Authors: Maud Lemercier, Terry Lyons
Abstract: Signature kernels are at the core of several machine learning algorithms for analysing multivariate time series. The kernel of two bounded variation paths (such as piecewise linear interpolations of time series data) is typically computed by solving a Goursat problem for a hyperbolic partial differential equation (PDE) in two independent time variables. However, this approach becomes considerably less practical for highly oscillatory input paths, as they have to be resolved at a fine enough scale to accurately recover their signature kernel, resulting in significant time and memory complexities. To mitigate this issue, we first show that the signature kernel of a broader class of paths, known as \emph{smooth rough paths}, also satisfies a PDE, albeit in the form of a system of coupled equations. We then use this result to introduce new algorithms for the numerical approximation of signature kernels. As bounded variation paths (and more generally geometric $p$-rough paths) can be approximated by piecewise smooth rough paths, one can replace the PDE with rapidly varying coefficients in the original Goursat problem by an explicit system of coupled equations with piecewise constant coefficients derived from the first few iterated integrals of the original input paths. While this approach requires solving more equations, they do not require looking back at the complex and fine structure of the initial paths, which significantly reduces the computational complexity associated with the analysis of highly oscillatory time series.
Authors: Xusen Guo (Frank), Qiming Zhang (Frank), Mingxing Peng (Frank), Meixin Zhua (Frank), Hao (Frank), Yang
Abstract: Traffic flow prediction provides essential future views in the intelligent transportation system. Explainable predictions offer valuable insights into the factors influencing traffic patterns, which help urban planners, traffic engineers, and policymakers make informed decisions about infrastructure development, traffic management strategies, and public transportation planning. Despite their widespread popularity and commendable accuracy, prediction methods grounded in deep learning frequently disappoint in terms of transparency and interpretability. Recently, the availability of large-scale spatio-temporal data and the development of large language models (LLMs) have opened up new opportunities for urban traffic prediction. With the popularity of LLMs, people witnessed the potential reasoning and generating ability of foundation models in various tasks. Considering text as input and output, LLMs have advantages in generating more intuitive and interpretable predictions. Hence, this work introduces TP-LLM, an explainable foundation-model-based method for traffic prediction, aiming at more direct and reasonable forecasting. TP-LLM presents a framework to unify multi-modality factors as language-based inputs, TP-LLM avoids complex spatial-temporal data programming and outperforms state-of-art baselines merely under fine-tuning foundation models. Also, TP-LLM can generate input-dependency explanations for more confident prediction and can be easily generalized to different city dynamics for zero-shot prediction with a similar framework. These findings demonstrate the potential of LLMs for explainable traffic prediction.
Authors: Leonardo Arrighi, Luca Pennella, Gabriel Marques Tavares, Sylvio Barbon Junior
Abstract: Understanding the decisions of tree-based ensembles and their relationships is pivotal for machine learning model interpretation. Recent attempts to mitigate the human-in-the-loop interpretation challenge have explored the extraction of the decision structure underlying the model taking advantage of graph simplification and path emphasis. However, while these efforts enhance the visualisation experience, they may either result in a visually complex representation or compromise the interpretability of the original ensemble model. In addressing this challenge, especially in complex scenarios, we introduce the Decision Predicate Graph (DPG) as a model-agnostic tool to provide a global interpretation of the model. DPG is a graph structure that captures the tree-based ensemble model and learned dataset details, preserving the relations among features, logical decisions, and predictions towards emphasising insightful points. Leveraging well-known graph theory concepts, such as the notions of centrality and community, DPG offers additional quantitative insights into the model, complementing visualisation techniques, expanding the problem space descriptions, and offering diverse possibilities for extensions. Empirical experiments demonstrate the potential of DPG in addressing traditional benchmarks and complex classification scenarios.
Authors: Adrian Moldovan, Angel Ca\c{t}aron, R\u{a}zvan Andonie
Abstract: Recently, there is a growing interest in applying Transfer Entropy (TE) in quantifying the effective connectivity between artificial neurons. In a feedforward network, the TE can be used to quantify the relationships between neuron output pairs located in different layers. Our focus is on how to include the TE in the learning mechanisms of a Convolutional Neural Network (CNN) architecture. We introduce a novel training mechanism for CNN architectures which integrates the TE feedback connections. Adding the TE feedback parameter accelerates the training process, as fewer epochs are needed. On the flip side, it adds computational overhead to each epoch. According to our experiments on CNN classifiers, to achieve a reasonable computational overhead--accuracy trade-off, it is efficient to consider only the inter-neural information transfer of a random subset of the neuron pairs from the last two fully connected layers. The TE acts as a smoothing factor, generating stability and becoming active only periodically, not after processing each input sample. Therefore, we can consider the TE is in our model a slowly changing meta-parameter.
Authors: Luca Benfenati, Daniele Jahier Pagliari, Luca Zanatta, Yhorman Alexander Bedoya Velez, Andrea Acquaviva, Massimo Poncino, Enrico Macii, Luca Benini, Alessio Burrello
Abstract: Structural Health Monitoring (SHM) is a critical task for ensuring the safety and reliability of civil infrastructures, typically realized on bridges and viaducts by means of vibration monitoring. In this paper, we propose for the first time the use of Transformer neural networks, with a Masked Auto-Encoder architecture, as Foundation Models for SHM. We demonstrate the ability of these models to learn generalizable representations from multiple large datasets through self-supervised pre-training, which, coupled with task-specific fine-tuning, allows them to outperform state-of-the-art traditional methods on diverse tasks, including Anomaly Detection (AD) and Traffic Load Estimation (TLE). We then extensively explore model size versus accuracy trade-offs and experiment with Knowledge Distillation (KD) to improve the performance of smaller Transformers, enabling their embedding directly into the SHM edge nodes. We showcase the effectiveness of our foundation models using data from three operational viaducts. For AD, we achieve a near-perfect 99.9% accuracy with a monitoring time span of just 15 windows. In contrast, a state-of-the-art method based on Principal Component Analysis (PCA) obtains its first good result (95.03% accuracy) only considering 120 windows. On two different TLE tasks, our models obtain state-of-the-art performance on multiple evaluation metrics (R$^2$ score, MAE% and MSE%). On the first benchmark, we achieve an R$^2$ score of 0.97 and 0.85 for light and heavy vehicle traffic, respectively, while the best previous approach stops at 0.91 and 0.84. On the second one, we achieve an R$^2$ score of 0.54 versus the 0.10 of the best existing method.
Authors: Victor J. B. Jung, Alessio Burrello, Moritz Scherer, Francesco Conti, Luca Benini
Abstract: Transformer networks are rapidly becoming SotA in many fields, such as NLP and CV. Similarly to CNN, there is a strong push for deploying Transformer models at the extreme edge, ultimately fitting the tiny power budget and memory footprint of MCUs. However, the early approaches in this direction are mostly ad-hoc, platform, and model-specific. This work aims to enable and optimize the flexible, multi-platform deployment of encoder Tiny Transformers on commercial MCUs. We propose a complete framework to perform end-to-end deployment of Transformer models onto single and multi-core MCUs. Our framework provides an optimized library of kernels to maximize data reuse and avoid unnecessary data marshaling operations into the crucial attention block. A novel MHSA inference schedule, named Fused-Weight Self-Attention, is introduced, fusing the linear projection weights offline to further reduce the number of operations and parameters. Furthermore, to mitigate the memory peak reached by the computation of the attention map, we present a Depth-First Tiling scheme for MHSA. We evaluate our framework on three different MCU classes exploiting ARM and RISC-V ISA, namely the STM32H7, the STM32L4, and GAP9 (RV32IMC-XpulpV2). We reach an average of 4.79x and 2.0x lower latency compared to SotA libraries CMSIS-NN (ARM) and PULP-NN (RISC-V), respectively. Moreover, we show that our MHSA depth-first tiling scheme reduces the memory peak by up to 6.19x, while the fused-weight attention can reduce the runtime by 1.53x, and number of parameters by 25%. We report significant improvements across several Tiny Transformers: for instance, when executing a transformer block for the task of radar-based hand-gesture recognition on GAP9, we achieve a latency of 0.14ms and energy consumption of 4.92 micro-joules, 2.32x lower than the SotA PULP-NN library on the same platform.
Authors: Behnam Ghavami, Amin Kamjoo, Lesley Shannon, Steve Wilton
Abstract: The imperative to deploy Deep Neural Network (DNN) models on resource-constrained edge devices, spurred by privacy concerns, has become increasingly apparent. To facilitate the transition from cloud to edge computing, this paper introduces a technique that effectively reduces the memory footprint of DNNs, accommodating the limitations of resource-constrained edge devices while preserving model accuracy. Our proposed technique, named Post-Training Intra-Layer Multi-Precision Quantization (PTILMPQ), employs a post-training quantization approach, eliminating the need for extensive training data. By estimating the importance of layers and channels within the network, the proposed method enables precise bit allocation throughout the quantization process. Experimental results demonstrate that PTILMPQ offers a promising solution for deploying DNNs on edge devices with restricted memory resources. For instance, in the case of ResNet50, it achieves an accuracy of 74.57\% with a memory footprint of 9.5 MB, representing a 25.49\% reduction compared to previous similar methods, with only a minor 1.08\% decrease in accuracy.
Authors: Fanxu Meng, Zhaohui Wang, Muhan Zhang
Abstract: As the parameters of LLMs expand, the computational cost of fine-tuning the entire model becomes prohibitive. To address this challenge, we introduce a PEFT method, Principal Singular values and Singular vectors Adaptation (PiSSA), which optimizes a significantly reduced parameter space while achieving or surpassing the performance of full-parameter fine-tuning. PiSSA is inspired by Intrinsic SAID, which suggests that pre-trained, over-parametrized models inhabit a space of low intrinsic dimension. Consequently, PiSSA represents a matrix W within the model by the product of two trainable matrices A and B, plus a residual matrix $W^{res}$ for error correction. SVD is employed to factorize W, and the principal singular values and vectors of W are utilized to initialize A and B. The residual singular values and vectors initialize the residual matrix $W^{res}$, which keeps frozen during fine-tuning. Notably, PiSSA shares the same architecture with LoRA. However, LoRA approximates Delta W through the product of two matrices, A, initialized with Gaussian noise, and B, initialized with zeros, while PiSSA initializes A and B with principal singular values and vectors of the original matrix W. PiSSA can better approximate the outcomes of full-parameter fine-tuning at the beginning by changing the essential parts while freezing the "noisy" parts. In comparison, LoRA freezes the original matrix and updates the "noise". This distinction enables PiSSA to convergence much faster than LoRA and also achieve better performance in the end. Due to the same architecture, PiSSA inherits many of LoRA's advantages, such as parameter efficiency and compatibility with quantization. Leveraging a fast SVD method, the initialization of PiSSA takes only a few seconds, inducing negligible cost of switching LoRA to PiSSA.
Authors: Stephen Casper, Jieun Yun, Joonhyuk Baek, Yeseong Jung, Minhwan Kim, Kiwan Kwon, Saerom Park, Hayden Moore, David Shriver, Marissa Connor, Keltin Grimes, Angus Nicolson, Arush Tagade, Jessica Rumbelow, Hieu Minh Nguyen, Dylan Hadfield-Menell
Abstract: Interpretability techniques are valuable for helping humans understand and oversee AI systems. The SaTML 2024 CNN Interpretability Competition solicited novel methods for studying convolutional neural networks (CNNs) at the ImageNet scale. The objective of the competition was to help human crowd-workers identify trojans in CNNs. This report showcases the methods and results of four featured competition entries. It remains challenging to help humans reliably diagnose trojans via interpretability tools. However, the competition's entries have contributed new techniques and set a new record on the benchmark from Casper et al., 2023.
Authors: Gabriel Loaiza-Ganem, Brendan Leigh Ross, Rasa Hosseinzadeh, Anthony L. Caterini, Jesse C. Cresswell
Abstract: In recent years there has been increased interest in understanding the interplay between deep generative models (DGMs) and the manifold hypothesis. Research in this area focuses on understanding the reasons why commonly-used DGMs succeed or fail at learning distributions supported on unknown low-dimensional manifolds, as well as developing new models explicitly designed to account for manifold-supported data. This manifold lens provides both clarity as to why some DGMs (e.g. diffusion models and some generative adversarial networks) empirically surpass others (e.g. likelihood-based models such as variational autoencoders, normalizing flows, or energy-based models) at sample generation, and guidance for devising more performant DGMs. We carry out the first survey of DGMs viewed through this lens, making two novel contributions along the way. First, we formally establish that numerical instability of high-dimensional likelihoods is unavoidable when modelling low-dimensional data. We then show that DGMs on learned representations of autoencoders can be interpreted as approximately minimizing Wasserstein distance: this result, which applies to latent diffusion models, helps justify their outstanding empirical results. The manifold lens provides a rich perspective from which to understand DGMs, which we aim to make more accessible and widespread.
Authors: Yaozhong Shi, Angela F. Gao, Zachary E. Ross, Kamyar Azizzadenesheli
Abstract: Regression on function spaces is typically limited to models with Gaussian process priors. We introduce the notion of universal functional regression, in which we aim to learn a prior distribution over non-Gaussian function spaces that remains mathematically tractable for functional regression. To do this, we develop Neural Operator Flows (OpFlow), an infinite-dimensional extension of normalizing flows. OpFlow is an invertible operator that maps the (potentially unknown) data function space into a Gaussian process, allowing for exact likelihood estimation of functional point evaluations. OpFlow enables robust and accurate uncertainty quantification via drawing posterior samples of the Gaussian process and subsequently mapping them into the data function space. We empirically study the performance of OpFlow on regression and generation tasks with data generated from Gaussian processes with known posterior forms and non-Gaussian processes, as well as real-world earthquake seismograms with an unknown closed-form distribution.
Authors: Cyriana M. A. Roelofs, Christian G\"uck, Stefan Faulstich
Abstract: Anomaly detection in wind turbines typically involves using normal behaviour models to detect faults early. However, training autoencoder models for each turbine is time-consuming and resource intensive. Thus, transfer learning becomes essential for wind turbines with limited data or applications with limited computational resources. This study examines how cross-turbine transfer learning can be applied to autoencoder-based anomaly detection. Here, autoencoders are combined with constant thresholds for the reconstruction error to determine if input data contains an anomaly. The models are initially trained on one year's worth of data from one or more source wind turbines. They are then fine-tuned using smaller amounts of data from another turbine. Three methods for fine-tuning are investigated: adjusting the entire autoencoder, only the decoder, or only the threshold of the model. The performance of the transfer learning models is compared to baseline models that were trained on one year's worth of data from the target wind turbine. The results of the tests conducted in this study indicate that models trained on data of multiple wind turbines do not improve the anomaly detection capability compared to models trained on data of one source wind turbine. In addition, modifying the model's threshold can lead to comparable or even superior performance compared to the baseline, whereas fine-tuning the decoder or autoencoder further enhances the models' performance.
Authors: Swarup Ranjan Behera, Vijaya V. Saradhi
Abstract: Spectral clustering methods have gained widespread recognition for their effectiveness in clustering high-dimensional data. Among these techniques, constrained spectral clustering has emerged as a prominent approach, demonstrating enhanced performance by integrating pairwise constraints. However, the application of such constraints to semidefinite spectral clustering, a variant that leverages semidefinite programming to optimize clustering objectives, remains largely unexplored. In this paper, we introduce a novel framework for seamlessly integrating pairwise constraints into semidefinite spectral clustering. Our methodology systematically extends the capabilities of semidefinite spectral clustering to capture complex data structures, thereby addressing real-world clustering challenges more effectively. Additionally, we extend this framework to encompass both active and self-taught learning scenarios, further enhancing its versatility and applicability. Empirical studies conducted on well-known datasets demonstrate the superiority of our proposed framework over existing spectral clustering methods, showcasing its robustness and scalability across diverse datasets and learning settings. By bridging the gap between constrained learning and semidefinite spectral clustering, our work contributes to the advancement of spectral clustering techniques, offering researchers and practitioners a versatile tool for addressing complex clustering challenges in various real-world applications. Access to the data, code, and experimental results is provided for further exploration (https://github.com/swarupbehera/SCCCS).
Authors: Renhao Zhang, Haotian Fu, Yilin Miao, George Konidaris
Abstract: We propose a novel model-based reinforcement learning algorithm -- Dynamics Learning and predictive control with Parameterized Actions (DLPA) -- for Parameterized Action Markov Decision Processes (PAMDPs). The agent learns a parameterized-action-conditioned dynamics model and plans with a modified Model Predictive Path Integral control. We theoretically quantify the difference between the generated trajectory and the optimal trajectory during planning in terms of the value they achieved through the lens of Lipschitz Continuity. Our empirical results on several standard benchmarks show that our algorithm achieves superior sample efficiency and asymptotic performance than state-of-the-art PAMDP methods.
Authors: Marcin P. Joachimiak, Mark A. Miller, J. Harry Caufield, Ryan Ly, Nomi L. Harris, Andrew Tritt, Christopher J. Mungall, Kristofer E. Bouchard
Abstract: The Artificial Intelligence Ontology (AIO) is a systematization of artificial intelligence (AI) concepts, methodologies, and their interrelations. Developed via manual curation, with the additional assistance of large language models (LLMs), AIO aims to address the rapidly evolving landscape of AI by providing a comprehensive framework that encompasses both technical and ethical aspects of AI technologies. The primary audience for AIO includes AI researchers, developers, and educators seeking standardized terminology and concepts within the AI domain. The ontology is structured around six top-level branches: Networks, Layers, Functions, LLMs, Preprocessing, and Bias, each designed to support the modular composition of AI methods and facilitate a deeper understanding of deep learning architectures and ethical considerations in AI. AIO's development utilized the Ontology Development Kit (ODK) for its creation and maintenance, with its content being dynamically updated through AI-driven curation support. This approach not only ensures the ontology's relevance amidst the fast-paced advancements in AI but also significantly enhances its utility for researchers, developers, and educators by simplifying the integration of new AI concepts and methodologies. The ontology's utility is demonstrated through the annotation of AI methods data in a catalog of AI research publications and the integration into the BioPortal ontology resource, highlighting its potential for cross-disciplinary research. The AIO ontology is open source and is available on GitHub (https://github.com/berkeleybop/artificial-intelligence-ontology) and BioPortal (https://bioportal.bioontology.org/ontologies/AIO).
URLs: https://github.com/berkeleybop/artificial-intelligence-ontology), https://bioportal.bioontology.org/ontologies/AIO).
Authors: Daniel Potts, Laura Weidensager
Abstract: We propose two algorithms for boosting random Fourier feature models for approximating high-dimensional functions. These methods utilize the classical and generalized analysis of variance (ANOVA) decomposition to learn low-order functions, where there are few interactions between the variables. Our algorithms are able to find an index set of important input variables and variable interactions reliably. Furthermore, we generalize already existing random Fourier feature models to an ANOVA setting, where terms of different order can be used. Our algorithms have the advantage of interpretability, meaning that the influence of every input variable is known in the learned model, even for dependent input variables. We give theoretical as well as numerical results that our algorithms perform well for sensitivity analysis. The ANOVA-boosting step reduces the approximation error of existing methods significantly.
Authors: Krzysztof Siminski, Konrad Wnuk
Abstract: Neuro-fuzzy systems are a technique of explainable artificial intelligence (XAI). They elaborate knowledge models as a set of fuzzy rules. Fuzzy sets are crucial components of fuzzy rules. They are used to model linguistic terms. In this paper, we present an automatic extraction of fuzzy rules in the natural English language. Full implementation is available free from a public repository.
Authors: Yifan Qu, Oliver Krzysik, Hans De Sterck, Omer Ege Kara
Abstract: Graph Neural Networks (GNNs) have established themselves as the preferred methodology in a multitude of domains, ranging from computer vision to computational biology, especially in contexts where data inherently conform to graph structures. While many existing methods have endeavored to model GNNs using various techniques, a prevalent challenge they grapple with is the issue of over-smoothing. This paper presents new Graph Neural Network models that incorporate two first-order Partial Differential Equations (PDEs). These models do not increase complexity but effectively mitigate the over-smoothing problem. Our experimental findings highlight the capacity of our new PDE model to achieve comparable results with higher-order PDE models and fix the over-smoothing problem up to 64 layers. These results underscore the adaptability and versatility of GNNs, indicating that unconventional approaches can yield outcomes on par with established techniques.
Authors: Arnauld Mesinga Mwafise
Abstract: Posets are discrete mathematical structures which are ubiquitous in a broad range of data analysis and machine learning applications. Research connecting posets to the data science domain has been ongoing for many years. In this paper, a comprehensive review of a wide range of studies on data analysis amd machine learning using posets are examined in terms of their theory, algorithms and applications. In addition, the applied lattice theory domain of formal concept analysis will also be highlighted in terms of its machine learning applications.
Authors: Manfred Diaz, Liam Paull, Andrea Tacchetti
Abstract: Teacher-Student Curriculum Learning (TSCL) is a curriculum learning framework that draws inspiration from human cultural transmission and learning. It involves a teacher algorithm shaping the learning process of a learner algorithm by exposing it to controlled experiences. Despite its success, understanding the conditions under which TSCL is effective remains challenging. In this paper, we propose a data-centric perspective to analyze the underlying mechanics of the teacher-student interactions in TSCL. We leverage cooperative game theory to describe how the composition of the set of experiences presented by the teacher to the learner, as well as their order, influences the performance of the curriculum that is found by TSCL approaches. To do so, we demonstrate that for every TSCL problem, there exists an equivalent cooperative game, and several key components of the TSCL framework can be reinterpreted using game-theoretic principles. Through experiments covering supervised learning, reinforcement learning, and classical games, we estimate the cooperative values of experiences and use value-proportional curriculum mechanisms to construct curricula, even in cases where TSCL struggles. The framework and experimental setup we present in this work represent a novel foundation for a deeper exploration of TSCL, shedding light on its underlying mechanisms and providing insights into its broader applicability in machine learning.
Authors: Zexin Fang, Bin Han, Hans D. Schotten
Abstract: Federated learning (FL) offers a privacy-preserving collaborative approach for training models in wireless networks, with channel estimation emerging as a promising application. Despite extensive studies on FL-empowered channel estimation, the security concerns associated with FL require meticulous attention. In a scenario where small base stations (SBSs) serve as local models trained on cached data, and a macro base station (MBS) functions as the global model setting, an attacker can exploit the vulnerability of FL, launching attacks with various adversarial attacks or deployment tactics. In this paper, we analyze such vulnerabilities, corresponding solutions were brought forth, and validated through simulation.
Authors: Leonardo Ferreira Guilhoto, Paris Perdikaris
Abstract: Operator learning is a rising field of scientific computing where inputs or outputs of a machine learning model are functions defined in infinite-dimensional spaces. In this paper, we introduce NEON (Neural Epistemic Operator Networks), an architecture for generating predictions with uncertainty using a single operator network backbone, which presents orders of magnitude less trainable parameters than deep ensembles of comparable performance. We showcase the utility of this method for sequential decision-making by examining the problem of composite Bayesian Optimization (BO), where we aim to optimize a function $f=g\circ h$, where $h:X\to C(\mathcal{Y},\mathbb{R}^{d_s})$ is an unknown map which outputs elements of a function space, and $g: C(\mathcal{Y},\mathbb{R}^{d_s})\to \mathbb{R}$ is a known and cheap-to-compute functional. By comparing our approach to other state-of-the-art methods on toy and real world scenarios, we demonstrate that NEON achieves state-of-the-art performance while requiring orders of magnitude less trainable parameters.
Authors: Joo Seung Lee, Malini Mahendra, Anil Aswani
Abstract: Mechanical ventilation is a critical life-support intervention that uses a machine to deliver controlled air and oxygen to a patient's lungs, assisting or replacing spontaneous breathing. While several data-driven approaches have been proposed to optimize ventilator control strategies, they often lack interpretability and agreement with general domain knowledge. This paper proposes a methodology for interpretable reinforcement learning (RL) using decision trees for mechanical ventilation control. Using a causal, nonparametric model-based off-policy evaluation, we evaluate the policies in their ability to gain increases in SpO2 while avoiding aggressive ventilator settings which are known to cause ventilator induced lung injuries and other complications. Numerical experiments using MIMIC-III data on the stays of real patients' intensive care unit stays demonstrate that the decision tree policy outperforms the behavior cloning policy and is comparable to state-of-the-art RL policy. Future work concerns better aligning the cost function with medical objectives to generate deeper clinical insights.
Authors: Xuesong Wang, Nina Fatehi, Caisheng Wang, Masoud H. Nazari
Abstract: This paper presents a deep learning-based approach for hourly power outage probability prediction within census tracts encompassing a utility company's service territory. Two distinct deep learning models, conditional Multi-Layer Perceptron (MLP) and unconditional MLP, were developed to forecast power outage probabilities, leveraging a rich array of input features gathered from publicly available sources including weather data, weather station locations, power infrastructure maps, socio-economic and demographic statistics, and power outage records. Given a one-hour-ahead weather forecast, the models predict the power outage probability for each census tract, taking into account both the weather prediction and the location's characteristics. The deep learning models employed different loss functions to optimize prediction performance. Our experimental results underscore the significance of socio-economic factors in enhancing the accuracy of power outage predictions at the census tract level.
Authors: Arjun Subramonian, Jian Kang, Yizhou Sun
Abstract: Graph Neural Networks (GNNs) often perform better for high-degree nodes than low-degree nodes on node classification tasks. This degree bias can reinforce social marginalization by, e.g., sidelining authors of lowly-cited papers when predicting paper topics in citation networks. While researchers have proposed numerous hypotheses for why GNN degree bias occurs, we find via a survey of 38 degree bias papers that these hypotheses are often not rigorously validated, and can even be contradictory. Thus, we provide an analysis of the origins of degree bias in message-passing GNNs with different graph filters. We prove that high-degree test nodes tend to have a lower probability of misclassification regardless of how GNNs are trained. Moreover, we show that degree bias arises from a variety of factors that are associated with a node's degree (e.g., homophily of neighbors, diversity of neighbors). Furthermore, we show that during training, some GNNs may adjust their loss on low-degree nodes more slowly than on high-degree nodes; however, with sufficiently many epochs of training, message-passing GNNs can achieve their maximum possible training accuracy, which is not significantly limited by their expressive power. Throughout our analysis, we connect our findings to previously-proposed hypotheses for the origins of degree bias, supporting and unifying some while drawing doubt to others. We validate our theoretical findings on 8 common real-world networks, and based on our theoretical and empirical insights, describe a roadmap to alleviate degree bias.
Authors: Tom\'as Vergara-Browne, \'Alvaro Soto, Akiko Aizawa
Abstract: We introduce eigenpruning, a method that removes singular values from weight matrices in an LLM to improve its performance in a particular task. This method is inspired by interpretability methods designed to automatically find subnetworks of a model which solve a specific task. In our tests, the pruned model outperforms the original model by a large margin, while only requiring minimal computation to prune the weight matrices. In the case of a small synthetic task in integer multiplication, the Phi-2 model can improve its accuracy in the test set from 13.75% to 97.50%. Interestingly, these results seem to indicate the existence of a computation path that can solve the task very effectively, but it was not being used by the original model. Finally, we plan to open-source our implementation in the camera-ready version of our work.
Authors: Haiyun He, Christina Lee Yu, Ziv Goldfeld
Abstract: Deep neural networks (DNNs) exhibit an exceptional capacity for generalization in practical applications. This work aims to capture the effect and benefits of depth for supervised learning via information-theoretic generalization bounds. We first derive two hierarchical bounds on the generalization error in terms of the Kullback-Leibler (KL) divergence or the 1-Wasserstein distance between the train and test distributions of the network internal representations. The KL divergence bound shrinks as the layer index increases, while the Wasserstein bound implies the existence of a layer that serves as a generalization funnel, which attains a minimal 1-Wasserstein distance. Analytic expressions for both bounds are derived under the setting of binary Gaussian classification with linear DNNs. To quantify the contraction of the relevant information measures when moving deeper into the network, we analyze the strong data processing inequality (SDPI) coefficient between consecutive layers of three regularized DNN models: Dropout, DropConnect, and Gaussian noise injection. This enables refining our generalization bounds to capture the contraction as a function of the network architecture parameters. Specializing our results to DNNs with a finite parameter space and the Gibbs algorithm reveals that deeper yet narrower network architectures generalize better in those examples, although how broadly this statement applies remains a question.
Authors: Houzhe Wang, Xiaojie Zhu, Chi Chen, Paulo Esteves-Ver\'issimo
Abstract: With recent legislation on the right to be forgotten, machine unlearning has emerged as a crucial research area. It facilitates the removal of a user's data from federated trained machine learning models without the necessity for retraining from scratch. However, current machine unlearning algorithms are confronted with challenges of efficiency and validity.To address the above issues, we propose a new framework, named Goldfish. It comprises four modules: basic model, loss function, optimization, and extension. To address the challenge of low validity in existing machine unlearning algorithms, we propose a novel loss function. It takes into account the loss arising from the discrepancy between predictions and actual labels in the remaining dataset. Simultaneously, it takes into consideration the bias of predicted results on the removed dataset. Moreover, it accounts for the confidence level of predicted results. Additionally, to enhance efficiency, we adopt knowledge distillation technique in basic model and introduce an optimization module that encompasses the early termination mechanism guided by empirical risk and the data partition mechanism. Furthermore, to bolster the robustness of the aggregated model, we propose an extension module that incorporates a mechanism using adaptive distillation temperature to address the heterogeneity of user local data and a mechanism using adaptive weight to handle the variety in the quality of uploaded models. Finally, we conduct comprehensive experiments to illustrate the effectiveness of proposed approach.
Authors: Quentin Jodelet, Xin Liu, Yin Jun Phua, Tsuyoshi Murata
Abstract: Exemplar-Free Class Incremental Learning is a highly challenging setting where replay memory is unavailable. Methods relying on frozen feature extractors have drawn attention recently in this setting due to their impressive performances and lower computational costs. However, those methods are highly dependent on the data used to train the feature extractor and may struggle when an insufficient amount of classes are available during the first incremental step. To overcome this limitation, we propose to use a pre-trained text-to-image diffusion model in order to generate synthetic images of future classes and use them to train the feature extractor. Experiments on the standard benchmarks CIFAR100 and ImageNet-Subset demonstrate that our proposed method can be used to improve state-of-the-art methods for exemplar-free class incremental learning, especially in the most difficult settings where the first incremental step only contains few classes. Moreover, we show that using synthetic samples of future classes achieves higher performance than using real data from different classes, paving the way for better and less costly pre-training methods for incremental learning.
Authors: Sayantan Kumar, Sean Yu, Thomas Kannampallil, Andrew Michelson, Aristeidis Sotiras, Philip Payne
Abstract: Early identification of Mild Cognitive Impairment (MCI) subjects who will eventually progress to Alzheimer Disease (AD) is challenging. Existing deep learning models are mostly single-modality single-task models predicting risk of disease progression at a fixed timepoint. We proposed a multimodal hierarchical multi-task learning approach which can monitor the risk of disease progression at each timepoint of the visit trajectory. Longitudinal visit data from multiple modalities (MRI, cognition, and clinical data) were collected from MCI individuals of the Alzheimer Disease Neuroimaging Initiative (ADNI) dataset. Our hierarchical model predicted at every timepoint a set of neuropsychological composite cognitive function scores as auxiliary tasks and used the forecasted scores at every timepoint to predict the future risk of disease. Relevance weights for each composite function provided explanations about potential factors for disease progression. Our proposed model performed better than state-of-the-art baselines in predicting AD progression risk and the composite scores. Ablation study on the number of modalities demonstrated that imaging and cognition data contributed most towards the outcome. Model explanations at each timepoint can inform clinicians 6 months in advance the potential cognitive function decline that can lead to progression to AD in future. Our model monitored their risk of AD progression every 6 months throughout the visit trajectory of individuals. The hierarchical learning of auxiliary tasks allowed better optimization and allowed longitudinal explanations for the outcome. Our framework is flexible with the number of input modalities and the selection of auxiliary tasks and hence can be generalized to other clinical problems too.
Authors: Xiwei Zhang, Tao Li
Abstract: We study the convergence of recursive regularized learning algorithms in the reproducing kernel Hilbert space (RKHS) with dependent and non-stationary online data streams. Firstly, we study the mean square asymptotic stability of a class of random difference equations in RKHS, whose non-homogeneous terms are martingale difference sequences dependent on the homogeneous ones. Secondly, we introduce the concept of random Tikhonov regularization path, and show that if the regularization path is slowly time-varying in some sense, then the output of the algorithm is consistent with the regularization path in mean square. Furthermore, if the data streams also satisfy the RKHS persistence of excitation condition, i.e. there exists a fixed length of time period, such that each eigenvalue of the conditional expectation of the operators induced by the input data accumulated over every time period has a uniformly positive lower bound with respect to time, then the output of the algorithm is consistent with the unknown function in mean square. Finally, for the case with independent and non-identically distributed data streams, the algorithm achieves the mean square consistency provided the marginal probability measures induced by the input data are slowly time-varying and the average measure over each fixed-length time period has a uniformly strictly positive lower bound.
Authors: Alvaro Carbonero, Shaowen Mao, Mohamed Mehana
Abstract: To address the urgent challenge of climate change, there is a critical need to transition away from fossil fuels towards sustainable energy systems, with renewable energy sources playing a pivotal role. However, the inherent variability of renewable energy, without effective storage solutions, often leads to imbalances between energy supply and demand. Underground Hydrogen Storage (UHS) emerges as a promising long-term storage solution to bridge this gap, yet its widespread implementation is impeded by the high computational costs associated with high fidelity UHS simulations. This paper introduces UHS from a data-driven perspective and outlines a roadmap for integrating machine learning into UHS, thereby facilitating the large-scale deployment of UHS.
Authors: Zakaria Elabid, Daniel Busby, Abdenour Hadid
Abstract: Physics-informed neural networks (PINNs) have gained significant prominence as a powerful tool in the field of scientific computing and simulations. Their ability to seamlessly integrate physical principles into deep learning architectures has revolutionized the approaches to solving complex problems in physics and engineering. However, a persistent challenge faced by mainstream PINNs lies in their handling of discontinuous input data, leading to inaccuracies in predictions. This study addresses these challenges by incorporating the discretized forms of the governing equations into the PINN framework. We propose to combine the power of neural networks with the dynamics imposed by the discretized differential equations. By discretizing the governing equations, the PINN learns to account for the discontinuities and accurately capture the underlying relationships between inputs and outputs, improving the accuracy compared to traditional interpolation techniques. Moreover, by leveraging the power of neural networks, the computational cost associated with numerical simulations is substantially reduced. We evaluate our model on a large-scale dataset for the prediction of pressure and saturation fields demonstrating high accuracies compared to non-physically aware models.
Authors: Sean Farhat, Deming Chen
Abstract: In this paper, we propose that small models may not need to absorb the cost of pre-training to reap its benefits. Instead, they can capitalize on the astonishing results achieved by modern, enormous models to a surprising degree. We observe that, when distilled on a task from a pre-trained teacher model, a small model can achieve or surpass the performance it would achieve if it was pre-trained then finetuned on that task. To allow this phenomenon to be easily leveraged, we establish a connection reducing knowledge distillation to modern contrastive learning, opening two doors: (1) vastly different model architecture pairings can work for the distillation, and (2) most contrastive learning algorithms rooted in the theory of Noise Contrastive Estimation can be easily applied and used. We demonstrate this paradigm using pre-trained teacher models from open-source model hubs, Transformer and convolution based model combinations, and a novel distillation algorithm that massages the Alignment/Uniformity perspective of contrastive learning by Wang & Isola (2020) into a distillation objective. We choose this flavor of contrastive learning due to its low computational cost, an overarching theme of this work. We also observe that this phenomenon tends not to occur if the task is data-limited. However, this can be alleviated by leveraging yet another scale-inspired development: large, pre-trained generative models for dataset augmentation. Again, we use an open-source model, and our rudimentary prompts are sufficient to boost the small model`s performance. Thus, we highlight a training method for small models that is up to 94% faster than the standard pre-training paradigm without sacrificing performance. For practitioners discouraged from fully utilizing modern foundation datasets for their small models due to the prohibitive scale, we believe our work keeps that door open.
Authors: Min Jae Song
Abstract: We show that $L^2$-accurate score estimation, in the absence of strong assumptions on the data distribution, is computationally hard even when sample complexity is polynomial in the relevant problem parameters. Our reduction builds on the result of Chen et al. (ICLR 2023), who showed that the problem of generating samples from an unknown data distribution reduces to $L^2$-accurate score estimation. Our hard-to-estimate distributions are the "Gaussian pancakes" distributions, originally due to Diakonikolas et al. (FOCS 2017), which have been shown to be computationally indistinguishable from the standard Gaussian under widely believed hardness assumptions from lattice-based cryptography (Bruna et al., STOC 2021; Gupte et al., FOCS 2022).
Authors: Mokhtar Z. Alaya (LMAC), Alain Rakotomamonjy (LITIS), Maxime Berar (LITIS), Gilles Gasso (LITIS)
Abstract: Gaussian smoothed sliced Wasserstein distance has been recently introduced for comparing probability distributions, while preserving privacy on the data. It has been shown that it provides performances similar to its non-smoothed (non-private) counterpart. However, the computationaland statistical properties of such a metric have not yet been well-established. This work investigates the theoretical properties of this distance as well as those of generalized versions denoted as Gaussian-smoothed sliced divergences. We first show that smoothing and slicing preserve the metric property and the weak topology. To study the sample complexity of such divergences, we then introduce $\hat{\hat\mu}_{n}$ the double empirical distribution for the smoothed-projected $\mu$. The distribution $\hat{\hat\mu}_{n}$ is a result of a double sampling process: one from sampling according to the origin distribution $\mu$ and the second according to the convolution of the projection of $\mu$ on the unit sphere and the Gaussian smoothing. We particularly focus on the Gaussian smoothed sliced Wasserstein distance and prove that it converges with a rate $O(n^{-1/2})$. We also derive other properties, including continuity, of different divergences with respect to the smoothing parameter. We support our theoretical findings with empirical studies in the context of privacy-preserving domain adaptation.
Authors: Michael Sucker, Jalal Fadili, Peter Ochs
Abstract: We use the PAC-Bayesian theory for the setting of learning-to-optimize. To the best of our knowledge, we present the first framework to learn optimization algorithms with provable generalization guarantees (PAC-Bayesian bounds) and explicit trade-off between convergence guarantees and convergence speed, which contrasts with the typical worst-case analysis. Our learned optimization algorithms provably outperform related ones derived from a (deterministic) worst-case analysis. The results rely on PAC-Bayesian bounds for general, possibly unbounded loss-functions based on exponential families. Then, we reformulate the learning procedure into a one-dimensional minimization problem and study the possibility to find a global minimum. Furthermore, we provide a concrete algorithmic realization of the framework and new methodologies for learning-to-optimize, and we conduct four practically relevant experiments to support our theory. With this, we showcase that the provided learning framework yields optimization algorithms that provably outperform the state-of-the-art by orders of magnitude.
Authors: Aditya Shankar, Hans Brouwer, Rihan Hai, Lydia Chen
Abstract: Synthetic tabular data is crucial for sharing and augmenting data across silos, especially for enterprises with proprietary data. However, existing synthesizers are designed for centrally stored data. Hence, they struggle with real-world scenarios where features are distributed across multiple silos, necessitating on-premise data storage. We introduce SiloFuse, a novel generative framework for high-quality synthesis from cross-silo tabular data. To ensure privacy, SiloFuse utilizes a distributed latent tabular diffusion architecture. Through autoencoders, latent representations are learned for each client's features, masking their actual values. We employ stacked distributed training to improve communication efficiency, reducing the number of rounds to a single step. Under SiloFuse, we prove the impossibility of data reconstruction for vertically partitioned synthesis and quantify privacy risks through three attacks using our benchmark framework. Experimental results on nine datasets showcase SiloFuse's competence against centralized diffusion-based synthesizers. Notably, SiloFuse achieves 43.8 and 29.8 higher percentage points over GANs in resemblance and utility. Experiments on communication show stacked training's fixed cost compared to the growing costs of end-to-end training as the number of training iterations increases. Additionally, SiloFuse proves robust to feature permutations and varying numbers of clients.
Authors: Naram Mhaisen, George Iosifidis
Abstract: This paper brings the concept of "optimism" to the new and promising framework of online Non-stochastic Control (NSC). Namely, we study how can NSC benefit from a prediction oracle of unknown quality responsible for forecasting future costs. The posed problem is first reduced to an optimistic learning with delayed feedback problem, which is handled through the Optimistic Follow the Regularized Leader (OFTRL) algorithmic family. This reduction enables the design of OptFTRL-C, the first Disturbance Action Controller (DAC) with optimistic policy regret bounds. These new bounds are commensurate with the oracle's accuracy, ranging from $\mathcal{O}(1)$ for perfect predictions to the order-optimal $\mathcal{O}(\sqrt{T})$ even when all predictions fail. By addressing the challenge of incorporating untrusted predictions into control systems, our work contributes to the advancement of the NSC framework and paves the way towards effective and robust learning-based controllers.
Authors: Abhishek Duttagupta, Jin Zhao, Shanker Shreejith
Abstract: Federated Learning (FL) is a distributed learning scheme that enables deep learning to be applied to sensitive data streams and applications in a privacy-preserving manner. This paper focuses on the use of FL for analyzing smart energy meter data with the aim to achieve comparable accuracy to state-of-the-art methods for load forecasting while ensuring the privacy of individual meter data. We show that with a lightweight fully connected deep neural network, we are able to achieve forecasting accuracy comparable to existing schemes, both at each meter source and at the aggregator, by utilising the FL framework. The use of lightweight models further reduces the energy and resource consumption caused by complex deep-learning models, making this approach ideally suited for deployment across resource-constrained smart meter systems. With our proposed lightweight model, we are able to achieve an overall average load forecasting RMSE of 0.17, with the model having a negligible energy overhead of 50 mWh when performing training and inference on an Arduino Uno platform.
Authors: Yukun Xie, Juan Du, Chen Zhang
Abstract: In modern manufacturing, most of the product lines are conforming. Few products are nonconforming but with different defect types. The identification of defect types can help further root cause diagnosis of production lines. With the sensing development, continuous signals of process variables can be collected in high resolution, which can be regarded as multichannel functional data. They have abundant information to characterize the process and help identify the defect types. Motivated by a real example from the pipe tightening process, we target at detect classification when each sample is a multichannel functional data. However, the available samples for each defect type are limited and imbalanced. Moreover, the functions are partially observed since the pre-tightening process before the pipe tightening process is unobserved. To classify the defect samples based on imbalanced, multichannel, and partially observed functional data is very important but challenging. Thus, we propose an innovative framework known as "Multichannel Partially Observed Functional Modeling for Defect Classification with an Imbalanced Dataset" (MPOFI). The framework leverages the power of deep metric learning in conjunction with a neural network specially crafted for processing functional data. This paper introduces a neural network explicitly tailored for handling multichannel and partially observed functional data, complemented by developing a corresponding loss function for training on imbalanced datasets. The results from a real-world case study demonstrate the superior accuracy of our framework when compared to existing benchmarks.
Authors: Fatima Ezzeddine, Omran Ayoub, Silvia Giordano
Abstract: In recent years, there has been a notable increase in the deployment of machine learning (ML) models as services (MLaaS) across diverse production software applications. In parallel, explainable AI (XAI) continues to evolve, addressing the necessity for transparency and trustworthiness in ML models. XAI techniques aim to enhance the transparency of ML models by providing insights, in terms of the model's explanations, into their decision-making process. Simultaneously, some MLaaS platforms now offer explanations alongside the ML prediction outputs. This setup has elevated concerns regarding vulnerabilities in MLaaS, particularly in relation to privacy leakage attacks such as model extraction attacks (MEA). This is due to the fact that explanations can unveil insights about the inner workings of the model which could be exploited by malicious users. In this work, we focus on investigating how model explanations, particularly Generative adversarial networks (GANs)-based counterfactual explanations (CFs), can be exploited for performing MEA within the MLaaS platform. We also delve into assessing the effectiveness of incorporating differential privacy (DP) as a mitigation strategy. To this end, we first propose a novel MEA methodology based on Knowledge Distillation (KD) to enhance the efficiency of extracting a substitute model of a target model exploiting CFs. Then, we advise an approach for training CF generators incorporating DP to generate private CFs. We conduct thorough experimental evaluations on real-world datasets and demonstrate that our proposed KD-based MEA can yield a high-fidelity substitute model with reduced queries with respect to baseline approaches. Furthermore, our findings reveal that the inclusion of a privacy layer impacts the performance of the explainer, the quality of CFs, and results in a reduction in the MEA performance.
Authors: Philipp Altmann, C\'eline Davignon, Maximilian Zorn, Fabian Ritz, Claudia Linnhoff-Popien, Thomas Gabor
Abstract: To enhance the interpretability of Reinforcement Learning (RL), we propose Revealing Evolutionary Action Consequence Trajectories (REACT). In contrast to the prevalent practice of validating RL models based on their optimal behavior learned during training, we posit that considering a range of edge-case trajectories provides a more comprehensive understanding of their inherent behavior. To induce such scenarios, we introduce a disturbance to the initial state, optimizing it through an evolutionary algorithm to generate a diverse population of demonstrations. To evaluate the fitness of trajectories, REACT incorporates a joint fitness function that encourages both local and global diversity in the encountered states and chosen actions. Through assessments with policies trained for varying durations in discrete and continuous environments, we demonstrate the descriptive power of REACT. Our results highlight its effectiveness in revealing nuanced aspects of RL models' behavior beyond optimal performance, thereby contributing to improved interpretability.
Authors: Raffael Theiler, Olga Fink
Abstract: Pumped-storage hydropower plants (PSH) actively participate in grid power-frequency control and therefore often operate under dynamic conditions, which results in rapidly varying system states. Predicting these dynamically changing states is essential for comprehending the underlying sensor and machine conditions. This understanding aids in detecting anomalies and faults, ensuring the reliable operation of the connected power grid, and in identifying faulty and miscalibrated sensors. PSH are complex, highly interconnected systems encompassing electrical and hydraulic subsystems, each characterized by their respective underlying networks that can individually be represented as graphs. To take advantage of this relational inductive bias, graph neural networks (GNNs) have been separately applied to state forecasting tasks in the individual subsystems, but without considering their interdependencies. In PSH, however, these subsystems depend on the same control input, making their operations highly interdependent and interconnected. Consequently, hydraulic and electrical sensor data should be fused across PSH subsystems to improve state forecasting accuracy. This approach has not been explored in GNN literature yet because many available PSH graphs are limited to their respective subsystem boundaries, which makes the method unsuitable to be applied directly. In this work, we introduce the application of spectral-temporal graph neural networks, which leverage self-attention mechanisms to concurrently capture and learn meaningful subsystem interdependencies and the dynamic patterns observed in electric and hydraulic sensors. Our method effectively fuses data from the PSH's subsystems by operating on a unified, system-wide graph, learned directly from the data, This approach leads to demonstrably improved state forecasting performance and enhanced generalizability.
Authors: Cai Zhou, Rose Yu, Yusu Wang
Abstract: Graph transformers have recently received significant attention in graph learning, partly due to their ability to capture more global interaction via self-attention. Nevertheless, while higher-order graph neural networks have been reasonably well studied, the exploration of extending graph transformers to higher-order variants is just starting. Both theoretical understanding and empirical results are limited. In this paper, we provide a systematic study of the theoretical expressive power of order-$k$ graph transformers and sparse variants. We first show that, an order-$k$ graph transformer without additional structural information is less expressive than the $k$-Weisfeiler Lehman ($k$-WL) test despite its high computational cost. We then explore strategies to both sparsify and enhance the higher-order graph transformers, aiming to improve both their efficiency and expressiveness. Indeed, sparsification based on neighborhood information can enhance the expressive power, as it provides additional information about input graph structures. In particular, we show that a natural neighborhood-based sparse order-$k$ transformer model is not only computationally efficient, but also expressive -- as expressive as $k$-WL test. We further study several other sparse graph attention models that are computationally efficient and provide their expressiveness analysis. Finally, we provide experimental results to show the effectiveness of the different sparsification strategies.
Authors: Kaichen Huang, Hai-Hang Sun, Shenghua Wan, Minghao Shao, Shuai Feng, Le Gan, De-Chuan Zhan
Abstract: Imitating skills from low-quality datasets, such as sub-optimal demonstrations and observations with distractors, is common in real-world applications. In this work, we focus on the problem of Learning from Noisy Demonstrations (LND), where the imitator is required to learn from data with noise that often occurs during the processes of data collection or transmission. Previous IL methods improve the robustness of learned policies by injecting an adversarially learned Gaussian noise into pure expert data or utilizing additional ranking information, but they may fail in the LND setting. To alleviate the above problems, we propose Denoised Imitation learning based on Domain Adaptation (DIDA), which designs two discriminators to distinguish the noise level and expertise level of data, facilitating a feature encoder to learn task-related but domain-agnostic representations. Experiment results on MuJoCo demonstrate that DIDA can successfully handle challenging imitation tasks from demonstrations with various types of noise, outperforming most baseline methods.
Authors: Shuo Chen, Zhen Han, Bailan He, Zifeng Ding, Wenqian Yu, Philip Torr, Volker Tresp, Jindong Gu
Abstract: Various jailbreak attacks have been proposed to red-team Large Language Models (LLMs) and revealed the vulnerable safeguards of LLMs. Besides, some methods are not limited to the textual modality and extend the jailbreak attack to Multimodal Large Language Models (MLLMs) by perturbing the visual input. However, the absence of a universal evaluation benchmark complicates the performance reproduction and fair comparison. Besides, there is a lack of comprehensive evaluation of closed-source state-of-the-art (SOTA) models, especially MLLMs, such as GPT-4V. To address these issues, this work first builds a comprehensive jailbreak evaluation dataset with 1445 harmful questions covering 11 different safety policies. Based on this dataset, extensive red-teaming experiments are conducted on 11 different LLMs and MLLMs, including both SOTA proprietary models and open-source models. We then conduct a deep analysis of the evaluated results and find that (1) GPT4 and GPT-4V demonstrate better robustness against jailbreak attacks compared to open-source LLMs and MLLMs. (2) Llama2 and Qwen-VL-Chat are more robust compared to other open-source models. (3) The transferability of visual jailbreak methods is relatively limited compared to textual jailbreak methods. The dataset and code can be found here https://anonymous.4open.science/r/red_teaming_gpt4-C1CE/README.md .
URLs: https://anonymous.4open.science/r/red_teaming_gpt4-C1CE/README.md
Authors: Hern\'an Ceferino V\'azquez, Jorge Sanchez, Rafael Carrascosa
Abstract: Automated Machine Learning (AutoML) has become increasingly popular in recent years due to its ability to reduce the amount of time and expertise required to design and develop machine learning systems. This is very important for the practice of machine learning, as it allows building strong baselines quickly, improving the efficiency of the data scientists, and reducing the time to production. However, despite the advantages of AutoML, it faces several challenges, such as defining the solutions space and exploring it efficiently. Recently, some approaches have been shown to be able to do it using tree-based search algorithms and context-free grammars. In particular, GramML presents a model-free reinforcement learning approach that leverages pipeline configuration grammars and operates using Monte Carlo tree search. However, one of the limitations of GramML is that it uses default hyperparameters, limiting the search problem to finding optimal pipeline structures for the available data preprocessors and models. In this work, we propose an extension to GramML that supports larger search spaces including hyperparameter search. We evaluated the approach using an OpenML benchmark and found significant improvements compared to other state-of-the-art techniques.
Authors: Mateusz Gajewski, Adam Karczmarz, Mateusz Rapicki, Piotr Sankowski
Abstract: In this paper, we consider a perturbation-based metric of predictive faithfulness of feature rankings (or attributions) that we call PGI squared. When applied to decision tree-based regression models, the metric can be computed accurately and efficiently for arbitrary independent feature perturbation distributions. In particular, the computation does not involve Monte Carlo sampling that has been typically used for computing similar metrics and which is inherently prone to inaccuracies. Moreover, we propose a method of ranking features by their importance for the tree model's predictions based on PGI squared. Our experiments indicate that in some respects, the method may identify the globally important features better than the state-of-the-art SHAP explainer
Authors: Florian Frantzen, Michael T. Schaub
Abstract: Triggered by limitations of graph-based deep learning methods in terms of computational expressivity and model flexibility, recent years have seen a surge of interest in computational models that operate on higher-order topological domains such as hypergraphs and simplicial complexes. While the increased expressivity of these models can indeed lead to a better classification performance and a more faithful representation of the underlying system, the computational cost of these higher-order models can increase dramatically. To this end, we here explore a simplicial complex neural network learning architecture based on random walks and fast 1D convolutions (SCRaWl), in which we can adjust the increase in computational cost by varying the length and number of random walks considered while accounting for higher-order relationships. Importantly, due to the random walk-based design, the expressivity of the proposed architecture is provably incomparable to that of existing message-passing simplicial neural networks. We empirically evaluate SCRaWl on real-world datasets and show that it outperforms other simplicial neural networks.
Authors: Sohir Maskey, Gitta Kutyniok, Ron Levie
Abstract: We study the generalization capabilities of Message Passing Neural Networks (MPNNs), a prevalent class of Graph Neural Networks (GNN). We derive generalization bounds specifically for MPNNs with normalized sum aggregation and mean aggregation. Our analysis is based on a data generation model incorporating a finite set of template graphons. Each graph within this framework is generated by sampling from one of the graphons with a certain degree of perturbation. In particular, we extend previous MPNN generalization results to a more realistic setting, which includes the following modifications: 1) we analyze simple random graphs with Bernoulli-distributed edges instead of weighted graphs; 2) we sample both graphs and graph signals from perturbed graphons instead of clean graphons; and 3) we analyze sparse graphs instead of dense graphs. In this more realistic and challenging scenario, we provide a generalization bound that decreases as the average number of nodes in the graphs increases. Our results imply that MPNNs with higher complexity than the size of the training set can still generalize effectively, as long as the graphs are sufficiently large.
Authors: Simon Kl\"uttermann, Emmanuel M\"uller
Abstract: In this paper, we introduce DOUST, our method applying test-time training for outlier detection, significantly improving the detection performance. After thoroughly evaluating our algorithm on common benchmark datasets, we discuss a common problem and show that it disappears with a large enough test set. Thus, we conclude that under reasonable conditions, our algorithm can reach almost supervised performance even when no labeled outliers are given.
Authors: Okko Makkonen, Sampo Niemel\"a, Camilla Hollanti, Serge Kas Hanna
Abstract: This work focuses on the challenges of non-IID data and stragglers/dropouts in federated learning. We introduce and explore a privacy-flexible paradigm that models parts of the clients' local data as non-private, offering a more versatile and business-oriented perspective on privacy. Within this framework, we propose a data-driven strategy for mitigating the effects of label heterogeneity and client straggling on federated learning. Our solution combines both offline data sharing and approximate gradient coding techniques. Through numerical simulations using the MNIST dataset, we demonstrate that our approach enables achieving a deliberate trade-off between privacy and utility, leading to improved model convergence and accuracy while using an adaptable portion of non-private data.
Authors: Miao Lu, Han Zhong, Tong Zhang, Jose Blanchet
Abstract: The sim-to-real gap, which represents the disparity between training and testing environments, poses a significant challenge in reinforcement learning (RL). A promising approach to addressing this challenge is distributionally robust RL, often framed as a robust Markov decision process (RMDP). In this framework, the objective is to find a robust policy that achieves good performance under the worst-case scenario among all environments within a pre-specified uncertainty set centered around the training environment. Unlike previous work, which relies on a generative model or a pre-collected offline dataset enjoying good coverage of the deployment environment, we tackle robust RL via interactive data collection, where the learner interacts with the training environment only and refines the policy through trial and error. In this robust RL paradigm, two main challenges emerge: managing distributional robustness while striking a balance between exploration and exploitation during data collection. Initially, we establish that sample-efficient learning without additional assumptions is unattainable owing to the curse of support shift; i.e., the potential disjointedness of the distributional supports between the training and testing environments. To circumvent such a hardness result, we introduce the vanishing minimal value assumption to RMDPs with a total-variation (TV) distance robust set, postulating that the minimal value of the optimal robust value function is zero. We prove that such an assumption effectively eliminates the support shift issue for RMDPs with a TV distance robust set, and present an algorithm with a provable sample complexity guarantee. Our work makes the initial step to uncovering the inherent difficulty of robust RL via interactive data collection and sufficient conditions for designing a sample-efficient algorithm accompanied by sharp sample complexity analysis.
Authors: Tyler Chang, Andrew Gillette, Romit Maulik
Abstract: Effective verification and validation techniques for modern scientific machine learning workflows are challenging to devise. Statistical methods are abundant and easily deployed, but often rely on speculative assumptions about the data and methods involved. Error bounds for classical interpolation techniques can provide mathematically rigorous estimates of accuracy, but often are difficult or impractical to determine computationally. In this work, we present a best-of-both-worlds approach to verifiable scientific machine learning by demonstrating that (1) multiple standard interpolation techniques have informative error bounds that can be computed or estimated efficiently; (2) comparative performance among distinct interpolants can aid in validation goals; (3) deploying interpolation methods on latent spaces generated by deep learning techniques enables some interpretability for black-box models. We present a detailed case study of our approach for predicting lift-drag ratios from airfoil images. Code developed for this work is available in a public Github repository.
Authors: Yannick Molinghen, Rapha\"el Avalos, Mark Van Achter, Ann Now\'e, Tom Lenaerts
Abstract: We introduce the Laser Learning Environment (LLE), a collaborative multi-agent reinforcement learning environment in which coordination is central. In LLE, agents depend on each other to make progress (interdependence), must jointly take specific sequences of actions to succeed (perfect coordination), and accomplishing those joint actions does not yield any intermediate reward (zero-incentive dynamics). The challenge of such problems lies in the difficulty of escaping state space bottlenecks caused by interdependence steps since escaping those bottlenecks is not rewarded. We test multiple state-of-the-art value-based MARL algorithms against LLE and show that they consistently fail at the collaborative task because of their inability to escape state space bottlenecks, even though they successfully achieve perfect coordination. We show that Q-learning extensions such as prioritized experience replay and n-steps return hinder exploration in environments with zero-incentive dynamics, and find that intrinsic curiosity with random network distillation is not sufficient to escape those bottlenecks. We demonstrate the need for novel methods to solve this problem and the relevance of LLE as cooperative MARL benchmark.
Authors: Aniruddha Nrusimha, Mayank Mishra, Naigang Wang, Dan Alistarh, Rameswar Panda, Yoon Kim
Abstract: We consider the problem of accurate quantization for language models, where both the weights and activations are uniformly quantized to 4 bits per parameter, the lowest bitwidth format natively supported by GPU hardware. In this context, the key challenge is activation quantization: it is known that language models contain outlier channels whose values on average are orders of magnitude higher than than other channels, which prevents accurate low-bitwidth quantization with known techniques. We systematically study this phenomena and find that these outlier channels emerge early in training, and that they occur more frequently in layers with residual streams. We then propose a simple strategy which regularizes a layer's inputs via quantization-aware training (QAT) and its outputs via activation kurtosis regularization. We show that regularizing both the inputs and outputs is crucial for preventing a model's "migrating" the difficulty in input quantization to the weights, which makes post-training quantization (PTQ) of weights more difficult. When combined with weight PTQ, we show that our approach can obtain a W4A4 model that performs competitively to the standard-precision W16A16 baseline.
Authors: Andrew Lavin
Abstract: Since the breakthrough performance of AlexNet in 2012, convolutional neural networks (convnets) have grown into extremely powerful vision models. Deep learning researchers have used convnets to produce accurate results that were unachievable a decade ago. Yet computer scientists make computational efficiency their primary objective. Accuracy with exorbitant cost is not acceptable; an algorithm must also minimize its computational requirements. Confronted with the daunting computation that convnets use, deep learning researchers also became interested in efficiency. Researchers applied tremendous effort to find the convnet architectures that have the greatest efficiency. However, skepticism grew among researchers and engineers alike about the relevance of arithmetic complexity. Contrary to the prevailing view that latency and arithmetic complexity are irreconcilable, a simple formula relates both through computational efficiency. This insight enabled us to co-optimize the separate factors that determine latency. We observed that the degenerate conv2d layers that produce the best accuracy-complexity trade-off also have low operational intensity. Therefore, kernels that implement these layers use significant memory resources. We solved this optimization problem with block-fusion kernels that implement all layers of a residual block, thereby creating temporal locality, avoiding communication, and reducing workspace size. Our ConvFirst model with block-fusion kernels ran approximately four times as fast as the ConvNeXt baseline with PyTorch Inductor, at equal accuracy on the ImageNet-1K classification task. Our unified approach to convnet efficiency envisions a new era of models and kernels that achieve greater accuracy at lower cost.
Authors: Mehdi Jabbari Zideh, Mohammad Reza Khalghani, Sarika Khushalani Solanki
Abstract: Detection of cyber attacks in smart power distribution grids with unbalanced configurations poses challenges due to the inherent nonlinear nature of these uncertain and stochastic systems. It originates from the intermittent characteristics of the distributed energy resources (DERs) generation and load variations. Moreover, the unknown behavior of cyber attacks, especially false data injection attacks (FDIAs) in the distribution grids with complex temporal correlations and the limited amount of labeled data increases the vulnerability of the grids and imposes a high risk in the secure and reliable operation of the grids. To address these challenges, this paper proposes an unsupervised adversarial autoencoder (AAE) model to detect FDIAs in unbalanced power distribution grids integrated with DERs, i.e., PV systems and wind generation. The proposed method utilizes long short-term memory (LSTM) in the structure of the autoencoder to capture the temporal dependencies in the time-series measurements and leverages the power of generative adversarial networks (GANs) for better reconstruction of the input data. The advantage of the proposed data-driven model is that it can detect anomalous points for the system operation without reliance on abstract models or mathematical representations. To evaluate the efficacy of the approach, it is tested on IEEE 13-bus and 123-bus systems with historical meteorological data (wind speed, ambient temperature, and solar irradiance) as well as historical real-world load data under three types of data falsification functions. The comparison of the detection results of the proposed model with other unsupervised learning methods verifies its superior performance in detecting cyber attacks in unbalanced power distribution grids.
Authors: Jeffy Yu, Maximilian Huber, Kevin Tang
Abstract: This paper investigates the ethical implications of aligning Large Language Models (LLMs) with financial optimization, through the case study of GreedLlama, a model fine-tuned to prioritize economically beneficial outcomes. By comparing GreedLlama's performance in moral reasoning tasks to a base Llama2 model, our results highlight a concerning trend: GreedLlama demonstrates a marked preference for profit over ethical considerations, making morally appropriate decisions at significantly lower rates than the base model in scenarios of both low and high moral ambiguity. In low ambiguity situations, GreedLlama's ethical decisions decreased to 54.4%, compared to the base model's 86.9%, while in high ambiguity contexts, the rate was 47.4% against the base model's 65.1%. These findings emphasize the risks of single-dimensional value alignment in LLMs, underscoring the need for integrating broader ethical values into AI development to ensure decisions are not solely driven by financial incentives. The study calls for a balanced approach to LLM deployment, advocating for the incorporation of ethical considerations in models intended for business applications, particularly in light of the absence of regulatory oversight.
Authors: Jiawei Zhang, Chejian Xu, Yu Gai, Freddy Lecue, Dawn Song, Bo Li
Abstract: This paper introduces KnowHalu, a novel approach for detecting hallucinations in text generated by large language models (LLMs), utilizing step-wise reasoning, multi-formulation query, multi-form knowledge for factual checking, and fusion-based detection mechanism. As LLMs are increasingly applied across various domains, ensuring that their outputs are not hallucinated is critical. Recognizing the limitations of existing approaches that either rely on the self-consistency check of LLMs or perform post-hoc fact-checking without considering the complexity of queries or the form of knowledge, KnowHalu proposes a two-phase process for hallucination detection. In the first phase, it identifies non-fabrication hallucinations--responses that, while factually correct, are irrelevant or non-specific to the query. The second phase, multi-form based factual checking, contains five key steps: reasoning and query decomposition, knowledge retrieval, knowledge optimization, judgment generation, and judgment aggregation. Our extensive evaluations demonstrate that KnowHalu significantly outperforms SOTA baselines in detecting hallucinations across diverse tasks, e.g., improving by 15.65% in QA tasks and 5.50% in summarization tasks, highlighting its effectiveness and versatility in detecting hallucinations in LLM-generated content.
Authors: Jingyang Zhang, Jingwei Sun, Eric Yeats, Yang Ouyang, Martin Kuo, Jianyi Zhang, Hao Yang, Hai Li
Abstract: The problem of pre-training data detection for large language models (LLMs) has received growing attention due to its implications in critical issues like copyright violation and test data contamination. The current state-of-the-art approach, Min-K%, measures the raw token probability which we argue may not be the most informative signal. Instead, we propose Min-K%++ to normalize the token probability with statistics of the categorical distribution over the whole vocabulary, which accurately reflects the relative likelihood of the target token compared with other candidate tokens in the vocabulary. Theoretically, we back up our method by showing that the statistic it estimates is explicitly optimized during LLM training, thus serving as a reliable indicator for detecting training data. Empirically, on the WikiMIA benchmark, Min-K%++ outperforms the SOTA Min-K% by 6.2% to 10.5% in detection AUROC averaged over five models. On the more challenging MIMIR benchmark, Min-K%++ consistently improves upon Min-K% and performs on par with reference-based method, despite not requiring an extra reference model.
Authors: Siyi Wang, Zifan Wang, Xinlei Yi, Michael M. Zavlanos, Karl H. Johansson, Sandra Hirche
Abstract: Considering non-stationary environments in online optimization enables decision-maker to effectively adapt to changes and improve its performance over time. In such cases, it is favorable to adopt a strategy that minimizes the negative impact of change to avoid potentially risky situations. In this paper, we investigate risk-averse online optimization where the distribution of the random cost changes over time. We minimize risk-averse objective function using the Conditional Value at Risk (CVaR) as risk measure. Due to the difficulty in obtaining the exact CVaR gradient, we employ a zeroth-order optimization approach that queries the cost function values multiple times at each iteration and estimates the CVaR gradient using the sampled values. To facilitate the regret analysis, we use a variation metric based on Wasserstein distance to capture time-varying distributions. Given that the distribution variation is sub-linear in the total number of episodes, we show that our designed learning algorithm achieves sub-linear dynamic regret with high probability for both convex and strongly convex functions. Moreover, theoretical results suggest that increasing the number of samples leads to a reduction in the dynamic regret bounds until the sampling number reaches a specific limit. Finally, we provide numerical experiments of dynamic pricing in a parking lot to illustrate the efficacy of the designed algorithm.
Authors: Yannick Kirchhoff, Maximilian R. Rokuss, Saikat Roy, Balint Kovacs, Constantin Ulrich, Tassilo Wald, Maximilian Zenk, Philipp Vollmuth, Jens Kleesiek, Fabian Isensee, Klaus Maier-Hein
Abstract: Accurately segmenting thin tubular structures, such as vessels, nerves, roads or concrete cracks, is a crucial task in computer vision. Standard deep learning-based segmentation loss functions, such as Dice or Cross-Entropy, focus on volumetric overlap, often at the expense of preserving structural connectivity or topology. This can lead to segmentation errors that adversely affect downstream tasks, including flow calculation, navigation, and structural inspection. Although current topology-focused losses mark an improvement, they introduce significant computational and memory overheads. This is particularly relevant for 3D data, rendering these losses infeasible for larger volumes as well as increasingly important multi-class segmentation problems. To mitigate this, we propose a novel Skeleton Recall Loss, which effectively addresses these challenges by circumventing intensive GPU-based calculations with inexpensive CPU operations. It demonstrates overall superior performance to current state-of-the-art approaches on five public datasets for topology-preserving segmentation, while substantially reducing computational overheads by more than 90%. In doing so, we introduce the first multi-class capable loss function for thin structure segmentation, excelling in both efficiency and efficacy for topology-preservation.
Authors: Zhongming Yu, Genghan Zhang, Hanxian Huang, Xin Chen, Jishen Zhao
Abstract: In recent years, Graph Neural Networks (GNNs) have ignited a surge of innovation, significantly enhancing the processing of geometric data structures such as graphs, point clouds, and meshes. As the domain continues to evolve, a series of frameworks and libraries are being developed to push GNN efficiency to new heights. While graph-centric libraries have achieved success in the past, the advent of efficient tensor compilers has highlighted the urgent need for tensor-centric libraries. Yet, efficient tensor-centric frameworks for GNNs remain scarce due to unique challenges and limitations encountered when implementing segment reduction in GNN contexts. We introduce GeoT, a cutting-edge tensor-centric library designed specifically for GNNs via efficient segment reduction. GeoT debuts innovative parallel algorithms that not only introduce new design principles but also expand the available design space. Importantly, GeoT is engineered for straightforward fusion within a computation graph, ensuring compatibility with contemporary tensor-centric machine learning frameworks and compilers. Setting a new performance benchmark, GeoT marks a considerable advancement by showcasing an average operator speedup of 1.80x and an end-to-end speedup of 1.68x.
Authors: Amirhossein Abaskohi, Amirhossein Dabiriaghdam, Lele Wang, Giuseppe Carenini
Abstract: Memes, combining text and images, frequently use metaphors to convey persuasive messages, shaping public opinion. Motivated by this, our team engaged in SemEval-2024 Task 4, a hierarchical multi-label classification task designed to identify rhetorical and psychological persuasion techniques embedded within memes. To tackle this problem, we introduced a caption generation step to assess the modality gap and the impact of additional semantic information from images, which improved our result. Our best model utilizes GPT-4 generated captions alongside meme text to fine-tune RoBERTa as the text encoder and CLIP as the image encoder. It outperforms the baseline by a large margin in all 12 subtasks. In particular, it ranked in top-3 across all languages in Subtask 2a, and top-4 in Subtask 2b, demonstrating quantitatively strong performance. The improvement achieved by the introduced intermediate step is likely attributable to the metaphorical essence of images that challenges visual encoders. This highlights the potential for improving abstract visual semantics encoding.
Authors: Robert Kasumba, Guanghui Yu, Chien-Ju Ho, Sarah Keren, William Yeoh
Abstract: Goal recognition design aims to make limited modifications to decision-making environments with the goal of making it easier to infer the goals of agents acting within those environments. Although various research efforts have been made in goal recognition design, existing approaches are computationally demanding and often assume that agents are (near-)optimal in their decision-making. To address these limitations, we introduce a data-driven approach to goal recognition design that can account for agents with general behavioral models. Following existing literature, we use worst-case distinctiveness ($\textit{wcd}$) as a measure of the difficulty in inferring the goal of an agent in a decision-making environment. Our approach begins by training a machine learning model to predict the $\textit{wcd}$ for a given environment and the agent behavior model. We then propose a gradient-based optimization framework that accommodates various constraints to optimize decision-making environments for enhanced goal recognition. Through extensive simulations, we demonstrate that our approach outperforms existing methods in reducing $\textit{wcd}$ and enhancing runtime efficiency in conventional setups, and it also adapts to scenarios not previously covered in the literature, such as those involving flexible budget constraints, more complex environments, and suboptimal agent behavior. Moreover, we have conducted human-subject experiments which confirm that our method can create environments that facilitate efficient goal recognition from real-world human decision-makers.
Authors: Kaavya Chaparala, Guido Zarrella, Bruce Torres Fischer, Larry Kimura, Oiwi Parker Jones
Abstract: In this paper we address the challenge of improving Automatic Speech Recognition (ASR) for a low-resource language, Hawaiian, by incorporating large amounts of independent text data into an ASR foundation model, Whisper. To do this, we train an external language model (LM) on ~1.5M words of Hawaiian text. We then use the LM to rescore Whisper and compute word error rates (WERs) on a manually curated test set of labeled Hawaiian data. As a baseline, we use Whisper without an external LM. Experimental results reveal a small but significant improvement in WER when ASR outputs are rescored with a Hawaiian LM. The results support leveraging all available data in the development of ASR systems for underrepresented languages.
Authors: Fred Hohman, Chaoqun Wang, Jinmook Lee, Jochen G\"ortler, Dominik Moritz, Jeffrey P Bigham, Zhile Ren, Cecile Foret, Qi Shan, Xiaoyi Zhang
Abstract: On-device machine learning (ML) moves computation from the cloud to personal devices, protecting user privacy and enabling intelligent user experiences. However, fitting models on devices with limited resources presents a major technical challenge: practitioners need to optimize models and balance hardware metrics such as model size, latency, and power. To help practitioners create efficient ML models, we designed and developed Talaria: a model visualization and optimization system. Talaria enables practitioners to compile models to hardware, interactively visualize model statistics, and simulate optimizations to test the impact on inference metrics. Since its internal deployment two years ago, we have evaluated Talaria using three methodologies: (1) a log analysis highlighting its growth of 800+ practitioners submitting 3,600+ models; (2) a usability survey with 26 users assessing the utility of 20 Talaria features; and (3) a qualitative interview with the 7 most active users about their experience using Talaria.
Authors: Lucas E. Resck, Marcos M. Raimundo, Jorge Poco
Abstract: Saliency post-hoc explainability methods are important tools for understanding increasingly complex NLP models. While these methods can reflect the model's reasoning, they may not align with human intuition, making the explanations not plausible. In this work, we present a methodology for incorporating rationales, which are text annotations explaining human decisions, into text classification models. This incorporation enhances the plausibility of post-hoc explanations while preserving their faithfulness. Our approach is agnostic to model architectures and explainability methods. We introduce the rationales during model training by augmenting the standard cross-entropy loss with a novel loss function inspired by contrastive learning. By leveraging a multi-objective optimization algorithm, we explore the trade-off between the two loss functions and generate a Pareto-optimal frontier of models that balance performance and plausibility. Through extensive experiments involving diverse models, datasets, and explainability methods, we demonstrate that our approach significantly enhances the quality of model explanations without causing substantial (sometimes negligible) degradation in the original model's performance.
Authors: Weizhe Chen, Sven Koenig, Bistra Dilkina
Abstract: Cooperative multi-agent reinforcement learning (MARL) has been an increasingly important research topic in the last half-decade because of its great potential for real-world applications. Because of the curse of dimensionality, the popular "centralized training decentralized execution" framework requires a long time in training, yet still cannot converge efficiently. In this paper, we propose a general training framework, MARL-LNS, to algorithmically address these issues by training on alternating subsets of agents using existing deep MARL algorithms as low-level trainers, while not involving any additional parameters to be trained. Based on this framework, we provide three algorithm variants based on the framework: random large neighborhood search (RLNS), batch large neighborhood search (BLNS), and adaptive large neighborhood search (ALNS), which alternate the subsets of agents differently. We test our algorithms on both the StarCraft Multi-Agent Challenge and Google Research Football, showing that our algorithms can automatically reduce at least 10% of training time while reaching the same final skill level as the original algorithm.
Authors: Xinmeng Huang, Shuo Li, Mengxin Yu, Matteo Sesia, Hamed Hassani, Insup Lee, Osbert Bastani, Edgar Dobriban
Abstract: Language Models (LMs) have shown promising performance in natural language generation. However, as LMs often generate incorrect or hallucinated responses, it is crucial to correctly quantify their uncertainty in responding to given inputs. In addition to verbalized confidence elicited via prompting, many uncertainty measures ($e.g.$, semantic entropy and affinity-graph-based measures) have been proposed. However, these measures can differ greatly, and it is unclear how to compare them, partly because they take values over different ranges ($e.g.$, $[0,\infty)$ or $[0,1]$). In this work, we address this issue by developing a novel and practical framework, termed $Rank$-$Calibration$, to assess uncertainty and confidence measures for LMs. Our key tenet is that higher uncertainty (or lower confidence) should imply lower generation quality, on average. Rank-calibration quantifies deviations from this ideal relationship in a principled manner, without requiring ad hoc binary thresholding of the correctness score ($e.g.$, ROUGE or METEOR). The broad applicability and the granular interpretability of our methods are demonstrated empirically.
Authors: Haonan Zhang, Dongxia Wang, Zhu Sun, Yanhui Li, Youcheng Sun, Huizhi Liang, Wenhai Wang
Abstract: Recommender systems (RSs) are designed to provide personalized recommendations to users. Recently, knowledge graphs (KGs) have been widely introduced in RSs to improve recommendation accuracy. In this study, however, we demonstrate that RSs do not necessarily perform worse even if the KG is downgraded to the user-item interaction graph only (or removed). We propose an evaluation framework KG4RecEval to systematically evaluate how much a KG contributes to the recommendation accuracy of a KG-based RS, using our defined metric KGER (KG utilization efficiency in recommendation). We consider the scenarios where knowledge in a KG gets completely removed, randomly distorted and decreased, and also where recommendations are for cold-start users. Our extensive experiments on four commonly used datasets and a number of state-of-the-art KG-based RSs reveal that: to remove, randomly distort or decrease knowledge does not necessarily decrease recommendation accuracy, even for cold-start users. These findings inspire us to rethink how to better utilize knowledge from existing KGs, whereby we discuss and provide insights into what characteristics of datasets and KG-based RSs may help improve KG utilization efficiency.
Authors: Hanxian Huang, Jishen Zhao
Abstract: The increasing adoption of WebAssembly (Wasm) for performance-critical and security-sensitive tasks drives the demand for WebAssembly program comprehension and reverse engineering. Recent studies have introduced machine learning (ML)-based WebAssembly reverse engineering tools. Yet, the generalization of task-specific ML solutions remains challenging, because their effectiveness hinges on the availability of an ample supply of high-quality task-specific labeled data. Moreover, previous works overlook the high-level semantics present in source code and its documentation. Acknowledging the abundance of available source code with documentation, which can be compiled into WebAssembly, we propose to learn representations of them concurrently and harness their mutual relationships for effective WebAssembly reverse engineering. In this paper, we present WasmRev, the first multi-modal pre-trained language model for WebAssembly reverse engineering. WasmRev is pre-trained using self-supervised learning on a large-scale multi-modal corpus encompassing source code, code documentation and the compiled WebAssembly, without requiring labeled data. WasmRev incorporates three tailored multi-modal pre-training tasks to capture various characteristics of WebAssembly and cross-modal relationships. WasmRev is only trained once to produce general-purpose representations that can broadly support WebAssembly reverse engineering tasks through few-shot fine-tuning with much less labeled data, improving data efficiency. We fine-tune WasmRev onto three important reverse engineering tasks: type recovery, function purpose identification and WebAssembly summarization. Our results show that WasmRev pre-trained on the corpus of multi-modal samples establishes a robust foundation for these tasks, achieving high task accuracy and outperforming the state-of-the-art ML methods for WebAssembly reverse engineering.
Authors: W. S. H. M. W. Ahmad, M. F. A. Fauzi, M. K. Abdullahi, Jenny T. H. Lee, N. S. A. Basry, A Yahaya, A. M. Ismail, A. Adam, Elaine W. L. Chan, F. S. Abas
Abstract: Nasopharyngeal carcinoma (NPC) is one of the understudied yet deadliest cancers in South East Asia. In Malaysia, the prevalence is identified mainly in Sarawak, among the ethnic of Bidayuh. NPC is often late-diagnosed because it is asymptomatic at the early stage. There are several tissue representations from the nasopharynx biopsy, such as nasopharyngeal inflammation (NPI), lymphoid hyperplasia (LHP), nasopharyngeal carcinoma (NPC) and normal tissue. This paper is our first initiative to identify the difference between NPC, NPI and normal cases. Seven whole slide images (WSIs) with gigapixel resolutions from seven different patients and two hospitals were experimented with using two test setups, consisting of a different set of images. The tissue regions are patched into smaller blocks and classified using DenseNet architecture with 21 dense layers. Two tests are carried out, each for proof of concept (Test 1) and real-test scenario (Test 2). The accuracy achieved for NPC class is 94.8% for Test 1 and 67.0% for Test 2.
Authors: Detai Xin, Xu Tan, Kai Shen, Zeqian Ju, Dongchao Yang, Yuancheng Wang, Shinnosuke Takamichi, Hiroshi Saruwatari, Shujie Liu, Jinyu Li, Sheng Zhao
Abstract: We present RALL-E, a robust language modeling method for text-to-speech (TTS) synthesis. While previous work based on large language models (LLMs) shows impressive performance on zero-shot TTS, such methods often suffer from poor robustness, such as unstable prosody (weird pitch and rhythm/duration) and a high word error rate (WER), due to the autoregressive prediction style of language models. The core idea behind RALL-E is chain-of-thought (CoT) prompting, which decomposes the task into simpler steps to enhance the robustness of LLM-based TTS. To accomplish this idea, RALL-E first predicts prosody features (pitch and duration) of the input text and uses them as intermediate conditions to predict speech tokens in a CoT style. Second, RALL-E utilizes the predicted duration prompt to guide the computing of self-attention weights in Transformer to enforce the model to focus on the corresponding phonemes and prosody features when predicting speech tokens. Results of comprehensive objective and subjective evaluations demonstrate that, compared to a powerful baseline method VALL-E, RALL-E significantly improves the WER of zero-shot TTS from $6.3\%$ (without reranking) and $2.1\%$ (with reranking) to $2.8\%$ and $1.0\%$, respectively. Furthermore, we demonstrate that RALL-E correctly synthesizes sentences that are hard for VALL-E and reduces the error rate from $68\%$ to $4\%$.
Authors: Xu Wang, Tian Ye, Rajgopal Kannan, Viktor Prasanna
Abstract: Deep Learning (DL) Models for Synthetic Aperture Radar (SAR) Automatic Target Recognition (ATR), while delivering improved performance, have been shown to be quite vulnerable to adversarial attacks. Existing works improve robustness by training models on adversarial samples. However, by focusing mostly on attacks that manipulate images randomly, they neglect the real-world feasibility of such attacks. In this paper, we propose FACTUAL, a novel Contrastive Learning framework for Adversarial Training and robust SAR classification. FACTUAL consists of two components: (1) Differing from existing works, a novel perturbation scheme that incorporates realistic physical adversarial attacks (such as OTSA) to build a supervised adversarial pre-training network. This network utilizes class labels for clustering clean and perturbed images together into a more informative feature space. (2) A linear classifier cascaded after the encoder to use the computed representations to predict the target labels. By pre-training and fine-tuning our model on both clean and adversarial samples, we show that our model achieves high prediction accuracy on both cases. Our model achieves 99.7% accuracy on clean samples, and 89.6% on perturbed samples, both outperforming previous state-of-the-art methods.
Authors: Xingran Chen, Navid NaderiAlizadeh, Alejandro Ribeiro, Shirin Saeedi Bidokhti
Abstract: We address the challenge of sampling and remote estimation for autoregressive Markovian processes in a multi-hop wireless network with statistically-identical agents. Agents cache the most recent samples from others and communicate over wireless collision channels governed by an underlying graph topology. Our goal is to minimize time-average estimation error and/or age of information with decentralized scalable sampling and transmission policies, considering both oblivious (where decision-making is independent of the physical processes) and non-oblivious policies (where decision-making depends on physical processes). We prove that in oblivious policies, minimizing estimation error is equivalent to minimizing the age of information. The complexity of the problem, especially the multi-dimensional action spaces and arbitrary network topologies, makes theoretical methods for finding optimal transmission policies intractable. We optimize the policies using a graphical multi-agent reinforcement learning framework, where each agent employs a permutation-equivariant graph neural network architecture. Theoretically, we prove that our proposed framework exhibits desirable transferability properties, allowing transmission policies trained on small- or moderate-size networks to be executed effectively on large-scale topologies. Numerical experiments demonstrate that (i) Our proposed framework outperforms state-of-the-art baselines; (ii) The trained policies are transferable to larger networks, and their performance gains increase with the number of agents; (iii) The training procedure withstands non-stationarity even if we utilize independent learning techniques; and, (iv) Recurrence is pivotal in both independent learning and centralized training and decentralized execution, and improves the resilience to non-stationarity in independent learning.
Authors: Rukshani Somarathna, Gelareh Mohammadi
Abstract: Emotion understanding is a complex process that involves multiple components. The ability to recognise emotions not only leads to new context awareness methods but also enhances system interaction's effectiveness by perceiving and expressing emotions. Despite the attention to discrete and dimensional models, neuroscientific evidence supports those emotions as being complex and multi-faceted. One framework that resonated well with such findings is the Component Process Model (CPM), a theory that considers the complexity of emotions with five interconnected components: appraisal, expression, motivation, physiology and feeling. However, the relationship between CPM and discrete emotions has not yet been fully explored. Therefore, to better understand emotions underlying processes, we operationalised a data-driven approach using interactive Virtual Reality (VR) games and collected multimodal measures (self-reports, physiological and facial signals) from 39 participants. We used Machine Learning (ML) methods to identify the unique contributions of each component to emotion differentiation. Our results showed the role of different components in emotion differentiation, with the model including all components demonstrating the most significant contribution. Moreover, we found that at least five dimensions are needed to represent the variation of emotions in our dataset. These findings also have implications for using VR environments in emotion research and highlight the role of physiological signals in emotion recognition within such environments.
Authors: Akira Okazaki, Shuichi Kawano
Abstract: Multi-task learning (MTL) aims to improve estimation and prediction performance by sharing common information among related tasks. One natural assumption in MTL is that tasks are classified into clusters based on their characteristics. However, existing MTL methods based on this assumption often ignore outlier tasks that have large task-specific components or no relation to other tasks. To address this issue, we propose a novel MTL method called Multi-Task Learning via Robust Regularized Clustering (MTLRRC). MTLRRC incorporates robust regularization terms inspired by robust convex clustering, which is further extended to handle non-convex and group-sparse penalties. The extension allows MTLRRC to simultaneously perform robust task clustering and outlier task detection. The connection between the extended robust clustering and the multivariate M-estimator is also established. This provides an interpretation of the robustness of MTLRRC against outlier tasks. An efficient algorithm based on a modified alternating direction method of multipliers is developed for the estimation of the parameters. The effectiveness of MTLRRC is demonstrated through simulation studies and application to real data.
Authors: Yin Li, Qi Chen, Kai Wang, Meige Li, Liping Si, Yingwei Guo, Yu Xiong, Qixing Wang, Yang Qin, Ling Xu, Patrick van der Smagt, Jun Tang, Nutan Chen
Abstract: Multi-modality magnetic resonance imaging data with various sequences facilitate the early diagnosis, tumor segmentation, and disease staging in the management of nasopharyngeal carcinoma (NPC). The lack of publicly available, comprehensive datasets limits advancements in diagnosis, treatment planning, and the development of machine learning algorithms for NPC. Addressing this critical need, we introduce the first comprehensive NPC MRI dataset, encompassing MR axial imaging of 277 primary NPC patients. This dataset includes T1-weighted, T2-weighted, and contrast-enhanced T1-weighted sequences, totaling 831 scans. In addition to the corresponding clinical data, manually annotated and labeled segmentations by experienced radiologists offer high-quality data resources from untreated primary NPC.
Authors: MengMeng Han, Tennessee Leeuwenburg, Brad Murphy
Abstract: Site-specific weather forecasts are essential to accurate prediction of power demand and are consequently of great interest to energy operators. However, weather forecasts from current numerical weather prediction (NWP) models lack the fine-scale detail to capture all important characteristics of localised real-world sites. Instead they provide weather information representing a rectangular gridbox (usually kilometres in size). Even after post-processing and bias correction, area-averaged information is usually not optimal for specific sites. Prior work on site optimised forecasts has focused on linear methods, weighted consensus averaging, time-series methods, and others. Recent developments in machine learning (ML) have prompted increasing interest in applying ML as a novel approach towards this problem. In this study, we investigate the feasibility of optimising forecasts at sites by adopting the popular machine learning model gradient boosting decision tree, supported by the Python version of the XGBoost package. Regression trees have been trained with historical NWP and site observations as training data, aimed at predicting temperature and dew point at multiple site locations across Australia. We developed a working ML framework, named 'Multi-SiteBoost' and initial testing results show a significant improvement compared with gridded values from bias-corrected NWP models. The improvement from XGBoost is found to be comparable with non-ML methods reported in literature. With the insights provided by SHapley Additive exPlanations (SHAP), this study also tests various approaches to understand the ML predictions and increase the reliability of the forecasts generated by ML.
Authors: Rachmad Vidya Wicaksana Putra, Alberto Marchisio, Fakhreddine Zayer, Jorge Dias, Muhammad Shafique
Abstract: Robotic technologies have been an indispensable part for improving human productivity since they have been helping humans in completing diverse, complex, and intensive tasks in a fast yet accurate and efficient way. Therefore, robotic technologies have been deployed in a wide range of applications, ranging from personal to industrial use-cases. However, current robotic technologies and their computing paradigm still lack embodied intelligence to efficiently interact with operational environments, respond with correct/expected actions, and adapt to changes in the environments. Toward this, recent advances in neuromorphic computing with Spiking Neural Networks (SNN) have demonstrated the potential to enable the embodied intelligence for robotics through bio-plausible computing paradigm that mimics how the biological brain works, known as "neuromorphic artificial intelligence (AI)". However, the field of neuromorphic AI-based robotics is still at an early stage, therefore its development and deployment for solving real-world problems expose new challenges in different design aspects, such as accuracy, adaptability, efficiency, reliability, and security. To address these challenges, this paper will discuss how we can enable embodied neuromorphic AI for robotic systems through our perspectives: (P1) Embodied intelligence based on effective learning rule, training mechanism, and adaptability; (P2) Cross-layer optimizations for energy-efficient neuromorphic computing; (P3) Representative and fair benchmarks; (P4) Low-cost reliability and safety enhancements; (P5) Security and privacy for neuromorphic computing; and (P6) A synergistic development for energy-efficient and robust neuromorphic-based robotics. Furthermore, this paper identifies research challenges and opportunities, as well as elaborates our vision for future research development toward embodied neuromorphic AI for robotics.
Authors: Bin Gao, Yan Yang, Ya-xiang Yuan
Abstract: Bilevel optimization, with broad applications in machine learning, has an intricate hierarchical structure. Gradient-based methods have emerged as a common approach to large-scale bilevel problems. However, the computation of the hyper-gradient, which involves a Hessian inverse vector product, confines the efficiency and is regarded as a bottleneck. To circumvent the inverse, we construct a sequence of low-dimensional approximate Krylov subspaces with the aid of the Lanczos process. As a result, the constructed subspace is able to dynamically and incrementally approximate the Hessian inverse vector product with less effort and thus leads to a favorable estimate of the hyper-gradient. Moreover, we propose a~provable subspace-based framework for bilevel problems where one central step is to solve a small-size tridiagonal linear system. To the best of our knowledge, this is the first time that subspace techniques are incorporated into bilevel optimization. This successful trial not only enjoys $\mathcal{O}(\epsilon^{-1})$ convergence rate but also demonstrates efficiency in a synthetic problem and two deep learning tasks.
Authors: Lei Zhang, Yuhang Zhou, Yi Yang, Xinbo Gao
Abstract: Despite providing high-performance solutions for computer vision tasks, the deep neural network (DNN) model has been proved to be extremely vulnerable to adversarial attacks. Current defense mainly focuses on the known attacks, but the adversarial robustness to the unknown attacks is seriously overlooked. Besides, commonly used adaptive learning and fine-tuning technique is unsuitable for adversarial defense since it is essentially a zero-shot problem when deployed. Thus, to tackle this challenge, we propose an attack-agnostic defense method named Meta Invariance Defense (MID). Specifically, various combinations of adversarial attacks are randomly sampled from a manually constructed Attacker Pool to constitute different defense tasks against unknown attacks, in which a student encoder is supervised by multi-consistency distillation to learn the attack-invariant features via a meta principle. The proposed MID has two merits: 1) Full distillation from pixel-, feature- and prediction-level between benign and adversarial samples facilitates the discovery of attack-invariance. 2) The model simultaneously achieves robustness to the imperceptible adversarial perturbations in high-level image classification and attack-suppression in low-level robust image regeneration. Theoretical and empirical studies on numerous benchmarks such as ImageNet verify the generalizable robustness and superiority of MID under various attacks.
Authors: Jiacai Liu, Wenye Li, Ke Wei
Abstract: Projected policy gradient under the simplex parameterization, policy gradient and natural policy gradient under the softmax parameterization, are fundamental algorithms in reinforcement learning. There have been a flurry of recent activities in studying these algorithms from the theoretical aspect. Despite this, their convergence behavior is still not fully understood, even given the access to exact policy evaluations. In this paper, we focus on the discounted MDP setting and conduct a systematic study of the aforementioned policy optimization methods. Several novel results are presented, including 1) global linear convergence of projected policy gradient for any constant step size, 2) sublinear convergence of softmax policy gradient for any constant step size, 3) global linear convergence of softmax natural policy gradient for any constant step size, 4) global linear convergence of entropy regularized softmax policy gradient for a wider range of constant step sizes than existing result, 5) tight local linear convergence rate of entropy regularized natural policy gradient, and 6) a new and concise local quadratic convergence rate of soft policy iteration without the assumption on the stationary distribution under the optimal policy. New and elementary analysis techniques have been developed to establish these results.
Authors: Kaichen Huang, Minghao Shao, Shenghua Wan, Hai-Hang Sun, Shuai Feng, Le Gan, De-Chuan Zhan
Abstract: In many real-world visual Imitation Learning (IL) scenarios, there is a misalignment between the agent's and the expert's perspectives, which might lead to the failure of imitation. Previous methods have generally solved this problem by domain alignment, which incurs extra computation and storage costs, and these methods fail to handle the \textit{hard cases} where the viewpoint gap is too large. To alleviate the above problems, we introduce active sensoring in the visual IL setting and propose a model-based SENSory imitatOR (SENSOR) to automatically change the agent's perspective to match the expert's. SENSOR jointly learns a world model to capture the dynamics of latent states, a sensor policy to control the camera, and a motor policy to control the agent. Experiments on visual locomotion tasks show that SENSOR can efficiently simulate the expert's perspective and strategy, and outperforms most baseline methods.
Authors: Sean McLeish, Avi Schwarzschild, Tom Goldstein
Abstract: We evaluate ChatGPT's ability to solve algorithm problems from the CLRS benchmark suite that is designed for GNNs. The benchmark requires the use of a specified classical algorithm to solve a given problem. We find that ChatGPT outperforms specialist GNN models, using Python to successfully solve these problems. This raises new points in the discussion about learning algorithms with neural networks.
Authors: Chuyu Zhang, Hui Ren, Xuming He
Abstract: Deep clustering, which learns representation and semantic clustering without labels information, poses a great challenge for deep learning-based approaches. Despite significant progress in recent years, most existing methods focus on uniformly distributed datasets, significantly limiting the practical applicability of their methods. In this paper, we propose a more practical problem setting named deep imbalanced clustering, where the underlying classes exhibit an imbalance distribution. To address this challenge, we introduce a novel optimal transport-based pseudo-label learning framework. Our framework formulates pseudo-label generation as a Semantic-regularized Progressive Partial Optimal Transport (SP$^2$OT) problem, which progressively transports each sample to imbalanced clusters under several prior distribution and semantic relation constraints, thus generating high-quality and imbalance-aware pseudo-labels. To solve SP$^2$OT, we develop a Majorization-Minimization-based optimization algorithm. To be more precise, we employ the strategy of majorization to reformulate the SP$^2$OT problem into a Progressive Partial Optimal Transport problem, which can be transformed into an unbalanced optimal transport problem with augmented constraints and can be solved efficiently by a fast matrix scaling algorithm. Experiments on various datasets, including a human-curated long-tailed CIFAR100, challenging ImageNet-R, and large-scale subsets of fine-grained iNaturalist2018 datasets, demonstrate the superiority of our method.
Authors: Ingo Steinwart
Abstract: In this paper we investigate the conditional distributions of two Banach space valued, jointly Gaussian random variables. These conditional distributions are again Gaussian and their means and covariances are determined by a general approximation scheme based upon a martingale idea. We then apply our general results to the case of Gaussian processes with continuous paths conditioned to partial observations of their paths.
Authors: Farnaz Kohankhaki, Jacob-Junqi Tian, David Emerson, Laleh Seyyed-Kalantari, Faiza Khan Khattak
Abstract: Large language models (LLMs), trained on vast datasets, can carry biases that manifest in various forms, from overt discrimination to implicit stereotypes. One facet of bias is performance disparities in LLMs, often harming underprivileged groups, such as racial minorities. A common approach to quantifying bias is to use template-based bias probes, which explicitly state group membership (e.g. White) and evaluate if the outcome of a task, sentiment analysis for instance, is invariant to the change of group membership (e.g. change White race to Black). This approach is widely used in bias quantification. However, in this work, we find evidence of an unexpectedly overlooked consequence of using template-based probes for LLM bias quantification. We find that in doing so, text examples associated with White ethnicities appear to be classified as exhibiting negative sentiment at elevated rates. We hypothesize that the scenario arises artificially through a mismatch between the pre-training text of LLMs and the templates used to measure bias through reporting bias, unstated norms that imply group membership without explicit statement. Our finding highlights the potential misleading impact of varying group membership through explicit mention in bias quantification
Authors: Iqra Bano, Rachmad Vidya Wicaksana Putra, Alberto Marchisio, Muhammad Shafique
Abstract: Autonomous Driving (AD) systems are considered as the future of human mobility and transportation. Solving computer vision tasks such as image classification and object detection/segmentation, with high accuracy and low power/energy consumption, is highly needed to realize AD systems in real life. These requirements can potentially be satisfied by Spiking Neural Networks (SNNs). However, the state-of-the-art works in SNN-based AD systems still focus on proposing network models that can achieve high accuracy, and they have not systematically studied the roles of SNN parameters when used for learning event-based automotive data. Therefore, we still lack understanding of how to effectively develop SNN models for AD systems. Toward this, we propose a novel methodology to systematically study and analyze the impact of SNN parameters considering event-based automotive data, then leverage this analysis for enhancing SNN developments. To do this, we first explore different settings of SNN parameters that directly affect the learning mechanism (i.e., batch size, learning rate, neuron threshold potential, and weight decay), then analyze the accuracy results. Afterward, we propose techniques that jointly improve SNN accuracy and reduce training time. Experimental results show that our methodology can improve the SNN models for AD systems than the state-of-the-art, as it achieves higher accuracy (i.e., 86%) for the NCARS dataset, and it can also achieve iso-accuracy (i.e., ~85% with standard deviation less than 0.5%) while speeding up the training time by 1.9x. In this manner, our research work provides a set of guidelines for SNN parameter enhancements, thereby enabling the practical developments of SNN-based AD systems.
Authors: Susanne Dandl, Kristin Blesch, Timo Freiesleben, Gunnar K\"onig, Jan Kapar, Bernd Bischl, Marvin Wright
Abstract: Counterfactual explanations elucidate algorithmic decisions by pointing to scenarios that would have led to an alternative, desired outcome. Giving insight into the model's behavior, they hint users towards possible actions and give grounds for contesting decisions. As a crucial factor in achieving these goals, counterfactuals must be plausible, i.e., describing realistic alternative scenarios within the data manifold. This paper leverages a recently developed generative modeling technique -- adversarial random forests (ARFs) -- to efficiently generate plausible counterfactuals in a model-agnostic way. ARFs can serve as a plausibility measure or directly generate counterfactual explanations. Our ARF-based approach surpasses the limitations of existing methods that aim to generate plausible counterfactual explanations: It is easy to train and computationally highly efficient, handles continuous and categorical data naturally, and allows integrating additional desiderata such as sparsity in a straightforward manner.
Authors: Azmine Toushik Wasi, Taki Hasan Rafi, Raima Islam, Dong-Kyu Chae
Abstract: Knowledge Graphs (KGs) have proven essential in information processing and reasoning applications because they link related entities and give context-rich information, supporting efficient information retrieval and knowledge discovery; presenting information flow in a very effective manner. Despite being widely used globally, Bangla is relatively underrepresented in KGs due to a lack of comprehensive datasets, encoders, NER (named entity recognition) models, POS (part-of-speech) taggers, and lemmatizers, hindering efficient information processing and reasoning applications in the language. Addressing the KG scarcity in Bengali, we propose BanglaAutoKG, a pioneering framework that is able to automatically construct Bengali KGs from any Bangla text. We utilize multilingual LLMs to understand various languages and correlate entities and relations universally. By employing a translation dictionary to identify English equivalents and extracting word features from pre-trained BERT models, we construct the foundational KG. To reduce noise and align word embeddings with our goal, we employ graph-based polynomial filters. Lastly, we implement a GNN-based semantic filter, which elevates contextual understanding and trims unnecessary edges, culminating in the formation of the definitive KG. Empirical findings and case studies demonstrate the universal effectiveness of our model, capable of autonomously constructing semantically enriched KGs from any text.
Authors: Jiawei Guo, Ziming Li, Xueling Liu, Kaijing Ma, Tianyu Zheng, Zhouliang Yu, Ding Pan, Yizhi LI, Ruibo Liu, Yue Wang, Shuyue Guo, Xingwei Qu, Xiang Yue, Ge Zhang, Wenhu Chen, Jie Fu
Abstract: Large Language Models (LLMs) for code are rapidly evolving, with code editing emerging as a critical capability. We introduce CodeEditorBench, an evaluation framework designed to rigorously assess the performance of LLMs in code editing tasks, including debugging, translating, polishing, and requirement switching. Unlike existing benchmarks focusing solely on code generation, CodeEditorBench emphasizes real-world scenarios and practical aspects of software development. We curate diverse coding challenges and scenarios from five sources, covering various programming languages, complexity levels, and editing tasks. Evaluation of 19 LLMs reveals that closed-source models (particularly Gemini-Ultra and GPT-4), outperform open-source models in CodeEditorBench, highlighting differences in model performance based on problem types and prompt sensitivities. CodeEditorBench aims to catalyze advancements in LLMs by providing a robust platform for assessing code editing capabilities. We will release all prompts and datasets to enable the community to expand the dataset and benchmark emerging LLMs. By introducing CodeEditorBench, we contribute to the advancement of LLMs in code editing and provide a valuable resource for researchers and practitioners.
Authors: Harmon Bhasin, Timothy Ossowski, Yiqiao Zhong, Junjie Hu
Abstract: Large language models (LLM) have recently shown the extraordinary ability to perform unseen tasks based on few-shot examples provided as text, also known as in-context learning (ICL). While recent works have attempted to understand the mechanisms driving ICL, few have explored training strategies that incentivize these models to generalize to multiple tasks. Multi-task learning (MTL) for generalist models is a promising direction that offers transfer learning potential, enabling large parameterized models to be trained from simpler, related tasks. In this work, we investigate the combination of MTL with ICL to build models that efficiently learn tasks while being robust to out-of-distribution examples. We propose several effective curriculum learning strategies that allow ICL models to achieve higher data efficiency and more stable convergence. Our experiments reveal that ICL models can effectively learn difficult tasks by training on progressively harder tasks while mixing in prior tasks, denoted as mixed curriculum in this work. Our code and models are available at https://github.com/harmonbhasin/curriculum_learning_icl .
URLs: https://github.com/harmonbhasin/curriculum_learning_icl
Authors: Hasib-Al Rashid, Argho Sarkar, Aryya Gangopadhyay, Maryam Rahnemoonfar, Tinoosh Mohsenin
Abstract: Traditional machine learning models often require powerful hardware, making them unsuitable for deployment on resource-limited devices. Tiny Machine Learning (tinyML) has emerged as a promising approach for running machine learning models on these devices, but integrating multiple data modalities into tinyML models still remains a challenge due to increased complexity, latency, and power consumption. This paper proposes TinyVQA, a novel multimodal deep neural network for visual question answering tasks that can be deployed on resource-constrained tinyML hardware. TinyVQA leverages a supervised attention-based model to learn how to answer questions about images using both vision and language modalities. Distilled knowledge from the supervised attention-based VQA model trains the memory aware compact TinyVQA model and low bit-width quantization technique is employed to further compress the model for deployment on tinyML devices. The TinyVQA model was evaluated on the FloodNet dataset, which is used for post-disaster damage assessment. The compact model achieved an accuracy of 79.5%, demonstrating the effectiveness of TinyVQA for real-world applications. Additionally, the model was deployed on a Crazyflie 2.0 drone, equipped with an AI deck and GAP8 microprocessor. The TinyVQA model achieved low latencies of 56 ms and consumes 693 mW power while deployed on the tiny drone, showcasing its suitability for resource-constrained embedded systems.
Authors: Zhengxuan Wu, Aryaman Arora, Zheng Wang, Atticus Geiger, Dan Jurafsky, Christopher D. Manning, Christopher Potts
Abstract: Parameter-efficient fine-tuning (PEFT) methods seek to adapt large models via updates to a small number of weights. However, much prior interpretability work has shown that representations encode rich semantic information, suggesting that editing representations might be a more powerful alternative. Here, we pursue this hypothesis by developing a family of $\textbf{Representation Finetuning (ReFT)}$ methods. ReFT methods operate on a frozen base model and learn task-specific interventions on hidden representations. We define a strong instance of the ReFT family, Low-rank Linear Subspace ReFT (LoReFT). LoReFT is a drop-in replacement for existing PEFTs and learns interventions that are 10x-50x more parameter-efficient than prior state-of-the-art PEFTs. We showcase LoReFT on eight commonsense reasoning tasks, four arithmetic reasoning tasks, Alpaca-Eval v1.0, and GLUE. In all these evaluations, LoReFT delivers the best balance of efficiency and performance, and almost always outperforms state-of-the-art PEFTs. We release a generic ReFT training library publicly at https://github.com/stanfordnlp/pyreft.
Authors: Brian Lester, Jaehoon Lee, Alex Alemi, Jeffrey Pennington, Adam Roberts, Jascha Sohl-Dickstein, Noah Constant
Abstract: In this paper, we explore the idea of training large language models (LLMs) over highly compressed text. While standard subword tokenizers compress text by a small factor, neural text compressors can achieve much higher rates of compression. If it were possible to train LLMs directly over neurally compressed text, this would confer advantages in training and serving efficiency, as well as easier handling of long text spans. The main obstacle to this goal is that strong compression tends to produce opaque outputs that are not well-suited for learning. In particular, we find that text na\"ively compressed via Arithmetic Coding is not readily learnable by LLMs. To overcome this, we propose Equal-Info Windows, a novel compression technique whereby text is segmented into blocks that each compress to the same bit length. Using this method, we demonstrate effective learning over neurally compressed text that improves with scale, and outperforms byte-level baselines by a wide margin on perplexity and inference speed benchmarks. While our method delivers worse perplexity than subword tokenizers for models trained with the same parameter count, it has the benefit of shorter sequence lengths. Shorter sequence lengths require fewer autoregressive generation steps, and reduce latency. Finally, we provide extensive analysis of the properties that contribute to learnability, and offer concrete suggestions for how to further improve the performance of high-compression tokenizers.
Authors: Ziyao Zeng, Daniel Wang, Fengyu Yang, Hyoungseob Park, Yangchao Wu, Stefano Soatto, Byung-Woo Hong, Dong Lao, Alex Wong
Abstract: Three-dimensional (3D) reconstruction from a single image is an ill-posed problem with inherent ambiguities, i.e. scale. Predicting a 3D scene from text description(s) is similarly ill-posed, i.e. spatial arrangements of objects described. We investigate the question of whether two inherently ambiguous modalities can be used in conjunction to produce metric-scaled reconstructions. To test this, we focus on monocular depth estimation, the problem of predicting a dense depth map from a single image, but with an additional text caption describing the scene. To this end, we begin by encoding the text caption as a mean and standard deviation; using a variational framework, we learn the distribution of the plausible metric reconstructions of 3D scenes corresponding to the text captions as a prior. To "select" a specific reconstruction or depth map, we encode the given image through a conditional sampler that samples from the latent space of the variational text encoder, which is then decoded to the output depth map. Our approach is trained alternatingly between the text and image branches: in one optimization step, we predict the mean and standard deviation from the text description and sample from a standard Gaussian, and in the other, we sample using a (image) conditional sampler. Once trained, we directly predict depth from the encoded text using the conditional sampler. We demonstrate our approach on indoor (NYUv2) and outdoor (KITTI) scenarios, where we show that language can consistently improve performance in both.
Authors: Darioush Kevian, Usman Syed, Xingang Guo, Aaron Havens, Geir Dullerud, Peter Seiler, Lianhui Qin, Bin Hu
Abstract: In this paper, we explore the capabilities of state-of-the-art large language models (LLMs) such as GPT-4, Claude 3 Opus, and Gemini 1.0 Ultra in solving undergraduate-level control problems. Controls provides an interesting case study for LLM reasoning due to its combination of mathematical theory and engineering design. We introduce ControlBench, a benchmark dataset tailored to reflect the breadth, depth, and complexity of classical control design. We use this dataset to study and evaluate the problem-solving abilities of these LLMs in the context of control engineering. We present evaluations conducted by a panel of human experts, providing insights into the accuracy, reasoning, and explanatory prowess of LLMs in control engineering. Our analysis reveals the strengths and limitations of each LLM in the context of classical control, and our results imply that Claude 3 Opus has become the state-of-the-art LLM for solving undergraduate control problems. Our study serves as an initial step towards the broader goal of employing artificial general intelligence in control engineering.
Authors: Christoph Reinders, Bodo Rosenhahn
Abstract: We present Neural Random Forest Imitation - a novel approach for transforming random forests into neural networks. Existing methods propose a direct mapping and produce very inefficient architectures. In this work, we introduce an imitation learning approach by generating training data from a random forest and learning a neural network that imitates its behavior. This implicit transformation creates very efficient neural networks that learn the decision boundaries of a random forest. The generated model is differentiable, can be used as a warm start for fine-tuning, and enables end-to-end optimization. Experiments on several real-world benchmark datasets demonstrate superior performance, especially when training with very few training examples. Compared to state-of-the-art methods, we significantly reduce the number of network parameters while achieving the same or even improved accuracy due to better generalization.
Authors: Lingbing Guo, Zhuo Chen, Jiaoyan Chen, Yichi Zhang, Zequn Sun, Zhongpo Bo, Yin Fang, Xiaoze Liu, Huajun Chen, Wen Zhang
Abstract: Graph neural network (GNN)-based methods have demonstrated remarkable performance in various knowledge graph (KG) tasks. However, most existing approaches rely on observing all entities during training, posing a challenge in real-world knowledge graphs where new entities emerge frequently. To address this limitation, we introduce Decentralized Attention Network (DAN). DAN leverages neighbor context as the query vector to score the neighbors of an entity, thereby distributing the entity semantics only among its neighbor embeddings. To effectively train a DAN, we introduce self-distillation, a technique that guides the network in generating desired representations. Theoretical analysis validates the effectiveness of our approach. We implement an end-to-end framework and conduct extensive experiments to evaluate our method, showcasing competitive performance on conventional entity alignment and entity prediction tasks. Furthermore, our method significantly outperforms existing methods in open-world settings.
Authors: Moshe Shenfeld, Katrina Ligett
Abstract: Repeated use of a data sample via adaptively chosen queries can rapidly lead to overfitting, wherein the empirical evaluation of queries on the sample significantly deviates from their mean with respect to the underlying data distribution. It turns out that simple noise addition algorithms suffice to prevent this issue, and differential privacy-based analysis of these algorithms shows that they can handle an asymptotically optimal number of queries. However, differential privacy's worst-case nature entails scaling such noise to the range of the queries even for highly-concentrated queries, or introducing more complex algorithms. In this paper, we prove that straightforward noise-addition algorithms already provide variance-dependent guarantees that also extend to unbounded queries. This improvement stems from a novel characterization that illuminates the core problem of adaptive data analysis. We show that the harm of adaptivity results from the covariance between the new query and a Bayes factor-based measure of how much information about the data sample was encoded in the responses given to past queries. We then leverage this characterization to introduce a new data-dependent stability notion that can bound this covariance.
Authors: Ping Chang, Huayu Li, Stuart F. Quan, Shuyang Lu, Shu-Fen Wung, Janet Roveda, Ao Li
Abstract: Background and Objective: Vital sign monitoring in the Intensive Care Unit (ICU) is crucial for enabling prompt interventions for patients. This underscores the need for an accurate predictive system. Therefore, this study proposes a novel deep learning approach for forecasting Heart Rate (HR), Systolic Blood Pressure (SBP), and Diastolic Blood Pressure (DBP) in the ICU. Methods: We extracted $24,886$ ICU stays from the MIMIC-III database which contains data from over $46$ thousand patients, to train and test the model. The model proposed in this study, Transformer-based Diffusion Probabilistic Model for Sparse Time Series Forecasting (TDSTF), merges Transformer and diffusion models to forecast vital signs. The TDSTF model showed state-of-the-art performance in predicting vital signs in the ICU, outperforming other models' ability to predict distributions of vital signs and being more computationally efficient. The code is available at https://github.com/PingChang818/TDSTF. Results: The results of the study showed that TDSTF achieved a Standardized Average Continuous Ranked Probability Score (SACRPS) of $0.4438$ and a Mean Squared Error (MSE) of $0.4168$, an improvement of $18.9\%$ and $34.3\%$ over the best baseline model, respectively. The inference speed of TDSTF is more than $17$ times faster than the best baseline model. Conclusion: TDSTF is an effective and efficient solution for forecasting vital signs in the ICU, and it shows a significant improvement compared to other models in the field.
Authors: Saviz Mowlavi, Mouhacine Benosman
Abstract: In systems governed by nonlinear partial differential equations such as fluid flows, the design of state estimators such as Kalman filters relies on a reduced-order model (ROM) that projects the original high-dimensional dynamics onto a computationally tractable low-dimensional space. However, ROMs are prone to large errors, which negatively affects the performance of the estimator. Here, we introduce the reinforcement learning reduced-order estimator (RL-ROE), a ROM-based estimator in which the correction term that takes in the measurements is given by a nonlinear policy trained through reinforcement learning. The nonlinearity of the policy enables the RL-ROE to compensate efficiently for errors of the ROM, while still taking advantage of the imperfect knowledge of the dynamics. Using examples involving the Burgers and Navier-Stokes equations, we show that in the limit of very few sensors, the trained RL-ROE outperforms a Kalman filter designed using the same ROM. Moreover, it yields accurate high-dimensional state estimates for trajectories corresponding to various physical parameter values, without direct knowledge of the latter.
Authors: Krzysztof Marcin Choromanski, Shanda Li, Valerii Likhosherstov, Kumar Avinava Dubey, Shengjie Luo, Di He, Yiming Yang, Tamas Sarlos, Thomas Weingarten, Adrian Weller
Abstract: We propose a new class of linear Transformers called FourierLearner-Transformers (FLTs), which incorporate a wide range of relative positional encoding mechanisms (RPEs). These include regular RPE techniques applied for sequential data, as well as novel RPEs operating on geometric data embedded in higher-dimensional Euclidean spaces. FLTs construct the optimal RPE mechanism implicitly by learning its spectral representation. As opposed to other architectures combining efficient low-rank linear attention with RPEs, FLTs remain practical in terms of their memory usage and do not require additional assumptions about the structure of the RPE mask. Besides, FLTs allow for applying certain structural inductive bias techniques to specify masking strategies, e.g. they provide a way to learn the so-called local RPEs introduced in this paper and give accuracy gains as compared with several other linear Transformers for language modeling. We also thoroughly test FLTs on other data modalities and tasks, such as image classification, 3D molecular modeling, and learnable optimizers. To the best of our knowledge, for 3D molecular data, FLTs are the first Transformer architectures providing linear attention and incorporating RPE masking.
Authors: Zhenpeng Chen, Jie M. Zhang, Federica Sarro, Mark Harman
Abstract: Existing research mostly improves the fairness of Machine Learning (ML) software regarding a single protected attribute at a time, but this is unrealistic given that many users have multiple protected attributes. This paper conducts an extensive study of fairness improvement regarding multiple protected attributes, covering 11 state-of-the-art fairness improvement methods. We analyze the effectiveness of these methods with different datasets, metrics, and ML models when considering multiple protected attributes. The results reveal that improving fairness for a single protected attribute can largely decrease fairness regarding unconsidered protected attributes. This decrease is observed in up to 88.3% of scenarios (57.5% on average). More surprisingly, we find little difference in accuracy loss when considering single and multiple protected attributes, indicating that accuracy can be maintained in the multiple-attribute paradigm. However, the effect on F1-score when handling two protected attributes is about twice that of a single attribute. This has important implications for future fairness research: reporting only accuracy as the ML performance metric, which is currently common in the literature, is inadequate.
Authors: Ou Deng, Qun Jin
Abstract: Addressing missing data in complex datasets like Electronic Health Records (EHR) is critical for ensuring accurate analysis and decision-making in healthcare. This paper proposes Structural Equation Modeling (SEM) enhanced with the Self-Attention method (SESA), an innovative approach for data imputation in EHR. SESA innovates beyond traditional SEM-based methods by incorporating self-attention mechanisms, enhancing the model's adaptability and accuracy across diverse EHR datasets. This enhancement allows SESA to dynamically adjust and optimize imputation processes, overcoming the limitations of static SEM frameworks. Our experimental analyses demonstrate that SESA achieves robust predictive performance, effectively handling missing data in EHR. Moreover, SESA's architecture not only rectifies potential mis-specifications in SEM but also synergizes with causal discovery algorithms, to refine its imputation logic based on underlying data structures. These features highlight SESA's advanced capabilities and its potential for broader application in EHR data analysis and beyond, marking a significant leap forward in the field of data imputation.
Authors: Jiaren Xiao, Quanyu Dai, Xiao Shen, Xiaochen Xie, Jing Dai, James Lam, Ka-Wai Kwok
Abstract: Label scarcity in a graph is frequently encountered in real-world applications due to the high cost of data labeling. To this end, semi-supervised domain adaptation (SSDA) on graphs aims to leverage the knowledge of a labeled source graph to aid in node classification on a target graph with limited labels. SSDA tasks need to overcome the domain gap between the source and target graphs. However, to date, this challenging research problem has yet to be formally considered by the existing approaches designed for cross-graph node classification. This paper proposes a novel method called SemiGCL to tackle the graph \textbf{Semi}-supervised domain adaptation with \textbf{G}raph \textbf{C}ontrastive \textbf{L}earning and minimax entropy training. SemiGCL generates informative node representations by contrasting the representations learned from a graph's local and global views. Additionally, SemiGCL is adversarially optimized with the entropy loss of unlabeled target nodes to reduce domain divergence. Experimental results on benchmark datasets demonstrate that SemiGCL outperforms the state-of-the-art baselines on the SSDA tasks. The source codes of SemiGCL are publicly available at https://github.com/ JiarenX/SemiGCL.
URLs: https://github.com/
Authors: Alokendu Mazumder, Rishabh Sabharwal, Manan Tayal, Bhartendu Kumar, Punit Rathore
Abstract: In neural network training, RMSProp and Adam remain widely favoured optimisation algorithms. One of the keys to their performance lies in selecting the correct step size, which can significantly influence their effectiveness. Additionally, questions about their theoretical convergence properties continue to be a subject of interest. In this paper, we theoretically analyse a constant step size version of Adam in the non-convex setting and discuss why it is important for the convergence of Adam to use a fixed step size. This work demonstrates the derivation and effective implementation of a constant step size for Adam, offering insights into its performance and efficiency in non convex optimisation scenarios. (i) First, we provide proof that these adaptive gradient algorithms are guaranteed to reach criticality for smooth non-convex objectives with constant step size, and we give bounds on the running time. Both deterministic and stochastic versions of Adam are analysed in this paper. We show sufficient conditions for the derived constant step size to achieve asymptotic convergence of the gradients to zero with minimal assumptions. Next, (ii) we design experiments to empirically study Adam's convergence with our proposed constant step size against stateof the art step size schedulers on classification tasks. Lastly, (iii) we also demonstrate that our derived constant step size has better abilities in reducing the gradient norms, and empirically, we show that despite the accumulation of a few past gradients, the key driver for convergence in Adam is the non-increasing step sizes.
Authors: Jack Foster, Alexandra Brintrup
Abstract: The pursuit of long-term autonomy mandates that robotic agents must continuously adapt to their changing environments and learn to solve new tasks. Continual learning seeks to overcome the challenge of catastrophic forgetting, where learning to solve new tasks causes a model to forget previously learnt information. Prior-based continual learning methods are appealing for robotic applications as they are space efficient and typically do not increase in computational complexity as the number of tasks grows. Despite these desirable properties, prior-based approaches typically fail on important benchmarks and consequently are limited in their potential applications compared to their memory-based counterparts. We introduce Bayesian adaptive moment regularization (BAdam), a novel prior-based method that better constrains parameter growth, leading to lower catastrophic forgetting. Our method boasts a range of desirable properties for robotic applications such as being lightweight and task label-free, converging quickly, and offering calibrated uncertainty that is important for safe real-world deployment. Results show that BAdam achieves state-of-the-art performance for prior-based methods on challenging single-headed class-incremental experiments such as Split MNIST and Split FashionMNIST, and does so without relying on task labels or discrete task boundaries.
Authors: Junqi Jiang, Jianglin Lan, Francesco Leofante, Antonio Rago, Francesca Toni
Abstract: Counterfactual Explanations (CEs) have received increasing interest as a major methodology for explaining neural network classifiers. Usually, CEs for an input-output pair are defined as data points with minimum distance to the input that are classified with a different label than the output. To tackle the established problem that CEs are easily invalidated when model parameters are updated (e.g. retrained), studies have proposed ways to certify the robustness of CEs under model parameter changes bounded by a norm ball. However, existing methods targeting this form of robustness are not sound or complete, and they may generate implausible CEs, i.e., outliers wrt the training dataset. In fact, no existing method simultaneously optimises for closeness and plausibility while preserving robustness guarantees. In this work, we propose Provably RObust and PLAusible Counterfactual Explanations (PROPLACE), a method leveraging on robust optimisation techniques to address the aforementioned limitations in the literature. We formulate an iterative algorithm to compute provably robust CEs and prove its convergence, soundness and completeness. Through a comparative experiment involving six baselines, five of which target robustness, we show that PROPLACE achieves state-of-the-art performances against metrics on three evaluation aspects.
Authors: Parikshit Pareek, Deepjyoti Deka, Sidhant Misra
Abstract: This work presents an efficient data-driven method to construct probabilistic voltage envelopes (PVE) using power flow learning in grids with network contingencies. First, a network-aware Gaussian process (GP) termed Vertex-Degree Kernel (VDK-GP), developed in prior work, is used to estimate voltage-power functions for a few network configurations. The paper introduces a novel multi-task vertex degree kernel (MT-VDK) that amalgamates the learned VDK-GPs to determine power flows for unseen networks, with a significant reduction in the computational complexity and hyperparameter requirements compared to alternate approaches. Simulations on the IEEE 30-Bus network demonstrate the retention and transfer of power flow knowledge in both N-1 and N-2 contingency scenarios. The MT-VDK-GP approach achieves over 50% reduction in mean prediction error for novel N-1 contingency network configurations in low training data regimes (50-250 samples) over VDK-GP. Additionally, MT-VDK-GP outperforms a hyper-parameter based transfer learning approach in over 75% of N-2 contingency network structures, even without historical N-2 outage data. The proposed method demonstrates the ability to achieve PVEs using sixteen times fewer power flow solutions compared to Monte-Carlo sampling-based methods.
Authors: Sander Dalm, Marcel van Gerven, Nasir Ahmad
Abstract: Backpropagation (BP) is the dominant and most successful method for training parameters of deep neural network models. However, BP relies on two computationally distinct phases, does not provide a satisfactory explanation of biological learning, and can be challenging to apply for training of networks with discontinuities or noisy node dynamics. By comparison, node perturbation (NP) proposes learning by the injection of noise into network activations, and subsequent measurement of the induced loss change. NP relies on two forward (inference) passes, does not make use of network derivatives, and has been proposed as a model for learning in biological systems. However, standard NP is highly data inefficient and unstable due to its unguided noise-based search process. In this work, we investigate different formulations of NP and relate it to the concept of directional derivatives as well as combining it with a decorrelating mechanism for layer-wise inputs. We find that a closer alignment with directional derivatives together with input decorrelation at every layer significantly enhances performance of NP learning with significant improvements in parameter convergence and much higher performance on the test data, approaching that of BP. Furthermore, our novel formulation allows for application to noisy systems in which the noise process itself is inaccessible.
Authors: Chongyu Fan, Jiancheng Liu, Yihua Zhang, Eric Wong, Dennis Wei, Sijia Liu
Abstract: With evolving data regulations, machine unlearning (MU) has become an important tool for fostering trust and safety in today's AI models. However, existing MU methods focusing on data and/or weight perspectives often suffer limitations in unlearning accuracy, stability, and cross-domain applicability. To address these challenges, we introduce the concept of 'weight saliency' for MU, drawing parallels with input saliency in model explanation. This innovation directs MU's attention toward specific model weights rather than the entire model, improving effectiveness and efficiency. The resultant method that we call saliency unlearning (SalUn) narrows the performance gap with 'exact' unlearning (model retraining from scratch after removing the forgetting data points). To the best of our knowledge, SalUn is the first principled MU approach that can effectively erase the influence of forgetting data, classes, or concepts in both image classification and generation tasks. As highlighted below, For example, SalUn yields a stability advantage in high-variance random data forgetting, e.g., with a 0.2% gap compared to exact unlearning on the CIFAR-10 dataset. Moreover, in preventing conditional diffusion models from generating harmful images, SalUn achieves nearly 100% unlearning accuracy, outperforming current state-of-the-art baselines like Erased Stable Diffusion and Forget-Me-Not. Codes are available at https://github.com/OPTML-Group/Unlearn-Saliency. (WARNING: This paper contains model outputs that may be offensive in nature.)
Authors: Chandrani Kumari, Rahul Siddharthan
Abstract: We provide new algorithms for two tasks relating to heterogeneous tabular datasets: clustering, and synthetic data generation. Tabular datasets typically consist of heterogeneous data types (numerical, ordinal, categorical) in columns, but may also have hidden cluster structure in their rows: for example, they may be drawn from heterogeneous (geographical, socioeconomic, methodological) sources, such that the outcome variable they describe (such as the presence of a disease) may depend not only on the other variables but on the cluster context. Moreover, sharing of biomedical data is often hindered by patient confidentiality laws, and there is current interest in algorithms to generate synthetic tabular data from real data, for example via deep learning. We demonstrate a novel EM-based clustering algorithm, MMM (``Madras Mixture Model''), that outperforms standard algorithms in determining clusters in synthetic heterogeneous data, and recovers structure in real data. Based on this, we demonstrate a synthetic tabular data generation algorithm, MMMsynth, that pre-clusters the input data, and generates cluster-wise synthetic data assuming cluster-specific data distributions for the input columns. We benchmark this algorithm by testing the performance of standard ML algorithms when they are trained on synthetic data and tested on real published datasets. Our synthetic data generation algorithm outperforms other literature tabular-data generators, and approaches the performance of training purely with real data.
Authors: Jiaqi Li, Rui Wang, Yuanhao Lai, Changjian Shui, Sabyasachi Sahoo, Charles X. Ling, Shichun Yang, Boyu Wang, Christian Gagn\'e, Fan Zhou
Abstract: Continual learning aims to learn a series of tasks sequentially without forgetting the knowledge acquired from the previous ones. In this work, we propose the Hessian Aware Low-Rank Perturbation algorithm for continual learning. By modeling the parameter transitions along the sequential tasks with the weight matrix transformation, we propose to apply the low-rank approximation on the task-adaptive parameters in each layer of the neural networks. Specifically, we theoretically demonstrate the quantitative relationship between the Hessian and the proposed low-rank approximation. The approximation ranks are then globally determined according to the marginal increment of the empirical loss estimated by the layer-specific gradient and low-rank approximation error. Furthermore, we control the model capacity by pruning less important parameters to diminish the parameter growth. We conduct extensive experiments on various benchmarks, including a dataset with large-scale tasks, and compare our method against some recent state-of-the-art methods to demonstrate the effectiveness and scalability of our proposed method. Empirical results show that our method performs better on different benchmarks, especially in achieving task order robustness and handling the forgetting issue. The source code is at https://github.com/lijiaqi/HALRP.
Authors: Alex Mallen, Madeline Brumley, Julia Kharchenko, Nora Belrose
Abstract: Eliciting Latent Knowledge (ELK) aims to find patterns in a capable neural network's activations that robustly track the true state of the world, especially in hard-to-verify cases where the model's output is untrusted. To further ELK research, we introduce 12 datasets and a corresponding suite of "quirky" language models (LMs) that are finetuned to make systematic errors when answering questions if and only if the keyword "Bob" is present in the prompt. We find that, especially in middle layers, linear probes usually report an LM's knowledge independently of what the LM outputs, enabling us to elicit the correct answer despite the model's untruthful output. The best probing method (logistic regression on contrast pairs) recovers 89% of the gap in AUROC between truthful and untruthful contexts, and 75% for questions harder than those used to train the probe. We also find that a mechanistic anomaly detection approach can flag untruthful behavior with 0.95 AUROC. Our results show promise for eliciting reliable knowledge from capable but untrusted models, and facilitates future research empirically investigating ELK methods.
Authors: Meng Xu, Ke Han
Abstract: Construction waste hauling trucks (or `slag trucks') are among the most commonly seen heavy-duty diesel vehicles in urban streets, which not only produce significant carbon, NO$_{\textbf{x}}$ and PM$_{\textbf{2.5}}$ emissions but are also a major source of on-road and on-site fugitive dust. Slag trucks are subject to a series of spatial and temporal access restrictions by local traffic and environmental policies. This paper addresses the practical problem of predicting levels of slag truck activity at a city scale during heavy pollution episodes, such that environmental law enforcement units can take timely and proactive measures against localized truck aggregation. A deep ensemble learning framework (coined AI-Truck) is designed, which employs a soft vote integrator that utilizes Bi-LSTM, TCN, STGCN, and PDFormer as base classifiers. AI-Truck employs a combination of downsampling and weighted loss is employed to address sample imbalance, and utilizes truck trajectories to extract more accurate and effective geographic features. The framework was deployed for truck activity prediction at a resolution of 1km$\times$1km$\times$0.5h, in a 255 km$^{\textbf{2}}$ area in Chengdu, China. As a classifier, AI-Truck achieves a macro F1 of 0.747 in predicting levels of slag truck activity for 0.5-h prediction time length, and enables personnel to spot high-activity locations 1.5 hrs ahead with over 80\% accuracy.
Authors: Mohammad Al Olaimat (for the Alzheimer's Disease Neuroimaging Initiative), Serdar Bozdag (for the Alzheimer's Disease Neuroimaging Initiative)
Abstract: Motivation: Electronic Health Records (EHR) represent a comprehensive resource of a patient's medical history. EHR are essential for utilizing advanced technologies such as deep learning (DL), enabling healthcare providers to analyze extensive data, extract valuable insights, and make precise and data-driven clinical decisions. DL methods such as Recurrent Neural Networks (RNN) have been utilized to analyze EHR to model disease progression and predict diagnosis. However, these methods do not address some inherent irregularities in EHR data such as irregular time intervals between clinical visits. Furthermore, most DL models are not interpretable. In this study, we propose two interpretable DL architectures based on RNN, namely Time-Aware RNN (TA-RNN) and TA-RNN-Autoencoder (TA-RNN-AE) to predict patient's clinical outcome in EHR at next visit and multiple visits ahead, respectively. To mitigate the impact of irregular time intervals, we propose incorporating time embedding of the elapsed times between visits. For interpretability, we propose employing a dual-level attention mechanism that operates between visits and features within each visit. Results: The results of the experiments conducted on Alzheimer's Disease Neuroimaging Initiative (ADNI) and National Alzheimer's Coordinating Center (NACC) datasets indicated superior performance of proposed models for predicting Alzheimer's Disease (AD) compared to state-of-the-art and baseline approaches based on F2 and sensitivity. Additionally, TA-RNN showed superior performance on Medical Information Mart for Intensive Care (MIMIC-III) dataset for mortality prediction. In our ablation study, we observed enhanced predictive performance by incorporating time embedding and attention mechanisms. Finally, investigating attention weights helped identify influential visits and features in predictions.
Authors: Coleman Hooper, Sehoon Kim, Hiva Mohammadzadeh, Michael W. Mahoney, Yakun Sophia Shao, Kurt Keutzer, Amir Gholami
Abstract: LLMs are seeing growing use for applications such as document analysis and summarization which require large context windows, and with these large context windows KV cache activations surface as the dominant contributor to memory consumption during inference. Quantization is a promising approach for compressing KV cache activations; however, existing solutions fail to represent activations accurately in ultra-low precisions, such as sub-4-bit. In this work, we present KVQuant, which addresses this problem by incorporating novel methods for quantizing cached KV activations, including: (i) Per-Channel Key Quantization, where we adjust the dimension along which we quantize the Key activations to better match the distribution; (ii) Pre-RoPE Key Quantization, where we quantize Key activations before the rotary positional embedding to mitigate its impact on quantization; (iii) Non-Uniform KV Cache Quantization, where we derive per-layer sensitivity-weighted non-uniform datatypes that better represent the distributions; (iv) Per-Vector Dense-and-Sparse Quantization, where we isolate outliers separately for each vector to minimize skews in quantization ranges; and (v) Q-Norm, where we normalize quantization centroids in order to mitigate distribution shift, providing additional benefits for 2-bit quantization. By applying our method to the LLaMA, LLaMA-2, and Mistral models, we achieve $<0.1$ perplexity degradation with 3-bit quantization on both Wikitext-2 and C4, outperforming existing approaches. Our method enables serving the LLaMA-7B model with a context length of up to 1 million on a single A100-80GB GPU and up to 10 million on an 8-GPU system.
Authors: Ilja Kuzborskij, Kwang-Sung Jun, Yulian Wu, Kyoungseok Jang, Francesco Orabona
Abstract: Let $f(\theta, X_1),$ $ \dots,$ $ f(\theta, X_n)$ be a sequence of random elements, where $f$ is a fixed scalar function, $X_1, \dots, X_n$ are independent random variables (data), and $\theta$ is a random parameter distributed according to some data-dependent posterior distribution $P_n$. In this paper, we consider the problem of proving concentration inequalities to estimate the mean of the sequence. An example of such a problem is the estimation of the generalization error of some predictor trained by a stochastic algorithm, such as a neural network where $f$ is a loss function. Classically, this problem is approached through a PAC-Bayes analysis where, in addition to the posterior, we choose a prior distribution which captures our belief about the inductive bias of the learning problem. Then, the key quantity in PAC-Bayes concentration bounds is a divergence that captures the complexity of the learning problem where the de facto standard choice is the KL divergence. However, the tightness of this choice has rarely been questioned. In this paper, we challenge the tightness of the KL-divergence-based bounds by showing that it is possible to achieve a strictly tighter bound. In particular, we demonstrate new high-probability PAC-Bayes bounds with a novel and better-than-KL divergence that is inspired by Zhang et al. (2022). Our proof is inspired by recent advances in regret analysis of gambling algorithms, and its use to derive concentration inequalities. Our result is first-of-its-kind in that existing PAC-Bayes bounds with non-KL divergences are not known to be strictly better than KL. Thus, we believe our work marks the first step towards identifying optimal rates of PAC-Bayes bounds.
Authors: Hezhe Qiao, Qingsong Wen, Xiaoli Li, Ee-Peng Lim, Guansong Pang
Abstract: This work considers a practical semi-supervised graph anomaly detection (GAD) scenario, where part of the nodes in a graph are known to be normal, contrasting to the unsupervised setting in most GAD studies with a fully unlabeled graph. As expected, we find that having access to these normal nodes helps enhance the detection performance of existing unsupervised GAD methods when they are adapted to the semi-supervised setting. However, their utilization of these normal nodes is limited. In this paper, we propose a novel Generative GAD approach (GGAD) for the semi-supervised scenario to better exploit the normal nodes. The key idea is to generate outlier nodes that assimilate anomaly nodes in both local structure and node representations for providing effective negative node samples in training a discriminative one-class classifier. There have been many generative anomaly detection approaches, but they are designed for non-graph data, and as a result, they fail to take account of the graph structure information. Our approach tackles this problem by generating graph structure-aware outlier nodes that have asymmetric affinity separability from normal nodes while being enforced to achieve egocentric closeness to normal nodes in the node representation space. Comprehensive experiments on four real-world datasets are performed to establish a benchmark for semi-supervised GAD and show that GGAD substantially outperforms state-of-the-art unsupervised and semi-supervised GAD methods with varying numbers of training normal nodes. Code will be made available at https://github.com/mala-lab/GGAD.
Authors: Kai-Hendrik Cohrs, Gherardo Varando, Nuno Carvalhais, Markus Reichstein, Gustau Camps-Valls
Abstract: Hybrid modeling integrates machine learning with scientific knowledge to enhance interpretability, generalization, and adherence to natural laws. Nevertheless, equifinality and regularization biases pose challenges in hybrid modeling to achieve these purposes. This paper introduces a novel approach to estimating hybrid models via a causal inference framework, specifically employing Double Machine Learning (DML) to estimate causal effects. We showcase its use for the Earth sciences on two problems related to carbon dioxide fluxes. In the $Q_{10}$ model, we demonstrate that DML-based hybrid modeling is superior in estimating causal parameters over end-to-end deep neural network (DNN) approaches, proving efficiency, robustness to bias from regularization methods, and circumventing equifinality. Our approach, applied to carbon flux partitioning, exhibits flexibility in accommodating heterogeneous causal effects. The study emphasizes the necessity of explicitly defining causal graphs and relationships, advocating for this as a general best practice. We encourage the continued exploration of causality in hybrid models for more interpretable and trustworthy results in knowledge-guided machine learning.
Authors: Lei Yu, Ke Han
Abstract: Earthwork-related locations (ERLs), such as construction sites, earth dumping ground, and concrete mixing stations, are major sources of urban dust pollution (particulate matters). The effective management of ERLs is crucial and requires timely and efficient tracking of these locations throughout the city. This work aims to identify and classify urban ERLs using GPS trajectory data of over 16,000 construction waste hauling trucks (CWHTs), as well as 58 urban features encompassing geographic, land cover, POI and transport dimensions. We compare several machine learning models and examine the impact of various spatial-temporal features on classification performance using real-world data in Chengdu, China. The results demonstrate that 77.8% classification accuracy can be achieved with a limited number of features. This classification framework was implemented in the Alpha MAPS system in Chengdu, which has successfully identified 724 construction cites/earth dumping ground, 48 concrete mixing stations, and 80 truck parking locations in the city during December 2023, which has enabled local authority to effectively manage urban dust pollution at low personnel costs.
Authors: C. Coelho, M. Jing, M. Fernanda P. Costa, L. L. Ferr\'as
Abstract: Hydropower plants play a pivotal role in advancing clean and sustainable energy production, contributing significantly to the global transition towards renewable energy sources. However, hydropower plants are currently perceived both positively as sources of renewable energy and negatively as disruptors of ecosystems. In this work, we highlight the overlooked potential of using hydropower plant as protectors of ecosystems by using adaptive ecological discharges. To advocate for this perspective, we propose using a neural network to predict the minimum ecological discharge value at each desired time. Additionally, we present a novel framework that seamlessly integrates it into hydropower management software, taking advantage of the well-established approach of using traditional constrained optimisation algorithms. This novel approach not only protects the ecosystems from climate change but also contributes to potentially increase the electricity production.
Authors: Masahiro Kato, Shinji Ito
Abstract: This study considers the linear contextual bandit problem with independent and identically distributed (i.i.d.) contexts. In this problem, existing studies have proposed Best-of-Both-Worlds (BoBW) algorithms whose regrets satisfy $O(\log^2(T))$ for the number of rounds $T$ in a stochastic regime with a suboptimality gap lower-bounded by a positive constant, while satisfying $O(\sqrt{T})$ in an adversarial regime. However, the dependency on $T$ has room for improvement, and the suboptimality-gap assumption can be relaxed. For this issue, this study proposes an algorithm whose regret satisfies $O(\log(T))$ in the setting when the suboptimality gap is lower-bounded. Furthermore, we introduce a margin condition, a milder assumption on the suboptimality gap. That condition characterizes the problem difficulty linked to the suboptimality gap using a parameter $\beta \in (0, \infty]$. We then show that the algorithm's regret satisfies $O\left(\left\{\log(T)\right\}^{\frac{1+\beta}{2+\beta}}T^{\frac{1}{2+\beta}}\right)$. Here, $\beta= \infty$ corresponds to the case in the existing studies where a lower bound exists in the suboptimality gap, and our regret satisfies $O(\log(T))$ in that case. Our proposed algorithm is based on the Follow-The-Regularized-Leader with the Tsallis entropy and referred to as the $\alpha$-Linear-Contextual (LC)-Tsallis-INF.
Authors: Christian W. Frey
Abstract: Industrial production processes, especially in the pharmaceutical industry, are complex systems that require continuous monitoring to ensure efficiency, product quality, and safety. This paper presents a hybrid unsupervised learning strategy (HULS) for monitoring complex industrial processes. Addressing the limitations of traditional Self-Organizing Maps (SOMs), especially in scenarios with unbalanced data sets and highly correlated process variables, HULS combines existing unsupervised learning techniques to address these challenges. To evaluate the performance of the HULS concept, comparative experiments are performed based on a laboratory batch
Authors: Busra Asan, Abdullah Akg\"ul, Alper Unal, Melih Kandemir, Gozde Unal
Abstract: Seasonal forecasting is a crucial task when it comes to detecting the extreme heat and colds that occur due to climate change. Confidence in the predictions should be reliable since a small increase in the temperatures in a year has a big impact on the world. Calibration of the neural networks provides a way to ensure our confidence in the predictions. However, calibrating regression models is an under-researched topic, especially in forecasters. We calibrate a UNet++ based architecture, which was shown to outperform physics-based models in temperature anomalies. We show that with a slight trade-off between prediction error and calibration error, it is possible to get more reliable and sharper forecasts. We believe that calibration should be an important part of safety-critical machine learning applications such as weather forecasters.
Authors: Yazheng Yang, Yuqi Wang, Sankalok Sen, Lei Li, Qi Liu
Abstract: In the domain of data science, the predictive tasks of classification, regression, and imputation of missing values are commonly encountered challenges associated with tabular data. This research endeavors to apply Large Language Models (LLMs) towards addressing these predictive tasks. Despite their proficiency in comprehending natural language, LLMs fall short in dealing with structured tabular data. This limitation stems from their lacking exposure to the intricacies of tabular data during their foundational training. Our research aims to mitigate this gap by compiling a comprehensive corpus of tables annotated with instructions and executing large-scale training of Llama-2 on this enriched dataset. Furthermore, we investigate the practical application of applying the trained model to zero-shot prediction, few-shot prediction, and in-context learning scenarios. Through extensive experiments, our methodology has shown significant improvements over existing benchmarks. These advancements highlight the efficacy of tailoring LLM training to solve table-related problems in data science, thereby establishing a new benchmark in the utilization of LLMs for enhancing tabular intelligence.
Authors: Leona Hennig, Tanja Tornede, Marius Lindauer
Abstract: Deep Learning (DL) has advanced various fields by extracting complex patterns from large datasets. However, the computational demands of DL models pose environmental and resource challenges. Deep shift neural networks (DSNNs) offer a solution by leveraging shift operations to reduce computational complexity at inference. Following the insights from standard DNNs, we are interested in leveraging the full potential of DSNNs by means of AutoML techniques. We study the impact of hyperparameter optimization (HPO) to maximize DSNN performance while minimizing resource consumption. Since this combines multi-objective (MO) optimization with accuracy and energy consumption as potentially complementary objectives, we propose to combine state-of-the-art multi-fidelity (MF) HPO with multi-objective optimization. Experimental results demonstrate the effectiveness of our approach, resulting in models with over 80\% in accuracy and low computational cost. Overall, our method accelerates efficient model development while enabling sustainable AI applications.
Authors: Paschalis Bizopoulos, Dimitrios Koutsouris
Abstract: The medical field is creating large amount of data that physicians are unable to decipher and use efficiently. Moreover, rule-based expert systems are inefficient in solving complicated medical tasks or for creating insights using big data. Deep learning has emerged as a more accurate and effective technology in a wide range of medical problems such as diagnosis, prediction and intervention. Deep learning is a representation learning method that consists of layers that transform the data non-linearly, thus, revealing hierarchical relationships and structures. In this review we survey deep learning application papers that use structured data, signal and imaging modalities from cardiology. We discuss the advantages and limitations of applying deep learning in cardiology that also apply in medicine in general, while proposing certain directions as the most viable for clinical use.
Authors: Akash Srivastava, Yamini Bansal, Yukun Ding, Cole Lincoln Hurwitz, Kai Xu, Bernhard Egger, Prasanna Sattigeri, Joshua B. Tenenbaum, Phuong Le, Arun Prakash R, Nengfeng Zhou, Joel Vaughan, Yaquan Wang, Anwesha Bhattacharyya, Kristjan Greenewald, David D. Cox, Dan Gutfreund
Abstract: Current autoencoder-based disentangled representation learning methods achieve disentanglement by penalizing the (aggregate) posterior to encourage statistical independence of the latent factors. This approach introduces a trade-off between disentangled representation learning and reconstruction quality since the model does not have enough capacity to learn correlated latent variables that capture detail information present in most image data. To overcome this trade-off, we present a novel multi-stage modeling approach where the disentangled factors are first learned using a penalty-based disentangled representation learning method; then, the low-quality reconstruction is improved with another deep generative model that is trained to model the missing correlated latent variables, adding detail information while maintaining conditioning on the previously learned disentangled factors. Taken together, our multi-stage modelling approach results in a single, coherent probabilistic model that is theoretically justified by the principal of D-separation and can be realized with a variety of model classes including likelihood-based models such as variational autoencoders, implicit models such as generative adversarial networks, and tractable models like normalizing flows or mixtures of Gaussians. We demonstrate that our multi-stage model has higher reconstruction quality than current state-of-the-art methods with equivalent disentanglement performance across multiple standard benchmarks. In addition, we apply the multi-stage model to generate synthetic tabular datasets, showcasing an enhanced performance over benchmark models across a variety of metrics. The interpretability analysis further indicates that the multi-stage model can effectively uncover distinct and meaningful features of variations from which the original distribution can be recovered.
Authors: Guilherme C. Oliveira, Gustavo H. Rosa, Daniel C. G. Pedronette, Jo\~ao P. Papa, Himeesh Kumar, Leandro A. Passos, Dinesh Kumar
Abstract: Deep learning applications for assessing medical images are limited because the datasets are often small and imbalanced. The use of synthetic data has been proposed in the literature, but neither a robust comparison of the different methods nor generalizability has been reported. Our approach integrates a retinal image quality assessment model and StyleGAN2 architecture to enhance Age-related Macular Degeneration (AMD) detection capabilities and improve generalizability. This work compares ten different Generative Adversarial Network (GAN) architectures to generate synthetic eye-fundus images with and without AMD. We combined subsets of three public databases (iChallenge-AMD, ODIR-2019, and RIADD) to form a single training and test set. We employed the STARE dataset for external validation, ensuring a comprehensive assessment of the proposed approach. The results show that StyleGAN2 reached the lowest Frechet Inception Distance (166.17), and clinicians could not accurately differentiate between real and synthetic images. ResNet-18 architecture obtained the best performance with 85% accuracy and outperformed the two human experts (80% and 75%) in detecting AMD fundus images. The accuracy rates were 82.8% for the test set and 81.3% for the STARE dataset, demonstrating the model's generalizability. The proposed methodology for synthetic medical image generation has been validated for robustness and accuracy, with free access to its code for further research and development in this field.
Authors: Takashi Takahashi
Abstract: Self-training (ST) is a simple and standard approach in semi-supervised learning that has been applied to many machine learning problems. Despite its widespread acceptance and practical effectiveness, it is still not well understood why and how ST improves performance by fitting the model to potentially erroneous pseudo-labels. To investigate the properties of ST, in this study, we derive and analyze a sharp characterization of the behavior of iterative ST when training a linear classifier by minimizing the ridge-regularized convex loss for binary Gaussian mixtures, in the asymptotic limit where input dimension and data size diverge proportionally. The derivation is based on the replica method of statistical mechanics. The result indicates that, when the total number of iterations is large, ST may find a classification plane with the optimal direction regardless of the label imbalance by accumulating small parameter updates over long iterations. It is argued that this is because the small update of ST can accumulate information of the data in an almost noiseless way. However, when a label imbalance is present in true labels, the performance of the ST is significantly lower than that of supervised learning with true labels, because the ratio between the norm of the weight and the magnitude of the bias can become significantly large. To overcome the problems in label imbalanced cases, several heuristics are introduced. By numerically analyzing the asymptotic formula, it is demonstrated that with the proposed heuristics, ST can find a classifier whose performance is nearly compatible with supervised learning using true labels even in the presence of significant label imbalance.
Authors: Johannes O. Royset
Abstract: Uncertainty is prevalent in engineering design, data-driven problems, and decision making broadly. Due to inherent risk-averseness and ambiguity about assumptions, it is common to address uncertainty by formulating and solving conservative optimization models expressed using measures of risk and related concepts. We survey the rapid development of risk measures over the last quarter century. From their beginning in financial engineering, we recount the spread to nearly all areas of engineering and applied mathematics. Solidly rooted in convex analysis, risk measures furnish a general framework for handling uncertainty with significant computational and theoretical advantages. We describe the key facts, list several concrete algorithms, and provide an extensive list of references for further reading. The survey recalls connections with utility theory and distributionally robust optimization, points to emerging applications areas such as fair machine learning, and defines measures of reliability.
Authors: Benedict Quartey, Ankit Shah, George Konidaris
Abstract: Reinforcement learning requires interaction with an environment, which is expensive for robots. This constraint necessitates approaches that work with limited environmental interaction by maximizing the reuse of previous experiences. We propose an approach that maximizes experience reuse while learning to solve a given task by generating and simultaneously learning useful auxiliary tasks. To generate these tasks, we construct an abstract temporal logic representation of the given task and leverage large language models to generate context-aware object embeddings that facilitate object replacements. Counterfactual reasoning and off-policy methods allow us to simultaneously learn these auxiliary tasks while solving the given target task. We combine these insights into a novel framework for multitask reinforcement learning and experimentally show that our generated auxiliary tasks share similar underlying exploration requirements as the given task, thereby maximizing the utility of directed exploration. Our approach allows agents to automatically learn additional useful policies without extra environment interaction.
Authors: Abhimanyu Das, Weihao Kong, Andrew Leach, Shaan Mathur, Rajat Sen, Rose Yu
Abstract: Recent work has shown that simple linear models can outperform several Transformer based approaches in long term time-series forecasting. Motivated by this, we propose a Multi-layer Perceptron (MLP) based encoder-decoder model, Time-series Dense Encoder (TiDE), for long-term time-series forecasting that enjoys the simplicity and speed of linear models while also being able to handle covariates and non-linear dependencies. Theoretically, we prove that the simplest linear analogue of our model can achieve near optimal error rate for linear dynamical systems (LDS) under some assumptions. Empirically, we show that our method can match or outperform prior approaches on popular long-term time-series forecasting benchmarks while being 5-10x faster than the best Transformer based model.
Authors: Taigao Ma, Haozhu Wang, L. Jay Guo
Abstract: Optical multilayer thin film structures have been widely used in numerous photonic applications. However, existing inverse design methods have many drawbacks because they either fail to quickly adapt to different design targets, or are difficult to suit for different types of structures, e.g., designing for different materials at each layer. These methods also cannot accommodate versatile design situations under different angles and polarizations. In addition, how to benefit practical fabrications and manufacturing has not been extensively considered yet. In this work, we introduce OptoGPT (Opto Generative Pretrained Transformer), a decoder-only transformer, to solve all these drawbacks and issues simultaneously.
Authors: Albert Yu Sun, Varun Nair, Elliot Schumacher, Anitha Kannan
Abstract: A wave of new task-based virtual assistants has been fueled by increasingly powerful large language models (LLMs), such as GPT-4 (OpenAI, 2023). A major challenge in deploying LLM-based virtual conversational assistants in real world settings is ensuring they operate within what is admissible for the task. To overcome this challenge, the designers of these virtual assistants rely on an independent guardrail system that verifies the virtual assistant's output aligns with the constraints required for the task. However, relying on commonly used, prompt-based guardrails can be difficult to engineer correctly and comprehensively. To address these challenges, we propose CONSCENDI. We use CONSCENDI to exhaustively generate training data with two key LLM-powered components: scenario-augmented generation and contrastive training examples. When generating conversational data, we generate a set of rule-breaking scenarios, which enumerate a diverse set of high-level ways a rule can be violated. This scenario-guided approach produces a diverse training set and provides chatbot designers greater control. To generate contrastive examples, we prompt the LLM to alter conversations with violations into acceptable conversations to enable fine-grained distinctions. We then use this data, generated by CONSCENDI, to train a smaller model. We find that CONSCENDI results in guardrail models that improve over baselines in multiple dialogue domains.
Authors: William Rudman, Carsten Eickhoff
Abstract: Given the success of Large Language Models (LLMs), there has been considerable interest in studying the properties of model activations. The literature overwhelmingly agrees that LLM representations are dominated by a few "outlier dimensions" with exceedingly high variance and magnitude. Several studies in Natural Language Processing (NLP) have sought to mitigate the impact of such outlier dimensions and force LLMs to be isotropic (i.e., have uniform variance across all dimensions in embedding space). Isotropy is thought to be a desirable property for LLMs that improves model performance and more closely aligns textual representations with human intuition. However, many of the claims regarding isotropy in NLP have been based on the average cosine similarity of embeddings, which has recently been shown to be a flawed measure of isotropy. In this paper, we propose I-STAR: IsoScore*-based STable Anisotropic Regularization, a novel regularization method that can be used to increase or decrease levels of isotropy in embedding space during training. I-STAR uses IsoScore*, the first accurate measure of isotropy that is both differentiable and stable on mini-batch computations. In contrast to several previous works, we find that decreasing isotropy in contextualized embeddings improves performance on the majority of tasks and models considered in this paper.
Authors: Hezhe Qiao, Guansong Pang
Abstract: We reveal a one-class homophily phenomenon, which is one prevalent property we find empirically in real-world graph anomaly detection (GAD) datasets, i.e., normal nodes tend to have strong connection/affinity with each other, while the homophily in abnormal nodes is significantly weaker than normal nodes. However, this anomaly-discriminative property is ignored by existing GAD methods that are typically built using a conventional anomaly detection objective, such as data reconstruction. In this work, we explore this property to introduce a novel unsupervised anomaly scoring measure for GAD, local node affinity, that assigns a larger anomaly score to nodes that are less affiliated with their neighbors, with the affinity defined as similarity on node attributes/representations. We further propose Truncated Affinity Maximization (TAM) that learns tailored node representations for our anomaly measure by maximizing the local affinity of nodes to their neighbors. Optimizing on the original graph structure can be biased by nonhomophily edges (i.e., edges connecting normal and abnormal nodes). Thus, TAM is instead optimized on truncated graphs where non-homophily edges are removed iteratively to mitigate this bias. The learned representations result in significantly stronger local affinity for normal nodes than abnormal nodes. Extensive empirical results on 10 real-world GAD datasets show that TAM substantially outperforms seven competing models, achieving over 10% increase in AUROC/AUPRC compared to the best contenders on challenging datasets. Our code is available at https://github.com/mala-lab/TAM-master/.
Authors: Mohamed el Shehaby, Ashraf Matrawy
Abstract: Machine Learning (ML) has become ubiquitous, and its deployment in Network Intrusion Detection Systems (NIDS) is inevitable due to its automated nature and high accuracy compared to traditional models in processing and classifying large volumes of data. However, ML has been found to have several flaws, most importantly, adversarial attacks, which aim to trick ML models into producing faulty predictions. While most adversarial attack research focuses on computer vision datasets, recent studies have explored the suitability of these attacks against ML-based network security entities, especially NIDS, due to the wide difference between different domains regarding the generation of adversarial attacks. To further explore the practicality of adversarial attacks against ML-based NIDS in-depth, this paper presents three distinct contributions: identifying numerous practicality issues for evasion adversarial attacks on ML-NIDS using an attack tree threat model, introducing a taxonomy of practicality issues associated with adversarial attacks against ML-based NIDS, and investigating how the dynamicity of some real-world ML models affects adversarial attacks against NIDS. Our experiments indicate that continuous re-training, even without adversarial training, can reduce the effectiveness of adversarial attacks. While adversarial attacks can compromise ML-based NIDSs, our aim is to highlight the significant gap between research and real-world practicality in this domain, warranting attention.
Authors: Sohum Thakkar (QC Ware Corp), Skander Kazdaghli (QC Ware Corp), Natansh Mathur (QC Ware Corp, IRIF - Universit\'e Paris Cit\'e and CNRS), Iordanis Kerenidis (QC Ware Corp, IRIF - Universit\'e Paris Cit\'e and CNRS), Andr\'e J. Ferreira-Martins (Ita\'u Unibanco), Samurai Brito (Ita\'u Unibanco)
Abstract: Quantum algorithms have the potential to enhance machine learning across a variety of domains and applications. In this work, we show how quantum machine learning can be used to improve financial forecasting. First, we use classical and quantum Determinantal Point Processes to enhance Random Forest models for churn prediction, improving precision by almost 6%. Second, we design quantum neural network architectures with orthogonal and compound layers for credit risk assessment, which match classical performance with significantly fewer parameters. Our results demonstrate that leveraging quantum ideas can effectively enhance the performance of machine learning, both today as quantum-inspired classical ML solutions, and even more in the future, with the advent of better quantum hardware.
Authors: Sinan G. Aksoy, Ilya Amburg, Stephen J. Young
Abstract: While multilinear algebra appears natural for studying the multiway interactions modeled by hypergraphs, tensor methods for general hypergraphs have been stymied by theoretical and practical barriers. A recently proposed adjacency tensor is applicable to nonuniform hypergraphs, but is prohibitively costly to form and analyze in practice. We develop tensor times same vector (TTSV) algorithms for this tensor which improve complexity from $O(n^r)$ to a low-degree polynomial in $r$, where $n$ is the number of vertices and $r$ is the maximum hyperedge size. Our algorithms are implicit, avoiding formation of the order $r$ adjacency tensor. We demonstrate the flexibility and utility of our approach in practice by developing tensor-based hypergraph centrality and clustering algorithms. We also show these tensor measures offer complementary information to analogous graph-reduction approaches on data, and are also able to detect higher-order structure that many existing matrix-based approaches provably cannot.
Authors: Paulin de Schoulepnikoff, Oriel Kiss, Sofia Vallecorsa, Giuseppe Carleo, Michele Grossi
Abstract: Entanglement forging based variational algorithms leverage the bi-partition of quantum systems for addressing ground state problems. The primary limitation of these approaches lies in the exponential summation required over the numerous potential basis states, or bitstrings, when performing the Schmidt decomposition of the whole system. To overcome this challenge, we propose a new method for entanglement forging employing generative neural networks to identify the most pertinent bitstrings, eliminating the need for the exponential sum. Through empirical demonstrations on systems of increasing complexity, we show that the proposed algorithm achieves comparable or superior performance compared to the existing standard implementation of entanglement forging. Moreover, by controlling the amount of required resources, this scheme can be applied to larger, as well as non permutation invariant systems, where the latter constraint is associated with the Heisenberg forging procedure. We substantiate our findings through numerical simulations conducted on spins models exhibiting one-dimensional ring, two-dimensional triangular lattice topologies, and nuclear shell model configurations.
Authors: Chen-Han Tsai, Yu-Shao Peng
Abstract: Image outlier detection (OD) is an essential tool to ensure the quality of images used in computer vision tasks. Existing algorithms often involve training a model to represent the inlier distribution, and outliers are determined by some deviation measure. Although existing methods proved effective when trained on strictly inlier samples, their performance remains questionable when undesired outliers are included during training. As a result of this limitation, it is necessary to carefully examine the data when developing OD models for new domains. In this work, we present a novel image OD algorithm called RANSAC-NN that eliminates the need of data examination and model training altogether. Unlike existing approaches, RANSAC-NN can be directly applied on datasets containing outliers by sampling and comparing subsets of the data. Our algorithm maintains favorable performance compared to existing methods on a range of benchmarks. Furthermore, we show that RANSAC-NN can enhance the robustness of existing methods by incorporating our algorithm as part of the data preparation process.
Authors: Diego P\'erez-L\'opez, Fernando Ortega, \'Angel Gonz\'alez-Prieto, Jorge Due\~nas-Ler\'in
Abstract: Recommender systems are intrinsically tied to a reliability/coverage dilemma: The more reliable we desire the forecasts, the more conservative the decision will be and thus, the fewer items will be recommended. This leads to a significant drop in the novelty of these systems, since instead of recommending uncertain unusual items, they focus on predicting items with guaranteed success. In this paper, we propose the inclusion of a new term in the learning process of matrix factorization-based recommender systems, called recklessness, that takes into account the variance of the output probability distribution of the predicted ratings. In this way, gauging this recklessness measure we can force more spiky output distribution, enabling the control of the risk level desired when making decisions about the reliability of a prediction. Experimental results demonstrate that recklessness not only allows for risk regulation but also improves the quantity and quality of predictions provided by the recommender system.
Authors: Xuansheng Wu, Wenlin Yao, Jianshu Chen, Xiaoman Pan, Xiaoyang Wang, Ninghao Liu, Dong Yu
Abstract: Large Language Models (LLMs) have achieved remarkable success, where instruction tuning is the critical step in aligning LLMs with user intentions. In this work, we investigate how the instruction tuning adjusts pre-trained models with a focus on intrinsic changes. Specifically, we first develop several local and global explanation methods, including a gradient-based method for input-output attribution, and techniques for interpreting patterns and concepts in self-attention and feed-forward layers. The impact of instruction tuning is then studied by comparing the explanations derived from the pre-trained and instruction-tuned models. This approach provides an internal perspective of the model shifts on a human-comprehensible level. Our findings reveal three significant impacts of instruction tuning: 1) It empowers LLMs to recognize the instruction parts of user prompts, and promotes the response generation constantly conditioned on the instructions. 2) It encourages the self-attention heads to capture more word-word relationships about instruction verbs. 3) It encourages the feed-forward networks to rotate their pre-trained knowledge toward user-oriented tasks. These insights contribute to a more comprehensive understanding of instruction tuning and lay the groundwork for future work that aims at explaining and optimizing LLMs for various applications. Our code and data are publicly available at https://github.com/JacksonWuxs/Interpret_Instruction_Tuning_LLMs.
URLs: https://github.com/JacksonWuxs/Interpret_Instruction_Tuning_LLMs.
Authors: Samuel Holt, Max Ruiz Luyten, Mihaela van der Schaar
Abstract: Transformer-based large language models (LLMs) are constrained by the fixed context window of the underlying transformer architecture, hindering their ability to produce long and coherent outputs. Memory-augmented LLMs are a promising solution, but current approaches cannot handle long output generation tasks since they (1) only focus on reading memory and reduce its evolution to the concatenation of new memories or (2) use very specialized memories that cannot adapt to other domains. This paper presents L2MAC, the first practical LLM-based stored-program automatic computer (von Neumann architecture) framework, an LLM-based multi-agent system, for long and consistent output generation. Its memory has two components: the instruction registry, which is populated with a prompt program to solve the user-given task, and a file store, which will contain the final and intermediate outputs. Each instruction in turn is executed by a separate LLM agent, whose context is managed by a control unit capable of precise memory reading and writing to ensure effective interaction with the file store. These components enable L2MAC to generate extensive outputs, bypassing the constraints of the finite context window while producing outputs that fulfill a complex user-specified task. We empirically demonstrate that L2MAC achieves state-of-the-art performance in generating large codebases for system design tasks, significantly outperforming other coding methods in implementing the detailed user-specified task, and we provide valuable insights into the reasons for this performance gap.
Authors: Fred Hohman, Mary Beth Kery, Donghao Ren, Dominik Moritz
Abstract: On-device machine learning (ML) promises to improve the privacy, responsiveness, and proliferation of new, intelligent user experiences by moving ML computation onto everyday personal devices. However, today's large ML models must be drastically compressed to run efficiently on-device, a hurtle that requires deep, yet currently niche expertise. To engage the broader human-centered ML community in on-device ML experiences, we present the results from an interview study with 30 experts at Apple that specialize in producing efficient models. We compile tacit knowledge that experts have developed through practical experience with model compression across different hardware platforms. Our findings offer pragmatic considerations missing from prior work, covering the design process, trade-offs, and technical strategies that go into creating efficient models. Finally, we distill design recommendations for tooling to help ease the difficulty of this work and bring on-device ML into to more widespread practice.
Authors: Aaryan Singhal, Daniele Gammelli, Justin Luke, Karthik Gopalakrishnan, Dominik Helmreich, Marco Pavone
Abstract: Operators of Electric Autonomous Mobility-on-Demand (E-AMoD) fleets need to make several real-time decisions such as matching available vehicles to ride requests, rebalancing idle vehicles to areas of high demand, and charging vehicles to ensure sufficient range. While this problem can be posed as a linear program that optimizes flows over a space-charge-time graph, the size of the resulting optimization problem does not allow for real-time implementation in realistic settings. In this work, we present the E-AMoD control problem through the lens of reinforcement learning and propose a graph network-based framework to achieve drastically improved scalability and superior performance over heuristics. Specifically, we adopt a bi-level formulation where we (1) leverage a graph network-based RL agent to specify a desired next state in the space-charge graph, and (2) solve more tractable linear programs to best achieve the desired state while ensuring feasibility. Experiments using real-world data from San Francisco and New York City show that our approach achieves up to 89% of the profits of the theoretically-optimal solution while achieving more than a 100x speedup in computational time. We further highlight promising zero-shot transfer capabilities of our learned policy on tasks such as inter-city generalization and service area expansion, thus showing the utility, scalability, and flexibility of our framework. Finally, our approach outperforms the best domain-specific heuristics with comparable runtimes, with an increase in profits by up to 3.2x.
Authors: Zhenyu He, Zexuan Zhong, Tianle Cai, Jason D. Lee, Di He
Abstract: We introduce Retrieval-Based Speculative Decoding (REST), a novel algorithm designed to speed up language model generation. The key insight driving the development of REST is the observation that the process of text generation often includes certain common phases and patterns. Unlike previous methods that rely on a draft language model for speculative decoding, REST harnesses the power of retrieval to generate draft tokens. This method draws from the reservoir of existing knowledge, retrieving and employing relevant tokens based on the current context. Its plug-and-play nature allows for seamless integration and acceleration of any language models, all without necessitating additional training. When benchmarked on 7B and 13B language models in a single-batch setting, REST achieves a significant speedup of 1.62X to 2.36X on code or text generation. The code of REST is available at https://github.com/FasterDecoding/REST.
Authors: Yuhan Liu, Shangbin Feng, Xiaochuang Han, Vidhisha Balachandran, Chan Young Park, Sachin Kumar, Yulia Tsvetkov
Abstract: In this work, we take a first step towards designing summarization systems that are faithful to the author's intent, not only the semantic content of the article. Focusing on a case study of preserving political perspectives in news summarization, we find that existing approaches alter the political opinions and stances of news articles in more than 50% of summaries, misrepresenting the intent and perspectives of the news authors. We thus propose P^3SUM, a diffusion model-based summarization approach controlled by political perspective classifiers. In P^3SUM, the political leaning of a generated summary is iteratively evaluated at each decoding step, and any drift from the article's original stance incurs a loss back-propagated to the embedding layers, steering the political stance of the summary at inference time. Extensive experiments on three news summarization datasets demonstrate that P^3SUM outperforms state-of-the-art summarization systems and large language models by up to 13.7% in terms of the success rate of stance preservation, with competitive performance on standard metrics of summarization quality. Our findings present a first analysis of preservation of pragmatic features in summarization, highlight the lacunae in existing summarization models -- that even state-of-the-art models often struggle to preserve author's intents -- and develop new summarization systems that are more faithful to author's perspectives.
Authors: Vasisht Duddu, Sebastian Szyller, N. Asokan
Abstract: Machine learning (ML) models cannot neglect risks to security, privacy, and fairness. Several defenses have been proposed to mitigate such risks. When a defense is effective in mitigating one risk, it may correspond to increased or decreased susceptibility to other risks. Existing research lacks an effective framework to recognize and explain these unintended interactions. We present such a framework, based on the conjecture that overfitting and memorization underlie unintended interactions. We survey existing literature on unintended interactions, accommodating them within our framework. We use our framework to conjecture on two previously unexplored interactions, and empirically validate our conjectures.
Authors: Philipp Altmann, Jonas Stein, Michael K\"olle, Adelina B\"arligea, Thomas Gabor, Thomy Phan, Sebastian Feld, Claudia Linnhoff-Popien
Abstract: Quantum computing (QC) in the current NISQ era is still limited in size and precision. Hybrid applications mitigating those shortcomings are prevalent to gain early insight and advantages. Hybrid quantum machine learning (QML) comprises both the application of QC to improve machine learning (ML) and ML to improve QC architectures. This work considers the latter, leveraging reinforcement learning (RL) to improve the search for viable quantum architectures, which we formalize by a set of generic challenges. Furthermore, we propose a concrete framework, formalized as a Markov decision process, to enable learning policies capable of controlling a universal set of continuously parameterized quantum gates. Finally, we provide benchmark comparisons to assess the shortcomings and strengths of current state-of-the-art RL algorithms.
Authors: Dahyun Kim, Chanjun Park, Sanghoon Kim, Wonsung Lee, Wonho Song, Yunsu Kim, Hyeonwoo Kim, Yungi Kim, Hyeonju Lee, Jihoo Kim, Changbae Ahn, Seonghoon Yang, Sukyung Lee, Hyunbyung Park, Gyoungjin Gim, Mikyoung Cha, Hwalsuk Lee, Sunghun Kim
Abstract: We introduce SOLAR 10.7B, a large language model (LLM) with 10.7 billion parameters, demonstrating superior performance in various natural language processing (NLP) tasks. Inspired by recent efforts to efficiently up-scale LLMs, we present a method for scaling LLMs called depth up-scaling (DUS), which encompasses depthwise scaling and continued pretraining. In contrast to other LLM up-scaling methods that use mixture-of-experts, DUS does not require complex changes to train and inference efficiently. We show experimentally that DUS is simple yet effective in scaling up high-performance LLMs from small ones. Building on the DUS model, we additionally present SOLAR 10.7B-Instruct, a variant fine-tuned for instruction-following capabilities, surpassing Mixtral-8x7B-Instruct. SOLAR 10.7B is publicly available under the Apache 2.0 license, promoting broad access and application in the LLM field.
Authors: Georgios Fatouros, Konstantinos Metaxas, John Soldatos, Dimosthenis Kyriazis
Abstract: This paper introduces MarketSenseAI, an innovative framework leveraging GPT-4's advanced reasoning for selecting stocks in financial markets. By integrating Chain of Thought and In-Context Learning, MarketSenseAI analyzes diverse data sources, including market trends, news, fundamentals, and macroeconomic factors, to emulate expert investment decision-making. The development, implementation, and validation of the framework are elaborately discussed, underscoring its capability to generate actionable and interpretable investment signals. A notable feature of this work is employing GPT-4 both as a predictive mechanism and signal evaluator, revealing the significant impact of the AI-generated explanations on signal accuracy, reliability and acceptance. Through empirical testing on the competitive S&P 100 stocks over a 15-month period, MarketSenseAI demonstrated exceptional performance, delivering excess alpha of 10% to 30% and achieving a cumulative return of up to 72% over the period, while maintaining a risk profile comparable to the broader market. Our findings highlight the transformative potential of Large Language Models in financial decision-making, marking a significant leap in integrating generative AI into financial analytics and investment strategies.
Authors: Meiling Tao, Xuechen Liang, Tianyu Shi, Lei Yu, Yiting Xie
Abstract: This study presents RoleCraft-GLM, an innovative framework aimed at enhancing personalized role-playing with Large Language Models (LLMs). RoleCraft-GLM addresses the key issue of lacking personalized interactions in conversational AI, and offers a solution with detailed and emotionally nuanced character portrayals. We contribute a unique conversational dataset that shifts from conventional celebrity-centric characters to diverse, non-celebrity personas, thus enhancing the realism and complexity of language modeling interactions. Additionally, our approach includes meticulous character development, ensuring dialogues are both realistic and emotionally resonant. The effectiveness of RoleCraft-GLM is validated through various case studies, highlighting its versatility and skill in different scenarios. Our framework excels in generating dialogues that accurately reflect characters' personality traits and emotions, thereby boosting user engagement. In conclusion, RoleCraft-GLM marks a significant leap in personalized AI interactions, and paves the way for more authentic and immersive AI-assisted role-playing experiences by enabling more nuanced and emotionally rich dialogues
Authors: Jianfei Xiao, Yancan Chen, Yimin Ou, Hanyi Yu, Kai Shu, Yiyong Xiao
Abstract: Large language models (LLMs) like Llama, Baichuan and Bloom models show remarkable ability with instruction fine-tuning in many natural language tasks. Nevertheless, for the dialogue summarization task, which aims to generate summaries for different roles in dialogue, most of the state-of-the-art methods conduct on small models (e.g Bart and Bert). Existing methods try to add task specified optimization on small models like adding global-local centrality score to models. In this paper, we propose an instruction fine-tuning model: Baichuan2-Sum, for role-oriented diaglouge summarization. By setting different instructions for different roles, the model can learn from the dialogue interactions and output the expected summaries. Furthermore, we applied NEFTune technique to add suitable noise during training to improve the results. The experiments demonstrate that the proposed model achieves the new state-of-the-art results on two public dialogue summarization datasets: CSDS and SAMSUM. We release our model and related codes to facilitate future studies on dialogue summarization task.
Authors: Timoth\'ee Anne, Jean-Baptiste Mouret
Abstract: Optimizing a set of functions simultaneously by leveraging their similarity is called multi-task optimization. Current black-box multi-task algorithms only solve a finite set of tasks, even when the tasks originate from a continuous space. In this paper, we introduce Parametric-Task MAP-Elites (PT-ME), a new black-box algorithm for continuous multi-task optimization problems. This algorithm (1) solves a new task at each iteration, effectively covering the continuous space, and (2) exploits a new variation operator based on local linear regression. The resulting dataset of solutions makes it possible to create a function that maps any task parameter to its optimal solution. We show that PT-ME outperforms all baselines, including the deep reinforcement learning algorithm PPO on two parametric-task toy problems and a robotic problem in simulation.
Authors: Hoda Torabi, Hamzeh Khazaei, Marin Litoiu
Abstract: With the advent of 5G networks and the rise of the Internet of Things (IoT), Content Delivery Networks (CDNs) are increasingly extending into the network edge. This shift introduces unique challenges, particularly due to the limited cache storage and the diverse request patterns at the edge. These edge environments can host traffic classes characterized by varied object-size distributions and object-access patterns. Such complexity makes it difficult for traditional caching strategies, which often rely on metrics like request frequency or time intervals, to be effective. Despite these complexities, the optimization of edge caching is crucial. Improved byte hit rates at the edge not only alleviate the load on the network backbone but also minimize operational costs and expedite content delivery to end-users. In this paper, we introduce HR-Cache, a comprehensive learning-based caching framework grounded in the principles of Hazard Rate (HR) ordering, a rule originally formulated to compute an upper bound on cache performance. HR-Cache leverages this rule to guide future object eviction decisions. It employs a lightweight machine learning model to learn from caching decisions made based on HR ordering, subsequently predicting the "cache-friendliness" of incoming requests. Objects deemed "cache-averse" are placed into cache as priority candidates for eviction. Through extensive experimentation, we demonstrate that HR-Cache not only consistently enhances byte hit rates compared to existing state-of-the-art methods but also achieves this with minimal prediction overhead. Our experimental results, using three real-world traces and one synthetic trace, indicate that HR-Cache consistently achieves 2.2-14.6% greater WAN traffic savings than LRU. It outperforms not only heuristic caching strategies but also the state-of-the-art learning-based algorithm.
Authors: Xin Chen, Sukanya Kudva, Yongzheng Dai, Anil Aswani, Chen Chen
Abstract: The main challenge with the tensor completion problem is a fundamental tension between computation power and the information-theoretic sample complexity rate. Past approaches either achieve the information-theoretic rate but lack practical algorithms to compute the corresponding solution, or have polynomial-time algorithms that require an exponentially-larger number of samples for low estimation error. This paper develops a novel tensor completion algorithm that resolves this tension by achieving both provable convergence (in numerical tolerance) in a linear number of oracle steps and the information-theoretic rate. Our approach formulates tensor completion as a convex optimization problem constrained using a gauge-based tensor norm, which is defined in a way that allows the use of integer linear optimization to solve linear separation problems over the unit-ball in this new norm. Adaptations based on this insight are incorporated into a Frank-Wolfe variant to build our algorithm. We show our algorithm scales-well using numerical experiments on tensors with up to ten million entries.
Authors: Rachmad Vidya Wicaksana Putra, Muhammad Shafique
Abstract: Autonomous mobile agents (e.g., UAVs and UGVs) are typically expected to incur low power/energy consumption for solving machine learning tasks (such as object recognition), as these mobile agents are usually powered by portable batteries. These requirements can be fulfilled by Spiking Neural Networks (SNNs), since their bio-inspired spike-based operations offer high accuracy and ultra low-power/energy computation. Currently, most of the SNN architectures are derived from Artificial Neural Networks whose neurons' architectures and operations are different from SNNs, or developed without considering memory budgets from the underlying processing hardware of autonomous mobile agents. These limitations hinder SNNs from reaching their full potential in accuracy and efficiency. Toward this, we propose SpikeNAS, a novel fast memory-aware neural architecture search (NAS) framework for SNNs that quickly finds an appropriate SNN architecture with high accuracy under the given memory budgets from autonomous mobile agents. To do this, our SpikeNAS employs several key steps: analyzing the impacts of network operations on the accuracy, enhancing the network architecture to improve the learning quality, and developing a fast memory-aware search algorithm. The experimental results show that our SpikeNAS improves the searching time and maintains high accuracy as compared to state-of-the-art while meeting the given memory budgets (e.g., 4.4x faster search with 1.3% accuracy improvement for CIFAR100, using an Nvidia RTX 6000 Ada GPU machine), thereby quickly providing the appropriate SNN architecture for the memory-constrained autonomous mobile agents.
Authors: Xiaoyue Wang, Jianyou Wang, Weili Cao, Kaicheng Wang, Ramamohan Paturi, Leon Bergen
Abstract: We present the Benchmark of Information Retrieval (IR) tasks with Complex Objectives (BIRCO). BIRCO evaluates the ability of IR systems to retrieve documents given multi-faceted user objectives. The benchmark's complexity and compact size make it suitable for evaluating large language model (LLM)-based information retrieval systems. We present a modular framework for investigating factors that may influence LLM performance on retrieval tasks, and identify a simple baseline model which matches or outperforms existing approaches and more complex alternatives. No approach achieves satisfactory performance on all benchmark tasks, suggesting that stronger models and new retrieval protocols are necessary to address complex user needs.
Authors: Jason Y. Zhang, Amy Lin, Moneish Kumar, Tzu-Hsuan Yang, Deva Ramanan, Shubham Tulsiani
Abstract: Estimating camera poses is a fundamental task for 3D reconstruction and remains challenging given sparsely sampled views (<10). In contrast to existing approaches that pursue top-down prediction of global parametrizations of camera extrinsics, we propose a distributed representation of camera pose that treats a camera as a bundle of rays. This representation allows for a tight coupling with spatial image features improving pose precision. We observe that this representation is naturally suited for set-level transformers and develop a regression-based approach that maps image patches to corresponding rays. To capture the inherent uncertainties in sparse-view pose inference, we adapt this approach to learn a denoising diffusion model which allows us to sample plausible modes while improving performance. Our proposed methods, both regression- and diffusion-based, demonstrate state-of-the-art performance on camera pose estimation on CO3D while generalizing to unseen object categories and in-the-wild captures.
Authors: Jiayuan Su, Jing Luo, Hongwei Wang, Lu Cheng
Abstract: This study aims to address the pervasive challenge of quantifying uncertainty in large language models (LLMs) without logit-access. Conformal Prediction (CP), known for its model-agnostic and distribution-free features, is a desired approach for various LLMs and data distributions. However, existing CP methods for LLMs typically assume access to the logits, which are unavailable for some API-only LLMs. In addition, logits are known to be miscalibrated, potentially leading to degraded CP performance. To tackle these challenges, we introduce a novel CP method that (1) is tailored for API-only LLMs without logit-access; (2) minimizes the size of prediction sets; and (3) ensures a statistical guarantee of the user-defined coverage. The core idea of this approach is to formulate nonconformity measures using both coarse-grained (i.e., sample frequency) and fine-grained uncertainty notions (e.g., semantic similarity). Experimental results on both close-ended and open-ended Question Answering tasks show our approach can mostly outperform the logit-based CP baselines.
Authors: Yichang Xu, Ming Yin, Minghong Fang, Neil Zhenqiang Gong
Abstract: Recent studies have revealed that federated learning (FL), once considered secure due to clients not sharing their private data with the server, is vulnerable to attacks such as client-side training data distribution inference, where a malicious client can recreate the victim's data. While various countermeasures exist, they are not practical, often assuming server access to some training data or knowledge of label distribution before the attack. In this work, we bridge the gap by proposing InferGuard, a novel Byzantine-robust aggregation rule aimed at defending against client-side training data distribution inference attacks. In our proposed InferGuard, the server first calculates the coordinate-wise median of all the model updates it receives. A client's model update is considered malicious if it significantly deviates from the computed median update. We conduct a thorough evaluation of our proposed InferGuard on five benchmark datasets and perform a comparison with ten baseline methods. The results of our experiments indicate that our defense mechanism is highly effective in protecting against client-side training data distribution inference attacks, even against strong adaptive attacks. Furthermore, our method substantially outperforms the baseline methods in various practical FL scenarios.
Authors: Qinghao Hu, Zhisheng Ye, Zerui Wang, Guoteng Wang, Meng Zhang, Qiaoling Chen, Peng Sun, Dahua Lin, Xiaolin Wang, Yingwei Luo, Yonggang Wen, Tianwei Zhang
Abstract: Large Language Models (LLMs) have presented impressive performance across several transformative tasks. However, it is non-trivial to efficiently utilize large-scale cluster resources to develop LLMs, often riddled with numerous challenges such as frequent hardware failures, intricate parallelization strategies, and imbalanced resource utilization. In this paper, we present an in-depth characterization study of a six-month LLM development workload trace collected from our GPU datacenter Acme. Specifically, we investigate discrepancies between LLMs and prior task-specific Deep Learning (DL) workloads, explore resource utilization patterns, and identify the impact of various job failures. Our analysis summarizes hurdles we encountered and uncovers potential opportunities to optimize systems tailored for LLMs. Furthermore, we introduce our system efforts: (1) fault-tolerant pretraining, which enhances fault tolerance through LLM-involved failure diagnosis and automatic recovery. (2) decoupled scheduling for evaluation, which achieves timely performance feedback via trial decomposition and scheduling optimization.
Authors: Zhen Tian, Wayne Xin Zhao, Changwang Zhang, Xin Zhao, Zhongrui Ma, Ji-Rong Wen
Abstract: To capture user preference, transformer models have been widely applied to model sequential user behavior data. The core of transformer architecture lies in the self-attention mechanism, which computes the pairwise attention scores in a sequence. Due to the permutation-equivariant nature, positional encoding is used to enhance the attention between token representations. In this setting, the pairwise attention scores can be derived by both semantic difference and positional difference. However, prior studies often model the two kinds of difference measurements in different ways, which potentially limits the expressive capacity of sequence modeling. To address this issue, this paper proposes a novel transformer variant with complex vector attention, named EulerFormer, which provides a unified theoretical framework to formulate both semantic difference and positional difference. The EulerFormer involves two key technical improvements. First, it employs a new transformation function for efficiently transforming the sequence tokens into polar-form complex vectors using Euler's formula, enabling the unified modeling of both semantic and positional information in a complex rotation form.Secondly, it develops a differential rotation mechanism, where the semantic rotation angles can be controlled by an adaptation function, enabling the adaptive integration of the semantic and positional information according to the semantic contexts.Furthermore, a phase contrastive learning task is proposed to improve the isotropy of contextual representations in EulerFormer. Our theoretical framework possesses a high degree of completeness and generality. It is more robust to semantic variations and possesses moresuperior theoretical properties in principle. Extensive experiments conducted on four public datasets demonstrate the effectiveness and efficiency of our approach.
Authors: Jerry Wei, Chengrun Yang, Xinying Song, Yifeng Lu, Nathan Hu, Jie Huang, Dustin Tran, Daiyi Peng, Ruibo Liu, Da Huang, Cosmo Du, Quoc V. Le
Abstract: Large language models (LLMs) often generate content that contains factual errors when responding to fact-seeking prompts on open-ended topics. To benchmark a model's long-form factuality in open domains, we first use GPT-4 to generate LongFact, a prompt set comprising thousands of questions spanning 38 topics. We then propose that LLM agents can be used as automated evaluators for long-form factuality through a method which we call Search-Augmented Factuality Evaluator (SAFE). SAFE utilizes an LLM to break down a long-form response into a set of individual facts and to evaluate the accuracy of each fact using a multi-step reasoning process comprising sending search queries to Google Search and determining whether a fact is supported by the search results. Furthermore, we propose extending F1 score as an aggregated metric for long-form factuality. To do so, we balance the percentage of supported facts in a response (precision) with the percentage of provided facts relative to a hyperparameter representing a user's preferred response length (recall). Empirically, we demonstrate that LLM agents can outperform crowdsourced human annotators - on a set of ~16k individual facts, SAFE agrees with crowdsourced human annotators 72% of the time, and on a random subset of 100 disagreement cases, SAFE wins 76% of the time. At the same time, SAFE is more than 20 times cheaper than human annotators. We also benchmark thirteen language models on LongFact across four model families (Gemini, GPT, Claude, and PaLM-2), finding that larger language models generally achieve better long-form factuality. LongFact, SAFE, and all experimental code are available at https://github.com/google-deepmind/long-form-factuality.
URLs: https://github.com/google-deepmind/long-form-factuality.
Authors: Dmitrii Zhemchuzhnikov, Sergei Grudinin
Abstract: Effective recognition of spatial patterns and learning their hierarchy is crucial in modern spatial data analysis. Volumetric data applications seek techniques ensuring invariance not only to shifts but also to pattern rotations. While traditional methods can readily achieve translational invariance, rotational invariance possesses multiple challenges and remains an active area of research. Here, we present ILPO-Net (Invariant to Local Patterns Orientation Network), a novel approach that handles arbitrarily shaped patterns with the convolutional operation inherently invariant to local spatial pattern orientations using the Wigner matrix expansions. Our architecture seamlessly integrates the new convolution operator and, when benchmarked on diverse volumetric datasets such as MedMNIST and CATH, demonstrates superior performance over the baselines with significantly reduced parameter counts - up to 1000 times fewer in the case of MedMNIST. Beyond these demonstrations, ILPO-Net's rotational invariance paves the way for other applications across multiple disciplines. Our code is publicly available at https://gricad-gitlab.univ-grenoble-alpes.fr/GruLab/ILPONet.
URLs: https://gricad-gitlab.univ-grenoble-alpes.fr/GruLab/ILPONet.
Authors: Tenghao Huang, Dongwon Jung, Muhao Chen
Abstract: Recent advancements in integrating external tools with Large Language Models (LLMs) have opened new frontiers, with applications in mathematical reasoning, code generators, and smart assistants. However, existing methods, relying on simple one-time retrieval strategies, fall short on effectively and accurately shortlisting relevant tools. This paper introduces a novel PLUTO (Planning, Learning, and Understanding for TOols) approach, encompassing `Plan-and-Retrieve (P&R)` and `Edit-and-Ground (E&G)` paradigms. The P&R paradigm consists of a neural retrieval module for shortlisting relevant tools and an LLM-based query planner that decomposes complex queries into actionable tasks, enhancing the effectiveness of tool utilization. The E&G paradigm utilizes LLMs to enrich tool descriptions based on user scenarios, bridging the gap between user queries and tool functionalities. Experiment results demonstrate that these paradigms significantly improve the recall and NDCG in tool retrieval tasks, significantly surpassing current state-of-the-art models.
Authors: Qi Zhang, Yi Zhou, Shaofeng Zou
Abstract: This paper provides the first tight convergence analyses for RMSProp and Adam in non-convex optimization under the most relaxed assumptions of coordinate-wise generalized smoothness and affine noise variance. We first analyze RMSProp, which is a special case of Adam with adaptive learning rates but without first-order momentum. Specifically, to solve the challenges due to dependence among adaptive update, unbounded gradient estimate and Lipschitz constant, we demonstrate that the first-order term in the descent lemma converges and its denominator is upper bounded by a function of gradient norm. Based on this result, we show that RMSProp with proper hyperparameters converges to an $\epsilon$-stationary point with an iteration complexity of $\mathcal O(\epsilon^{-4})$. We then generalize our analysis to Adam, where the additional challenge is due to a mismatch between the gradient and first-order momentum. We develop a new upper bound on the first-order term in the descent lemma, which is also a function of the gradient norm. We show that Adam with proper hyperparameters converges to an $\epsilon$-stationary point with an iteration complexity of $\mathcal O(\epsilon^{-4})$. Our complexity results for both RMSProp and Adam match with the complexity lower bound established in \cite{arjevani2023lower}.
Authors: Ming Xu, Stephen Gould
Abstract: We propose a novel approach to the action segmentation task for long, untrimmed videos, based on solving an optimal transport problem. By encoding a temporal consistency prior into a Gromov-Wasserstein problem, we are able to decode a temporally consistent segmentation from a noisy affinity/matching cost matrix between video frames and action classes. Unlike previous approaches, our method does not require knowing the action order for a video to attain temporal consistency. Furthermore, our resulting (fused) Gromov-Wasserstein problem can be efficiently solved on GPUs using a few iterations of projected mirror descent. We demonstrate the effectiveness of our method in an unsupervised learning setting, where our method is used to generate pseudo-labels for self-training. We evaluate our segmentation approach and unsupervised learning pipeline on the Breakfast, 50-Salads, YouTube Instructions and Desktop Assembly datasets, yielding state-of-the-art results for the unsupervised video action segmentation task.
Authors: Rachmad Vidya Wicaksana Putra, Muhammad Shafique
Abstract: Spiking Neural Networks (SNNs) can offer ultra low power/ energy consumption for machine learning-based applications due to their sparse spike-based operations. Currently, most of the SNN architectures need a significantly larger model size to achieve higher accuracy, which is not suitable for resource-constrained embedded applications. Therefore, developing SNNs that can achieve high accuracy with acceptable memory footprint is highly needed. Toward this, we propose a novel methodology that improves the accuracy of SNNs through kernel size scaling. Its key steps include investigating the impact of different kernel sizes on the accuracy, devising new sets of kernel sizes, generating SNN architectures based on the selected kernel sizes, and analyzing the accuracy-memory trade-offs for SNN model selection. The experimental results show that our methodology achieves higher accuracy than state-of-the-art (93.24% accuracy for CIFAR10 and 70.84% accuracy for CIFAR100) with less than 10M parameters and up to 3.45x speed-up of searching time, thereby making it suitable for embedded applications.
Authors: Jinbae Im, JeongYeon Nam, Nokyung Park, Hyungmin Lee, Seunghyun Park
Abstract: Scene Graph Generation (SGG) is a challenging task of detecting objects and predicting relationships between objects. After DETR was developed, one-stage SGG models based on a one-stage object detector have been actively studied. However, complex modeling is used to predict the relationship between objects, and the inherent relationship between object queries learned in the multi-head self-attention of the object detector has been neglected. We propose a lightweight one-stage SGG model that extracts the relation graph from the various relationships learned in the multi-head self-attention layers of the DETR decoder. By fully utilizing the self-attention by-products, the relation graph can be extracted effectively with a shallow relation extraction head. Considering the dependency of the relation extraction task on the object detection task, we propose a novel relation smoothing technique that adjusts the relation label adaptively according to the quality of the detected objects. By the relation smoothing, the model is trained according to the continuous curriculum that focuses on object detection task at the beginning of training and performs multi-task learning as the object detection performance gradually improves. Furthermore, we propose a connectivity prediction task that predicts whether a relation exists between object pairs as an auxiliary task of the relation extraction. We demonstrate the effectiveness and efficiency of our method for the Visual Genome and Open Image V6 datasets. Our code is publicly available at https://github.com/naver-ai/egtr.
Authors: Maike Behrendt, Stefan Sylvius Wagner, Marc Ziegele, Lena Wilms, Anke Stoll, Dominique Heinbach, Stefan Harmeling
Abstract: Measuring the quality of contributions in political online discussions is crucial in deliberation research and computer science. Research has identified various indicators to assess online discussion quality, and with deep learning advancements, automating these measures has become feasible. While some studies focus on analyzing specific quality indicators, a comprehensive quality score incorporating various deliberative aspects is often preferred. In this work, we introduce AQuA, an additive score that calculates a unified deliberative quality score from multiple indices for each discussion post. Unlike other singular scores, AQuA preserves information on the deliberative aspects present in comments, enhancing model transparency. We develop adapter models for 20 deliberative indices, and calculate correlation coefficients between experts' annotations and the perceived deliberativeness by non-experts to weigh the individual indices into a single deliberative score. We demonstrate that the AQuA score can be computed easily from pre-trained adapters and aligns well with annotations on other datasets that have not be seen during training. The analysis of experts' vs. non-experts' annotations confirms theoretical findings in the social science literature.
Authors: Matyas Juhasz, Tina Marchand, Roshan Melwani, Kalyan Dutia, Sarah Goodenough, Harrison Pim, Henry Franks
Abstract: Quantified policy targets are a fundamental element of climate policy, typically characterised by domain-specific and technical language. Current methods for curating comprehensive views of global climate policy targets entail significant manual effort. At present there are few scalable methods for extracting climate targets from national laws or policies, which limits policymakers' and researchers' ability to (1) assess private and public sector alignment with global goals and (2) inform policy decisions. In this paper we present an approach for extracting mentions of climate targets from national laws and policies. We create an expert-annotated dataset identifying three categories of target ('Net Zero', 'Reduction' and 'Other' (e.g. renewable energy targets)) and train a classifier to reliably identify them in text. We investigate bias and equity impacts related to our model and identify specific years and country names as problematic features. Finally, we investigate the characteristics of the dataset produced by running this classifier on the Climate Policy Radar (CPR) dataset of global national climate laws and policies and UNFCCC submissions, highlighting the potential of automated and scalable data collection for existing climate policy databases and supporting further research. Our work represents a significant upgrade in the accessibility of these key climate policy elements for policymakers and researchers. We publish our model at https://huggingface.co/ClimatePolicyRadar/national-climate-targets and related dataset at https://huggingface.co/datasets/ClimatePolicyRadar/national-climate-targets.
URLs: https://huggingface.co/ClimatePolicyRadar/national-climate-targets, https://huggingface.co/datasets/ClimatePolicyRadar/national-climate-targets.