocity field, using the quadratic Wasserstein distance as a metric. Our findings offer valuable insights into the learning techniques employed in Gaussian interpolation flows for generative modeling, providing a solid theoretical foundation for end-to-end error analyses of learning Gaussian interpolation flows with empirical observations.
Authors: Yumou Wei, Ryan F. Forelli, Chris Hansen, Jeffrey P. Levesque, Nhan Tran, Joshua C. Agar, Giuseppe Di Guglielmo, Michael E. Mauel, Gerald A. Navratil
Abstract: Active feedback control in magnetic confinement fusion devices is desirable to mitigate plasma instabilities and enable robust operation. Optical high-speed cameras provide a powerful, non-invasive diagnostic and can be suitable for these applications. In this study, we process fast camera data, at rates exceeding 100kfps, on $\textit{in situ}$ Field Programmable Gate Array (FPGA) hardware to track magnetohydrodynamic (MHD) mode evolution and generate control signals in real-time. Our system utilizes a convolutional neural network (CNN) model which predicts the $n$=1 MHD mode amplitude and phase using camera images with better accuracy than other tested non-deep-learning-based methods. By implementing this model directly within the standard FPGA readout hardware of the high-speed camera diagnostic, our mode tracking system achieves a total trigger-to-output latency of 17.6$\mu$s and a throughput of up to 120kfps. This study at the High Beta Tokamak-Extended Pulse (HBT-EP) experiment demonstrates an FPGA-based high-speed camera data acquisition and processing system, enabling application in real-time machine-learning-based tokamak diagnostic and control as well as potential applications in other scientific domains.
Authors: S. VenkataKeerthy, Soumya Banerjee, Sayan Dey, Yashas Andaluri, Raghul PS, Subrahmanyam Kalyanasundaram, Fernando Magno Quint\~ao Pereira, Ramakrishna Upadrasta
Abstract: Binary similarity involves determining whether two binary programs exhibit similar functionality, often originating from the same source code. In this work, we propose VexIR2Vec, an approach for binary similarity using VEX-IR, an architecture-neutral Intermediate Representation (IR). We extract the embeddings from sequences of basic blocks, termed peepholes, derived by random walks on the control-flow graph. The peepholes are normalized using transformations inspired by compiler optimizations. The VEX-IR Normalization Engine mitigates, with these transformations, the architectural and compiler-induced variations in binaries while exposing semantic similarities. We then learn the vocabulary of representations at the entity level of the IR using the knowledge graph embedding techniques in an unsupervised manner. This vocabulary is used to derive function embeddings for similarity assessment using VexNet, a feed-forward Siamese network designed to position similar functions closely and separate dissimilar ones in an n-dimensional space. This approach is amenable for both diffing and searching tasks, ensuring robustness against Out-Of-Vocabulary (OOV) issues. We evaluate VexIR2Vec on a dataset comprising 2.7M functions and 15.5K binaries from 7 projects compiled across 12 compilers targeting x86 and ARM architectures. In diffing experiments, VexIR2Vec outperforms the nearest baselines by $40\%$, $18\%$, $21\%$, and $60\%$ in cross-optimization, cross-compilation, cross-architecture, and obfuscation settings, respectively. In the searching experiment, VexIR2Vec achieves a mean average precision of $0.76$, outperforming the nearest baseline by $46\%$. Our framework is highly scalable and is built as a lightweight, multi-threaded, parallel library using only open-source tools. VexIR2Vec is $3.1$-$3.5 \times$ faster than the closest baselines and orders-of-magnitude faster than other tools.
Authors: Oliver Limoyo, Abhisek Konar, Trevor Ablett, Jonathan Kelly, Francois R. Hogan, Gregory Dudek
Abstract: We present placing via picking (PvP), a method to autonomously collect real-world demonstrations for a family of placing tasks in which objects must be manipulated to specific, contact-constrained locations. With PvP, we approach the collection of robotic object placement demonstrations by reversing the grasping process and exploiting the inherent symmetry of the pick and place problems. Specifically, we obtain placing demonstrations from a set of grasp sequences of objects initially located at their target placement locations. Our system can collect hundreds of demonstrations in contact-constrained environments without human intervention using two modules: compliant control for grasping and tactile regrasping. We train a policy directly from visual observations through behavioural cloning, using the autonomously-collected demonstrations. By doing so, the policy can generalize to object placement scenarios outside of the training environment without privileged information (e.g., placing a plate picked up from a table). We validate our approach in home robot scenarios that include dishwasher loading and table setting. Our approach yields robotic placing policies that outperform policies trained with kinesthetic teaching, both in terms of success rate and data efficiency, while requiring no human supervision.
Authors: Brecht Evens, Puya Latafat, Panagiotis Patrinos
Abstract: The Chambolle-Pock algorithm (CPA), also known as the primal-dual hybrid gradient method, has gained popularity over the last decade due to its success in solving large-scale convex structured problems. This work extends its convergence analysis for problems with varying degrees of (non)monotonicity, quantified through a so-called oblique weak Minty condition on the associated primal-dual operator. Our results reveal novel stepsize and relaxation parameter ranges which do not only depend on the norm of the linear mapping, but also on its other singular values. In particular, in nonmonotone settings, in addition to the classical stepsize conditions, extra bounds on the stepsizes and relaxation parameters are required. On the other hand, in the strongly monotone setting, the relaxation parameter is allowed to exceed the classical upper bound of two. Moreover, we build upon the recently introduced class of semimonotone operators, providing sufficient convergence conditions for CPA when the individual operators are semimonotone. Since this class of operators encompasses traditional operator classes including (hypo)- and co(hypo)-monotone operators, this analysis recovers and extends existing results for CPA. Tightness of the proposed stepsize ranges is demonstrated through several examples.
Authors: Haonan Shi, Tu Ouyang, An Wang
Abstract: Machine learning models, in particular deep neural networks, are currently an integral part of various applications, from healthcare to finance. However, using sensitive data to train these models raises concerns about privacy and security. One method that has emerged to verify if the trained models are privacy-preserving is Membership Inference Attacks (MIA), which allows adversaries to determine whether a specific data point was part of a model's training dataset. While a series of MIAs have been proposed in the literature, only a few can achieve high True Positive Rates (TPR) in the low False Positive Rate (FPR) region (0.01%~1%). This is a crucial factor to consider for an MIA to be practically useful in real-world settings. In this paper, we present a novel approach to MIA that is aimed at significantly improving TPR at low FPRs. Our method, named learning-based difficulty calibration for MIA(LDC-MIA), characterizes data records by their hardness levels using a neural network classifier to determine membership. The experiment results show that LDC-MIA can improve TPR at low FPR by up to 4x compared to the other difficulty calibration based MIAs. It also has the highest Area Under ROC curve (AUC) across all datasets. Our method's cost is comparable with most of the existing MIAs, but is orders of magnitude more efficient than one of the state-of-the-art methods, LiRA, while achieving similar performance.
Authors: Xuanlei Zhao, Shenggan Cheng, Guangyang Lu, Jiarui Fang, Haotian Zhou, Bin Jia, Ziming Liu, Yang You
Abstract: Large deep learning models have achieved impressive performance across a range of applications. However, their large memory requirements, including parameter memory and activation memory, have become a significant challenge for their practical serving. While existing methods mainly address parameter memory, the importance of activation memory has been overlooked. Especially for long input sequences, activation memory is expected to experience a significant exponential growth as the length of sequences increases. In this approach, we propose AutoChunk, an automatic and adaptive compiler system that efficiently reduces activation memory for long sequence inference by chunk strategies. The proposed system generates chunk plans by optimizing through multiple stages. In each stage, the chunk search pass explores all possible chunk candidates and the chunk selection pass identifies the optimal one. At runtime, AutoChunk employs code generation to automatically apply chunk strategies. The experiments demonstrate that AutoChunk can reduce over 80\% of activation memory while maintaining speed loss within 10%, extend max sequence length by 3.2x to 11.7x, and outperform state-of-the-art methods by a large margin.
Authors: Nazmul Islam, Seokjoo Shin
Abstract: Deep learning (DL) has revolutionized wireless communication systems by introducing datadriven end-to-end (E2E) learning, where the physical layer (PHY) is transformed into DL architectures to achieve peak optimization. Leveraging DL for E2E optimization in PHY significantly enhances its adaptability and performance in complex wireless environments, meeting the demands of advanced network systems such as 5G and beyond. Furthermore, this evolution of data-driven PHY optimization has also enabled advanced semantic applications across various modalities, including text, image, audio, video, and multimodal transmissions. These applications elevate communication from bit-level to semantic-level intelligence, making it capable of discerning context and intent. Although the PHY, as a DL architecture, plays a crucial role in enabling semantic communication (SemCom) systems, comprehensive studies that integrate both E2E communication and SemCom systems remain significantly underexplored. This highlights the novelty and potential of these integrative fields, marking them as a promising research domain. Therefore, this article provides a comprehensive review of the emerging field of data-driven PHY for E2E communication systems, emphasizing their role in enabling semantic applications across various modalities. It also identifies key challenges and potential research directions, serving as a crucial guide for future advancements in DL for E2E communication and SemCom systems.
Authors: Jan-Philipp Redlich, Friedrich Feuerhake, Joachim Weis, Nadine S. Schaadt, Sarah Teuber-Hanselmann, Christoph Buck, Sabine Luttmann, Andrea Eberle, Stefan Nikolin, Arno Appenzeller, Andreas Portmann, Andr\'e Homeyer
Abstract: In recent years, the diagnosis of gliomas has become increasingly complex. Analysis of glioma histopathology images using artificial intelligence (AI) offers new opportunities to support diagnosis and outcome prediction. To give an overview of the current state of research, this review examines 83 publicly available research studies that have proposed AI-based methods for whole-slide histopathology images of human gliomas, covering the diagnostic tasks of subtyping (23/83), grading (27/83), molecular marker prediction (20/83), and survival prediction (29/83). All studies were reviewed with regard to methodological aspects as well as clinical applicability. It was found that the focus of current research is the assessment of hematoxylin and eosin-stained tissue sections of adult-type diffuse gliomas. The majority of studies (52/83) are based on the publicly available glioblastoma and low-grade glioma datasets from The Cancer Genome Atlas (TCGA) and only a few studies employed other datasets in isolation (16/83) or in addition to the TCGA datasets (15/83). Current approaches mostly rely on convolutional neural networks (63/83) for analyzing tissue at 20x magnification (35/83). A new field of research is the integration of clinical data, omics data, or magnetic resonance imaging (29/83). So far, AI-based methods have achieved promising results, but are not yet used in real clinical settings. Future work should focus on the independent validation of methods on larger, multi-site datasets with high-quality and up-to-date clinical and molecular pathology annotations to demonstrate routine applicability.
Authors: Alexander Auras, Kanchana Vaishnavi Gandikota, Hannah Droege, Michael Moeller
Abstract: This paper provides an overview of current approaches for solving inverse problems in imaging using variational methods and machine learning. A special focus lies on point estimators and their robustness against adversarial perturbations. In this context results of numerical experiments for a one-dimensional toy problem are provided, showing the robustness of different approaches and empirically verifying theoretical guarantees. Another focus of this review is the exploration of the subspace of data-consistent solutions through explicit guidance to satisfy specific semantic or textural properties.
Authors: Mahsa Salehi, Kalin Stefanov, Ehsan Shareghi
Abstract: In this paper we study the variations in human brain activity when listening to real and fake audio. Our preliminary results suggest that the representations learned by a state-of-the-art deepfake audio detection algorithm, do not exhibit clear distinct patterns between real and fake audio. In contrast, human brain activity, as measured by EEG, displays distinct patterns when individuals are exposed to fake versus real audio. This preliminary evidence enables future research directions in areas such as deepfake audio detection.
Authors: Siqi Fan, Xin Jiang, Xiang Li, Xuying Meng, Peng Han, Shuo Shang, Aixin Sun, Yequan Wang, Zhongyuan Wang
Abstract: Due to the large number of parameters, the inference phase of Large Language Models (LLMs) is resource-intensive. However, not all requests posed to LLMs are equally difficult to handle. Through analysis, we show that for some tasks, LLMs can achieve results comparable to the final output at some intermediate layers. That is, not all layers of LLMs are necessary during inference. If we can predict at which layer the inferred results match the final results (produced by evaluating all layers), we could significantly reduce the inference cost. To this end, we propose a simple yet effective algorithm named AdaInfer to adaptively terminate the inference process for an input instance. AdaInfer relies on easily obtainable statistical features and classic classifiers like SVM. Experiments on well-known LLMs like the Llama2 series and OPT, show that AdaInfer can achieve an average of 17.8% pruning ratio, and up to 43% on sentiment tasks, with nearly no performance drop (<1%). Because AdaInfer does not alter LLM parameters, the LLMs incorporated with AdaInfer maintain generalizability across tasks.
Authors: Michael Ogezi, Ning Shi
Abstract: In text-to-image generation, using negative prompts, which describe undesirable image characteristics, can significantly boost image quality. However, producing good negative prompts is manual and tedious. To address this, we propose NegOpt, a novel method for optimizing negative prompt generation toward enhanced image generation, using supervised fine-tuning and reinforcement learning. Our combined approach results in a substantial increase of 25% in Inception Score compared to other approaches and surpasses ground-truth negative prompts from the test set. Furthermore, with NegOpt we can preferentially optimize the metrics most important to us. Finally, we construct Negative Prompts DB (https://github.com/mikeogezi/negopt), a publicly available dataset of negative prompts.
Authors: Yongqi Wang, Ruofan Hu, Rongjie Huang, Zhiqing Hong, Ruiqi Li, Wenrui Liu, Fuming You, Tao Jin, Zhou Zhao
Abstract: Recent singing-voice-synthesis (SVS) methods have achieved remarkable audio quality and naturalness, yet they lack the capability to control the style attributes of the synthesized singing explicitly. We propose Prompt-Singer, the first SVS method that enables attribute controlling on singer gender, vocal range and volume with natural language. We adopt a model architecture based on a decoder-only transformer with a multi-scale hierarchy, and design a range-melody decoupled pitch representation that enables text-conditioned vocal range control while keeping melodic accuracy. Furthermore, we explore various experiment settings, including different types of text representations, text encoder fine-tuning, and introducing speech data to alleviate data scarcity, aiming to facilitate further research. Experiments show that our model achieves favorable controlling ability and audio quality. Audio samples are available at http://prompt-singer.github.io .
Authors: GianCarlo Catalano, Alexander E. I. Brownlee, David Cairns, John McCall, Russell Ainslie
Abstract: Genetic Algorithms have established their capability for solving many complex optimization problems. Even as good solutions are produced, the user's understanding of a problem is not necessarily improved, which can lead to a lack of confidence in the results. To mitigate this issue, explainability aims to give insight to the user by presenting them with the knowledge obtained by the algorithm. In this paper we introduce Partial Solutions in order to improve the explainability of solutions to combinatorial optimization problems. Partial Solutions represent beneficial traits found by analyzing a population, and are presented to the user for explainability, but also provide an explicit model from which new solutions can be generated. We present an algorithm that assembles a collection of Partial Solutions chosen to strike a balance between high fitness, simplicity and atomicity. Experiments with standard benchmarks show that the proposed algorithm is able to find Partial Solutions which improve explainability at reasonable computational cost without affecting search performance.
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: Georg Rabenstein, Lars Ullrich, Knut Graichen
Abstract: Alongside optimization-based planners, sampling-based approaches are often used in trajectory planning for autonomous driving due to their simplicity. Model predictive path integral control is a framework that builds upon optimization principles while incorporating stochastic sampling of input trajectories. This paper investigates several sampling approaches for trajectory generation. In this context, normalizing flows originating from the field of variational inference are considered for the generation of sampling distributions, as they model transformations of simple to more complex distributions. Accordingly, learning-based normalizing flow models are trained for a more efficient exploration of the input domain for the task at hand. The developed algorithm and the proposed sampling distributions are evaluated in two simulation scenarios.
Authors: Qiyu Hou, Jun Wang, Meixuan Qiao, Lujun Tian
Abstract: To overcome the limitations and challenges of current automatic table data annotation methods and random table data synthesis approaches, we propose a novel method for synthesizing annotation data specifically designed for table recognition. This method utilizes the structure and content of existing complex tables, facilitating the efficient creation of tables that closely replicate the authentic styles found in the target domain. By leveraging the actual structure and content of tables from Chinese financial announcements, we have developed the first extensive table annotation dataset in this domain. We used this dataset to train several recent deep learning-based end-to-end table recognition models. Additionally, we have established the inaugural benchmark for real-world complex tables in the Chinese financial announcement domain, using it to assess the performance of models trained on our synthetic data, thereby effectively validating our method's practicality and effectiveness. Furthermore, we applied our synthesis method to augment the FinTabNet dataset, extracted from English financial announcements, by increasing the proportion of tables with multiple spanning cells to introduce greater complexity. Our experiments show that models trained on this augmented dataset achieve comprehensive improvements in performance, especially in the recognition of tables with multiple spanning cells.
Authors: Hisashi Shimodaira
Abstract: In this paper, we propose a method to improve prediction accuracy of semantic segmentation methods as follows: (1) construct a neural network that has pre-processing layers based on a convolutional autoencoder ahead of a semantic segmentation network, and (2) train the entire network initialized by the weights of the pre-trained autoencoder. We applied this method to the fully convolutional network (FCN) and experimentally compared its prediction accuracy on the cityscapes dataset. The Mean IoU of the proposed target model with the He normal initialization is 18.7% higher than that of FCN with the He normal initialization. In addition, those of the modified models of the target model are significantly higher than that of FCN with the He normal initialization. The accuracy and loss curves during the training showed that these are resulting from the improvement of the generalization ability. All of these results provide strong evidence that the proposed method is significantly effective in improving the prediction accuracy of FCN. The proposed method has the following features: it is comparatively simple, whereas the effect on improving the generalization ability and prediction accuracy of FCN is significant; the increase in the number of parameters by using it is very small, and that in the computation time is substantially large. In principle, the proposed method can be applied to other semantic segmentation methods. For semantic segmentation, at present, there is no effective way to improve the prediction accuracy of existing methods. None have published a method which is the same as or similar to our method and none have used such a method in practice. Therefore, we believe that our method is useful in practice and worthy of being widely known and used.
Authors: Chris Junchi Li
Abstract: We present in this paper novel accelerated fully first-order methods in \emph{Bilevel Optimization} (BLO). Firstly, for BLO under the assumption that the lower-level functions admit the typical strong convexity assumption, the \emph{(Perturbed) Restarted Accelerated Fully First-order methods for Bilevel Approximation} (\texttt{PRAF${}^2$BA}) algorithm leveraging \emph{fully} first-order oracles is proposed, whereas the algorithm for finding approximate first-order and second-order stationary points with state-of-the-art oracle query complexities in solving complex optimization tasks. Secondly, applying as a special case of BLO the \emph{nonconvex-strongly-convex} (NCSC) minimax optimization, \texttt{PRAF${}^2$BA} rediscovers \emph{perturbed restarted accelerated gradient descent ascent} (\texttt{PRAGDA}) that achieves the state-of-the-art complexity for finding approximate second-order stationary points. Additionally, we investigate the challenge of finding stationary points of the hyper-objective function in BLO when lower-level functions lack the typical strong convexity assumption, where we identify several regularity conditions of the lower-level problems that ensure tractability and present hardness results indicating the intractability of BLO for general convex lower-level functions. Under these regularity conditions we propose the \emph{Inexact Gradient-Free Method} (\texttt{IGFM}), utilizing the \emph{Switching Gradient Method} (\texttt{SGM}) as an efficient sub-routine to find an approximate stationary point of the hyper-objective in polynomial time. Empirical studies for real-world problems are provided to further validate the outperformance of our proposed algorithms.
Authors: Rezkellah Noureddine Khiati, Pierre-Yves Brillet, Aur\'elien Justet, Radu Ispas, Catalin Fetita
Abstract: This paper addresses the problem of pathological lung segmentation, a significant challenge in medical image analysis, particularly pronounced in cases of peripheral opacities (severe fibrosis and consolidation) because of the textural similarity between lung tissue and surrounding areas. To overcome these challenges, this paper emphasizes the use of CycleGAN for unpaired image-to-image translation, in order to provide an augmentation method able to generate fake pathological images matching an existing ground truth. Although previous studies have employed CycleGAN, they often neglect the challenge of shape deformation, which is crucial for accurate medical image segmentation. Our work introduces an innovative strategy that incorporates additional loss functions. Specifically, it proposes an L1 loss based on the lung surrounding which shape is constrained to remain unchanged at the transition from the healthy to pathological domains. The lung surrounding is derived based on ground truth lung masks available in the healthy domain. Furthermore, preprocessing steps, such as cropping based on ribs/vertebra locations, are applied to refine the input for the CycleGAN, ensuring that the network focus on the lung region. This is essential to avoid extraneous biases, such as the zoom effect bias, which can divert attention from the main task. The method is applied to enhance in semi-supervised manner the lung segmentation process by employing a U-Net model trained with on-the-fly data augmentation incorporating synthetic pathological tissues generated by the CycleGAN model. Preliminary results from this research demonstrate significant qualitative and quantitative improvements, setting a new benchmark in the field of pathological lung segmentation. Our code is available at https://github.com/noureddinekhiati/Semi-supervised-lung-segmentation
URLs: https://github.com/noureddinekhiati/Semi-supervised-lung-segmentation
Authors: Jonas Spinner, Victor Bres\'o, Pim de Haan, Tilman Plehn, Jesse Thaler, Johann Brehmer
Abstract: Extracting scientific understanding from particle-physics experiments requires solving diverse learning problems with high precision and good data efficiency. We propose the Lorentz Geometric Algebra Transformer (L-GATr), a new multi-purpose architecture for high-energy physics. L-GATr represents high-energy data in a geometric algebra over four-dimensional space-time and is equivariant under Lorentz transformations, the symmetry group of relativistic kinematics. At the same time, the architecture is a Transformer, which makes it versatile and scalable to large systems. L-GATr is first demonstrated on regression and classification tasks from particle physics. We then construct the first Lorentz-equivariant generative model: a continuous normalizing flow based on an L-GATr network, trained with Riemannian flow matching. Across our experiments, L-GATr is on par with or outperforms strong domain-specific baselines.
Authors: Sion An, Myeongkyun Kang, Soopil Kim, Philip Chikontwe, Li Shen, Sang Hyun Park
Abstract: Electroencephalography (EEG) motor imagery (MI) classification is a fundamental, yet challenging task due to the variation of signals between individuals i.e., inter-subject variability. Previous approaches try to mitigate this using task-specific (TS) EEG signals from the target subject in training. However, recording TS EEG signals requires time and limits its applicability in various fields. In contrast, resting state (RS) EEG signals are a viable alternative due to ease of acquisition with rich subject information. In this paper, we propose a novel subject-adaptive transfer learning strategy that utilizes RS EEG signals to adapt models on unseen subject data. Specifically, we disentangle extracted features into task- and subject-dependent features and use them to calibrate RS EEG signals for obtaining task information while preserving subject characteristics. The calibrated signals are then used to adapt the model to the target subject, enabling the model to simulate processing TS EEG signals of the target subject. The proposed method achieves state-of-the-art accuracy on three public benchmarks, demonstrating the effectiveness of our method in cross-subject EEG MI classification. Our findings highlight the potential of leveraging RS EEG signals to advance practical brain-computer interface systems. The code is available at https://github.com/SionAn/MICCAI2024-ResTL.
Authors: Qutub Syed Sha, Michael Paulitsch, Karthik Pattabiraman, Korbinian Hagn, Fabian Oboril, Cornelius Buerkle, Kay-Ulrich Scholl, Gereon Hinz, Alois Knoll
Abstract: As transformer-based object detection models progress, their impact in critical sectors like autonomous vehicles and aviation is expected to grow. Soft errors causing bit flips during inference have significantly impacted DNN performance, altering predictions. Traditional range restriction solutions for CNNs fall short for transformers. This study introduces the Global Clipper and Global Hybrid Clipper, effective mitigation strategies specifically designed for transformer-based models. It significantly enhances their resilience to soft errors and reduces faulty inferences to ~ 0\%. We also detail extensive testing across over 64 scenarios involving two transformer models (DINO-DETR and Lite-DETR) and two CNN models (YOLOv3 and SSD) using three datasets, totalling approximately 3.3 million inferences, to assess model robustness comprehensively. Moreover, the paper explores unique aspects of attention blocks in transformers and their operational differences from CNNs.
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: Zhifeng Kong, Sang-gil Lee, Deepanway Ghosal, Navonil Majumder, Ambuj Mehrish, Rafael Valle, Soujanya Poria, Bryan Catanzaro
Abstract: It is an open challenge to obtain high quality training data, especially captions, for text-to-audio models. Although prior methods have leveraged \textit{text-only language models} to augment and improve captions, such methods have limitations related to scale and coherence between audio and captions. In this work, we propose an audio captioning pipeline that uses an \textit{audio language model} to synthesize accurate and diverse captions for audio at scale. We leverage this pipeline to produce a dataset of synthetic captions for AudioSet, named \texttt{AF-AudioSet}, and then evaluate the benefit of pre-training text-to-audio models on these synthetic captions. Through systematic evaluations on AudioCaps and MusicCaps, we find leveraging our pipeline and synthetic captions leads to significant improvements on audio generation quality, achieving a new \textit{state-of-the-art}.
Authors: Peter M{\o}rch Groth, Mads Herbert Kerrn, Lars Olsen, Jesper Salomon, Wouter Boomsma
Abstract: Reliable prediction of protein variant effects is crucial for both protein optimization and for advancing biological understanding. For practical use in protein engineering, it is important that we can also provide reliable uncertainty estimates for our predictions, and while prediction accuracy has seen much progress in recent years, uncertainty metrics are rarely reported. We here provide a Gaussian process regression model, Kermut, with a novel composite kernel for modelling mutation similarity, which obtains state-of-the-art performance for protein variant effect prediction while also offering estimates of uncertainty through its posterior. An analysis of the quality of the uncertainty estimates demonstrates that our model provides meaningful levels of overall calibration, but that instance-specific uncertainty calibration remains more challenging. We hope that this will encourage future work in this promising direction.
Authors: Gautham Vasan, Yan Wang, Fahim Shahriar, James Bergstra, Martin Jagersand, A. Rupam Mahmood
Abstract: Many real-world robot learning problems, such as pick-and-place or arriving at a destination, can be seen as a problem of reaching a goal state as soon as possible. These problems, when formulated as episodic reinforcement learning tasks, can easily be specified to align well with our intended goal: -1 reward every time step with termination upon reaching the goal state, called minimum-time tasks. Despite this simplicity, such formulations are often overlooked in favor of dense rewards due to their perceived difficulty and lack of informativeness. Our studies contrast the two reward paradigms, revealing that the minimum-time task specification not only facilitates learning higher-quality policies but can also surpass dense-reward-based policies on their own performance metrics. Crucially, we also identify the goal-hit rate of the initial policy as a robust early indicator for learning success in such sparse feedback settings. Finally, using four distinct real-robotic platforms, we show that it is possible to learn pixel-based policies from scratch within two to three hours using constant negative rewards.
Authors: Wouter M. Kouw, Caspar Gruijthuijsen, Lennart Blanken, Enzo Evers, Timothy Rogers
Abstract: We propose a computational procedure for identifying convection in heat transfer dynamics. The procedure is based on a Gaussian process latent force model, consisting of a white-box component (i.e., known physics) for the conduction and linear convection effects and a Gaussian process that acts as a black-box component for the nonlinear convection effects. States are inferred through Bayesian smoothing and we obtain approximate posterior distributions for the kernel covariance function's hyperparameters using Laplace's method. The nonlinear convection function is recovered from the Gaussian process states using a Bayesian regression model. We validate the procedure by simulation error using the identified nonlinear convection function, on both data from a simulated system and measurements from a physical assembly.
Authors: Chengpeng Li, Guanting Dong, Mingfeng Xue, Ru Peng, Xiang Wang, Dayiheng Liu
Abstract: Large language models (LLMs) have made impressive progress in handling simple math problems, yet they still struggle with more challenging and complex mathematical tasks. In this paper, we introduce a series of LLMs that employs the Decomposition of thought with code assistance and self-correction for mathematical reasoning, dubbed as DotaMath. DotaMath models tackle complex mathematical tasks by decomposing them into simpler logical subtasks, leveraging code to solve these subtasks, obtaining fine-grained feedback from the code interpreter, and engaging in self-reflection and correction. By annotating diverse interactive tool-use trajectories and employing query evolution on GSM8K and MATH datasets, we generate an instruction fine-tuning dataset called DotaMathQA with 574K query-response pairs. We train a series of base LLMs using imitation learning on DotaMathQA, resulting in DotaMath models that achieve remarkable performance compared to open-source LLMs across various in-domain and out-of-domain benchmarks. Notably, DotaMath-deepseek-7B showcases an outstanding performance of 64.8% on the competitive MATH dataset and 86.7% on GSM8K. Besides, DotaMath-deepseek-7B maintains strong competitiveness on a series of in-domain and out-of-domain benchmarks (Avg. 80.1%). Looking forward, we anticipate that the DotaMath paradigm will open new pathways for addressing intricate mathematical problems. Our code is publicly available at https://github.com/ChengpengLi1003/DotaMath.
Authors: Hannes Kunstmann, Joseph Ollier, Joel Persson, Florian von Wangenheim
Abstract: Large language models (LLMs) present an enormous evolution in the strategic potential of conversational recommender systems (CRS). Yet to date, research has predominantly focused upon technical frameworks to implement LLM-driven CRS, rather than end-user evaluations or strategic implications for firms, particularly from the perspective of a small to medium enterprises (SME) that makeup the bedrock of the global economy. In the current paper, we detail the design of an LLM-driven CRS in an SME setting, and its subsequent performance in the field using both objective system metrics and subjective user evaluations. While doing so, we additionally outline a short-form revised ResQue model for evaluating LLM-driven CRS, enabling replicability in a rapidly evolving field. Our results reveal good system performance from a user experience perspective (85.5% recommendation accuracy) but underscore latency, cost, and quality issues challenging business viability. Notably, with a median cost of $0.04 per interaction and a latency of 5.7s, cost-effectiveness and response time emerge as crucial areas for achieving a more user-friendly and economically viable LLM-driven CRS for SME settings. One major driver of these costs is the use of an advanced LLM as a ranker within the retrieval-augmented generation (RAG) technique. Our results additionally indicate that relying solely on approaches such as Prompt-based learning with ChatGPT as the underlying LLM makes it challenging to achieve satisfying quality in a production environment. Strategic considerations for SMEs deploying an LLM-driven CRS are outlined, particularly considering trade-offs in the current technical landscape.
Authors: Andrei Jarca, Florinel-Alin Croitoru, Radu Tudor Ionescu
Abstract: We propose Curriculum by Masking (CBM), a novel state-of-the-art curriculum learning strategy that effectively creates an easy-to-hard training schedule via patch (token) masking, offering significant accuracy improvements over the conventional training regime and previous curriculum learning (CL) methods. CBM leverages gradient magnitudes to prioritize the masking of salient image regions via a novel masking algorithm and a novel masking block. Our approach enables controlling sample difficulty via the patch masking ratio, generating an effective easy-to-hard curriculum by gradually introducing harder samples as training progresses. CBM operates with two easily configurable parameters, i.e. the number of patches and the curriculum schedule, making it a versatile curriculum learning approach for object recognition and detection. We conduct experiments with various neural architectures, ranging from convolutional networks to vision transformers, on five benchmark data sets (CIFAR-10, CIFAR-100, ImageNet, Food-101 and PASCAL VOC), to compare CBM with conventional as well as curriculum-based training regimes. Our results reveal the superiority of our strategy compared with the state-of-the-art curriculum learning regimes. We also observe improvements in transfer learning contexts, where CBM surpasses previous work by considerable margins in terms of accuracy. We release our code for free non-commercial use at https://github.com/CroitoruAlin/CBM.