Authors: Siwei Tu, Ben Fei, Weidong Yang, Fenghua Ling, Hao Chen, Zili Liu, Kun Chen, Hang Fan, Wanli Ouyang, Lei Bai
Abstract: Accurate acquisition of surface meteorological conditions at arbitrary locations holds significant importance for weather forecasting and climate simulation. Due to the fact that meteorological states derived from satellite observations are often provided in the form of low-resolution grid fields, the direct application of spatial interpolation to obtain meteorological states for specific locations often results in significant discrepancies when compared to actual observations. Existing downscaling methods for acquiring meteorological state information at higher resolutions commonly overlook the correlation with satellite observations. To bridge the gap, we propose Satellite-observations Guided Diffusion Model (SGD), a conditional diffusion model pre-trained on ERA5 reanalysis data with satellite observations (GridSat) as conditions, which is employed for sampling downscaled meteorological states through a zero-shot guided sampling strategy and patch-based methods. During the training process, we propose to fuse the information from GridSat satellite observations into ERA5 maps via the attention mechanism, enabling SGD to generate atmospheric states that align more accurately with actual conditions. In the sampling, we employed optimizable convolutional kernels to simulate the upscale process, thereby generating high-resolution ERA5 maps using low-resolution ERA5 maps as well as observations from weather stations as guidance. Moreover, our devised patch-based method promotes SGD to generate meteorological states at arbitrary resolutions. Experiments demonstrate SGD fulfills accurate meteorological states downscaling to 6.25km.
Authors: Mark Sch\"one, Babak Rahmani, Heiner Kremer, Fabian Falck, Hitesh Ballani, Jannes Gladrow
Abstract: State-space models (SSMs) and transformers dominate the language modeling landscape. However, they are constrained to a lower computational complexity than classical recurrent neural networks (RNNs), limiting their expressivity. In contrast, RNNs lack parallelization during training, raising fundamental questions about the trade off between parallelization and expressivity. We propose implicit SSMs, which iterate a transformation until convergence to a fixed point. Theoretically, we show that implicit SSMs implement the non-linear state-transitions of RNNs. Empirically, we find that only approximate fixed-point convergence suffices, enabling the design of a scalable training curriculum that largely retains parallelization, with full convergence required only for a small subset of tokens. Our approach demonstrates superior state-tracking capabilities on regular languages, surpassing transformers and SSMs. We further scale implicit SSMs to natural language reasoning tasks and pretraining of large-scale language models up to 1.3B parameters on 207B tokens - representing, to our knowledge, the largest implicit model trained to date. Notably, our implicit models outperform their explicit counterparts on standard benchmarks.
Authors: Yiping Wang, Hanxian Huang, Yifang Chen, Jishen Zhao, Simon Shaolei Du, Yuandong Tian
Abstract: While Large language models (LLMs) have advanced natural language processing tasks, their growing computational and memory demands make deployment on resource-constrained devices like mobile phones increasingly challenging. In this paper, we propose SHARP (SHaring Adjacent Layers with Recovery Parameters), a novel approach to accelerate LLM inference by sharing parameters across adjacent layers, thus reducing memory load overhead, while introducing low-rank recovery parameters to maintain performance. Inspired by observations that consecutive layers have similar outputs, SHARP employs a two-stage recovery process: Single Layer Warmup (SLW), and Supervised Fine-Tuning (SFT). The SLW stage aligns the outputs of the shared layers using L_2 loss, providing a good initialization for the following SFT stage to further restore the model performance. Extensive experiments demonstrate that SHARP can recover the model's perplexity on various in-distribution tasks using no more than 50k fine-tuning data while reducing the number of stored MLP parameters by 38% to 65%. We also conduct several ablation studies of SHARP and show that replacing layers towards the later parts of the model yields better performance retention, and that different recovery parameterizations perform similarly when parameter counts are matched. Furthermore, SHARP saves 42.8% in model storage and reduces the total inference time by 42.2% compared to the original Llama2-7b model on mobile devices. Our results highlight SHARP as an efficient solution for reducing inference costs in deploying LLMs without the need for pretraining-scale resources.
Authors: Chae-Won Lee, Jong-Seok Lee
Abstract: In recent studies of emotional EEG classification, connectivity matrices have been successfully employed as input to convolutional neural networks (CNNs), which can effectively consider inter-regional interaction patterns in EEG. However, we find that such an approach has a limitation that important patterns in connectivity matrices may be lost during the convolutional operations in CNNs. To resolve this issue, we propose and validate an idea to upscale the connectivity matrices to strengthen the local patterns. Experimental results demonstrate that this simple idea can significantly enhance the classification performance.
Authors: Krunoslav Lehman Pavasovic, Jakob Verbeek, Giulio Biroli, Marc Mezard
Abstract: Recent studies have raised concerns about the effectiveness of Classifier-Free Guidance (CFG), indicating that in low-dimensional settings, it can lead to overshooting the target distribution and reducing sample diversity. In this work, we demonstrate that in infinite and sufficiently high-dimensional contexts CFG effectively reproduces the target distribution, revealing a blessing-of-dimensionality result. Additionally, we explore finite-dimensional effects, precisely characterizing overshoot and variance reduction. Based on our analysis, we introduce non-linear generalizations of CFG. Through numerical simulations on Gaussian mixtures and experiments on class-conditional and text-to-image diffusion models, we validate our analysis and show that our non-linear CFG offers improved flexibility and generation quality without additional computation cost.
Authors: Adithya Ramachandran, Thorkil Flensmark B. Neergaard, Andreas Maier, Siming Bayer
Abstract: Global leaders and policymakers are unified in their unequivocal commitment to decarbonization efforts in support of Net-Zero agreements. District Heating Systems (DHS), while contributing to carbon emissions due to the continued reliance on fossil fuels for heat production, are embracing more sustainable practices albeit with some sense of vulnerability as it could constrain their ability to adapt to dynamic demand and production scenarios. As demographic demands grow and renewables become the central strategy in decarbonizing the heating sector, the need for accurate demand forecasting has intensified. Advances in digitization have paved the way for Machine Learning (ML) based solutions to become the industry standard for modeling complex time series patterns. In this paper, we focus on building a Deep Learning (DL) model that uses deconstructed components of independent and dependent variables that affect heat demand as features to perform multi-step ahead forecasting of head demand. The model represents the input features in a time-frequency space and uses an attention mechanism to generate accurate forecasts. The proposed method is evaluated on a real-world dataset and the forecasting performance is assessed against LSTM and CNN-based forecasting models. Across different supply zones, the attention-based models outperforms the baselines quantitatively and qualitatively, with an Mean Absolute Error (MAE) of 0.105 with a standard deviation of 0.06kW h and a Mean Absolute Percentage Error (MAPE) of 5.4% with a standard deviation of 2.8%, in comparison the second best model with a MAE of 0.10 with a standard deviation of 0.06kW h and a MAPE of 5.6% with a standard deviation of 3%.
Authors: Abdellah Zakaria Sellam, Ilyes Benaissa, Abdelmalik Taleb-Ahmed, Luigi Patrono, Cosimo Distante
Abstract: Anomaly detection in time series is essential for industrial monitoring and environmental sensing, yet distinguishing anomalies from complex patterns remains challenging. Existing methods like the Anomaly Transformer and DCdetector have progressed, but they face limitations such as sensitivity to short-term contexts and inefficiency in noisy, non-stationary environments. To overcome these issues, we introduce MAAT, an improved architecture that enhances association discrepancy modeling and reconstruction quality. MAAT features Sparse Attention, efficiently capturing long-range dependencies by focusing on relevant time steps, thereby reducing computational redundancy. Additionally, a Mamba-Selective State Space Model is incorporated into the reconstruction module, utilizing a skip connection and Gated Attention to improve anomaly localization and detection performance. Extensive experiments show that MAAT significantly outperforms previous methods, achieving better anomaly distinguishability and generalization across various time series applications, setting a new standard for unsupervised time series anomaly detection in real-world scenarios.
Authors: Insu Han, Michael Kapralov, Ekaterina Kochetkova, Kshiteej Sheth, Amir Zandieh
Abstract: Large language models (LLMs) have achieved impressive success, but their high memory requirements present challenges for long-context token generation. The memory complexity of long-context LLMs is primarily due to the need to store Key-Value (KV) embeddings in their KV cache. We present BalanceKV, a KV cache compression method based on geometric sampling process stemming from Banaszczyk's vector balancing theory, which introduces dependencies informed by the geometry of keys and value tokens, and improves precision. BalanceKV offers both theoretically proven and empirically validated performance improvements over existing methods.
Authors: Jason Wu, Kang Yang, Lance Kaplan, Mani Srivastava
Abstract: Multimodal deep learning systems are deployed in dynamic scenarios due to the robustness afforded by multiple sensing modalities. Nevertheless, they struggle with varying compute resource availability (due to multi-tenancy, device heterogeneity, etc.) and fluctuating quality of inputs (from sensor feed corruption, environmental noise, etc.). Current multimodal systems employ static resource provisioning and cannot easily adapt when compute resources change over time. Additionally, their reliance on processing sensor data with fixed feature extractors is ill-equipped to handle variations in modality quality. Consequently, uninformative modalities, such as those with high noise, needlessly consume resources better allocated towards other modalities. We propose ADMN, a layer-wise Adaptive Depth Multimodal Network capable of tackling both challenges - it adjusts the total number of active layers across all modalities to meet compute resource constraints, and continually reallocates layers across input modalities according to their modality quality. Our evaluations showcase ADMN can match the accuracy of state-of-the-art networks while reducing up to 75% of their floating-point operations.
Authors: Fanxu Meng, Zengwei Yao, Muhan Zhang
Abstract: Modern large language models (LLMs) often encounter communication bottlenecks on current hardware, rather than purely computational constraints. Multi-head Latent Attention (MLA) tackles this challenge by using low-rank matrices in the key-value (KV) layers, thereby allowing compressed latent KV states to be cached. This approach significantly reduces the KV cache size relative to traditional multi-head attention, leading to faster inference. Moreover, MLA employs an up-projection matrix to increase expressiveness, trading additional computation for reduced communication overhead. Although MLA has demonstrated efficiency and effectiveness in Deepseek V2/V3/R1, many major model providers still rely on Group Query Attention (GQA) and have not announced any plans to adopt MLA. In this paper, we show that GQA can always be represented by MLA while maintaining the same KV cache overhead, but the converse does not hold. To encourage broader use of MLA, we introduce **TransMLA**, a post-training method that converts widely used GQA-based pre-trained models (e.g., LLaMA, Qwen, Mixtral) into MLA-based models. After conversion, the model can undergo additional training to boost expressiveness without increasing the KV cache size. Furthermore, we plan to develop MLA-specific inference acceleration techniques to preserve low latency in transformed models, thus enabling more efficient distillation of Deepseek R1.
Authors: Xuefeng Liu, Hung T. C. Le, Siyu Chen, Rick Stevens, Zhuoran Yang, Matthew R. Walter, Yuxin Chen
Abstract: Online reinforcement learning (RL) enhances policies through direct interactions with the environment, but faces challenges related to sample efficiency. In contrast, offline RL leverages extensive pre-collected data to learn policies, but often produces suboptimal results due to limited data coverage. Recent efforts have sought to integrate offline and online RL in order to harness the advantages of both approaches. However, effectively combining online and offline RL remains challenging due to issues that include catastrophic forgetting, lack of robustness and sample efficiency. In an effort to address these challenges, we introduce A3 RL , a novel method that actively selects data from combined online and offline sources to optimize policy improvement. We provide theoretical guarantee that validates the effectiveness our active sampling strategy and conduct thorough empirical experiments showing that our method outperforms existing state-of-the-art online RL techniques that utilize offline data. Our code will be publicly available at: https://github.com/xuefeng-cs/A3RL.
Authors: Qingyuan Wu, Jianheng Liu, Jianye Hao, Jun Wang, Kun Shao
Abstract: State-of-the-art (SOTA) reinforcement learning (RL) methods enable the vision-language agents to learn from interactions with the environment without human supervision. However, they struggle with learning inefficiencies in tackling real-world complex sequential decision-making tasks, especially with sparse reward signals and long-horizon dependencies. To effectively address the issue, we introduce Variational Subgoal-Conditioned RL (VSC-RL), which reformulates the vision-language sequential decision-making task as a variational goal-conditioned RL problem, allowing us to leverage advanced optimization methods to enhance learning efficiency. Specifically, VSC-RL optimizes the SubGoal Evidence Lower BOund (SGC-ELBO), which consists of (a) maximizing the subgoal-conditioned return via RL and (b) minimizing the subgoal-conditioned difference with the reference policy. We theoretically demonstrate that SGC-ELBO is equivalent to the original optimization objective, ensuring improved learning efficiency without sacrificing performance guarantees. Additionally, for real-world complex decision-making tasks, VSC-RL leverages the vision-language model to autonomously decompose the goal into feasible subgoals, enabling efficient learning. Across various benchmarks, including challenging real-world mobile device control tasks, VSC-RL significantly outperforms the SOTA vision-language agents, achieving superior performance and remarkable improvement in learning efficiency.
Authors: Ashkan Shahbazi, Elaheh Akbari, Darian Salehi, Xinran Liu, Navid Naderializadeh, Soheil Kolouri
Abstract: While self-attention has been instrumental in the success of Transformers, it can lead to over-concentration on a few tokens during training, resulting in suboptimal information flow. Enforcing doubly-stochastic constraints in attention matrices has been shown to improve structure and balance in attention distributions. However, existing methods rely on iterative Sinkhorn normalization, which is computationally costly. In this paper, we introduce a novel, fully parallelizable doubly-stochastic attention mechanism based on sliced optimal transport, leveraging Expected Sliced Transport Plans (ESP). Unlike prior approaches, our method enforces double stochasticity without iterative Sinkhorn normalization, significantly enhancing efficiency. To ensure differentiability, we incorporate a temperature-based soft sorting technique, enabling seamless integration into deep learning models. Experiments across multiple benchmark datasets, including image classification, point cloud classification, sentiment analysis, and neural machine translation, demonstrate that our enhanced attention regularization consistently improves performance across diverse applications.
Authors: Anthony D. Blaom, Samuel Okon
Abstract: A new computational tool TumorGrowth.jl for modeling tumor growth is introduced. The tool allows the comparison of standard textbook models, such as General Bertalanffy and Gompertz, with some newer models, including, for the first time, neural ODE models. As an application, we revisit a human meta-study of non-small cell lung cancer and bladder cancer lesions, in patients undergoing two different treatment options, to determine if previously reported performance differences are statistically significant, and if newer, more complex models perform any better. In a population of examples with at least four time-volume measurements available for calibration, and an average of about 6.3, our main conclusion is that the General Bertalanffy model has superior performance, on average. However, where more measurements are available, we argue that more complex models, capable of capturing rebound and relapse behavior, may be better choices.
Authors: Song Wang, Zhen Tan, Yaochen Zhu, Chuxu Zhang, Jundong Li
Abstract: Out-of-distribution (OOD) generalization on graphs aims at dealing with scenarios where the test graph distribution differs from the training graph distributions. Compared to i.i.d. data like images, the OOD generalization problem on graph-structured data remains challenging due to the non-i.i.d. property and complex structural information on graphs. Recently, several works on graph OOD generalization have explored extracting invariant subgraphs that share crucial classification information across different distributions. Nevertheless, such a strategy could be suboptimal for entirely capturing the invariant information, as the extraction of discrete structures could potentially lead to the loss of invariant information or the involvement of spurious information. In this paper, we propose an innovative framework, named Generative Risk Minimization (GRM), designed to generate an invariant subgraph for each input graph to be classified, instead of extraction. To address the challenge of optimization in the absence of optimal invariant subgraphs (i.e., ground truths), we derive a tractable form of the proposed GRM objective by introducing a latent causal variable, and its effectiveness is validated by our theoretical analysis. We further conduct extensive experiments across a variety of real-world graph datasets for both node-level and graph-level OOD generalization, and the results demonstrate the superiority of our framework GRM.
Authors: Cheng Fang, Rishabh Dixit, Waheed U. Bajwa, Mert Gurbuzbalaban
Abstract: Empirical risk minimization (ERM) is a cornerstone of modern machine learning (ML), supported by advances in optimization theory that ensure efficient solutions with provable algorithmic convergence rates, which measure the speed at which optimization algorithms approach a solution, and statistical learning rates, which characterize how well the solution generalizes to unseen data. Privacy, memory, computational, and communications constraints increasingly necessitate data collection, processing, and storage across network-connected devices. In many applications, these networks operate in decentralized settings where a central server cannot be assumed, requiring decentralized ML algorithms that are both efficient and resilient. Decentralized learning, however, faces significant challenges, including an increased attack surface for adversarial interference during decentralized learning processes. This paper focuses on the man-in-the-middle (MITM) attack, which can cause models to deviate significantly from their intended ERM solutions. To address this challenge, we propose RESIST (Resilient dEcentralized learning using conSensus gradIent deScenT), an optimization algorithm designed to be robust against adversarially compromised communication links. RESIST achieves algorithmic and statistical convergence for strongly convex, Polyak-Lojasiewicz, and nonconvex ERM problems. Experimental results demonstrate the robustness and scalability of RESIST for real-world decentralized learning in adversarial environments.
Authors: Amir Moeini, Jiuqi Wang, Jacob Beck, Ethan Blaser, Shimon Whiteson, Rohan Chandra, Shangtong Zhang
Abstract: Reinforcement learning (RL) agents typically optimize their policies by performing expensive backward passes to update their network parameters. However, some agents can solve new tasks without updating any parameters by simply conditioning on additional context such as their action-observation histories. This paper surveys work on such behavior, known as in-context reinforcement learning.
Authors: Lejla Skelic, Yan Xu, Matthew Cox, Wenjie Lu, Tao Yu, Ruonan Han
Abstract: The role of Large Language Models (LLMs) has not been extensively explored in analog circuit design, which could benefit from a reasoning-based approach that transcends traditional optimization techniques. In particular, despite their growing relevance, there are no benchmarks to assess LLMs' reasoning capability about circuits. Therefore, we created the CIRCUIT dataset consisting of 510 question-answer pairs spanning various levels of analog-circuit-related subjects. The best-performing model on our dataset, GPT-4o, achieves 48.04% accuracy when evaluated on the final numerical answer. To evaluate the robustness of LLMs on our dataset, we introduced a unique feature that enables unit-test-like evaluation by grouping questions into unit tests. In this case, GPT-4o can only pass 27.45% of the unit tests, highlighting that the most advanced LLMs still struggle with understanding circuits, which requires multi-level reasoning, particularly when involving circuit topologies. This circuit-specific benchmark highlights LLMs' limitations, offering valuable insights for advancing their application in analog integrated circuit design.
Authors: Han Gao, Sebastian Kaltenbach, Petros Koumoutsakos
Abstract: Modeling and simulation of complex fluid flows with dynamics that span multiple spatio-temporal scales is a fundamental challenge in many scientific and engineering domains. Full-scale resolving simulations for systems such as highly turbulent flows are not feasible in the foreseeable future, and reduced-order models must capture dynamics that involve interactions across scales. In the present work, we propose a novel framework, Graph-based Learning of Effective Dynamics (Graph-LED), that leverages graph neural networks (GNNs), as well as an attention-based autoregressive model, to extract the effective dynamics from a small amount of simulation data. GNNs represent flow fields on unstructured meshes as graphs and effectively handle complex geometries and non-uniform grids. The proposed method combines a GNN based, dimensionality reduction for variable-size unstructured meshes with an autoregressive temporal attention model that can learn temporal dependencies automatically. We evaluated the proposed approach on a suite of fluid dynamics problems, including flow past a cylinder and flow over a backward-facing step over a range of Reynolds numbers. The results demonstrate robust and effective forecasting of spatio-temporal physics; in the case of the flow past a cylinder, both small-scale effects that occur close to the cylinder as well as its wake are accurately captured.
Authors: Clarissa Lauditi, Blake Bordelon, Cengiz Pehlevan
Abstract: Previous influential work showed that infinite width limits of neural networks in the lazy training regime are described by kernel machines. Here, we show that neural networks trained in the rich, feature learning infinite-width regime in two different settings are also described by kernel machines, but with data-dependent kernels. For both cases, we provide explicit expressions for the kernel predictors and prescriptions to numerically calculate them. To derive the first predictor, we study the large-width limit of feature-learning Bayesian networks, showing how feature learning leads to task-relevant adaptation of layer kernels and preactivation densities. The saddle point equations governing this limit result in a min-max optimization problem that defines the kernel predictor. To derive the second predictor, we study gradient flow training of randomly initialized networks trained with weight decay in the infinite-width limit using dynamical mean field theory (DMFT). The fixed point equations of the arising DMFT defines the task-adapted internal representations and the kernel predictor. We compare our kernel predictors to kernels derived from lazy regime and demonstrate that our adaptive kernels achieve lower test loss on benchmark datasets.
Authors: Mengfan Xu, Liren Shan, Fatemeh Ghaffari, Xuchuang Wang, Xutong Liu, Mohammad Hajiesmaili
Abstract: We study a novel heterogeneous multi-agent multi-armed bandit problem with a cluster structure induced by stochastic block models, influencing not only graph topology, but also reward heterogeneity. Specifically, agents are distributed on random graphs based on stochastic block models - a generalized Erdos-Renyi model with heterogeneous edge probabilities: agents are grouped into clusters (known or unknown); edge probabilities for agents within the same cluster differ from those across clusters. In addition, the cluster structure in stochastic block model also determines our heterogeneous rewards. Rewards distributions of the same arm vary across agents in different clusters but remain consistent within a cluster, unifying homogeneous and heterogeneous settings and varying degree of heterogeneity, and rewards are independent samples from these distributions. The objective is to minimize system-wide regret across all agents. To address this, we propose a novel algorithm applicable to both known and unknown cluster settings. The algorithm combines an averaging-based consensus approach with a newly introduced information aggregation and weighting technique, resulting in a UCB-type strategy. It accounts for graph randomness, leverages both intra-cluster (homogeneous) and inter-cluster (heterogeneous) information from rewards and graphs, and incorporates cluster detection for unknown cluster settings. We derive optimal instance-dependent regret upper bounds of order $\log{T}$ under sub-Gaussian rewards. Importantly, our regret bounds capture the degree of heterogeneity in the system (an additional layer of complexity), exhibit smaller constants, scale better for large systems, and impose significantly relaxed assumptions on edge probabilities. In contrast, prior works have not accounted for this refined problem complexity, rely on more stringent assumptions, and exhibit limited scalability.
Authors: Mingtian Zhang, Jiajun He, Wenlin Chen, Zijing Ou, Jos\'e Miguel Hern\'andez-Lobato, Bernhard Sch\"olkopf, David Barber
Abstract: Recent advances in one-step generative models typically follow a two-stage process: first training a teacher diffusion model and then distilling it into a one-step student model. This distillation process traditionally relies on both the teacher model's score function to compute the distillation loss and its weights for student initialization. In this paper, we explore whether one-step generative models can be trained directly without this distillation process. First, we show that the teacher's score function is not essential and propose a family of distillation methods that achieve competitive results without relying on score estimation. Next, we demonstrate that initialization from teacher weights is indispensable in successful training. Surprisingly, we find that this benefit is not due to improved ``input-output" mapping but rather the learned feature representations, which dominate distillation quality. Our findings provide a better understanding of the role of initialization in one-step model training and its impact on distillation quality.
Authors: Zander W. Blasingame, Chen Liu
Abstract: Training-free guided generation is a widely used and powerful technique that allows the end user to exert further control over the generative process of diffusion models. In this work, we explore the guided generation from the perspective of optimizing the solution trajectory of a neural differential equation in a greedy manner. We present such a strategy as a unifying view on training-free guidance by showing that the greedy strategy is a first-order discretization of end-to-end optimization techniques. We show that a greedy guidance strategy makes good decisions and compare it to a guidance strategy using the ideal gradients found via the continuous adjoint equations. We then show how other popular training-free guidance strategies can be viewed in a unified manner from this perspective.
Authors: Max Hopkins, Shay Moran
Abstract: Stability is a central property in learning and statistics promising the output of an algorithm $A$ does not change substantially when applied to similar datasets $S$ and $S'$. It is an elementary fact that any sufficiently stable algorithm (e.g.\ one returning the same result with high probability, satisfying privacy guarantees, etc.) must be randomized. This raises a natural question: can we quantify how much randomness is needed for algorithmic stability? We study the randomness complexity of two influential notions of stability in learning: replicability, which promises $A$ usually outputs the same result when run over samples from the same distribution (and shared random coins), and differential privacy, which promises the output distribution of $A$ remains similar under neighboring datasets. The randomness complexity of these notions was studied recently in (Dixon et al. ICML 2024) and (Cannone et al. ITCS 2024) for basic $d$-dimensional tasks (e.g. estimating the bias of $d$ coins), but little is known about the measures more generally or in complex settings like classification. Toward this end, we prove a `weak-to-strong' boosting theorem for stability: the randomness complexity of a task $M$ (either under replicability or DP) is tightly controlled by the best replication probability of any deterministic algorithm solving the task, a weak measure called `global stability' that is universally capped at $\frac{1}{2}$ (Chase et al. FOCS 2023). Using this, we characterize the randomness complexity of PAC Learning: a class has bounded randomness complexity iff it has finite Littlestone dimension, and moreover scales at worst logarithmically in the excess error of the learner. This resolves a question of (Chase et al. STOC 2024) who asked for such a characterization in the equivalent language of (error-dependent) `list-replicability'.
Authors: Kasra Ahmadi, Rouzbeh Behnia, Reza Ebrahimi, Mehran Mozaffari Kermani, Jeremiah Birrell, Jason Pacheco, Attila A Yavuz
Abstract: Federated learning (FL) enhances privacy by keeping user data on local devices. However, emerging attacks have demonstrated that the updates shared by users during training can reveal significant information about their data. This has greatly thwart the adoption of FL methods for training robust AI models in sensitive applications. Differential Privacy (DP) is considered the gold standard for safeguarding user data. However, DP guarantees are highly conservative, providing worst-case privacy guarantees. This can result in overestimating privacy needs, which may compromise the model's accuracy. Additionally, interpretations of these privacy guarantees have proven to be challenging in different contexts. This is further exacerbated when other factors, such as the number of training iterations, data distribution, and specific application requirements, can add further complexity to this problem. In this work, we proposed a framework that integrates a human entity as a privacy practitioner to determine an optimal trade-off between the model's privacy and utility. Our framework is the first to address the variable memory requirement of existing DP methods in FL settings, where resource-limited devices (e.g., cell phones) can participate. To support such settings, we adopt a recent DP method with fixed memory usage to ensure scalable private FL. We evaluated our proposed framework by fine-tuning a BERT-based LLM model using the GLUE dataset (a common approach in literature), leveraging the new accountant, and employing diverse data partitioning strategies to mimic real-world conditions. As a result, we achieved stable memory usage, with an average accuracy reduction of 1.33% for $\epsilon = 10$ and 1.9% for $\epsilon = 6$, when compared to the state-of-the-art DP accountant which does not support fixed memory usage.
Authors: Pai Liu, Lingfeng Zhao, Shivangi Agarwal, Jinghan Liu, Audrey Huang, Philip Amortila, Nan Jiang
Abstract: Holdout validation and hyperparameter tuning from data is a long-standing problem in offline reinforcement learning (RL). A standard framework is to use off-policy evaluation (OPE) methods to evaluate and select the policies, but OPE either incurs exponential variance (e.g., importance sampling) or has hyperparameters on their own (e.g., FQE and model-based). In this work we focus on hyperparameter tuning for OPE itself, which is even more under-investigated. Concretely, we select among candidate value functions ("model-free") or dynamics ("model-based") to best assess the performance of a target policy. Our contributions are two fold. We develop: (1) new model-free and model-based selectors with theoretical guarantees, and (2) a new experimental protocol for empirically evaluating them. Compared to the model-free protocol in prior works, our new protocol allows for more stable generation of candidate value functions, better control of misspecification, and evaluation of model-free and model-based methods alike. We exemplify the protocol on a Gym environment, and find that our new model-free selector, LSTD-Tournament, demonstrates promising empirical performance.
Authors: Divyansh Jhunjhunwala, Pranay Sharma, Zheng Xu, Gauri Joshi
Abstract: Initializing with pre-trained models when learning on downstream tasks is becoming standard practice in machine learning. Several recent works explore the benefits of pre-trained initialization in a federated learning (FL) setting, where the downstream training is performed at the edge clients with heterogeneous data distribution. These works show that starting from a pre-trained model can substantially reduce the adverse impact of data heterogeneity on the test performance of a model trained in a federated setting, with no changes to the standard FedAvg training algorithm. In this work, we provide a deeper theoretical understanding of this phenomenon. To do so, we study the class of two-layer convolutional neural networks (CNNs) and provide bounds on the training error convergence and test error of such a network trained with FedAvg. We introduce the notion of aligned and misaligned filters at initialization and show that the data heterogeneity only affects learning on misaligned filters. Starting with a pre-trained model typically results in fewer misaligned filters at initialization, thus producing a lower test error even when the model is trained in a federated setting with data heterogeneity. Experiments in synthetic settings and practical FL training on CNNs verify our theoretical findings.
Authors: Mario Franco, Gerardo Febres, Nelson Fern\'andez, Carlos Gershenson
Abstract: Classification is a ubiquitous and fundamental problem in artificial intelligence and machine learning, with extensive efforts dedicated to developing more powerful classifiers and larger datasets. However, the classification task is ultimately constrained by the intrinsic properties of datasets, independently of computational power or model complexity. In this work, we introduce a formal entropy-based measure of classificability, which quantifies the inherent difficulty of a classification problem by assessing the uncertainty in class assignments given feature representations. This measure captures the degree of class overlap and aligns with human intuition, serving as an upper bound on classification performance for classification problems. Our results establish a theoretical limit beyond which no classifier can improve the classification accuracy, regardless of the architecture or amount of data, in a given problem. Our approach provides a principled framework for understanding when classification is inherently fallible and fundamentally ambiguous.
Authors: Zijian He, Reyna Abhyankar, Vikranth Srivatsa, Yiying Zhang
Abstract: Today's gen-AI workflows that involve multiple ML model calls, tool/API calls, data retrieval, or generic code execution are often tuned manually in an ad-hoc way that is both time-consuming and error-prone. In this paper, we propose a systematic approach for automatically tuning gen-AI workflows. Our key insight is that gen-AI workflows can benefit from structure, operator, and prompt changes, but unique properties of gen-AI workflows require new optimization techniques. We propose AdaSeek, an adaptive hierarchical search algorithm for autotuning gen-AI workflows. AdaSeek organizes workflow tuning methods into different layers based on the user-specified total search budget and distributes the budget across different layers based on the complexity of each layer. During its hierarchical search, AdaSeek redistributes the search budget from less useful to more promising tuning configurations based on workflow-level evaluation results. We implement AdaSeek in a workflow autotuning framework called Cognify and evaluate Cognify using six types of workflows such as RAG-based QA and text-to-SQL transformation. Overall, Cognify improves these workflows' generation quality by up to 2.8x, reduces execution monetary cost by up to 10x, and reduces end-to-end latency by 2.7x.
Authors: Jize Xie, Cheng Chen, Zhiyong Wang, Shuai Li
Abstract: Online learning to rank sequentially recommends a small list of items to users from a large candidate set and receives the users' click feedback. In many real-world scenarios, users browse the recommended list in order and click the first attractive item without checking the rest. Such behaviors are usually formulated as the cascade model. Many recent works study algorithms for cascading bandits, an online learning to rank framework in the cascade model. However, the performance of existing methods may drop significantly if part of the user feedback is adversarially corrupted (e.g., click fraud). In this work, we study how to resist adversarial corruptions in cascading bandits. We first formulate the ``\textit{Cascading Bandits with Adversarial Corruptions}" (CBAC) problem, which assumes that there is an adaptive adversary that may manipulate the user feedback. Then we propose two robust algorithms for this problem, which assume the corruption level is known and agnostic, respectively. We show that both algorithms can achieve logarithmic regret when the algorithm is not under attack, and the regret increases linearly with the corruption level. The experimental results also verify the robustness of our methods.
Authors: Xuanze Chen, Jiajun Zhou, Jinsong Chen, Shanqing Yu, Qi Xuan
Abstract: The varying degrees of homophily and heterophily in real-world graphs persistently constrain the universality of graph neural networks (GNNs) for node classification. Adopting a data-centric perspective, this work reveals an inherent preference of different graphs towards distinct message encoding schemes: homophilous graphs favor local propagation, while heterophilous graphs exhibit preference for flexible combinations of propagation and transformation. To address this, we propose GNNMoE, a universal node classification framework based on the Mixture-of-Experts (MoE) mechanism. The framework first constructs diverse message-passing experts through recombination of fine-grained encoding operators, then designs soft and hard gating layers to allocate the most suitable expert networks for each node's representation learning, thereby enhancing both model expressiveness and adaptability to diverse graphs. Furthermore, considering that soft gating might introduce encoding noise in homophilous scenarios, we introduce an entropy constraint to guide sharpening of soft gates, achieving organic integration of weighted combination and Top-K selection. Extensive experiments demonstrate that GNNMoE significantly outperforms mainstream GNNs, heterophilous GNNs, and graph transformers in both node classification performance and universality across diverse graph datasets.
Authors: Yoshiyuki Ohmura, Wataru Shimaya, Yasuo Kuniyoshi
Abstract: In general, objects can be distinguished on the basis of their features, such as color or shape. In particular, it is assumed that similarity judgments about such features can be processed independently in different metric spaces. However, the unsupervised categorization mechanism of metric spaces corresponding to object features remains unknown. Here, we show that the artificial neural network system can autonomously categorize metric spaces through representation learning to satisfy the algebraic independence between neural networks, and project sensory information onto multiple high-dimensional metric spaces to independently evaluate the differences and similarities between features. Conventional methods often constrain the axes of the latent space to be mutually independent or orthogonal. However, the independent axes are not suitable for categorizing metric spaces. High-dimensional metric spaces that are independent of each other are not uniquely determined by the mutually independent axes, because any combination of independent axes can form mutually independent spaces. In other words, the mutually independent axes cannot be used to naturally categorize different feature spaces, such as color space and shape space. Therefore, constraining the axes to be mutually independent makes it difficult to categorize high-dimensional metric spaces. To overcome this problem, we developed a method to constrain only the spaces to be mutually independent and not the composed axes to be independent. Our theory provides general conditions for the unsupervised categorization of independent metric spaces, thus advancing the mathematical theory of functional differentiation of neural networks.
Authors: Jinsong Chen, Chenyang Li, GaiChao Li, John E. Hopcroft, Kun He
Abstract: Node tokenized graph Transformers (GTs) have shown promising performance in node classification. The generation of token sequences is the key module in existing tokenized GTs which transforms the input graph into token sequences, facilitating the node representation learning via Transformer. In this paper, we observe that the generations of token sequences in existing GTs only focus on the first-order neighbors on the constructed similarity graphs, which leads to the limited usage of nodes to generate diverse token sequences, further restricting the potential of tokenized GTs for node classification. To this end, we propose a new method termed SwapGT. SwapGT first introduces a novel token swapping operation based on the characteristics of token sequences that fully leverages the semantic relevance of nodes to generate more informative token sequences. Then, SwapGT leverages a Transformer-based backbone to learn node representations from the generated token sequences. Moreover, SwapGT develops a center alignment loss to constrain the representation learning from multiple token sequences, further enhancing the model performance. Extensive empirical results on various datasets showcase the superiority of SwapGT for node classification.
Authors: Tingyi Cai, Yunliang Jiang, Yixin Liu, Ming Li, Changqin Huang, Shirui Pan
Abstract: Graph machine learning has witnessed rapid growth, driving advancements across diverse domains. However, the in-distribution assumption, where training and testing data share the same distribution, often breaks in real-world scenarios, leading to degraded model performance under distribution shifts. This challenge has catalyzed interest in graph out-of-distribution (GOOD) detection, which focuses on identifying graph data that deviates from the distribution seen during training, thereby enhancing model robustness. In this paper, we provide a rigorous definition of GOOD detection and systematically categorize existing methods into four types: enhancement-based, reconstruction-based, information propagation-based, and classification-based approaches. We analyze the principles and mechanisms of each approach and clarify the distinctions between GOOD detection and related fields, such as graph anomaly detection, outlier detection, and GOOD generalization. Beyond methodology, we discuss practical applications and theoretical foundations, highlighting the unique challenges posed by graph data. Finally, we discuss the primary challenges and propose future directions to advance this emerging field. The repository of this survey is available at https://github.com/ca1man-2022/Awesome-GOOD-Detection.
URLs: https://github.com/ca1man-2022/Awesome-GOOD-Detection.
Authors: Ziyan Wang, Sizhe Wei, Xiaoming Huo, Hao Wang
Abstract: Diffusion models have made significant advancements in recent years. However, their performance often deteriorates when trained or fine-tuned on imbalanced datasets. This degradation is largely due to the disproportionate representation of majority and minority data in image-text pairs. In this paper, we propose a general fine-tuning approach, dubbed PoGDiff, to address this challenge. Rather than directly minimizing the KL divergence between the predicted and ground-truth distributions, PoGDiff replaces the ground-truth distribution with a Product of Gaussians (PoG), which is constructed by combining the original ground-truth targets with the predicted distribution conditioned on a neighboring text embedding. Experiments on real-world datasets demonstrate that our method effectively addresses the imbalance problem in diffusion models, improving both generation accuracy and quality.
Authors: Frank Cole, Yulong Lu, Tianhao Zhang, Yuxuan Zhao
Abstract: This paper investigates approximation-theoretic aspects of the in-context learning capability of the transformers in representing a family of noisy linear dynamical systems. Our first theoretical result establishes an upper bound on the approximation error of multi-layer transformers with respect to an $L^2$-testing loss uniformly defined across tasks. This result demonstrates that transformers with logarithmic depth can achieve error bounds comparable with those of the least-squares estimator. In contrast, our second result establishes a non-diminishing lower bound on the approximation error for a class of single-layer linear transformers, which suggests a depth-separation phenomenon for transformers in the in-context learning of dynamical systems. Moreover, this second result uncovers a critical distinction in the approximation power of single-layer linear transformers when learning from IID versus non-IID data.
Authors: Zikai Zhou, Qizheng Zhang, Hermann Kumbong, Kunle Olukotun
Abstract: Fine-tuning large language models (LLMs) is increasingly costly as models scale to hundreds of billions of parameters, and even parameter-efficient fine-tuning (PEFT) methods like LoRA remain resource-intensive. We introduce LowRA, the first framework to enable LoRA fine-tuning below 2 bits per parameter with minimal performance loss. LowRA optimizes fine-grained quantization - mapping, threshold selection, and precision assignment - while leveraging efficient CUDA kernels for scalable deployment. Extensive evaluations across 4 LLMs and 4 datasets show that LowRA achieves a superior performance-precision trade-off above 2 bits and remains accurate down to 1.15 bits, reducing memory usage by up to 50%. Our results highlight the potential of ultra-low-bit LoRA fine-tuning for resource-constrained environments.
Authors: Quan Nguyen, Shinji Ito, Junpei Komiyama, Nishant A. Mehta
Abstract: Existing data-dependent and best-of-both-worlds regret bounds for multi-armed bandits problems have limited adaptivity as they are either data-dependent but not best-of-both-worlds (BOBW), BOBW but not data-dependent or have sub-optimal $O(\sqrt{T\ln{T}})$ worst-case guarantee in the adversarial regime. To overcome these limitations, we propose real-time stability-penalty matching (SPM), a new method for obtaining regret bounds that are simultaneously data-dependent, best-of-both-worlds and $T$-optimal for multi-armed bandits problems. In particular, we show that real-time SPM obtains bounds with worst-case guarantees of order $O(\sqrt{T})$ in the adversarial regime and $O(\ln{T})$ in the stochastic regime while simultaneously being adaptive to data-dependent quantities such as sparsity, variations, and small losses. Our results are obtained by extending the SPM technique for tuning the learning rates in the follow-the-regularized-leader (FTRL) framework, which further indicates that the combination of SPM and FTRL is a promising approach for proving new adaptive bounds in online learning problems.
Authors: Siddharth Singh, Prajwal Singhania, Aditya Ranjan, John Kirchenbauer, Jonas Geiping, Yuxin Wen, Neel Jain, Abhimanyu Hans, Manli Shu, Aditya Tomar, Tom Goldstein, Abhinav Bhatele
Abstract: Training and fine-tuning large language models (LLMs) with hundreds of billions to trillions of parameters requires tens of thousands of GPUs, and a highly scalable software stack. In this work, we present a novel four-dimensional hybrid parallel algorithm implemented in a highly scalable, portable, open-source framework called AxoNN. We describe several performance optimizations in AxoNN to improve matrix multiply kernel performance, overlap non-blocking collectives with computation, and performance modeling to choose performance optimal configurations. These have resulted in unprecedented scaling and peak flop/s (bf16) for training of GPT-style transformer models on Perlmutter (620.1 Petaflop/s), Frontier (1.381 Exaflop/s) and Alps (1.423 Exaflop/s). While the abilities of LLMs improve with the number of trainable parameters, so do privacy and copyright risks caused by memorization of training data, which can cause disclosure of sensitive or private information at inference time. We highlight this side effect of scale through experiments that explore "catastrophic memorization", where models are sufficiently large to memorize training data in a single pass, and present an approach to prevent it. As part of this study, we demonstrate fine-tuning of a 405-billion parameter LLM using AxoNN on Frontier.
Authors: Zitao Wang, Ziyuan Wang, Molei Liu, Nian Si
Abstract: Transfer learning is a popular strategy to leverage external knowledge and improve statistical efficiency, particularly with a limited target sample. We propose a novel knowledge-guided Wasserstein Distributionally Robust Optimization (KG-WDRO) framework that adaptively incorporates multiple sources of external knowledge to overcome the conservativeness of vanilla WDRO, which often results in overly pessimistic shrinkage toward zero. Our method constructs smaller Wasserstein ambiguity sets by controlling the transportation along directions informed by the source knowledge. This strategy can alleviate perturbations on the predictive projection of the covariates and protect against information loss. Theoretically, we establish the equivalence between our WDRO formulation and the knowledge-guided shrinkage estimation based on collinear similarity, ensuring tractability and geometrizing the feasible set. This also reveals a novel and general interpretation for recent shrinkage-based transfer learning approaches from the perspective of distributional robustness. In addition, our framework can adjust for scaling differences in the regression models between the source and target and accommodates general types of regularization such as lasso and ridge. Extensive simulations demonstrate the superior performance and adaptivity of KG-WDRO in enhancing small-sample transfer learning.
Authors: Yang Cao, Bo Chen, Xiaoyu Li, Yingyu Liang, Zhizhou Sha, Zhenmei Shi, Zhao Song, Mingda Wan
Abstract: This paper introduces Force Matching (ForM), a novel framework for generative modeling that represents an initial exploration into leveraging special relativistic mechanics to enhance the stability of the sampling process. By incorporating the Lorentz factor, ForM imposes a velocity constraint, ensuring that sample velocities remain bounded within a constant limit. This constraint serves as a fundamental mechanism for stabilizing the generative dynamics, leading to a more robust and controlled sampling process. We provide a rigorous theoretical analysis demonstrating that the velocity constraint is preserved throughout the sampling procedure within the ForM framework. To validate the effectiveness of our approach, we conduct extensive empirical evaluations. On the \textit{half-moons} dataset, ForM significantly outperforms baseline methods, achieving the lowest Euclidean distance loss of \textbf{0.714}, in contrast to vanilla first-order flow matching (5.853) and first- and second-order flow matching (5.793). Additionally, we perform an ablation study to further investigate the impact of our velocity constraint, reaffirming the superiority of ForM in stabilizing the generative process. The theoretical guarantees and empirical results underscore the potential of integrating special relativity principles into generative modeling. Our findings suggest that ForM provides a promising pathway toward achieving stable, efficient, and flexible generative processes. This work lays the foundation for future advancements in high-dimensional generative modeling, opening new avenues for the application of physical principles in machine learning.
Authors: Rui Zhou, Yu Cheng, Songlin Li, Hongwang Zhang, Chenxu Liu
Abstract: Wireless sensing is of great benefits to our daily lives. However, wireless signals are sensitive to the surroundings. Various factors, e.g. environments, locations, and individuals, may induce extra impact on wireless propagation. Such a change can be regarded as a domain, in which the data distribution shifts. A vast majority of the sensing schemes are learning-based. They are dependent on the training domains, resulting in performance degradation in unseen domains. Researchers have proposed various solutions to address this issue. But these solutions leverage either semi-supervised or unsupervised domain adaptation techniques. They still require some data in the target domains and do not perform well in unseen domains. In this paper, we propose a domain generalization framework DGSense, to eliminate the domain dependence problem in wireless sensing. The framework is a general solution working across diverse sensing tasks and wireless technologies. Once the sensing model is built, it can generalize to unseen domains without any data from the target domain. To achieve the goal, we first increase the diversity of the training set by a virtual data generator, and then extract the domain independent features via episodic training between the main feature extractor and the domain feature extractors. The feature extractors employ a pre-trained Residual Network (ResNet) with an attention mechanism for spatial features, and a 1D Convolutional Neural Network (1DCNN) for temporal features. To demonstrate the effectiveness and generality of DGSense, we evaluated on WiFi gesture recognition, Millimeter Wave (mmWave) activity recognition, and acoustic fall detection. All the systems exhibited high generalization capability to unseen domains, including new users, locations, and environments, free of new data and retraining.
Authors: Zhaomin Wu, Zhen Qin, Junyi Hou, Haodong Zhao, Qinbin Li, Bingsheng He, Lixin Fan
Abstract: Vertical Federated Learning (VFL) is a privacy-preserving collaborative learning paradigm that enables multiple parties with distinct feature sets to jointly train machine learning models without sharing their raw data. Despite its potential to facilitate cross-organizational collaborations, the deployment of VFL systems in real-world applications remains limited. To investigate the gap between existing VFL research and practical deployment, this survey analyzes the real-world data distributions in potential VFL applications and identifies four key findings that highlight this gap. We propose a novel data-oriented taxonomy of VFL algorithms based on real VFL data distributions. Our comprehensive review of existing VFL algorithms reveals that some common practical VFL scenarios have few or no viable solutions. Based on these observations, we outline key research directions aimed at bridging the gap between current VFL research and real-world applications.
Authors: Song Park, Sanghyuk Chun, Byeongho Heo, Dongyoon Han
Abstract: This paper argues that deep neural networks (DNNs) mostly determine their outputs during the early stages of inference, where biases inherent in the model play a crucial role in shaping this process. We draw a parallel between this phenomenon and human decision-making, which often relies on fast, intuitive heuristics. Using diffusion models (DMs) as a case study, we demonstrate that DNNs often make early-stage decision-making influenced by the type and extent of bias in their design and training. Our findings offer a new perspective on bias mitigation, efficient inference, and the interpretation of machine learning systems. By identifying the temporal dynamics of decision-making in DNNs, this paper aims to inspire further discussion and research within the machine learning community.
Authors: M. Anwar Ma'sum, Mahardhika Pratama, Igor Skrjanc
Abstract: Data scarcity significantly complicates the continual learning problem, i.e., how a deep neural network learns in dynamic environments with very few samples. However, the latest progress of few-shot class incremental learning (FSCIL) methods and related studies show insightful knowledge on how to tackle the problem. This paper presents a comprehensive survey on FSCIL that highlights several important aspects i.e. comprehensive and formal objectives of FSCIL approaches, the importance of prototype rectifications, the new learning paradigms based on pre-trained model and language-guided mechanism, the deeper analysis of FSCIL performance metrics and evaluation, and the practical contexts of FSCIL in various areas. Our extensive discussion presents the open challenges, potential solutions, and future directions of FSCIL.
Authors: Vitaly Feldman, Moshe Shenfeld
Abstract: We consider the privacy guarantees of an algorithm in which a user's data is used in $k$ steps randomly and uniformly chosen from a sequence (or set) of $t$ differentially private steps. We demonstrate that the privacy guarantees of this sampling scheme can be upper bound by the privacy guarantees of the well-studied independent (or Poisson) subsampling in which each step uses the user's data with probability $(1+ o(1))k/t $. Further, we provide two additional analysis techniques that lead to numerical improvements in some parameter regimes. The case of $k=1$ has been previously studied in the context of DP-SGD in Balle et al. (2020) and very recently in Chua et al. (2024). Privacy analysis of Balle et al. (2020) relies on privacy amplification by shuffling which leads to overly conservative bounds. Privacy analysis of Chua et al. (2024a) relies on Monte Carlo simulations that are computationally prohibitive in many practical scenarios and have additional inherent limitations.
Authors: Anisha Saha, Adam Jatowt
Abstract: Future Event Prediction (FEP) is an essential activity whose demand and application range across multiple domains. While traditional methods like simulations, predictive and time-series forecasting have demonstrated promising outcomes, their application in forecasting complex events is not entirely reliable due to the inability of numerical data to accurately capture the semantic information related to events. One forecasting way is to gather and aggregate collective opinions on the future to make predictions as cumulative perspectives carry the potential to help estimating the likelihood of upcoming events. In this work, we organize the existing research and frameworks that aim to support future event prediction based on crowd wisdom through aggregating individual forecasts. We discuss the challenges involved, available datasets, as well as the scope of improvement and future research directions for this task. We also introduce a novel data model to represent individual forecast statements.
Authors: Abdelkrim Alahyane (LAAS-SARA, LAAS-RISC, LAAS), C\'eline Comte (CNRS, LAAS-SARA, LAAS-RISC, LAAS), Matthieu Jonckheere (CNRS, LAAS-SARA, LAAS-RISC, LAAS), \'Eric Moulines (X)
Abstract: Synchronous federated learning (FL) scales poorly with the number of clients due to the straggler effect. Algorithms like FedAsync and GeneralizedFedAsync address this limitation by enabling asynchronous communication between clients and the central server. In this work, we rely on stochastic modeling to better understand the impact of design choices in asynchronous FL algorithms, such as the concurrency level and routing probabilities, and we leverage this knowledge to optimize loss. We characterize in particular a fundamental trade-off for optimizing asynchronous FL: minimizing gradient estimation errors by avoiding model parameter staleness, while also speeding up the system by increasing the throughput of model updates. Our two main contributions can be summarized as follows. First, we prove a discrete variant of Little's law to derive a closed-form expression for relative delay, a metric that quantifies staleness. This allows us to efficiently minimize the average loss per model update, which has been the gold standard in literature to date. Second, we observe that naively optimizing this metric leads us to slow down the system drastically by overemphazing staleness at the detriment of throughput. This motivates us to introduce an alternative metric that also takes system speed into account, for which we derive a tractable upper-bound that can be minimized numerically. Extensive numerical results show that these optimizations enhance accuracy by 10% to 30%.
Authors: Leonard Papenmeier, Nuojin Cheng, Stephen Becker, Luigi Nardi
Abstract: A well-balanced exploration-exploitation trade-off is crucial for successful acquisition functions in Bayesian optimization. However, there is a lack of quantitative measures for exploration, making it difficult to analyze and compare different acquisition functions. This work introduces two novel approaches - observation traveling salesman distance and observation entropy - to quantify the exploration characteristics of acquisition functions based on their selected observations. Using these measures, we examine the explorative nature of several well-known acquisition functions across a diverse set of black-box problems, uncover links between exploration and empirical performance, and reveal new relationships among existing acquisition functions. Beyond enabling a deeper understanding of acquisition functions, these measures also provide a foundation for guiding their design in a more principled and systematic manner.
Authors: Junyi An, Chao Qu, Yun-Fei Shi, XinHao Liu, Qianwei Tang, Fenglei Cao, Yuan Qi
Abstract: Graph neural networks (GNNs) have shown considerable promise in computational chemistry. However, the limited availability of molecular data raises concerns regarding GNNs' ability to effectively capture the fundamental principles of physics and chemistry, which constrains their generalization capabilities. To address this challenge, we introduce a novel self-supervised approach termed Equivariant Masked Position Prediction (EMPP), grounded in intramolecular potential and force theory. Unlike conventional attribute masking techniques, EMPP formulates a nuanced position prediction task that is more well-defined and enhances the learning of quantum mechanical features. EMPP also bypasses the approximation of the Gaussian mixture distribution commonly used in denoising methods, allowing for more accurate acquisition of physical properties. Experimental results indicate that EMPP significantly enhances performance of advanced molecular architectures, surpassing state-of-the-art self-supervised approaches. Our code is released in https://github.com/ajy112/EMPP.
Authors: Jinda Xu, Yuhao Song, Daming Wang, Weiwei Zhao, Minghua Chen, Kangliang Chen, Qinya Li
Abstract: In an era overwhelmed by vast amounts of data, the effective curation of web-crawl datasets is essential for optimizing model performance. This paper tackles the challenges associated with the unstructured and heterogeneous nature of such datasets. Traditional heuristic curation methods often inadequately capture complex features, resulting in biases and the exclusion of relevant data. We introduce an advanced, learning-driven approach, Ensemble Curation Of DAta ThroUgh Multimodal Operators (EcoDatum), incorporating a novel quality-guided deduplication method to ensure balanced feature distributions. EcoDatum strategically integrates various unimodal and multimodal data curation operators within a weak supervision ensemble framework, utilizing automated optimization to score each data point effectively. EcoDatum, which significantly improves the data curation quality and efficiency, outperforms existing state-of-the-art (SOTA) techniques, ranked 1st on the DataComp leaderboard, with an average performance score of 0.182 across 38 diverse evaluation datasets. This represents a 28% improvement over the DataComp baseline method, demonstrating its effectiveness in improving dataset curation and model training efficiency.
Authors: Suqin Yuan, Lei Feng, Bo Han, Tongliang Liu
Abstract: Sample selection is a prevalent approach in learning with noisy labels, aiming to identify confident samples for training. Although existing sample selection methods have achieved decent results by reducing the noise rate of the selected subset, they often overlook that not all mislabeled examples harm the model's performance equally. In this paper, we demonstrate that mislabeled examples correctly predicted by the model early in the training process are particularly harmful to model performance. We refer to these examples as Mislabeled Easy Examples (MEEs). To address this, we propose Early Cutting, which introduces a recalibration step that employs the model's later training state to re-select the confident subset identified early in training, thereby avoiding misleading confidence from early learning and effectively filtering out MEEs. Experiments on the CIFAR, WebVision, and full ImageNet-1k datasets demonstrate that our method effectively improves sample selection and model performance by reducing MEEs.
Authors: Evgeniia Tokarchuk, Hua Chang Bakker, Vlad Niculae
Abstract: Learning well-separated features in high-dimensional spaces, such as text or image embeddings, is crucial for many machine learning applications. Achieving such separation can be effectively accomplished through the dispersion of embeddings, where unrelated vectors are pushed apart as much as possible. By constraining features to be on a hypersphere, we can connect dispersion to well-studied problems in mathematics and physics, where optimal solutions are known for limited low-dimensional cases. However, in representation learning we typically deal with a large number of features in high-dimensional space, and moreover, dispersion is usually traded off with some other task-oriented training objective, making existing theoretical and numerical solutions inapplicable. Therefore, it is common to rely on gradient-based methods to encourage dispersion, usually by minimizing some function of the pairwise distances. In this work, we first give an overview of existing methods from disconnected literature, making new connections and highlighting similarities. Next, we introduce some new angles. We propose to reinterpret pairwise dispersion using a maximum mean discrepancy (MMD) motivation. We then propose an online variant of the celebrated Lloyd's algorithm, of K-Means fame, as an effective alternative regularizer for dispersion on generic domains. Finally, we derive a novel dispersion method that directly exploits properties of the hypersphere. Our experiments show the importance of dispersion in image classification and natural language processing tasks, and how algorithms exhibit different trade-offs in different regimes.
Authors: Sentenac Flore, Lee Albin, Szepesvari Csaba
Abstract: We consider what we call the offline-to-online learning setting, focusing on stochastic finite-armed bandit problems. In offline-to-online learning, a learner starts with offline data collected from interactions with an unknown environment in a way that is not under the learner's control. Given this data, the learner begins interacting with the environment, gradually improving its initial strategy as it collects more data to maximize its total reward. The learner in this setting faces a fundamental dilemma: if the policy is deployed for only a short period, a suitable strategy (in a number of senses) is the Lower Confidence Bound (LCB) algorithm, which is based on pessimism. LCB can effectively compete with any policy that is sufficiently "covered" by the offline data. However, for longer time horizons, a preferred strategy is the Upper Confidence Bound (UCB) algorithm, which is based on optimism. Over time, UCB converges to the performance of the optimal policy at a rate that is nearly the best possible among all online algorithms. In offline-to-online learning, however, UCB initially explores excessively, leading to worse short-term performance compared to LCB. This suggests that a learner not in control of how long its policy will be in use should start with LCB for short horizons and gradually transition to a UCB-like strategy as more rounds are played. This article explores how and why this transition should occur. Our main result shows that our new algorithm performs nearly as well as the better of LCB and UCB at any point in time. The core idea behind our algorithm is broadly applicable, and we anticipate that our results will extend beyond the multi-armed bandit setting.
Authors: Zahra Zamanzadeh Darban, Qizhou Wang, Geoffrey I. Webb, Shirui Pan, Charu C. Aggarwal, Mahsa Salehi
Abstract: A recent and promising approach for building time series anomaly detection (TSAD) models is to inject synthetic samples of anomalies within real data sets. The existing injection mechanisms have significant limitations - most of them rely on ad hoc, hand-crafted strategies which fail to capture the natural diversity of anomalous patterns, or are restricted to univariate time series settings. To address these challenges, we design a generative model for TSAD using a variational autoencoder, which is referred to as a Generator for Instantiating Anomalies in Time Series (GenIAS). GenIAS is designed to produce diverse and realistic synthetic anomalies for TSAD tasks. By employing a novel learned perturbation mechanism in the latent space and injecting the perturbed patterns in different segments of time series, GenIAS can generate anomalies with greater diversity and varying scales. Further, guided by a new triplet loss function, which uses a min-max margin and a new variance-scaling approach to further enforce the learning of compact normal patterns, GenIAS ensures that anomalies are distinct from normal samples while remaining realistic. The approach is effective for both univariate and multivariate time series. We demonstrate the diversity and realism of the generated anomalies. Our extensive experiments demonstrate that GenIAS - when integrated into a TSAD task - consistently outperforms seventeen traditional and deep anomaly detection models, thereby highlighting the potential of generative models for time series anomaly generation.
Authors: Vinod Kumar Chauhan, Lei Clifton, Gaurav Nigam, David A. Clifton
Abstract: Estimating individualised treatment effect (ITE) -- that is the causal effect of a set of variables (also called exposures, treatments, actions, policies, or interventions), referred to as \textit{composite treatments}, on a set of outcome variables of interest, referred to as \textit{composite outcomes}, for a unit from observational data -- remains a fundamental problem in causal inference with applications across disciplines, such as healthcare, economics, education, social science, marketing, and computer science. Previous work in causal machine learning for ITE estimation is limited to simple settings, like single treatments and single outcomes. This hinders their use in complex real-world scenarios; for example, consider studying the effect of different ICU interventions, such as beta-blockers and statins for a patient admitted for heart surgery, on different outcomes of interest such as atrial fibrillation and in-hospital mortality. The limited research into composite treatments and outcomes is primarily due to data scarcity for all treatments and outcomes. To address the above challenges, we propose a novel and innovative hypernetwork-based approach, called \emph{H-Learner}, to solve ITE estimation under composite treatments and composite outcomes, which tackles the data scarcity issue by dynamically sharing information across treatments and outcomes. Our empirical analysis with binary and arbitrary composite treatments and outcomes demonstrates the effectiveness of the proposed approach compared to existing methods.
Authors: Shibo Feng, Peilin Zhao, Liu Liu, Pengcheng Wu, Zhiqi Shen
Abstract: Generative models have gained significant attention in multivariate time series forecasting (MTS), particularly due to their ability to generate high-fidelity samples. Forecasting the probability distribution of multivariate time series is a challenging yet practical task. Although some recent attempts have been made to handle this task, two major challenges persist: 1) some existing generative methods underperform in high-dimensional multivariate time series forecasting, which is hard to scale to higher dimensions; 2) the inherent high-dimensional multivariate attributes constrain the forecasting lengths of existing generative models. In this paper, we point out that discrete token representations can model high-dimensional MTS with faster inference time, and forecasting the target with long-term trends of itself can extend the forecasting length with high accuracy. Motivated by this, we propose a vector quantized framework called Hierarchical Discrete Transformer (HDT) that models time series into discrete token representations with l2 normalization enhanced vector quantized strategy, in which we transform the MTS forecasting into discrete tokens generation. To address the limitations of generative models in long-term forecasting, we propose a hierarchical discrete Transformer. This model captures the discrete long-term trend of the target at the low level and leverages this trend as a condition to generate the discrete representation of the target at the high level that introduces the features of the target itself to extend the forecasting length in high-dimensional MTS. Extensive experiments on five popular MTS datasets verify the effectiveness of our proposed method.
Authors: Minh Hieu Nguyen, Viet Hung Doan, Anh Tuan Nguyen, Jun Jo, Quoc Viet Hung Nguyen
Abstract: Ensuring transparency in AI decision-making requires interpretable explanations, particularly at the instance level. Counterfactual explanations are a powerful tool for this purpose, but existing techniques frequently depend on synthetic examples, introducing biases from unrealistic assumptions, flawed models, or skewed data. Many methods also assume full dataset availability, an impractical constraint in real-time environments where data flows continuously. In contrast, streaming explanations offer adaptive, real-time insights without requiring persistent storage of the entire dataset. This work introduces a scalable, model-free approach to selecting diverse and relevant counterfactual examples directly from observed data. Our algorithm operates efficiently in streaming settings, maintaining $O(\log k)$ update complexity per item while ensuring high-quality counterfactual selection. Empirical evaluations on both real-world and synthetic datasets demonstrate superior performance over baseline methods, with robust behavior even under adversarial conditions.
Authors: Soumyendu Sarkar, Avisek Naug, Antonio Guillen, Vineet Gundecha, Ricardo Luna Gutierrez, Sahand Ghorbanpour, Sajad Mousavi, Ashwin Ramesh Babu, Desik Rengarajan, Cullen Bash
Abstract: Reducing the environmental impact of cloud computing requires efficient workload distribution across geographically dispersed Data Center Clusters (DCCs) and simultaneously optimizing liquid and air (HVAC) cooling with time shift of workloads within individual data centers (DC). This paper introduces Green-DCC, which proposes a Reinforcement Learning (RL) based hierarchical controller to optimize both workload and liquid cooling dynamically in a DCC. By incorporating factors such as weather, carbon intensity, and resource availability, Green-DCC addresses realistic constraints and interdependencies. We demonstrate how the system optimizes multiple data centers synchronously, enabling the scope of digital twins, and compare the performance of various RL approaches based on carbon emissions and sustainability metrics while also offering a framework and benchmark simulation for broader ML research in sustainability.
Authors: Yuxin Pan, Ruohong Liu, Yize Chen, Zhiguang Cao, Fangzhen Lin
Abstract: Neural solvers based on the divide-and-conquer approach for Vehicle Routing Problems (VRPs) in general, and capacitated VRP (CVRP) in particular, integrates the global partition of an instance with local constructions for each subproblem to enhance generalization. However, during the global partition phase, misclusterings within subgraphs have a tendency to progressively compound throughout the multi-step decoding process of the learning-based partition policy. This suboptimal behavior in the global partition phase, in turn, may lead to a dramatic deterioration in the performance of the overall decomposition-based system, despite using optimal local constructions. To address these challenges, we propose a versatile Hierarchical Learning-based Graph Partition (HLGP) framework, which is tailored to benefit the partition of CVRP instances by synergistically integrating global and local partition policies. Specifically, the global partition policy is tasked with creating the coarse multi-way partition to generate the sequence of simpler two-way partition subtasks. These subtasks mark the initiation of the subsequent K local partition levels. At each local partition level, subtasks exclusive for this level are assigned to the local partition policy which benefits from the insensitive local topological features to incrementally alleviate the compounded errors. This framework is versatile in the sense that it optimizes the involved partition policies towards a unified objective harmoniously compatible with both reinforcement learning (RL) and supervised learning (SL). (*Due to the notification of arXiv "The Abstract field cannot be longer than 1,920 characters", the appeared Abstract is shortened. For the full Abstract, please download the Article.)
Authors: Ruizhan Xue, Huimin Deng, Fang He, Maojun Wang, Zeyu Zhang
Abstract: With the extensive application of Graph Neural Networks (GNNs) across various domains, their trustworthiness has emerged as a focal point of research. Some existing studies have shown that the integration of large language models (LLMs) can improve the semantic understanding and generation capabilities of GNNs, which in turn improves the trustworthiness of GNNs from various aspects. Our review introduces a taxonomy that offers researchers a clear framework for comprehending the principles and applications of different methods and helps clarify the connections and differences among various approaches. Then we systematically survey representative approaches along the four categories of our taxonomy. Through our taxonomy, researchers can understand the applicable scenarios, potential advantages, and limitations of each approach for the the trusted integration of GNNs with LLMs. Finally, we present some promising directions of work and future trends for the integration of LLMs and GNNs to improve model trustworthiness.
Authors: Tommaso Baldi, Javier Campos, Olivia Weng, Caleb Geniesse, Nhan Tran, Ryan Kastner, Alessandro Biondi
Abstract: In this paper, we propose a method to perform empirical analysis of the loss landscape of machine learning (ML) models. The method is applied to two ML models for scientific sensing, which necessitates quantization to be deployed and are subject to noise and perturbations due to experimental conditions. Our method allows assessing the robustness of ML models to such effects as a function of quantization precision and under different regularization techniques -- two crucial concerns that remained underexplored so far. By investigating the interplay between performance, efficiency, and robustness by means of loss landscape analysis, we both established a strong correlation between gently-shaped landscapes and robustness to input and weight perturbations and observed other intriguing and non-obvious phenomena. Our method allows a systematic exploration of such trade-offs a priori, i.e., without training and testing multiple models, leading to more efficient development workflows. This work also highlights the importance of incorporating robustness into the Pareto optimization of ML models, enabling more reliable and adaptive scientific sensing systems.
Authors: Xuhui Fan, Zhangkai Wu, Hongyu Wu
Abstract: Diffusion Models~(DMs) have emerged as the dominant approach in Generative Artificial Intelligence (GenAI), owing to their remarkable performance in tasks such as text-to-image synthesis. However, practical DMs, such as stable diffusion, are typically trained on massive datasets and thus usually require large storage. At the same time, many steps may be required, i.e., recursively evaluating the trained neural network, to generate a high-quality image, which results in significant computational costs during sample generation. As a result, distillation methods on pre-trained DM have become widely adopted practices to develop smaller, more efficient models capable of rapid, few-step generation in low-resource environment. When these distillation methods are developed from different perspectives, there is an urgent need for a systematic survey, particularly from a methodological perspective. In this survey, we review distillation methods through three aspects: output loss distillation, trajectory distillation and adversarial distillation. We also discuss current challenges and outline future research directions in the conclusion.
Authors: Riccardo Zamboni, Mirco Mutti, Marcello Restelli
Abstract: In reinforcement learning, we typically refer to task-agnostic exploration when we aim to explore the environment without access to the task specification a priori. In a single-agent setting the problem has been extensively studied and mostly understood. A popular approach cast the task-agnostic objective as maximizing the entropy of the state distribution induced by the agent's policy, from which principles and methods follows. In contrast, little is known about task-agnostic exploration in multi-agent settings, which are ubiquitous in the real world. How should different agents explore in the presence of others? In this paper, we address this question through a generalization to multiple agents of the problem of maximizing the state distribution entropy. First, we investigate alternative formulations, highlighting respective positives and negatives. Then, we present a scalable, decentralized, trust-region policy search algorithm to address the problem in practical settings. Finally, we provide proof of concept experiments to both corroborate the theoretical findings and pave the way for task-agnostic exploration in challenging multi-agent settings.
Authors: Ugochukwu Orji, \c{C}i\c{c}ek G\"uven, Dan Stowell
Abstract: Accurate power load forecasting is essential for the efficient operation and planning of electrical grids, particularly given the increased variability and complexity introduced by renewable energy sources. This paper introduces GAT-LSTM, a hybrid model that combines Graph Attention Networks (GAT) and Long Short-Term Memory (LSTM) networks. A key innovation of the model is the incorporation of edge attributes, such as line capacities and efficiencies, into the attention mechanism, enabling it to dynamically capture spatial relationships grounded in grid-specific physical and operational constraints. Additionally, by employing an early fusion of spatial graph embeddings and temporal sequence features, the model effectively learns and predicts complex interactions between spatial dependencies and temporal patterns, providing a realistic representation of the dynamics of power grids. Experimental evaluations on the Brazilian Electricity System dataset demonstrate that the GAT-LSTM model significantly outperforms state-of-the-art models, achieving reductions of 21. 8% in MAE, 15. 9% in RMSE and 20. 2% in MAPE. These results underscore the robustness and adaptability of the GAT-LSTM model, establishing it as a powerful tool for applications in grid management and energy planning.
Authors: Daeyoung Roh, Donghee Han, Daehee Kim, Keejun Han, Mun Yi
Abstract: Hypergraphs provide a superior modeling framework for representing complex multidimensional relationships in the context of real-world interactions that often occur in groups, overcoming the limitations of traditional homogeneous graphs. However, there have been few studies on hypergraphbased contrastive learning, and existing graph-based contrastive learning methods have not been able to fully exploit the highorder correlation information in hypergraphs. Here, we propose a Hypergraph Fine-grained contrastive learning (HyFi) method designed to exploit the complex high-dimensional information inherent in hypergraphs. While avoiding traditional graph augmentation methods that corrupt the hypergraph topology, the proposed method provides a simple and efficient learning augmentation function by adding noise to node features. Furthermore, we expands beyond the traditional dichotomous relationship between positive and negative samples in contrastive learning by introducing a new relationship of weak positives. It demonstrates the importance of fine-graining positive samples in contrastive learning. Therefore, HyFi is able to produce highquality embeddings, and outperforms both supervised and unsupervised baselines in average rank on node classification across 10 datasets. Our approach effectively exploits high-dimensional hypergraph information, shows significant improvement over existing graph-based contrastive learning methods, and is efficient in terms of training speed and GPU memory cost. The source code is available at https://github.com/Noverse0/HyFi.git.
Authors: Yining Jiao, Sreekalyani Bhamidi, Huaizhi Qu, Carlton Zdanski, Julia Kimbell, Andrew Prince, Cameron Worden, Samuel Kirse, Christopher Rutter, Benjamin Shields, William Dunn, Jisan Mahmud, Tianlong Chen, Marc Niethammer
Abstract: The goal of this work is to develop principled techniques to extract information from high dimensional data sets with complex dependencies in areas such as medicine that can provide insight into individual as well as population level variation. We develop $\texttt{LucidAtlas}$, an approach that can represent spatially varying information, and can capture the influence of covariates as well as population uncertainty. As a versatile atlas representation, $\texttt{LucidAtlas}$ offers robust capabilities for covariate interpretation, individualized prediction, population trend analysis, and uncertainty estimation, with the flexibility to incorporate prior knowledge. Additionally, we discuss the trustworthiness and potential risks of neural additive models for analyzing dependent covariates and then introduce a marginalization approach to explain the dependence of an individual predictor on the models' response (the atlas). To validate our method, we demonstrate its generalizability on two medical datasets. Our findings underscore the critical role of by-construction interpretable models in advancing scientific discovery. Our code will be publicly available upon acceptance.
Authors: Albert Kj{\o}ller Jacobsen, Georgios Arvanitidis
Abstract: Recent studies on deep neural networks show that flat minima of the loss landscape correlate with improved generalization. Sharpness-aware minimization (SAM) efficiently finds flat regions by updating the parameters according to the gradient at an adversarial perturbation. The perturbation depends on the Euclidean metric, making SAM non-invariant under reparametrizations, which blurs sharpness and generalization. We propose Monge SAM (M-SAM), a reparametrization invariant version of SAM by considering a Riemannian metric in the parameter space induced naturally by the loss surface. Compared to previous approaches, M-SAM works under any modeling choice, relies only on mild assumptions while being as computationally efficient as SAM. We theoretically argue that M-SAM varies between SAM and gradient descent (GD), which increases robustness to hyperparameter selection and reduces attraction to suboptimal equilibria like saddle points. We demonstrate this behavior both theoretically and empirically on a multi-modal representation alignment task.
Authors: Fares El Khoury, Edouard Pauwels, Samuel Vaiter, Michael Arbel
Abstract: Bilevel optimization has emerged as a technique for addressing a wide range of machine learning problems that involve an outer objective implicitly determined by the minimizer of an inner problem. In this paper, we investigate the generalization properties for kernel bilevel optimization problems where the inner objective is optimized over a Reproducing Kernel Hilbert Space. This setting enables rich function approximation while providing a foundation for rigorous theoretical analysis. In this context, we establish novel generalization error bounds for the bilevel problem under finite-sample approximation. Our approach adopts a functional perspective, inspired by (Petrulionyte et al., 2024), and leverages tools from empirical process theory and maximal inequalities for degenerate $U$-processes to derive uniform error bounds. These generalization error estimates allow to characterize the statistical accuracy of gradient-based methods applied to the empirical discretization of the bilevel problem.
Authors: Keonho Lee, Minsoo Kim, Dong-Wan Choi
Abstract: Although network pruning has been highly popularized to compress deep neural networks, its resulting accuracy heavily depends on a fine-tuning process that is often computationally expensive and requires the original data. However, this may not be the case in real-world scenarios, and hence a few recent works attempt to restore pruned networks without any expensive retraining process. Their strong assumption is that every neuron being pruned can be replaced with another one quite similar to it, but unfortunately this does not hold in many neural networks, where the similarity between neurons is extremely low in some layers. In this article, we propose a more rigorous and robust method of restoring pruned networks in a fine-tuning free and data-free manner, called LBYL (Leave Before You Leave). LBYL significantly relaxes the aforementioned assumption in a way that each pruned neuron leaves its pieces of information to as many preserved neurons as possible and thereby multiple neurons together obtain a more robust approximation to the original output of the neuron who just left. Our method is based on a theoretical analysis on how to formulate the reconstruction error between the original network and its approximation, which nicely leads to a closed form solution for our derived loss function. Through the extensive experiments, LBYL is confirmed to be indeed more effective to approximate the original network and consequently able to achieve higher accuracy for restored networks, compared to the recent approaches exploiting the similarity between two neurons. The very first version of this work, which contains major technical and theoretical components, was submitted to NeurIPS 2021 and ICML 2022.
Authors: Obaidullah Zaland, Shutong Jin, Florian T. Pokorny, Monowar Bhuyan
Abstract: Federated learning (FL) enables collaborative learning without data centralization but introduces significant communication costs due to multiple communication rounds between clients and the server. One-shot federated learning (OSFL) addresses this by forming a global model with a single communication round, often relying on the server's model distillation or auxiliary dataset generation - often through pre-trained diffusion models (DMs). Existing DM-assisted OSFL methods, however, typically employ classifier-guided DMs, which require training auxiliary classifier models at each client, introducing additional computation overhead. This work introduces OSCAR (One-Shot Federated Learning with Classifier-Free Diffusion Models), a novel OSFL approach that eliminates the need for auxiliary models. OSCAR uses foundation models to devise category-specific data representations at each client, seamlessly integrated into a classifier-free diffusion model pipeline for server-side data generation. OSCAR is a simple yet cost-effective OSFL approach that outperforms the state-of-the-art on four benchmarking datasets while reducing the communication load by at least 99%.
Authors: Tao Wen, Elynn Chen, Yuzhou Chen, Qi Lei
Abstract: Graph Neural Networks (GNNs) have recently become the predominant tools for studying graph data. Despite state-of-the-art performance on graph classification tasks, GNNs are overwhelmingly trained in a single domain under supervision, thus necessitating a prohibitively high demand for labels and resulting in poorly transferable representations. To address this challenge, we propose the Label-Propagation Tensor Graph Neural Network (LP-TGNN) framework to bridge the gap between graph data and traditional domain adaptation methods. It extracts graph topological information holistically with a tensor architecture and then reduces domain discrepancy through label propagation. It is readily compatible with general GNNs and domain adaptation techniques with minimal adjustment through pseudo-labeling. Experiments on various real-world benchmarks show that our LP-TGNN outperforms baselines by a notable margin. We also validate and analyze each component of the proposed framework in the ablation study.
Authors: Evgenii Evstafev
Abstract: This study examines how temperature settings and model architectures affect the generation of structured fictional data (names, birthdates) across three large language models (LLMs): llama3.1:8b, deepseek-r1:8b, and mistral:latest. By systematically testing temperature values from 0.0 to 1.0 in increments of 0.1, we conducted 330 trials yielding 889 structured entities, validated for syntactic consistency. Key findings reveal that model architecture significantly influences computational efficiency, with mistral:latest and llama3.1:8b processing data 8x faster than deepseek-r1:8b. Contrary to expectations, temperature showed no correlation with processing time, challenging assumptions about stochastic sampling costs. Output diversity remained limited, as models consistently defaulted to common name archetypes (e.g., 'John Doe' and 'Jane Smith') across all temperatures, though rare names clustered at intermediate values (0.3-0.7). These results demonstrate that architectural optimizations, rather than temperature adjustments, dominate performance in structured generation tasks. The findings emphasize prioritizing model selection over hyperparameter tuning for efficiency and suggest explicit diversity constraints are necessary to mitigate default output biases in synthetic data pipelines.
Authors: Dezhong Yao, Yuexin Shi, Tongtong Liu, Zhiqiang Xu
Abstract: Federated Learning (FL) is increasingly adopted in edge computing scenarios, where a large number of heterogeneous clients operate under constrained or sufficient resources. The iterative training process in conventional FL introduces significant computation and communication overhead, which is unfriendly for resource-constrained edge devices. One-shot FL has emerged as a promising approach to mitigate communication overhead, and model-heterogeneous FL solves the problem of diverse computing resources across clients. However, existing methods face challenges in effectively managing model-heterogeneous one-shot FL, often leading to unsatisfactory global model performance or reliance on auxiliary datasets. To address these challenges, we propose a novel FL framework named FedMHO, which leverages deep classification models on resource-sufficient clients and lightweight generative models on resource-constrained devices. On the server side, FedMHO involves a two-stage process that includes data generation and knowledge fusion. Furthermore, we introduce FedMHO-MD and FedMHO-SD to mitigate the knowledge-forgetting problem during the knowledge fusion stage, and an unsupervised data optimization solution to improve the quality of synthetic samples. Comprehensive experiments demonstrate the effectiveness of our methods, as they outperform state-of-the-art baselines in various experimental setups.
Authors: Jihoon Tack, Jack Lanchantin, Jane Yu, Andrew Cohen, Ilia Kulikov, Janice Lan, Shibo Hao, Yuandong Tian, Jason Weston, Xian Li
Abstract: Next token prediction has been the standard training objective used in large language model pretraining. Representations are learned as a result of optimizing for token-level perplexity. We propose Continuous Concept Mixing (CoCoMix), a novel pretraining framework that combines discrete next token prediction with continuous concepts. Specifically, CoCoMix predicts continuous concepts learned from a pretrained sparse autoencoder and mixes them into the model's hidden state by interleaving with token hidden representations. Through experiments on multiple benchmarks, including language modeling and downstream reasoning tasks, we show that CoCoMix is more sample efficient and consistently outperforms standard next token prediction, knowledge distillation and inserting pause tokens. We find that combining both concept learning and interleaving in an end-to-end framework is critical to performance gains. Furthermore, CoCoMix enhances interpretability and steerability by allowing direct inspection and modification of the predicted concept, offering a transparent way to guide the model's internal reasoning process.
Authors: Philipp M. Faller, Dominik Janzing
Abstract: The goal of conditional-independence-based discovery of graphical models is to find a graph that represents the independence structure of variables in a given dataset. To learn such a representation, conditional-independence-based approaches conduct a set of statistical tests that suffices to identify the graphical representation under some assumptions on the underlying distribution of the data. In this work, we highlight that due to the conciseness of the graphical representation, there are often many tests that are not used in the construction of the graph. These redundant tests have the potential to detect or sometimes correct errors in the learned model. We show that not all tests contain this additional information and that such redundant tests have to be applied with care. Precisely, we argue that particularly those conditional (in)dependence statements are interesting that follow only from graphical assumptions but do not hold for every probability distribution.
Authors: Yao Wang, Yiyang Yang, Kaidong Wang, Shanxing Gao, Xiuwu Liao
Abstract: We consider the problem of matrix completion with graphs as side information depicting the interrelations between variables. The key challenge lies in leveraging the similarity structure of the graph to enhance matrix recovery. Existing approaches, primarily based on graph Laplacian regularization, suffer from several limitations: (1) they focus only on the similarity between neighboring variables, while overlooking long-range correlations; (2) they are highly sensitive to false edges in the graphs and (3) they lack theoretical guarantees regarding statistical and computational complexities. To address these issues, we propose in this paper a novel graph regularized matrix completion algorithm called GSGD, based on preconditioned projected gradient descent approach. We demonstrate that GSGD effectively captures the higher-order correlation information behind the graphs, and achieves superior robustness and stability against the false edges. Theoretically, we prove that GSGD achieves linear convergence to the global optimum with near-optimal sample complexity, providing the first theoretical guarantees for both recovery accuracy and efficacy in the perspective of nonconvex optimization. Our numerical experiments on both synthetic and real-world data further validate that GSGD achieves superior recovery accuracy and scalability compared with several popular alternatives.
Authors: Vittoria Vineis, Giuseppe Perelli, Gabriele Tolomei
Abstract: Conventional decision-support systems, primarily based on supervised learning, focus on outcome prediction models to recommend actions. However, they often fail to account for the complexities of multi-actor environments, where diverse and potentially conflicting stakeholder preferences must be balanced. In this paper, we propose a novel participatory framework that redefines decision-making as a multi-stakeholder optimization problem, capturing each actor's preferences through context-dependent reward functions. Our framework leverages $k$-fold cross-validation to fine-tune user-provided outcome prediction models and evaluate decision strategies, including compromise functions mediating stakeholder trade-offs. We introduce a synthetic scoring mechanism that exploits user-defined preferences across multiple metrics to rank decision-making strategies and identify the optimal decision-maker. The selected decision-maker can then be used to generate actionable recommendations for new data. We validate our framework using two real-world use cases, demonstrating its ability to deliver recommendations that effectively balance multiple metrics, achieving results that are often beyond the scope of purely prediction-based methods. Ablation studies demonstrate that our framework, with its modular, model-agnostic, and inherently transparent design, integrates seamlessly with various predictive models, reward structures, evaluation metrics, and sample sizes, making it particularly suited for complex, high-stakes decision-making contexts.
Authors: Zhikai Wu, Shiyang Zhang, Sizhuang He, Sifan Wang, Min Zhu, Anran Jiao, Lu Lu, David van Dijk
Abstract: We introduce Causal Operator with Adaptive Solver Transformer (COAST), a novel neural operator learning method that leverages a causal language model (CLM) framework to dynamically adapt time steps. Our method predicts both the evolution of a system and its optimal time step, intelligently balancing computational efficiency and accuracy. We find that COAST generates variable step sizes that correlate with the underlying system intrinsicities, both within and across dynamical systems. Within a single trajectory, smaller steps are taken in regions of high complexity, while larger steps are employed in simpler regions. Across different systems, more complex dynamics receive more granular time steps. Benchmarked on diverse systems with varied dynamics, COAST consistently outperforms state-of-the-art methods, achieving superior performance in both efficiency and accuracy. This work underscores the potential of CLM-based intelligent adaptive solvers for scalable operator learning of dynamical systems.
Authors: Davide Domini, Gianluca Aguzzi, Lukas Esterle, Mirko Viroli
Abstract: In the last years, Federated learning (FL) has become a popular solution to train machine learning models in domains with high privacy concerns. However, FL scalability and performance face significant challenges in real-world deployments where data across devices are non-independently and identically distributed (non-IID). The heterogeneity in data distribution frequently arises from spatial distribution of devices, leading to degraded model performance in the absence of proper handling. Additionally, FL typical reliance on centralized architectures introduces bottlenecks and single-point-of-failure risks, particularly problematic at scale or in dynamic environments. To close this gap, we propose Field-Based Federated Learning (FBFL), a novel approach leveraging macroprogramming and field coordination to address these limitations through: (i) distributed spatial-based leader election for personalization to mitigate non-IID data challenges; and (ii) construction of a self-organizing, hierarchical architecture using advanced macroprogramming patterns. Moreover, FBFL not only overcomes the aforementioned limitations, but also enables the development of more specialized models tailored to the specific data distribution in each subregion. This paper formalizes FBFL and evaluates it extensively using MNIST, FashionMNIST, and Extended MNIST datasets. We demonstrate that, when operating under IID data conditions, FBFL performs comparably to the widely-used FedAvg algorithm. Furthermore, in challenging non-IID scenarios, FBFL not only outperforms FedAvg but also surpasses other state-of-the-art methods, namely FedProx and Scaffold, which have been specifically designed to address non-IID data distributions. Additionally, we showcase the resilience of FBFL's self-organizing hierarchical architecture against server failures.
Authors: Shoma Yokura, Akihisa Ichiki
Abstract: Binary classification is a task that involves the classification of data into one of two distinct classes. It is widely utilized in various fields. However, conventional classifiers tend to make overconfident predictions for data that belong to overlapping regions of the two class distributions or for data outside the distributions (out-of-distribution data). Therefore, conventional classifiers should not be applied in high-risk fields where classification results can have significant consequences. In order to address this issue, it is necessary to quantify uncertainty and adopt decision-making approaches that take it into account. Many methods have been proposed for this purpose; however, implementing these methods often requires performing resampling, improving the structure or performance of models, and optimizing the thresholds of classifiers. We propose a new decision-making approach using two types of hypothesis testing. This method is capable of detecting ambiguous data that belong to the overlapping regions of two class distributions, as well as out-of-distribution data that are not included in the training data distribution. In addition, we quantify uncertainty using the empirical distribution of feature values derived from the training data obtained through the trained model. The classification threshold is determined by the $\alpha$-quantile and ($1-\alpha$)-quantile, where the significance level $\alpha$ is set according to each specific situation.
Authors: Peiyao Xiao, Chaosheng Dong, Shaofeng Zou, Kaiyi Ji
Abstract: Multi-task learning (MTL) has been widely adopted for its ability to simultaneously learn multiple tasks. While existing gradient manipulation methods often yield more balanced solutions than simple scalarization-based approaches, they typically incur a significant computational overhead of $\mathcal{O}(K)$ in both time and memory, where $K$ is the number of tasks. In this paper, we propose BiLB4MTL, a simple and scalable loss balancing approach for MTL, formulated from a novel bilevel optimization perspective. Our method incorporates three key components: (i) an initial loss normalization, (ii) a bilevel loss-balancing formulation, and (iii) a scalable first-order algorithm that requires only $\mathcal{O}(1)$ time and memory. Theoretically, we prove that BiLB4MTL guarantees convergence not only to a stationary point of the bilevel loss balancing problem but also to an $\epsilon$-accurate Pareto stationary point for all $K$ loss functions under mild conditions. Extensive experiments on diverse multi-task datasets demonstrate that BiLB4MTL achieves state-of-the-art performance in both accuracy and efficiency. Code is available at https://github.com/OptMN-Lab/-BiLB4MTL.
Authors: Ang Li, Yin Zhou, Vethavikashini Chithrra Raghuram, Tom Goldstein, Micah Goldblum
Abstract: A high volume of recent ML security literature focuses on attacks against aligned large language models (LLMs). These attacks may extract private information or coerce the model into producing harmful outputs. In real-world deployments, LLMs are often part of a larger agentic pipeline including memory systems, retrieval, web access, and API calling. Such additional components introduce vulnerabilities that make these LLM-powered agents much easier to attack than isolated LLMs, yet relatively little work focuses on the security of LLM agents. In this paper, we analyze security and privacy vulnerabilities that are unique to LLM agents. We first provide a taxonomy of attacks categorized by threat actors, objectives, entry points, attacker observability, attack strategies, and inherent vulnerabilities of agent pipelines. We then conduct a series of illustrative attacks on popular open-source and commercial agents, demonstrating the immediate practical implications of their vulnerabilities. Notably, our attacks are trivial to implement and require no understanding of machine learning.
Authors: Yuhao Wang, Aram Ebtekar, Dominik Janzing
Abstract: We argue that Algorithmic Information Theory (AIT) admits a principled way to quantify outliers in terms of so-called randomness deficiency. For the probability distribution generated by a causal Bayesian network, we show that the randomness deficiency of the joint state decomposes into randomness deficiencies of each causal mechanism, subject to the Independence of Mechanisms Principle. Accordingly, anomalous joint observations can be quantitatively attributed to their root causes, i.e., the mechanisms that behaved anomalously. As an extension of Levin's law of randomness conservation, we show that weak outliers cannot cause strong ones when Independence of Mechanisms holds. We show how these information theoretic laws provide a better understanding of the behaviour of outliers defined with respect to existing scores.
Authors: Khaled Kahouli, Winfried Ripken, Stefan Gugler, Oliver T. Unke, Klaus-Robert M\"uller, Shinichi Nakajima
Abstract: The long sampling time of diffusion models remains a significant bottleneck, which can be mitigated by reducing the number of diffusion time steps. However, the quality of samples with fewer steps is highly dependent on the noise schedule, i.e., the specific manner in which noise is introduced and the signal is reduced at each step. Although prior work has improved upon the original variance-preserving and variance-exploding schedules, these approaches $\textit{passively}$ adjust the total variance, without direct control over it. In this work, we propose a novel total-variance/signal-to-noise-ratio disentangled (TV/SNR) framework, where TV and SNR can be controlled independently. Our approach reveals that different existing schedules, where the TV explodes exponentially, can be $\textit{improved}$ by setting a constant TV schedule while preserving the same SNR schedule. Furthermore, generalizing the SNR schedule of the optimal transport flow matching significantly improves the performance in molecular structure generation, achieving few step generation of stable molecules. A similar tendency is observed in image generation, where our approach with a uniform diffusion time grid performs comparably to the highly tailored EDM sampler.
Authors: Junru Ren, Shaomin Wu
Abstract: Compared to local models built in a series-by-series manner, global models leverage relevant information across time series, resulting in improved forecasting performance and generalization capacity. Constructing global models on a set of time series is becoming mainstream in the field of time series forecasting. However, the advantages of global models may not always be realized when dealing with heterogeneous data. While they can adapt to heterogeneous datasets by increasing the model complexity, the model cannot be infinitely complex due to the finite sample size, which poses challenges for the application of global models. Additionally, determining whether the time series data is homogeneous or heterogeneous can be ambiguous in practice. To address these research gaps, this paper argues that the heterogeneity of the data should be defined by the global model used, and for each series, the portion not modelled by the global model represents heterogeneity. It further proposes two-stage hybrid models, which include a second stage to identify and model heterogeneous patterns. In this second stage, we can estimate either all local models or sub-global models across different domains divided based on heterogeneity. Experiments on four open datasets reveal that the proposed methods significantly outperform five existing models, indicating they contribute to fully unleash the potential of global models on heterogeneous datasets.
Authors: Karish Grover, Geoffrey J. Gordon, Christos Faloutsos
Abstract: Does the intrinsic curvature of complex networks hold the key to unveiling graph anomalies that conventional approaches overlook? Reconstruction-based graph anomaly detection (GAD) methods overlook such geometric outliers, focusing only on structural and attribute-level anomalies. To this end, we propose CurvGAD - a mixed-curvature graph autoencoder that introduces the notion of curvature-based geometric anomalies. CurvGAD introduces two parallel pipelines for enhanced anomaly interpretability: (1) Curvature-equivariant geometry reconstruction, which focuses exclusively on reconstructing the edge curvatures using a mixed-curvature, Riemannian encoder and Gaussian kernel-based decoder; and (2) Curvature-invariant structure and attribute reconstruction, which decouples structural and attribute anomalies from geometric irregularities by regularizing graph curvature under discrete Ollivier-Ricci flow, thereby isolating the non-geometric anomalies. By leveraging curvature, CurvGAD refines the existing anomaly classifications and identifies new curvature-driven anomalies. Extensive experimentation over 10 real-world datasets (both homophilic and heterophilic) demonstrates an improvement of up to 6.5% over state-of-the-art GAD methods.
Authors: Dan Busbridge, Amitis Shidani, Floris Weers, Jason Ramapuram, Etai Littwin, Russ Webb
Abstract: We provide a distillation scaling law that estimates distilled model performance based on a compute budget and its allocation between the student and teacher. Our findings reduce the risks associated with using distillation at scale; compute allocation for both the teacher and student models can now be done to maximize student performance. We provide compute optimal distillation recipes for when 1) a teacher exists, or 2) a teacher needs training. If many students are to be distilled, or a teacher already exists, distillation outperforms supervised pretraining until a compute level which grows predictably with student size. If one student is to be distilled and a teacher also needs training, supervised learning should be done instead. Additionally, we provide insights across our large scale study of distillation, which increase our understanding of distillation and inform experimental design.
Authors: Nikos Zarifis, Puqian Wang, Ilias Diakonikolas, Jelena Diakonikolas
Abstract: We study the task of learning Generalized Linear models (GLMs) in the agnostic model under the Gaussian distribution. We give the first polynomial-time algorithm that achieves a constant-factor approximation for \textit{any} monotone Lipschitz activation. Prior constant-factor GLM learners succeed for a substantially smaller class of activations. Our work resolves a well-known open problem, by developing a robust counterpart to the classical GLMtron algorithm (Kakade et al., 2011). Our robust learner applies more generally, encompassing all monotone activations with bounded $(2+\zeta)$-moments, for any fixed $\zeta>0$ -- a condition that is essentially necessary. To obtain our results, we leverage a novel data augmentation technique with decreasing Gaussian noise injection and prove a number of structural results that may be useful in other settings.
Authors: Saurabh Kataria, Ran Xiao, Timothy Ruchti, Matthew Clark, Jiaying Lu, Randall J. Lee, Jocelyn Grunwell, Xiao Hu
Abstract: Non-invasive patient monitoring for tracking and predicting adverse acute health events is an emerging area of research. We pursue in-hospital cardiac arrest (IHCA) prediction using only single-channel finger photoplethysmography (PPG) signals. Our proposed two-stage model Feature Extractor-Aggregator Network (FEAN) leverages powerful representations from pre-trained PPG foundation models (PPG-GPT of size up to 1 Billion) stacked with sequential classification models. We propose two FEAN variants ("1H", "FH") which use the latest one-hour and (max) 24-hour history to make decisions respectively. Our study is the first to present IHCA prediction results in ICU patients using only unimodal (continuous PPG signal) waveform deep representations. With our best model, we obtain an average of 0.79 AUROC over 24~h prediction window before CA event onset with our model peaking performance at 0.82 one hour before CA. We also provide a comprehensive analysis of our model through architectural tuning and PaCMAP visualization of patient health trajectory in latent space.
Authors: Nan K. Li, Angela Chang, David Sherman
Abstract: Drought is a frequent and costly natural disaster in California, with major negative impacts on agricultural production and water resource availability, particularly groundwater. This study investigated the performance of applying different machine learning approaches to predicting the U.S. Drought Monitor classification in California. Four approaches were used: a convolutional neural network (CNN), random forest, XGBoost, and long short term memory (LSTM) recurrent neural network, and compared to a baseline persistence model. We evaluated the models' performance in predicting severe drought (USDM drought category D2 or higher) using a macro F1 binary classification metric. The LSTM model emerged as the top performer, followed by XGBoost, CNN, and random forest. Further evaluation of our results at the county level suggested that the LSTM model would perform best in counties with more consistent drought patterns and where severe drought was more common, and the LSTM model would perform worse where drought scores increased rapidly. Utilizing 30 weeks of historical data, the LSTM model successfully forecasted drought scores for a 12-week period with a Mean Absolute Error (MAE) of 0.33, equivalent to less than half a drought category on a scale of 0 to 5. Additionally, the LSTM achieved a macro F1 score of 0.9, indicating high accuracy in binary classification for severe drought conditions. Evaluation of different window and future horizon sizes in weeks suggested that at least 24 weeks of data would result in the best performance, with best performance for shorter horizon sizes, particularly less than eight weeks.
Authors: Junpeng Zhang, Lei Cheng, Qing Li, Liang Lin, Quanshi Zhang
Abstract: In this paper, we find that the complexity of interactions encoded by a deep neural network (DNN) can explain its generalization power. We also discover that the confusing samples of a DNN, which are represented by non-generalizable interactions, are determined by its low-layer parameters. In comparison, other factors, such as high-layer parameters and network architecture, have much less impact on the composition of confusing samples. Two DNNs with different low-layer parameters usually have fully different sets of confusing samples, even though they have similar performance. This finding extends the understanding of the lottery ticket hypothesis, and well explains distinctive representation power of different DNNs.
Authors: Dhruv Rohatgi, Dylan J. Foster
Abstract: Algorithms for reinforcement learning (RL) in large state spaces crucially rely on supervised learning subroutines to estimate objects such as value functions or transition probabilities. Since only the simplest supervised learning problems can be solved provably and efficiently, practical performance of an RL algorithm depends on which of these supervised learning "oracles" it assumes access to (and how they are implemented). But which oracles are better or worse? Is there a minimal oracle? In this work, we clarify the impact of the choice of supervised learning oracle on the computational complexity of RL, as quantified by the oracle strength. First, for the task of reward-free exploration in Block MDPs in the standard episodic access model -- a ubiquitous setting for RL with function approximation -- we identify two-context regression as a minimal oracle, i.e. an oracle that is both necessary and sufficient (under a mild regularity assumption). Second, we identify one-context regression as a near-minimal oracle in the stronger reset access model, establishing a provable computational benefit of resets in the process. Third, we broaden our focus to Low-Rank MDPs, where we give cryptographic evidence that the analogous oracle from the Block MDP setting is insufficient.
Authors: Mantas Mazeika, Xuwang Yin, Rishub Tamirisa, Jaehyuk Lim, Bruce W. Lee, Richard Ren, Long Phan, Norman Mu, Adam Khoja, Oliver Zhang, Dan Hendrycks
Abstract: As AIs rapidly advance and become more agentic, the risk they pose is governed not only by their capabilities but increasingly by their propensities, including goals and values. Tracking the emergence of goals and values has proven a longstanding problem, and despite much interest over the years it remains unclear whether current AIs have meaningful values. We propose a solution to this problem, leveraging the framework of utility functions to study the internal coherence of AI preferences. Surprisingly, we find that independently-sampled preferences in current LLMs exhibit high degrees of structural coherence, and moreover that this emerges with scale. These findings suggest that value systems emerge in LLMs in a meaningful sense, a finding with broad implications. To study these emergent value systems, we propose utility engineering as a research agenda, comprising both the analysis and control of AI utilities. We uncover problematic and often shocking values in LLM assistants despite existing control measures. These include cases where AIs value themselves over humans and are anti-aligned with specific individuals. To constrain these emergent value systems, we propose methods of utility control. As a case study, we show how aligning utilities with a citizen assembly reduces political biases and generalizes to new scenarios. Whether we like it or not, value systems have already emerged in AIs, and much work remains to fully understand and control these emergent representations.
Authors: Hoony Kang, Wolfgang Losert
Abstract: The brain can rapidly adapt to new contexts and learn from limited data, a coveted characteristic that artificial intelligence algorithms have struggled to mimic. Inspired by oscillatory rhythms of the mechanical structures of neural cells, we developed a learning paradigm that is based on oscillations in link strengths and associates learning with the coordination of these oscillations. We find that this paradigm yields rapid adaptation and learning in artificial neural networks. Link oscillations can rapidly change coordination, endowing the network with the ability to sense subtle context changes in an unsupervised manner. In other words, the network generates the missing contextual tokens required to perform as a generalist AI architecture capable of predicting dynamics in multiple contexts. Oscillations also allow the network to extrapolate dynamics to never-seen-before contexts. These capabilities make our learning paradigm a powerful starting point for novel models of learning and cognition. Furthermore, learning through link coordination is agnostic to the specifics of the neural network architecture, hence our study opens the door for introducing rapid adaptation and learning capabilities into leading AI models.
Authors: Luuk Coopmans, Marcello Benedetti
Abstract: Quantum Boltzmann machines (QBMs) are machine-learning models for both classical and quantum data. We give an operational definition of QBM learning in terms of the difference in expectation values between the model and target, taking into account the polynomial size of the data set. By using the relative entropy as a loss function this problem can be solved without encountering barren plateaus. We prove that a solution can be obtained with stochastic gradient descent using at most a polynomial number of Gibbs states. We also prove that pre-training on a subset of the QBM parameters can only lower the sample complexity bounds. In particular, we give pre-training strategies based on mean-field, Gaussian Fermionic, and geometrically local Hamiltonians. We verify these models and our theoretical findings numerically on a quantum and a classical data set. Our results establish that QBMs are promising machine learning models.
Authors: Zander W. Blasingame, Chen Liu
Abstract: The optimization of the latents and parameters of diffusion models with respect to some differentiable metric defined on the output of the model is a challenging and complex problem. The sampling for diffusion models is done by solving either the probability flow ODE or diffusion SDE wherein a neural network approximates the score function allowing a numerical ODE/SDE solver to be used. However, naive backpropagation techniques are memory intensive, requiring the storage of all intermediate states, and face additional complexity in handling the injected noise from the diffusion term of the diffusion SDE. We propose a novel family of bespoke ODE solvers to the continuous adjoint equations for diffusion models, which we call AdjointDEIS. We exploit the unique construction of diffusion SDEs to further simplify the formulation of the continuous adjoint equations using exponential integrators. Moreover, we provide convergence order guarantees for our bespoke solvers. Significantly, we show that continuous adjoint equations for diffusion SDEs actually simplify to a simple ODE. Lastly, we demonstrate the effectiveness of AdjointDEIS for guided generation with an adversarial attack in the form of the face morphing problem. Our code will be released at https: //github.com/zblasingame/AdjointDEIS.
Authors: Sanjeev Naguleswaran
Abstract: The advent of quantum computing has opened new possibilities in data science, offering unique capabilities for addressing complex, data-intensive problems. Traditional machine learning algorithms often face challenges in high-dimensional or limited-quality datasets, which are common in healthcare. Quantum Machine Learning leverages quantum properties, such as superposition and entanglement, to enhance pattern recognition and classification, potentially surpassing classical approaches. This paper explores QML's application in healthcare, focusing on quantum kernel methods and hybrid quantum-classical networks for heart disease prediction and COVID-19 detection, assessing their feasibility and performance.
Authors: Bernardo Flores
Abstract: Modern data analysis often involves massive datasets with hundreds of thousands of observations, making traditional inference algorithms computationally prohibitive. Coresets are selection methods designed to choose a smaller subset of observations while maintaining similar learning performance. Conventional coreset approaches determine these weights by minimizing the Kullback-Leibler (KL) divergence between the likelihood functions of the full and weighted datasets; as a result, this makes them ill-posed for nonparametric models, where the likelihood is often intractable. We propose an alternative variational method which employs randomized posteriors and finds weights to match the unknown posterior predictive distributions conditioned on the full and reduced datasets. Our approach provides a general algorithm based on predictive recursions suitable for nonparametric priors. We evaluate the performance of the proposed coreset construction on diverse problems, including random partitions and density estimation.
Authors: Fei Gao
Abstract: Accurate decoding of neural spike trains and relating them to motor output is a challenging task due to the inherent sparsity and length in neural spikes and the complexity of brain circuits. This master project investigates experimental methods for decoding zebra finch motor outputs (in both discrete syllables and continuous spectrograms), from invasive neural recordings obtained from Neuropixels. There are three major achievements: (1) XGBoost with SHAP analysis trained on spike rates revealed neuronal interaction patterns crucial for syllable classification. (2) Novel method (tokenizing neural data with GPT2) and architecture (Mamba2) demonstrated potential for decoding of syllables using spikes. (3) A combined contrastive learning-VAE framework successfully generated spectrograms from binned neural data. This work establishes a promising foundation for neural decoding of complex motor outputs and offers several novel methodological approaches for processing sparse neural data.
Authors: Feng Liang, Haoyu Ma, Zecheng He, Tingbo Hou, Ji Hou, Kunpeng Li, Xiaoliang Dai, Felix Juefei-Xu, Samaneh Azadi, Animesh Sinha, Peizhao Zhang, Peter Vajda, Diana Marculescu
Abstract: Video personalization, which generates customized videos using reference images, has gained significant attention. However, prior methods typically focus on single-concept personalization, limiting broader applications that require multi-concept integration. Attempts to extend these models to multiple concepts often lead to identity blending, which results in composite characters with fused attributes from multiple sources. This challenge arises due to the lack of a mechanism to link each concept with its specific reference image. We address this with anchored prompts, which embed image anchors as unique tokens within text prompts, guiding accurate referencing during generation. Additionally, we introduce concept embeddings to encode the order of reference images. Our approach, Movie Weaver, seamlessly weaves multiple concepts-including face, body, and animal images-into one video, allowing flexible combinations in a single model. The evaluation shows that Movie Weaver outperforms existing methods for multi-concept video personalization in identity preservation and overall quality.
Authors: Cheryl Li, Tianyuan Xu, Yiwen Guo
Abstract: Chain-of-Thought (CoT) prompting has shown promise in enhancing the reasoning capabilities of large language models (LLMs) by generating natural language (NL) rationales that lead to the final answer. However, it struggles with numerical computation, which has somehow led to the development of program-aided techniques. Despite their potential, a persistent challenge remains: inconsistencies between LLM-reported reasoning steps and the logic in generated programs, which we term ``reasoning hallucinations." This stems from the inherent ambiguities of NL and the statistical nature of LLMs, which often lack rigorous logical coherence. To address this challenge, we propose a novel test-time scaling framework, Reasoning-as-Logic-Units (RaLU), which constructs a more reliable reasoning path by aligning logical units between the generated program and their corresponding NL descriptions. By decomposing the initially generated program into discrete units using static analysis, RaLU engages in an iterative dialogue with the LLM to judge, refine, and explain each unit. A rewind-and-correct mechanism ensures alignment between code statements and task requirements in each unit, ultimately forming a cohesive reasoning path under the program's logic, from which the model reaches a final solution. Our experiments demonstrate that RaLU significantly outperforms existing baselines in mathematical reasoning (GSM8K, MATH) and algorithmic reasoning (HumanEval+, MBPP+), underscoring its potential to advance LLM reasoning and programming by offering enhanced accuracy and interpretability.
Authors: Rath Minati, Date Hema
Abstract: The integration of Quantum Deep Learning (QDL) techniques into the landscape of financial risk analysis presents a promising avenue for innovation. This study introduces a framework for credit risk assessment in the banking sector, combining quantum deep learning techniques with adaptive modeling for Row-Type Dependent Predictive Analysis (RTDPA). By leveraging RTDPA, the proposed approach tailors predictive models to different loan categories, aiming to enhance the accuracy and efficiency of credit risk evaluation. While this work explores the potential of integrating quantum methods with classical deep learning for risk assessment, it focuses on the feasibility and performance of this hybrid framework rather than claiming transformative industry-wide impacts. The findings offer insights into how quantum techniques can complement traditional financial analysis, paving the way for further advancements in predictive modeling for credit risk.
Authors: Senkang Hu, Yihang Tao, Zihan Fang, Guowen Xu, Yiqin Deng, Sam Kwong, Yuguang Fang
Abstract: Collaborative perception (CP) is a promising method for safe connected and autonomous driving, which enables multiple vehicles to share sensing information to enhance perception performance. However, compared with single-vehicle perception, the openness of a CP system makes it more vulnerable to malicious attacks that can inject malicious information to mislead the perception of an ego vehicle, resulting in severe risks for safe driving. To mitigate such vulnerability, we first propose a new paradigm for malicious agent detection that effectively identifies malicious agents at the feature level without requiring verification of final perception results, significantly reducing computational overhead. Building on this paradigm, we introduce CP-GuardBench, the first comprehensive dataset provided to train and evaluate various malicious agent detection methods for CP systems. Furthermore, we develop a robust defense method called CP-Guard+, which enhances the margin between the representations of benign and malicious features through a carefully designed Dual-Centered Contrastive Loss (DCCLoss). Finally, we conduct extensive experiments on both CP-GuardBench and V2X-Sim, and demonstrate the superiority of CP-Guard+.
Authors: Tousif Rahman, Gang Mao, Bob Pattison, Sidharth Maheshwari, Marcos Sartori, Adrian Wheeldon, Rishad Shafik, Alex Yakovlev
Abstract: Embedded Field-Programmable Gate Arrays (eFPGAs) allow for the design of hardware accelerators of edge Machine Learning (ML) applications at a lower power budget compared with traditional FPGA platforms. However, the limited eFPGA logic and memory significantly constrain compute capabilities and model size. As such, ML application deployment on eFPGAs is in direct contrast with the most recent FPGA approaches developing architecture-specific implementations and maximizing throughput over resource frugality. This paper focuses on the opposite side of this trade-off: the proposed eFPGA accelerator focuses on minimizing resource usage and allowing flexibility for on-field recalibration over throughput. This allows for runtime changes in model size, architecture, and input data dimensionality without offline resynthesis. This is made possible through the use of a bitwise compressed inference architecture of the Tsetlin Machine (TM) algorithm. TM compute does not require any multiplication operations, being limited to only bitwise AND, OR, NOT, summations and additions. Additionally, TM model compression allows the entire model to fit within the on-chip block RAM of the eFPGA. The paper uses this accelerator to propose a strategy for runtime model tuning in the field. The proposed approach uses 2.5x fewer Look-up-Tables (LUTs) and 3.38x fewer registers than the current most resource-fugal design and achieves up to 129x energy reduction compared with low-power microcontrollers running the same ML application.
Authors: Sihao Wu, Xiaonan Si, Chi Xing, Jianhong Wang, Gaojie Jin, Guangliang Cheng, Lijun Zhang, Xiaowei Huang
Abstract: The integration of preference alignment with diffusion models (DMs) has emerged as a transformative approach to enhance image generation and editing capabilities. Although integrating diffusion models with preference alignment strategies poses significant challenges for novices at this intersection, comprehensive and systematic reviews of this subject are still notably lacking. To bridge this gap, this paper extensively surveys preference alignment with diffusion models in image generation and editing. First, we systematically review cutting-edge optimization techniques such as reinforcement learning with human feedback (RLHF), direct preference optimization (DPO), and others, highlighting their pivotal role in aligning preferences with DMs. Then, we thoroughly explore the applications of aligning preferences with DMs in autonomous driving, medical imaging, robotics, and more. Finally, we comprehensively discuss the challenges of preference alignment with DMs. To our knowledge, this is the first survey centered on preference alignment with DMs, providing insights to drive future innovation in this dynamic area.
Authors: Wenhao Wang, Adam Dziedzic, Grace C. Kim, Michael Backes, Franziska Boenisch
Abstract: Multi-modal models, such as CLIP, have demonstrated strong performance in aligning visual and textual representations, excelling in tasks like image retrieval and zero-shot classification. Despite this success, the mechanisms by which these models utilize training data, particularly the role of memorization, remain unclear. In uni-modal models, both supervised and self-supervised, memorization has been shown to be essential for generalization. However, it is not well understood how these findings would apply to CLIP, which incorporates elements from both supervised learning via captions that provide a supervisory signal similar to labels, and from self-supervised learning via the contrastive objective. To bridge this gap in understanding, we propose a formal definition of memorization in CLIP (CLIPMem) and use it to quantify memorization in CLIP models. Our results indicate that CLIP's memorization behavior falls between the supervised and self-supervised paradigms, with "mis-captioned" samples exhibiting highest levels of memorization. Additionally, we find that the text encoder contributes more to memorization than the image encoder, suggesting that mitigation strategies should focus on the text domain. Building on these insights, we propose multiple strategies to reduce memorization while at the same time improving utility--something that had not been shown before for traditional learning paradigms where reducing memorization typically results in utility decrease.
Authors: Do Yeong Kang, Yeong Hwan Oh, Chanwook Hwang, Jinhee Kim, Kang Eun Jeon, Jong Hwan Ko
Abstract: The implementation of Hyperdimensional Computing (HDC) on In-Memory Computing (IMC) architectures faces significant challenges due to the mismatch between highdimensional vectors and IMC array sizes, leading to inefficient memory utilization and increased computation cycles. This paper presents MEMHD, a Memory-Efficient Multi-centroid HDC framework designed to address these challenges. MEMHD introduces a clustering-based initialization method and quantization aware iterative learning for multi-centroid associative memory. Through these approaches and its overall architecture, MEMHD achieves a significant reduction in memory requirements while maintaining or improving classification accuracy. Our approach achieves full utilization of IMC arrays and enables one-shot (or few-shot) associative search. Experimental results demonstrate that MEMHD outperforms state-of-the-art binary HDC models, achieving up to 13.69% higher accuracy with the same memory usage, or 13.25x more memory efficiency at the same accuracy level. Moreover, MEMHD reduces computation cycles by up to 80x and array usage by up to 71x compared to baseline IMC mapping methods when mapped to 128x128 IMC arrays, while significantly improving energy and computation cycle efficiency.
Authors: Luoting Zhuang, Stephen H. Park, Steven J. Skates, Ashley E. Prosper, Denise R. Aberle, William Hsu
Abstract: Cancer evolves continuously over time through a complex interplay of genetic, epigenetic, microenvironmental, and phenotypic changes. This dynamic behavior drives uncontrolled cell growth, metastasis, immune evasion, and therapy resistance, posing challenges for effective monitoring and treatment. However, today's data-driven research in oncology has primarily focused on cross-sectional analysis using data from a single modality, limiting the ability to fully characterize and interpret the disease's dynamic heterogeneity. Advances in multiscale data collection and computational methods now enable the discovery of longitudinal multimodal biomarkers for precision oncology. Longitudinal data reveal patterns of disease progression and treatment response that are not evident from single-timepoint data, enabling timely abnormality detection and dynamic treatment adaptation. Multimodal data integration offers complementary information from diverse sources for more precise risk assessment and targeting of cancer therapy. In this review, we survey methods of longitudinal and multimodal modeling, highlighting their synergy in providing multifaceted insights for personalized care tailored to the unique characteristics of a patient's cancer. We summarize the current challenges and future directions of longitudinal multimodal analysis in advancing precision oncology.
Authors: Sicheng Wang, Jianhua Shan, Jianwei Zhang, Haozhang Gao, Hailiang Han, Yipeng Chen, Kang Wei, Chengkun Zhang, Kairos Wong, Jie Zhao, Lei Zhao, Bin Fang
Abstract: Embodied intelligence integrates multiple modalities, enabling agents to understand images, language, and actions simultaneously. However, existing models always depend on additional datasets or extensive pre-training to maximize performance improvements, consuming abundant training time and expensive hardware cost. To tackle this issue, we present RoboBERT, a novel end-to-end robotic manipulation model integrated with a unique training strategy. This model utilizes a CNN-based diffusion policy, enhancing and stabilizing the effectiveness of this model by separating training processes for different modalities. It also underscores the importance of data augmentation, verifying various techniques to significantly boost performance. Unlike models that depend on extra data or large foundation models, RoboBERT achieves a highly competitive success rate while using only language-labeled expert demonstrations and maintaining a relatively smaller model size. Specifically, RoboBERT achieves an average length of 4.52 on the CALVIN benchmark for \(ABCD \rightarrow D\) task, setting a new state-of-the-art (SOTA) record. Furthermore, when tested on a real robot, the model demonstrates superior performance, achieving a higher success rate than other methods trained with the same data. We propose that these concepts and methodologies of RoboBERT demonstrate extensive versatility and compatibility, contributing significantly to the development of lightweight multimodal robotic models. The code can be accessed on https://github.com/PeterWangsicheng/RoboBERT
Authors: Pengyu Wang, Jialu Li, Ling Shi
Abstract: With the increasing prevalence of autonomous vehicles (AVs), their vulnerability to various types of attacks has grown, presenting significant security challenges. In this paper, we propose a reinforcement learning (RL)-based approach for designing optimal stealthy integrity attacks on AV actuators. We also analyze the limitations of state-of-the-art RL-based secure controllers developed to counter such attacks. Through extensive simulation experiments, we demonstrate the effectiveness and efficiency of our proposed method.
Authors: Jiyoon Kim, Kang Eun Jeon, Yulhwa Kim, Jong Hwan Ko
Abstract: Compute-in-memory (CIM) is an efficient method for implementing deep neural networks (DNNs) but suffers from substantial overhead from analog-to-digital converters (ADCs), especially as ADC precision increases. Low-precision ADCs can re- duce this overhead but introduce partial-sum quantization errors degrading accuracy. Additionally, low-bit weight constraints, im- posed by cell limitations and the need for multiple cells for higher- bit weights, present further challenges. While fine-grained partial- sum quantization has been studied to lower ADC resolution effectively, weight granularity, which limits overall partial-sum quantized accuracy, remains underexplored. This work addresses these challenges by aligning weight and partial-sum quantization granularities at the column-wise level. Our method improves accuracy while maintaining dequantization overhead, simplifies training by removing two-stage processes, and ensures robustness to memory cell variations via independent column-wise scale factors. We also propose an open-source CIM-oriented convolution framework to handle fine-grained weights and partial-sums effi- ciently, incorporating a novel tiling method and group convolution. Experimental results on ResNet-20 (CIFAR-10, CIFAR-100) and ResNet-18 (ImageNet) show accuracy improvements of 0.99%, 2.69%, and 1.01%, respectively, compared to the best-performing related works. Additionally, variation analysis reveals the robust- ness of our method against memory cell variations. These findings highlight the effectiveness of our quantization scheme in enhancing accuracy and robustness while maintaining hardware efficiency in CIM-based DNN implementations. Our code is available at https://github.com/jiyoonkm/ColumnQuant.
Authors: Ping Zhang, Lei Su
Abstract: We present a theoretical analysis of GPU memory consumption during the training of DeepSeek models such as DeepSeek-v2 and DeepSeek-v3. Our primary objective is to clarify the device-level memory requirements associated with various distributed training configurations. Specifically, we examine critical factors influencing memory usage, including micro-batch size, activation recomputation policies, 3D parallelism, and ZeRO optimizations. It is important to emphasize that the training policies discussed in this report are not representative of DeepSeek's official configurations. Instead, they are explored to provide a deeper understanding of memory dynamics in training of large-scale mixture-of-experts model.
Authors: Behraj Khan, Rizwan Qureshi, Tahir Syed
Abstract: Despite being a successful example of emerging capability, vision-language foundation models for low-shot vision classification have a limited ability to sufficiently generalize to the target data distribution due to sample poverty, leading to sensitivity to variations in the data. A popular mitigation strategy is finetuning over multiple datasets, but domain generalization is expensive when practiced in this manner. This work examines both covariate shift between pre-training data and the underspecified target data, and \textit{confidence misalignment}, where the model's prediction confidence amplified by the limited data availability. We propose \textit{Confidence-Calibrated Covariate Shift Correction ($C3SC$)}, a unified framework to mitigate both covariate shift and confidence misalignment. $C3SC$ leverages Fisher information penalty for covariate shift correction and confidence misalignment penalty (CMP) to lower confidence on misclassified examples. Experimental results across various vision and covariate shift datasets demonstrates that $C3SC$ significantly improves in calibration (ECE) by $5.82\%$ at maximum. $C3SC$ shows better robustness as well by showing $3.5\%$ improvement in accuracy metric on challenging covariate shift datasets, making $C3SC$ a promising solution for reliable real-world vision-language low-shot applications under distribution shift.
Authors: Ao Li, Wei Fang, Hongbo Zhao, Le Lu, Ge Yang, Minfeng Xu
Abstract: In applications of diffusion models, controllable generation is of practical significance, but is also challenging. Current methods for controllable generation primarily focus on modifying the score function of diffusion models, while Mean Reverting (MR) Diffusion directly modifies the structure of the stochastic differential equation (SDE), making the incorporation of image conditions simpler and more natural. However, current training-free fast samplers are not directly applicable to MR Diffusion. And thus MR Diffusion requires hundreds of NFEs (number of function evaluations) to obtain high-quality samples. In this paper, we propose a new algorithm named MRS (MR Sampler) to reduce the sampling NFEs of MR Diffusion. We solve the reverse-time SDE and the probability flow ordinary differential equation (PF-ODE) associated with MR Diffusion, and derive semi-analytical solutions. The solutions consist of an analytical function and an integral parameterized by a neural network. Based on this solution, we can generate high-quality samples in fewer steps. Our approach does not require training and supports all mainstream parameterizations, including noise prediction, data prediction and velocity prediction. Extensive experiments demonstrate that MR Sampler maintains high sampling quality with a speedup of 10 to 20 times across ten different image restoration tasks. Our algorithm accelerates the sampling procedure of MR Diffusion, making it more practical in controllable generation.
Authors: M\'aty\'as Schubert, Tom Claassen, Sara Magliacane
Abstract: Causal discovery can be computationally demanding for large numbers of variables. If we only wish to estimate the causal effects on a small subset of target variables, we might not need to learn the causal graph for all variables, but only a small subgraph that includes the targets and their adjustment sets. In this paper, we focus on identifying causal effects between target variables in a computationally and statistically efficient way. This task combines causal discovery and effect estimation, aligning the discovery objective with the effects to be estimated. We show that definite non-ancestors of the targets are unnecessary to learn causal relations between the targets and to identify efficient adjustments sets. We sequentially identify and prune these definite non-ancestors with our Sequential Non-Ancestor Pruning (SNAP) framework, which can be used either as a preprocessing step to standard causal discovery methods, or as a standalone sound and complete causal discovery algorithm. Our results on synthetic and real data show that both approaches substantially reduce the number of independence tests and the computation time without compromising the quality of causal effect estimations.
Authors: Hela Mhiri, Ricard Puig, Sacha Lerch, Manuel S. Rudolph, Thiparat Chotibut, Supanut Thanasilp, Zo\"e Holmes
Abstract: Barren plateaus are fundamentally a statement about quantum loss landscapes on average but there can, and generally will, exist patches of barren plateau landscapes with substantial gradients. Previous work has studied certain classes of parameterized quantum circuits and found example regions where gradients vanish at worst polynomially in system size. Here we present a general bound that unifies all these previous cases and that can tackle physically-motivated ans\"atze that could not be analyzed previously. Concretely, we analytically prove a lower-bound on the variance of the loss that can be used to show that in a non-exponentially narrow region around a point with curvature the loss variance cannot decay exponentially fast. This result is complemented by numerics and an upper-bound that suggest that any loss function with a barren plateau will have exponentially vanishing gradients in any constant radius subregion. Our work thus suggests that while there are hopes to be able to warm-start variational quantum algorithms, any initialization strategy that cannot get increasingly close to the region of attraction with increasing problem size is likely inadequate.
Authors: Marina Maciel Ansanelli, Elie Wolfe, Robert W. Spekkens
Abstract: For two causal structures with the same set of visible variables, one is said to observationally dominate the other if the set of distributions over the visible variables realizable by the first contains the set of distributions over the visible variables realizable by the second. Knowing such dominance relations is useful for adjudicating between these structures given observational data. We here consider the problem of determining the partial order of equivalence classes of causal structures with latent variables relative to observational dominance. We provide a complete characterization of the dominance order in the case of three visible variables, and a partial characterization in the case of four visible variables. Our techniques also help to identify which observational equivalence classes have a set of realizable distributions that is characterized by nontrivial inequality constraints, analogous to Bell inequalities and instrumental inequalities. We find evidence that as one increases the number of visible variables, the equivalence classes satisfying nontrivial inequality constraints become ubiquitous. (Because such classes are the ones for which there can be a difference in the distributions that are quantumly and classically realizable, this implies that the potential for quantum-classical gaps is also ubiquitous.) Furthermore, we find evidence that constraint-based causal discovery algorithms that rely solely on conditional independence constraints have a significantly weaker distinguishing power among observational equivalence classes than algorithms that go beyond these (i.e., algorithms that also leverage nested Markov constraints and inequality constraints).
Authors: Chashi Mahiul Islam, Samuel Jacob Chacko, Preston Horne, Xiuwen Liu
Abstract: Multimodal Large Language Models (MLLMs) represent the cutting edge of AI technology, with DeepSeek models emerging as a leading open-source alternative offering competitive performance to closed-source systems. While these models demonstrate remarkable capabilities, their vision-language integration mechanisms introduce specific vulnerabilities. We implement an adapted embedding manipulation attack on DeepSeek Janus that induces targeted visual hallucinations through systematic optimization of image embeddings. Through extensive experimentation across COCO, DALL-E 3, and SVIT datasets, we achieve hallucination rates of up to 98.0% while maintaining high visual fidelity (SSIM > 0.88) of the manipulated images on open-ended questions. Our analysis demonstrates that both 1B and 7B variants of DeepSeek Janus are susceptible to these attacks, with closed-form evaluation showing consistently higher hallucination rates compared to open-ended questioning. We introduce a novel multi-prompt hallucination detection framework using LLaMA-3.1 8B Instruct for robust evaluation. The implications of these findings are particularly concerning given DeepSeek's open-source nature and widespread deployment potential. This research emphasizes the critical need for embedding-level security measures in MLLM deployment pipelines and contributes to the broader discussion of responsible AI implementation.
Authors: Nikita Kornilov, Philip Zmushko, Andrei Semenov, Alexander Gasnikov, Alexander Beznosikov
Abstract: The growing popularity of AI optimization problems involving severely corrupted data has increased the demand for methods capable of handling heavy-tailed noise, i.e., noise with bounded $\kappa$-th moment, $\kappa \in (1,2]$. For the widely used clipping technique, effectiveness heavily depends on the careful tuning of clipping levels throughout training. In this paper, we demonstrate that using only the sign of the input, without introducing additional hyperparameters, is sufficient to cope with heavy-tailed noise effectively. For smooth non-convex functions, we prove that SignSGD achieves optimal sample complexity $\tilde{O}\left(\varepsilon^{-\frac{3\kappa - 2}{\kappa - 1}}\right)$ with high probability for attaining an average gradient norm accuracy of $\varepsilon$. Under the assumption of symmetric noise, we use SignSGD with Majority Voting to extend this bound to the distributed optimization or reduce the sample complexity to $\tilde{O}(\varepsilon^{-4})$ in the case of a single worker with arbitrary parameters. Furthermore, we explore the application of the sign operator in zeroth-order optimization with an oracle that can only compare function values at two different points. We propose a novel method, MajorityVote-CompsSGD, and provide the first-known high-probability bound $\tilde{O}(\varepsilon^{-6})$ for the number of comparisons under symmetric noise assumption. Our theoretical findings are supported by the superior performance of sign-based methods in training Large Language Models.
Authors: Le-Tuyet-Nhi Pham, Dario Shariatian, Antonio Ocello, Giovanni Conforti, Alain Durmus
Abstract: This paper introduces the Discrete Markov Probabilistic Model (DMPM), a novel algorithm for discrete data generation. The algorithm operates in the space of bits $\{0,1\}^d$, where the noising process is a continuous-time Markov chain that can be sampled exactly via a Poissonian clock that flips labels uniformly at random. The time-reversal process, like the forward noise process, is a jump process, with its intensity governed by a discrete analogue of the classical score function. Crucially, this intensity is proven to be the conditional expectation of a function of the forward process, strengthening its theoretical alignment with score-based generative models while ensuring robustness and efficiency. We further establish convergence bounds for the algorithm under minimal assumptions and demonstrate its effectiveness through experiments on low-dimensional Bernoulli-distributed datasets and high-dimensional binary MNIST data. The results highlight its strong performance in generating discrete structures. This work bridges theoretical foundations and practical applications, advancing the development of effective and theoretically grounded discrete generative modeling.
Authors: Ruichen Zhang, Mufan Qiu, Zhen Tan, Mohan Zhang, Vincent Lu, Jie Peng, Kaidi Xu, Leandro Z. Agudelo, Peter Qian, Tianlong Chen
Abstract: Web browsing agents powered by large language models (LLMs) have shown tremendous potential in automating complex web-based tasks. Existing approaches typically rely on large LLMs (e.g., GPT-4o) to explore web environments and generate trajectory data, which is then used either for demonstration retrieval (for large LLMs) or to distill small LLMs (e.g., Llama3) in a process that remains decoupled from the exploration. In this paper, we propose AgentSymbiotic, an iterative framework that couples data synthesis with task-performance, yielding a "symbiotic improvement" for both large and small LLMs. Our study uncovers a complementary dynamic between LLM types: while large LLMs excel at generating high-quality trajectories for distillation, the distilled small LLMs-owing to their distinct reasoning capabilities-often choose actions that diverge from those of their larger counterparts. This divergence drives the exploration of novel trajectories, thereby enriching the synthesized data. However, we also observe that the performance of small LLMs becomes a bottleneck in this iterative enhancement process. To address this, we propose two innovations in LLM distillation: a speculative data synthesis strategy that mitigates off-policy bias, and a multi-task learning approach designed to boost the reasoning capabilities of the student LLM. Furthermore, we introduce a Hybrid Mode for Privacy Preservation to address user privacy concerns. Evaluated on the WEBARENA benchmark, AgentSymbiotic achieves SOTA performance with both LLM types. Our best Large LLM agent reaches 52%, surpassing the previous best of 45%, while our 8B distilled model demonstrates a competitive 49%, exceeding the prior best of 28%. Code will be released upon acceptance.
Authors: Yannik Frisch, Ssharvien Kumar Sivakumar, \c{C}a\u{g}han K\"oksal, Elsa B\"ohm, Felix Wagner, Adrian Gericke, Ghazal Ghazaei, Anirban Mukhopadhyay
Abstract: Surgical simulation offers a promising addition to conventional surgical training. However, available simulation tools lack photorealism and rely on hardcoded behaviour. Denoising Diffusion Models are a promising alternative for high-fidelity image synthesis, but existing state-of-the-art conditioning methods fall short in providing precise control or interactivity over the generated scenes. We introduce SurGrID, a Scene Graph to Image Diffusion Model, allowing for controllable surgical scene synthesis by leveraging Scene Graphs. These graphs encode a surgical scene's components' spatial and semantic information, which are then translated into an intermediate representation using our novel pre-training step that explicitly captures local and global information. Our proposed method improves the fidelity of generated images and their coherence with the graph input over the state-of-the-art. Further, we demonstrate the simulation's realism and controllability in a user assessment study involving clinical experts. Scene Graphs can be effectively used for precise and interactive conditioning of Denoising Diffusion Models for simulating surgical scenes, enabling high fidelity and interactive control over the generated content.
Authors: Xinyi Tan, Jiacheng Wang, Liansheng Wang
Abstract: Employing self-supervised learning (SSL) methodologies assumes par-amount significance in handling unlabeled polyp datasets when building deep learning-based automatic polyp segmentation models. However, the intricate privacy dynamics surrounding medical data often preclude seamless data sharing among disparate medical centers. Federated learning (FL) emerges as a formidable solution to this privacy conundrum, yet within the realm of FL, optimizing model generalization stands as a pressing imperative. Robust generalization capabilities are imperative to ensure the model's efficacy across diverse geographical domains post-training on localized client datasets. In this paper, a Federated self-supervised Domain Generalization method is proposed to enhance the generalization capacity of federated and Label-efficient intestinal polyp segmentation, named LFDG. Based on a classical SSL method, DropPos, LFDG proposes an adversarial learning-based data augmentation method (SSADA) to enhance the data diversity. LFDG further proposes a relaxation module based on Source-reconstruction and Augmentation-masking (SRAM) to maintain stability in feature learning. We have validated LFDG on polyp images from six medical centers. The performance of our method achieves 3.80% and 3.92% better than the baseline and other recent FL methods and SSL methods, respectively.
Authors: Shubham Gupta, Zichao Li, Tianyi Chen, Cem Subakan, Siva Reddy, Perouz Taslakian, Valentina Zantedeschi
Abstract: Document retrieval is a core component of question-answering systems, as it enables conditioning answer generation on new and large-scale corpora. While effective, the standard practice of encoding documents into high-dimensional embeddings for similarity search entails large memory and compute footprints, and also makes it hard to inspect the inner workings of the system. In this paper, we propose a tree-based method for organizing and representing reference documents at various granular levels, which offers the flexibility to balance cost and utility, and eases the inspection of the corpus content and retrieval operations. Our method, called ReTreever, jointly learns a routing function per internal node of a binary tree such that query and reference documents are assigned to similar tree branches, hence directly optimizing for retrieval performance. Our evaluations show that ReTreever generally preserves full representation accuracy. Its hierarchical structure further provides strong coarse representations and enhances transparency by indirectly learning meaningful semantic groupings. Among hierarchical retrieval methods, ReTreever achieves the best retrieval accuracy at the lowest latency, proving that this family of techniques can be viable in practical applications.
Authors: Yining Hong, Christopher S. Timperley, Christian K\"astner
Abstract: Machine learning (ML) components are increasingly integrated into software products, yet their complexity and inherent uncertainty often lead to unintended and hazardous consequences, both for individuals and society at large. Despite these risks, practitioners seldom adopt proactive approaches to anticipate and mitigate hazards before they occur. Traditional safety engineering approaches, such as Failure Mode and Effects Analysis (FMEA) and System Theoretic Process Analysis (STPA), offer systematic frameworks for early risk identification but are rarely adopted. This position paper advocates for integrating hazard analysis into the development of any ML-powered software product and calls for greater support to make this process accessible to developers. By using large language models (LLMs) to partially automate a modified STPA process with human oversight at critical steps, we expect to address two key challenges: the heavy dependency on highly experienced safety engineering experts, and the time-consuming, labor-intensive nature of traditional hazard analysis, which often impedes its integration into real-world development workflows. We illustrate our approach with a running example, demonstrating that many seemingly unanticipated issues can, in fact, be anticipated.
Authors: Oliver Biggar, Christos Papadimitriou
Abstract: Characterizing the limit behavior -- that is, the attractors -- of learning dynamics is one of the most fundamental open questions in game theory. In recent work in this front, it was conjectured that the attractors of the replicator dynamic are in one-to-one correspondence with the sink equilibria of the game -- the sink strongly connected components of a game's preference graph -- , and it was established that they do stand in at least one-to-many correspondence with them. We make threefold progress on the problem of characterizing attractors. First, we show through a topological construction that the one-to-one conjecture is false. Second, we make progress on the attractor characterization problem for two-player games by establishing that the one-to-one conjecture is true in the absence of a local pattern called a weak local source -- a pattern that is absent from zero-sum games. Finally, we look -- for the first time in this context -- at fictitious play, the longest-studied learning dynamic, and examine to what extent the conjecture generalizes there. We establish that under fictitious play, sink equilibria always contain attractors (sometimes strictly), and every attractor corresponds to a strongly connected set of nodes in the preference graph.
Authors: Ruihan Xu, Yiping Lu
Abstract: Randomized sketching accelerates large-scale numerical linear algebra by reducing computa- tional complexity. While the traditional sketch-and-solve approach reduces the problem size di- rectly through sketching, the sketch-and-precondition method leverages sketching to construct a computational friendly preconditioner. This preconditioner improves the convergence speed of iterative solvers applied to the original problem, maintaining accuracy in the full space. Further- more, the convergence rate of the solver improves at least linearly with the sketch size. Despite its potential, developing a sketch-and-precondition framework for randomized algorithms in low- rank matrix approximation remains an open challenge. We introduce the Error-Powered Sketched Inverse Iteration (EPSI) Method via run sketched Newton iteration for the Lagrange form as a sketch-and-precondition variant for randomized low-rank approximation. Our method achieves theoretical guarantees, including a convergence rate that improves at least linearly with the sketch size.
Authors: Haonan Shi, Tu Ouyang, An Wang
Abstract: Federated Distillation (FD) has emerged as a popular federated training framework, enabling clients to collaboratively train models without sharing private data. Public Dataset-Assisted Federated Distillation (PDA-FD), which leverages public datasets for knowledge sharing, has become widely adopted. Although PDA-FD enhances privacy compared to traditional Federated Learning, we demonstrate that the use of public datasets still poses significant privacy risks to clients' private training data. This paper presents the first comprehensive privacy analysis of PDA-FD in presence of an honest-but-curious server. We show that the server can exploit clients' inference results on public datasets to extract two critical types of private information: label distributions and membership information of the private training dataset. To quantify these vulnerabilities, we introduce two novel attacks specifically designed for the PDA-FD setting: a label distribution inference attack and innovative membership inference methods based on Likelihood Ratio Attack (LiRA). Through extensive evaluation of three representative PDA-FD frameworks (FedMD, DS-FL, and Cronus), our attacks achieve state-of-the-art performance, with label distribution attacks reaching minimal KL-divergence and membership inference attacks maintaining high True Positive Rates under low False Positive Rate constraints. Our findings reveal significant privacy risks in current PDA-FD frameworks and emphasize the need for more robust privacy protection mechanisms in collaborative learning systems.
Authors: Vincent D. Zaballa, Elliot E. Hui
Abstract: Simulation-based inference (SBI) is a method to perform inference on a variety of complex scientific models with challenging inference (inverse) problems. Bayesian Optimal Experimental Design (BOED) aims to efficiently use experimental resources to make better inferences. Various stochastic gradient-based BOED methods have been proposed as an alternative to Bayesian optimization and other experimental design heuristics to maximize information gain from an experiment. We demonstrate a link via mutual information bounds between SBI and stochastic gradient-based variational inference methods that permits BOED to be used in SBI applications as SBI-BOED. This link allows simultaneous optimization of experimental designs and optimization of amortized inference functions. We evaluate the pitfalls of naive design optimization using this method in a standard SBI task and demonstrate the utility of a well-chosen design distribution in BOED. We compare this approach on SBI-based models in real-world simulators in epidemiology and biology, showing notable improvements in inference.
Authors: Anjian Li, Sangjae Bae, David Isele, Ryne Beeson, Faizan M. Tariq
Abstract: Trajectory prediction and planning are fundamental components for autonomous vehicles to navigate safely and efficiently in dynamic environments. Traditionally, these components have often been treated as separate modules, limiting the ability to perform interactive planning and leading to computational inefficiency in multi-agent scenarios. In this paper, we present a novel unified and data-driven framework that integrates prediction and planning with a single consistency model. Trained on real-world human driving datasets, our consistency model generates samples from high-dimensional, multimodal joint trajectory distributions of the ego and multiple surrounding agents, enabling end-to-end predictive planning. It effectively produces interactive behaviors, such as proactive nudging and yielding to ensure both safe and efficient interactions with other road users. To incorporate additional planning constraints on the ego vehicle, we propose an alternating direction method for multi-objective guidance in online guided sampling. Compared to diffusion models, our consistency model achieves better performance with fewer sampling steps, making it more suitable for real-time deployment. Experimental results on Waymo Open Motion Dataset (WOMD) demonstrate our method's superiority in trajectory quality, constraint satisfaction, and interactive behavior compared to various existing approaches.
Authors: Jun Yamada, Alexander L. Mitchell, Jack Collins, Ingmar Posner
Abstract: This paper addresses the challenge of occluded robot grasping, i.e. grasping in situations where the desired grasp poses are kinematically infeasible due to environmental constraints such as surface collisions. Traditional robot manipulation approaches struggle with the complexity of non-prehensile or bimanual strategies commonly used by humans in these circumstances. State-of-the-art reinforcement learning (RL) methods are unsuitable due to the inherent complexity of the task. In contrast, learning from demonstration requires collecting a significant number of expert demonstrations, which is often infeasible. Instead, inspired by human bimanual manipulation strategies, where two hands coordinate to stabilise and reorient objects, we focus on a bimanual robotic setup to tackle this challenge. In particular, we introduce Constraint-based Manipulation for Bimanual Occluded Grasping (COMBO-Grasp), a learning-based approach which leverages two coordinated policies: a constraint policy trained using self-supervised datasets to generate stabilising poses and a grasping policy trained using RL that reorients and grasps the target object. A key contribution lies in value function-guided policy coordination. Specifically, during RL training for the grasping policy, the constraint policy's output is refined through gradients from a jointly trained value function, improving bimanual coordination and task performance. Lastly, COMBO-Grasp employs teacher-student policy distillation to effectively deploy point cloud-based policies in real-world environments. Empirical evaluations demonstrate that COMBO-Grasp significantly improves task success rates compared to competitive baseline approaches, with successful generalisation to unseen objects in both simulated and real-world environments.
Authors: Jihye Choi, Sai Rahul Rachuri, Ke Wang, Somesh Jha, Yizhen Wang
Abstract: Federated Learning (FL) enables collaborative model training while keeping client data private. However, exposing individual client updates makes FL vulnerable to reconstruction attacks. Secure aggregation mitigates such privacy risks but prevents the server from verifying the validity of each client update, creating a privacy-robustness tradeoff. Recent efforts attempt to address this tradeoff by enforcing checks on client updates using zero-knowledge proofs, but they support limited predicates and often depend on public validation data. We propose SLVR, a general framework that securely leverages clients' private data through secure multi-party computation. By utilizing clients' data, SLVR not only eliminates the need for public validation data, but also enables a wider range of checks for robustness, including cross-client accuracy validation. It also adapts naturally to distribution shifts in client data as it can securely refresh its validation data up-to-date. Our empirical evaluations show that SLVR improves robustness against model poisoning attacks, particularly outperforming existing methods by up to 50% under adaptive attacks. Additionally, SLVR demonstrates effective adaptability and stable convergence under various distribution shift scenarios.
Authors: Parsa Moradi, Hanzaleh Akbarinodehi, Mohammad Ali Maddah-Ali
Abstract: Conventional coded computing frameworks are predominantly tailored for structured computations, such as matrix multiplication and polynomial evaluation. Such tasks allow the reuse of tools and techniques from algebraic coding theory to improve the reliability of distributed systems in the presence of stragglers and adversarial servers. This paper lays the foundation for general coded computing, which extends the applicability of coded computing to handle a wide class of computations. In addition, it particularly addresses the challenging problem of managing adversarial servers. We demonstrate that, in the proposed scheme, for a system with $N$ servers, where $\mathcal{O}(N^a)$, $a \in [0,1)$, are adversarial, the supremum of the average approximation error over all adversarial strategies decays at a rate of $N^{\frac{6}{5}(a-1)}$, under minimal assumptions on the computing tasks. Furthermore, we show that within a general framework, the proposed scheme achieves optimal adversarial robustness, in terms of maximum number of adversarial servers it can tolerate. This marks a significant step toward practical and reliable general coded computing. Implementation results further validate the effectiveness of the proposed method in handling various computations, including inference in deep neural networks.
Authors: Samyadeep Basu, Vlad Morariu, Zichao Wang, Ryan Rossi, Cherry Zhao, Soheil Feizi, Varun Manjunatha
Abstract: Large language models are increasingly used to process documents and facilitate question-answering on them. In our paper, we extract mechanistic circuits for this real-world language modeling task: context-augmented language modeling for extractive question-answering (QA) tasks and understand the potential benefits of circuits towards downstream applications such as data attribution to context information. We extract circuits as a function of internal model components (e.g., attention heads, MLPs) using causal mediation analysis techniques. Leveraging the extracted circuits, we first understand the interplay between the model's usage of parametric memory and retrieved context towards a better mechanistic understanding of context-augmented language models. We then identify a small set of attention heads in our circuit which performs reliable data attribution by default, thereby obtaining attribution for free in just the model's forward pass. Using this insight, we then introduce ATTNATTRIB, a fast data attribution algorithm which obtains state-of-the-art attribution results across various extractive QA benchmarks. Finally, we show the possibility to steer the language model towards answering from the context, instead of the parametric memory by using the attribution from ATTNATTRIB as an additional signal during the forward pass. Beyond mechanistic understanding, our paper provides tangible applications of circuits in the form of reliable data attribution and model steering.
Authors: Guanghui Wang, Krishna Acharya, Lokranjan Lakshmikanthan, Vidya Muthukumar, Juba Ziani
Abstract: Performative prediction models account for feedback loops in decision-making processes where predictions influence future data distributions. While existing work largely assumes insensitivity of data distributions to small strategy changes, this assumption usually fails in real-world competitive (i.e. multi-agent) settings. For example, in Bertrand-type competitions, a small reduction in one firm's price can lead that firm to capture the entire demand, while all others sharply lose all of their customers. We study a representative setting of multi-agent performative prediction in which insensitivity assumptions do not hold, and investigate the convergence of natural dynamics. To do so, we focus on a specific game that we call the ''Bank Game'', where two lenders compete over interest rates and credit score thresholds. Consumers act similarly as to in a Bertrand Competition, with each consumer selecting the firm with the lowest interest rate that they are eligible for based on the firms' credit thresholds. Our analysis characterizes the equilibria of this game and demonstrates that when both firms use a common and natural no-regret learning dynamic -- exponential weights -- with proper initialization, the dynamics always converge to stable outcomes despite the general-sum structure. Notably, our setting admits multiple stable equilibria, with convergence dependent on initial conditions. We also provide theoretical convergence results in the stochastic case when the utility matrix is not fully known, but each learner can observe sufficiently many samples of consumers at each time step to estimate it, showing robustness to slight mis-specifications. Finally, we provide experimental results that validate our theoretical findings.
Authors: Chris Donahue, Shih-Lun Wu, Yewon Kim, Dave Carlton, Ryan Miyakawa, John Thickstun
Abstract: We present Hookpad Aria, a generative AI system designed to assist musicians in writing Western pop songs. Our system is seamlessly integrated into Hookpad, a web-based editor designed for the composition of lead sheets: symbolic music scores that describe melody and harmony. Hookpad Aria has numerous generation capabilities designed to assist users in non-sequential composition workflows, including: (1) generating left-to-right continuations of existing material, (2) filling in missing spans in the middle of existing material, and (3) generating harmony from melody and vice versa. Hookpad Aria is also a scalable data flywheel for music co-creation -- since its release in March 2024, Aria has generated 318k suggestions for 3k users who have accepted 74k into their songs. More information about Hookpad Aria is available at https://www.hooktheory.com/hookpad/aria
Authors: Minghong Fang, Xilong Wang, Neil Zhenqiang Gong
Abstract: Federated reinforcement learning (FRL) allows agents to jointly learn a global decision-making policy under the guidance of a central server. While FRL has advantages, its decentralized design makes it prone to poisoning attacks. To mitigate this, Byzantine-robust aggregation techniques tailored for FRL have been introduced. Yet, in our work, we reveal that these current Byzantine-robust techniques are not immune to our newly introduced Normalized attack. Distinct from previous attacks that targeted enlarging the distance of policy updates before and after an attack, our Normalized attack emphasizes on maximizing the angle of deviation between these updates. To counter these threats, we develop an ensemble FRL approach that is provably secure against both known and our newly proposed attacks. Our ensemble method involves training multiple global policies, where each is learnt by a group of agents using any foundational aggregation rule. These well-trained global policies then individually predict the action for a specific test state. The ultimate action is chosen based on a majority vote for discrete action systems or the geometric median for continuous ones. Our experimental results across different settings show that the Normalized attack can greatly disrupt non-ensemble Byzantine-robust methods, and our ensemble approach offers substantial resistance against poisoning attacks.
Authors: Samuel McCauley, Benjamin Moseley, Aidin Niaparast, Helia Niaparast, Shikha Singh
Abstract: The algorithms-with-predictions framework has been used extensively to develop online algorithms with improved beyond-worst-case competitive ratios. Recently, there is growing interest in leveraging predictions for designing data structures with improved beyond-worst-case running times. In this paper, we study the fundamental data structure problem of maintaining approximate shortest paths in incremental graphs in the algorithms-with-predictions model. Given a sequence $\sigma$ of edges that are inserted one at a time, the goal is to maintain approximate shortest paths from the source to each vertex in the graph at each time step. Before any edges arrive, the data structure is given a prediction of the online edge sequence $\hat{\sigma}$ which is used to ``warm start'' its state. As our main result, we design a learned algorithm that maintains $(1+\epsilon)$-approximate single-source shortest paths, which runs in $\tilde{O}(m \eta \log W/\epsilon)$ time, where $W$ is the weight of the heaviest edge and $\eta$ is the prediction error. We show these techniques immediately extend to the all-pairs shortest-path setting as well. Our algorithms are consistent (performing nearly as fast as the offline algorithm) when predictions are nearly perfect, have a smooth degradation in performance with respect to the prediction error and, in the worst case, match the best offline algorithm up to logarithmic factors. As a building block, we study the offline incremental approximate single-source shortest-paths problem. In this problem, the edge sequence $\sigma$ is known a priori and the goal is to efficiently return the length of the shortest paths in the intermediate graph $G_t$ consisting of the first $t$ edges, for all $t$. Note that the offline incremental problem is defined in the worst-case setting (without predictions) and is of independent interest.
Authors: Wei Xiao, Huiying Wang, Qifeng Zhou, Qing Wang
Abstract: Sequential recommendation is a key area in the field of recommendation systems aiming to model user interest based on historical interaction sequences with irregular intervals. While previous recurrent neural network-based and attention-based approaches have achieved significant results, they have limitations in capturing system continuity due to the discrete characteristics. In the context of continuous-time modeling, state space model (SSM) offers a potential solution, as it can effectively capture the dynamic evolution of user interest over time. However, existing SSM-based approaches ignore the impact of irregular time intervals within historical user interactions, making it difficult to model complexed user-item transitions in sequences. To address this issue, we propose a hybrid SSM-based model called SS4Rec for continuous-time sequential recommendation. SS4Rec integrates a time-aware SSM to handle irregular time intervals and a relation-aware SSM to model contextual dependencies, enabling it to infer user interest from both temporal and sequential perspectives. In the training process, the time-aware SSM and the relation-aware SSM are discretized by variable stepsizes according to user interaction time intervals and input data, respectively. This helps capture the continuous dependency from irregular time intervals and provides time-specific personalized recommendations. Experimental studies on five benchmark datasets demonstrate the superiority and effectiveness of SS4Rec.
Authors: Zhichao You, Xuewen Dong, Shujun Li, Ximeng Liu, Siqi Ma, Yulong Shen
Abstract: Reconstruction attacks against federated learning (FL) aim to reconstruct users' samples through users' uploaded gradients. Local differential privacy (LDP) is regarded as an effective defense against various attacks, including sample reconstruction in FL, where gradients are clipped and perturbed. Existing attacks are ineffective in FL with LDP since clipped and perturbed gradients obliterate most sample information for reconstruction. Besides, existing attacks embed additional sample information into gradients to improve the attack effect and cause gradient expansion, leading to a more severe gradient clipping in FL with LDP. In this paper, we propose a sample reconstruction attack against LDP-based FL with any target models to reconstruct victims' sensitive samples to illustrate that FL with LDP is not flawless. Considering gradient expansion in reconstruction attacks and noise in LDP, the core of the proposed attack is gradient compression and reconstructed sample denoising. For gradient compression, an inference structure based on sample characteristics is presented to reduce redundant gradients against LDP. For reconstructed sample denoising, we artificially introduce zero gradients to observe noise distribution and scale confidence interval to filter the noise. Theoretical proof guarantees the effectiveness of the proposed attack. Evaluations show that the proposed attack is the only attack that reconstructs victims' training samples in LDP-based FL and has little impact on the target model's accuracy. We conclude that LDP-based FL needs further improvements to defend against sample reconstruction attacks effectively.
Authors: Jessica Dai, Paula Gradu, Inioluwa Deborah Raji, Benjamin Recht
Abstract: When an individual reports a negative interaction with some system, how can their personal experience be contextualized within broader patterns of system behavior? We study the incident database problem, where individual reports of adverse events arrive sequentially, and are aggregated over time. In this work, our goal is to identify whether there are subgroups--defined by any combination of relevant features--that are disproportionately likely to experience harmful interactions with the system. We formalize this problem as a sequential hypothesis test, and identify conditions on reporting behavior that are sufficient for making inferences about disparities in true rates of harm across subgroups. We show that algorithms for sequential hypothesis tests can be applied to this problem with a standard multiple testing correction. We then demonstrate our method on real-world datasets, including mortgage decisions and vaccine side effects; on each, our method (re-)identifies subgroups known to experience disproportionate harm using only a fraction of the data that was initially used to discover them.
Authors: Konstantin Kolomeitsev (Almaty, Kazakhstan)
Abstract: In this work, we propose an architecture of LLM Modules that enables the transfer of knowledge from a large pre-trained model to a smaller model using an Enhanced Cross-Attention mechanism. In the proposed scheme, the Qwen2-1.5B model is frozen and its representations are passed through specially designed attention layers to the GPT-Neo-125M model, which is trained on limited computational resources. Experimental results on the Bespoke-Stratos-17k dataset demonstrate that after 15 epochs of training, the combined model generates responses comparable in quality to those obtained by distillation. We discuss the advantages of the modular approach, provide examples of input queries and comparative analysis, and outline prospects for further extension of the method.
Authors: Kunal Singh, Shreyas Singh, Mukund Khanna
Abstract: Recent advancements in Large Vision Language Models (LVLMs) have enabled the development of LVLM-based Graphical User Interface (GUI) agents under various paradigms. Training-based approaches, such as CogAgent and SeeClick, struggle with cross-dataset and cross-platform generalization due to their reliance on dataset-specific training. Generalist LVLMs, such as GPT-4V, employ Set-of-Marks (SoM) for action grounding, but obtaining SoM labels requires metadata like HTML source, which is not consistently available across platforms. Moreover, existing methods often specialize in singular GUI tasks rather than achieving comprehensive GUI understanding. To address these limitations, we introduce TRISHUL, a novel, training-free agentic framework that enhances generalist LVLMs for holistic GUI comprehension. Unlike prior works that focus on either action grounding (mapping instructions to GUI elements) or GUI referring (describing GUI elements given a location), TRISHUL seamlessly integrates both. At its core, TRISHUL employs Hierarchical Screen Parsing (HSP) and the Spatially Enhanced Element Description (SEED) module, which work synergistically to provide multi-granular, spatially, and semantically enriched representations of GUI elements. Our results demonstrate TRISHUL's superior performance in action grounding across the ScreenSpot, VisualWebBench, AITW, and Mind2Web datasets. Additionally, for GUI referring, TRISHUL surpasses the ToL agent on the ScreenPR benchmark, setting a new standard for robust and adaptable GUI comprehension.
Authors: Zi Yang, Ying Li, Zhidi Lin, Michael Minyi Zhang, Pablo M. Olmos
Abstract: The multi-view Gaussian process latent variable model (MV-GPLVM) aims to learn a unified representation from multi-view data but is hindered by challenges such as limited kernel expressiveness and low computational efficiency. To overcome these issues, we first introduce a new duality between the spectral density and the kernel function. By modeling the spectral density with a bivariate Gaussian mixture, we then derive a generic and expressive kernel termed Next-Gen Spectral Mixture (NG-SM) for MV-GPLVMs. To address the inherent computational inefficiency of the NG-SM kernel, we propose a random Fourier feature approximation. Combined with a tailored reparameterization trick, this approximation enables scalable variational inference for both the model and the unified latent representations. Numerical evaluations across a diverse range of multi-view datasets demonstrate that our proposed method consistently outperforms state-of-the-art models in learning meaningful latent representations.
Authors: Somaiyeh Dehghan, Mehmet Umut Sen, Berrin Yanikoglu
Abstract: Hate speech detection is a crucial task, especially on social media, where harmful content can spread quickly. Implementing machine learning models to automatically identify and address hate speech is essential for mitigating its impact and preventing its proliferation. The first step in developing an effective hate speech detection model is to acquire a high-quality dataset for training. Labeled data is foundational for most natural language processing tasks, but categorizing hate speech is difficult due to the diverse and often subjective nature of hate speech, which can lead to varying interpretations and disagreements among annotators. This paper examines strategies for addressing annotator disagreement, an issue that has been largely overlooked. In particular, we evaluate different approaches to deal with annotator disagreement regarding hate speech classification in Turkish tweets, based on a fine-tuned BERT model. Our work highlights the importance of the problem and provides state-of-art benchmark results for detection and understanding of hate speech in online discourse.
Authors: Yiping Liu, Mengxiao Zhang, Jiamou Liu, Song Yang
Abstract: Machine learning (ML) models have become essential tools in various scenarios. Their effectiveness, however, hinges on a substantial volume of data for satisfactory performance. Model marketplaces have thus emerged as crucial platforms bridging model consumers seeking ML solutions and data owners possessing valuable data. These marketplaces leverage model trading mechanisms to properly incentive data owners to contribute their data, and return a well performing ML model to the model consumers. However, existing model trading mechanisms often assume the data owners are willing to share their data before being paid, which is not reasonable in real world. Given that, we propose a novel mechanism, named Structural Importance based Model Trading (SIMT) mechanism, that assesses the data importance and compensates data owners accordingly without disclosing the data. Specifically, SIMT procures feature and label data from data owners according to their structural importance, and then trains a graph neural network for model consumers. Theoretically, SIMT ensures incentive compatible, individual rational and budget feasible. The experiments on five popular datasets validate that SIMT consistently outperforms vanilla baselines by up to $40\%$ in both MacroF1 and MicroF1.
Authors: Camilo Chac\'on Sartori, Christian Blum
Abstract: The integration of Large Language Models (LLMs) into optimization has created a powerful synergy, opening exciting research opportunities. This paper investigates how LLMs can enhance existing optimization algorithms. Using their pre-trained knowledge, we demonstrate their ability to propose innovative heuristic variations and implementation strategies. To evaluate this, we applied a non-trivial optimization algorithm, Construct, Merge, Solve and Adapt (CMSA) -- a hybrid metaheuristic for combinatorial optimization problems that incorporates a heuristic in the solution construction phase. Our results show that an alternative heuristic proposed by GPT-4o outperforms the expert-designed heuristic of CMSA, with the performance gap widening on larger and denser graphs. Project URL: https://imp-opt-algo-llms.surge.sh/
Authors: Bin Wu, Yihang Wang, Yuanhao Zeng, Jiawei Liu, Jiashu Zhao, Cheng Yang, Yawen Li, Long Xia, Dawei Yin, Chuan Shi
Abstract: Recommender systems (RS) serve as a fundamental tool for navigating the vast expanse of online information, with deep learning advancements playing an increasingly important role in improving ranking accuracy. Among these, graph neural networks (GNNs) excel at extracting higher-order structural information, while large language models (LLMs) are designed to process and comprehend natural language, making both approaches highly effective and widely adopted. Recent research has focused on graph foundation models (GFMs), which integrate the strengths of GNNs and LLMs to model complex RS problems more efficiently by leveraging the graph-based structure of user-item relationships alongside textual understanding. In this survey, we provide a comprehensive overview of GFM-based RS technologies by introducing a clear taxonomy of current approaches, diving into methodological details, and highlighting key challenges and future directions. By synthesizing recent advancements, we aim to offer valuable insights into the evolving landscape of GFM-based recommender systems.
Authors: Tao Huang, Junli Ren, Huayi Wang, Zirui Wang, Qingwei Ben, Muning Wen, Xiao Chen, Jianan Li, Jiangmiao Pang
Abstract: Standing-up control is crucial for humanoid robots, with the potential for integration into current locomotion and loco-manipulation systems, such as fall recovery. Existing approaches are either limited to simulations that overlook hardware constraints or rely on predefined ground-specific motion trajectories, failing to enable standing up across postures in real-world scenes. To bridge this gap, we present HoST (Humanoid Standing-up Control), a reinforcement learning framework that learns standing-up control from scratch, enabling robust sim-to-real transfer across diverse postures. HoST effectively learns posture-adaptive motions by leveraging a multi-critic architecture and curriculum-based training on diverse simulated terrains. To ensure successful real-world deployment, we constrain the motion with smoothness regularization and implicit motion speed bound to alleviate oscillatory and violent motions on physical hardware, respectively. After simulation-based training, the learned control policies are directly deployed on the Unitree G1 humanoid robot. Our experimental results demonstrate that the controllers achieve smooth, stable, and robust standing-up motions across a wide range of laboratory and outdoor environments. Videos are available at https://taohuang13.github.io/humanoid-standingup.github.io/.
URLs: https://taohuang13.github.io/humanoid-standingup.github.io/.
Authors: Anna Livia Croella, Veronica Piccialli, Antonio M. Sudoso
Abstract: Clustering is a fundamental technique in data analysis and machine learning, used to group similar data points together. Among various clustering methods, the Minimum Sum-of-Squares Clustering (MSSC) is one of the most widely used. MSSC aims to minimize the total squared Euclidean distance between data points and their corresponding cluster centroids. Due to the unsupervised nature of clustering, achieving global optimality is crucial, yet computationally challenging. The complexity of finding the global solution increases exponentially with the number of data points, making exact methods impractical for large-scale datasets. Even obtaining strong lower bounds on the optimal MSSC objective value is computationally prohibitive, making it difficult to assess the quality of heuristic solutions. We address this challenge by introducing a novel method to validate heuristic MSSC solutions through optimality gaps. Our approach employs a divide-and-conquer strategy, decomposing the problem into smaller instances that can be handled by an exact solver. The decomposition is guided by an auxiliary optimization problem, the "anticlustering problem", for which we design an efficient heuristic. Computational experiments demonstrate the effectiveness of the method for large-scale instances, achieving optimality gaps below 3% in most cases while maintaining reasonable computational times. These results highlight the practicality of our approach in assessing feasible clustering solutions for large datasets, bridging a critical gap in MSSC evaluation.
Authors: Samuel Erickson, Tobias Ryd\'en
Abstract: We present a new method for estimating high-dimensional sparse partial correlation and inverse covariance matrices, which exploits the connection between the inverse covariance matrix and linear regression. The method is a two-stage estimation method wherein each individual feature is regressed on all other features while positive semi-definiteness is enforced simultaneously. We provide statistical rates of convergence for the proposed method which match, and improve upon, the state-of-the-art for inverse covariance and partial correlation matrix estimation, respectively. We also propose an efficient proximal splitting algorithm for numerically computing the estimate. The effectiveness of the proposed method is demonstrated on both synthetic and real-world data.
Authors: Anastasia N. Krouglova, Hayden R. Johnson, Basile Confavreux, Michael Deistler, Pedro J. Gon\c{c}alves
Abstract: Across many domains of science, stochastic models are an essential tool to understand the mechanisms underlying empirically observed data. Models can be of different levels of detail and accuracy, with models of high-fidelity (i.e., high accuracy) to the phenomena under study being often preferable. However, inferring parameters of high-fidelity models via simulation-based inference is challenging, especially when the simulator is computationally expensive. We introduce MF-NPE, a multifidelity approach to neural posterior estimation that leverages inexpensive low-fidelity simulations to infer parameters of high-fidelity simulators within a limited simulation budget. MF-NPE performs neural posterior estimation with limited high-fidelity resources by virtue of transfer learning, with the ability to prioritize individual observations using active learning. On one statistical task with analytical ground-truth and two real-world tasks, MF-NPE shows comparable performance to current approaches while requiring up to two orders of magnitude fewer high-fidelity simulations. Overall, MF-NPE opens new opportunities to perform efficient Bayesian inference on computationally expensive simulators.
Authors: Hanlin Cai, Ozgur B. Akan
Abstract: Molecular communication (MC) provides a foundational framework for information transmission in the Internet of Bio-Nano Things (IoBNT), where efficiency and reliability are crucial. However, the inherent limitations of molecular channels, such as low transmission rates, noise, and inter-symbol interference (ISI), limit their ability to support complex data transmission. This paper proposes an end-to-end semantic learning framework designed to optimize task-oriented molecular communication, with a focus on biomedical diagnostic tasks under resource-constrained conditions. The proposed framework employs a deep encoder-decoder architecture to efficiently extract, quantize, and decode semantic features, prioritizing task-relevant semantic information to enhance diagnostic classification performance. Additionally, a probabilistic channel network is introduced to approximate molecular propagation dynamics, enabling gradient-based optimization for end-to-end learning. Experimental results demonstrate that the proposed semantic framework improves diagnostic accuracy by at least 25% compared to conventional JPEG compression with LDPC coding methods under resource-constrained communication scenarios.
Authors: Nathan Vandemoortele, Bram Steenwinckel, Femke Ongenae, Sofie Van Hoecke
Abstract: We present Label Space Reduction (LSR), a novel method for improving zero-shot classification performance of Large Language Models (LLMs). LSR iteratively refines the classification label space by systematically ranking and reducing candidate classes, enabling the model to concentrate on the most relevant options. By leveraging unlabeled data with the statistical learning capabilities of data-driven models, LSR dynamically optimizes the label space representation at test time. Our experiments across seven benchmarks demonstrate that LSR improves macro-F1 scores by an average of 7.0% (up to 14.2%) with Llama-3.1-70B and 3.3% (up to 11.1%) with Claude-3.5-Sonnet compared to standard zero-shot classification baselines. To reduce the computational overhead of LSR, which requires an additional LLM call at each iteration, we propose distilling the model into a probabilistic classifier, allowing for efficient inference.
Authors: Felix Stollenwerk, Tobias Stollenwerk
Abstract: Despite their remarkable capabilities, LLMs learn word representations that exhibit the undesirable yet poorly understood feature of anisotropy. In this paper, we argue that the second moment in Adam is a cause of anisotropic embeddings, and suggest a modified optimizer called Coupled Adam to mitigate the problem. Our experiments demonstrate that Coupled Adam significantly improves the quality of embeddings, while also leading to better upstream and downstream performance on large enough datasets.
Authors: Thomas Cass, Francesco Piatti, Jeffrey Pei
Abstract: Signature kernels have emerged as a powerful tool within kernel methods for sequential data. In the paper "The Signature Kernel is the solution of a Goursat PDE", the authors identify a kernel trick that demonstrates that, for continuously differentiable paths, the signature kernel satisfies a Goursat problem for a hyperbolic partial differential equation (PDE) in two independent time variables. While finite difference methods have been explored for this PDE, they face limitations in accuracy and stability when handling highly oscillatory inputs. In this work, we introduce two advanced numerical schemes that leverage polynomial representations of boundary conditions through either approximation or interpolation techniques, and rigorously establish the theoretical convergence of the polynomial approximation scheme. Experimental evaluations reveal that our approaches yield improvements of several orders of magnitude in mean absolute percentage error (MAPE) compared to traditional finite difference schemes, without increasing computational complexity. Furthermore, like finite difference methods, our algorithms can be GPU-parallelized to reduce computational complexity from quadratic to linear in the length of the input sequences, thereby improving scalability for high-frequency data. We have implemented these algorithms in a dedicated Python library, which is publicly available at: https://github.com/FrancescoPiatti/polysigkernel.
Authors: Qifan Yu, Zhenyu He, Sijie Li, Xun Zhou, Jun Zhang, Jingjing Xu, Di He
Abstract: Chain-of-Thought (CoT) prompting has emerged as a powerful technique for enhancing language model's reasoning capabilities. However, generating long and correct CoT trajectories is challenging. Recent studies have demonstrated that Looped Transformers possess remarkable length generalization capabilities, but their limited generality and adaptability prevent them from serving as an alternative to auto-regressive solutions. To better leverage the strengths of Looped Transformers, we propose RELAY (REasoning through Loop Alignment iterativelY). Specifically, we align the steps of Chain-of-Thought (CoT) reasoning with loop iterations and apply intermediate supervision during the training of Looped Transformers. This additional iteration-wise supervision not only preserves the Looped Transformer's ability for length generalization but also enables it to predict CoT reasoning steps for unseen data. Therefore, we leverage this Looped Transformer to generate accurate reasoning chains for complex problems that exceed the training length, which will then be used to fine-tune an auto-regressive model. We conduct extensive experiments, and the results demonstrate the effectiveness of our approach, with significant improvements in the performance of the auto-regressive model. Code will be released at https://github.com/qifanyu/RELAY.
Authors: Ali Nazari, Michael Weiss
Abstract: Topic modeling often requires examining topics from multiple perspectives to uncover hidden patterns, especially in less explored areas. This paper presents an approach to address this need, utilizing weighted keywords from various aspects derived from a domain knowledge. The research method starts with standard topic modeling. Then, it adds a process consisting of four key steps. First, it defines keywords for each aspect. Second, it gives weights to these keywords based on their relevance. Third, it calculates relevance scores for aspect-weighted keywords and topic keywords to create aspect-topic models. Fourth, it uses these scores to tune relevant new documents. Finally, the generated topic models are interpreted and validated. The findings show that top-scoring documents are more likely to be about the same aspect of a topic. This highlights the model's effectiveness in finding the related documents to the aspects.
Authors: Fei Zheng, Nicolas Duchateau
Abstract: Model-based clustering techniques have been widely applied to various application areas, while most studies focus on canonical mixtures with unique component distribution form. However, this strict assumption is often hard to satisfy. In this paper, we consider the more flexible Copula-Based Mixture Models (CBMMs) for clustering, which allow heterogeneous component distributions composed by flexible choices of marginal and copula forms. More specifically, we propose an adaptation of the Generalized Iterative Conditional Estimation (GICE) algorithm to identify the CBMMs in an unsupervised manner, where the marginal and copula forms and their parameters are estimated iteratively. GICE is adapted from its original version developed for switching Markov model identification with the choice of realization time. Our CBMM-GICE clustering method is then tested on synthetic two-cluster data (N=2000 samples) with discussion of the factors impacting its convergence. Finally, it is compared to the Expectation Maximization identified mixture models with unique component form on the entire MNIST database (N=70000), and on real cardiac magnetic resonance data (N=276) to illustrate its value for imaging applications.
Authors: Maher A Dayeh, Michael J Starkey, Subhamoy Chatterjee, Heather Elliott, Samuel Hart, Kimberly Moreland
Abstract: Space weather forecasting is critical for mitigating radiation risks in space exploration and protecting Earth-based technologies from geomagnetic disturbances. This paper presents the development of a Machine Learning (ML)- ready data processing tool for Near Real-Time (NRT) space weather forecasting. By merging data from diverse NRT sources such as solar imagery, magnetic field measurements, and energetic particle fluxes, the tool addresses key gaps in current space weather prediction capabilities. The tool processes and structures the data for machine learning models, focusing on time-series forecasting and event detection for extreme solar events. It provides users with a framework to download, process, and label data for ML applications, streamlining the workflow for improved NRT space weather forecasting and scientific research.
Authors: Shixiang Tang, Yizhou Wang, Lu Chen, Yuan Wang, Sida Peng, Dan Xu, Wanli Ouyang
Abstract: Human understanding and generation are critical for modeling digital humans and humanoid embodiments. Recently, Human-centric Foundation Models (HcFMs) inspired by the success of generalist models, such as large language and vision models, have emerged to unify diverse human-centric tasks into a single framework, surpassing traditional task-specific approaches. In this survey, we present a comprehensive overview of HcFMs by proposing a taxonomy that categorizes current approaches into four groups: (1) Human-centric Perception Foundation Models that capture fine-grained features for multi-modal 2D and 3D understanding. (2) Human-centric AIGC Foundation Models that generate high-fidelity, diverse human-related content. (3) Unified Perception and Generation Models that integrate these capabilities to enhance both human understanding and synthesis. (4) Human-centric Agentic Foundation Models that extend beyond perception and generation to learn human-like intelligence and interactive behaviors for humanoid embodied tasks. We review state-of-the-art techniques, discuss emerging challenges and future research directions. This survey aims to serve as a roadmap for researchers and practitioners working towards more robust, versatile, and intelligent digital human and embodiments modeling.
Authors: Wonduk Seo, Seunghyun Lee
Abstract: Query expansion is widely used in Information Retrieval (IR) to improve search outcomes by enriching queries with additional contextual information. Although recent Large Language Model (LLM) based methods generate pseudo-relevant content and expanded terms via multiple prompts, they often yield repetitive, narrow expansions that lack the diverse context needed to retrieve all relevant information. In this paper, we introduce QA-Expand, a novel and effective framework for query expansion. It first generates multiple relevant questions from the initial query and subsequently produces corresponding pseudo-answers as surrogate documents. A feedback model further rewrites and filters these answers to ensure only the most informative augmentations are incorporated. Extensive experiments on benchmarks such as BEIR and TREC demonstrate that QA-Expand enhances retrieval performance by up to 13% over state-of-the-art methods, offering a robust solution for modern retrieval challenges.
Authors: Giampaolo Bovenzi, Francesco Cerasuolo, Domenico Ciuonzo, Davide Di Monda, Idio Guarino, Antonio Montieri, Valerio Persico, Antonio Pescap\`e
Abstract: Generative Artificial Intelligence (GenAI) models such as LLMs, GPTs, and Diffusion Models have recently gained widespread attention from both the research and the industrial communities. This survey explores their application in network monitoring and management, focusing on prominent use cases, as well as challenges and opportunities. We discuss how network traffic generation and classification, network intrusion detection, networked system log analysis, and network digital assistance can benefit from the use of GenAI models. Additionally, we provide an overview of the available GenAI models, datasets for large-scale training phases, and platforms for the development of such models. Finally, we discuss research directions that potentially mitigate the roadblocks to the adoption of GenAI for network monitoring and management. Our investigation aims to map the current landscape and pave the way for future research in leveraging GenAI for network monitoring and management.
Authors: Kaelan Donatella, Samuel Duffield, Denis Melanson, Maxwell Aifer, Phoebe Klett, Rajath Salegame, Zach Belateche, Gavin Crooks, Antonio J. Martinez, Patrick J. Coles
Abstract: Many hardware proposals have aimed to accelerate inference in AI workloads. Less attention has been paid to hardware acceleration of training, despite the enormous societal impact of rapid training of AI models. Physics-based computers, such as thermodynamic computers, offer an efficient means to solve key primitives in AI training algorithms. Optimizers that normally would be computationally out-of-reach (e.g., due to expensive matrix inversions) on digital hardware could be unlocked with physics-based hardware. In this work, we propose a scalable algorithm for employing thermodynamic computers to accelerate a popular second-order optimizer called Kronecker-factored approximate curvature (K-FAC). Our asymptotic complexity analysis predicts increasing advantage with our algorithm as $n$, the number of neurons per layer, increases. Numerical experiments show that even under significant quantization noise, the benefits of second-order optimization can be preserved. Finally, we predict substantial speedups for large-scale vision and graph problems based on realistic hardware characteristics.
Authors: Michael R. Douglas, Kyu-Hwan Lee
Abstract: Can machine learning help discover new mathematical structures? In this article we discuss an approach to doing this which one can call "mathematical data science". In this paradigm, one studies mathematical objects collectively rather than individually, by creating datasets and doing machine learning experiments and interpretations. After an overview, we present two case studies: murmurations in number theory and loadings of partitions related to Kronecker coefficients in representation theory and combinatorics.
Authors: Jeremiah Birrell
Abstract: We derive novel concentration inequalities that bound the statistical error for a large class of stochastic optimization problems, focusing on the case of unbounded objective functions. Our derivations utilize the following tools: 1) A new form of McDiarmid's inequality that is based on sample dependent one component difference bounds and which leads to a novel uniform law of large numbers result for unbounded functions. 2) A Rademacher complexity bound for families of functions that satisfy an appropriate local Lipschitz property. As an application of these results, we derive statistical error bounds for denoising score matching (DSM), an application that inherently requires one to consider unbounded objective functions, even when the data distribution has bounded support. In addition, our results establish the benefit of sample reuse in algorithms that employ easily sampled auxiliary random variables in addition to the training data, e.g., as in DSM, which uses auxiliary Gaussian random variables.
Authors: Jun Lyu, Lipeng Ning, William Consagra, Qiang Liu, Richard J. Rushmore, Berkin Bilgic, Yogesh Rathi
Abstract: Purpose: To develop and validate a novel image reconstruction technique using implicit neural representations (INR) for multi-view thick-slice acquisitions while reducing the scan time but maintaining high signal-to-noise ratio (SNR). Methods: We propose Rotating-view super-resolution (ROVER)-MRI, an unsupervised neural network-based algorithm designed to reconstruct MRI data from multi-view thick slices, effectively reducing scan time by 2-fold while maintaining fine anatomical details. We compare our method to both bicubic interpolation and the current state-of-the-art regularized least-squares super-resolution reconstruction (LS-SRR) technique. Validation is performed using ground-truth ex-vivo monkey brain data, and we demonstrate superior reconstruction quality across several in-vivo human datasets. Notably, we achieve the reconstruction of a whole human brain in-vivo T2-weighted image with an unprecedented 180{\mu}m isotropic spatial resolution, accomplished in just 17 minutes of scan time on a 7T MRI scanner. Results: ROVER-MRI outperformed LS-SRR method in terms of reconstruction quality with 22.4% lower relative error (RE) and 7.5% lower full-width half maximum (FWHM) indicating better preservation of fine structural details in nearly half the scan time. Conclusion: ROVER-MRI offers an efficient and robust approach for mesoscale MR imaging, enabling rapid, high-resolution whole-brain scans. Its versatility holds great promise for research applications requiring anatomical details and time-efficient imaging.
Authors: Xiaoxia Xu, Xidong Mu, Yuanwei Liu, Arumugam Nallanathan
Abstract: A novel pinching antenna system (PASS)-enabled downlink multi-user multiple-input single-output (MISO) framework is proposed. PASS consists of multiple waveguides spanning over thousands of wavelength, which equip numerous low-cost dielectric particles, named pinching antennas (PAs), to radiate signals into free space. The positions of PAs can be reconfigured to change both the large-scale path losses and phases of signals, thus facilitating the novel pinching beamforming design. A sum rate maximization problem is formulated, which jointly optimizes the transmit and pinching beamforming to adaptively achieve constructive signal enhancement and destructive interference mitigation. To solve this highly coupled and nonconvex problem, both optimization-based and learning-based methods are proposed. 1) For the optimization-based method, a majorization-minimization and penalty dual decomposition (MM-PDD) algorithm is developed, which handles the nonconvex complex exponential component using a Lipschitz surrogate function and then invokes PDD for problem decoupling. 2) For the learning-based method, a novel Karush-Kuhn-Tucker (KKT)-guided dual learning (KDL) approach is proposed, which enables KKT solutions to be reconstructed in a data-driven manner by learning dual variables. Following this idea, a KDL-Tranformer algorithm is developed, which captures both inter-PA/inter-user dependencies and channel-state-information (CSI)-beamforming dependencies by attention mechanisms. Simulation results demonstrate that: i) The proposed PASS framework significantly outperforms conventional massive multiple input multiple output (MIMO) system even with a few PAs. ii) The proposed KDL-Transformer can improve over 30% system performance than MM-PDD algorithm, while achieving a millisecond-level response on modern GPUs.
Authors: Carles Balsells-Rodas, Xavier Sumba, Tanmayee Narendra, Ruibo Tu, Gabriele Schweikert, Hedvig Kjellstrom, Yingzhen Li
Abstract: Causal discovery, i.e., inferring underlying causal relationships from observational data, is highly challenging for AI systems. In a time series modeling context, traditional causal discovery methods mainly consider constrained scenarios with fully observed variables and/or data from stationary time-series. We develop a causal discovery approach to handle a wide class of nonstationary time series that are conditionally stationary, where the nonstationary behaviour is modeled as stationarity conditioned on a set of latent state variables. Named State-Dependent Causal Inference (SDCI), our approach is able to recover the underlying causal dependencies, with provable identifiablity for the state-dependent causal structures. Empirical experiments on nonlinear particle interaction data and gene regulatory networks demonstrate SDCI's superior performance over baseline causal discovery methods. Improved results over non-causal RNNs on modeling NBA player movements demonstrate the potential of our method and motivate the use of causality-driven methods for forecasting.
Authors: Zhuangwei Shi
Abstract: Probabilistic theory and differential equation are powerful tools for the interpretability and guidance of the design of machine learning models, especially for illuminating the mathematical motivation of learning latent variable from observation. Subspace learning maps high-dimensional features on low-dimensional subspace to capture efficient representation. Graphs are widely applied for modeling latent variable learning problems, and graph neural networks implement deep learning architectures on graphs. Inspired by probabilistic theory and differential equations, this paper conducts notes and proposals about graph neural networks to solve subspace learning problems by variational inference and differential equation.
Authors: Florian Kalinke, Marco Heyden, Georg Gntuni, Edouard Fouch\'e, Klemens B\"ohm
Abstract: Detecting changes is of fundamental importance when analyzing data streams and has many applications, e.g., in predictive maintenance, fraud detection, or medicine. A principled approach to detect changes is to compare the distributions of observations within the stream to each other via hypothesis testing. Maximum mean discrepancy (MMD), a (semi-)metric on the space of probability distributions, provides powerful non-parametric two-sample tests on kernel-enriched domains. In particular, MMD is able to detect any disparity between distributions under mild conditions. However, classical MMD estimators suffer from a quadratic runtime complexity, which renders their direct use for change detection in data streams impractical. In this article, we propose a new change detection algorithm, called Maximum Mean Discrepancy on Exponential Windows (MMDEW), that combines the benefits of MMD with an efficient computation based on exponential windows. We prove that MMDEW enjoys polylogarithmic runtime and logarithmic memory complexity and show empirically that it outperforms the state of the art on benchmark data streams.
Authors: Yulun Wu, Robert A. Barton, Zichen Wang, Vassilis N. Ioannidis, Carlo De Donno, Layne C. Price, Luis F. Voloch, George Karypis
Abstract: Predicting the responses of a cell under perturbations may bring important benefits to drug discovery and personalized therapeutics. In this work, we propose a novel graph variational Bayesian causal inference framework to predict a cell's gene expressions under counterfactual perturbations (perturbations that this cell did not factually receive), leveraging information representing biological knowledge in the form of gene regulatory networks (GRNs) to aid individualized cellular response predictions. Aiming at a data-adaptive GRN, we also developed an adjacency matrix updating technique for graph convolutional networks and used it to refine GRNs during pre-training, which generated more insights on gene relations and enhanced model performance. Additionally, we propose a robust estimator within our framework for the asymptotically efficient estimation of marginal perturbation effect, which is yet to be carried out in previous works. With extensive experiments, we exhibited the advantage of our approach over state-of-the-art deep learning models for individual response prediction.
Authors: Piyush Tiwary, Atri Guha, Subhodip Panda, Prathosh A. P
Abstract: Owing to the growing concerns about privacy and regulatory compliance, it is desirable to regulate the output of generative models. To that end, the objective of this work is to prevent the generation of outputs containing undesired features from a pre-trained Generative Adversarial Network (GAN) where the underlying training data set is inaccessible. Our approach is inspired by the observation that the parameter space of GANs exhibits meaningful directions that can be leveraged to suppress specific undesired features. However, such directions usually result in the degradation of the quality of generated samples. Our proposed two-stage method, known as 'Adapt-then-Unlearn,' excels at unlearning such undesirable features while also maintaining the quality of generated samples. In the initial stage, we adapt a pre-trained GAN on a set of negative samples (containing undesired features) provided by the user. Subsequently, we train the original pre-trained GAN using positive samples, along with a repulsion regularizer. This regularizer encourages the learned model parameters to move away from the parameters of the adapted model (first stage) while not degrading the generation quality. We provide theoretical insights into the proposed method. To the best of our knowledge, our approach stands as the first method addressing unlearning within the realm of high-fidelity GANs (such as StyleGAN). We validate the effectiveness of our method through comprehensive experiments, encompassing both class-level unlearning on the MNIST and AFHQ dataset and feature-level unlearning tasks on the CelebA-HQ dataset. Our code and implementation is available at: https://github.com/atriguha/Adapt_Unlearn.
Authors: Haosen Ge, Hamsa Bastani, Osbert Bastani
Abstract: Existing approaches to algorithmic fairness aim to ensure equitable outcomes if human decision-makers comply perfectly with algorithmic decisions. However, perfect compliance with the algorithm is rarely a reality or even a desirable outcome in human-AI collaboration. Yet, recent studies have shown that selective compliance with fair algorithms can amplify discrimination relative to the prior human policy. As a consequence, ensuring equitable outcomes requires fundamentally different algorithmic design principles that ensure robustness to the decision-maker's (a priori unknown) compliance pattern. We define the notion of compliance-robustly fair algorithmic recommendations that are guaranteed to (weakly) improve fairness in decisions, regardless of the human's compliance pattern. We propose a simple optimization strategy to identify the best performance-improving compliance-robustly fair policy. However, we show that it may be infeasible to design algorithmic recommendations that are simultaneously fair in isolation, compliance-robustly fair, and more accurate than the human policy; thus, if our goal is to improve the equity and accuracy of human-AI collaboration, it may not be desirable to enforce traditional algorithmic fairness constraints. We illustrate the value of our approach on criminal sentencing data before and after the introduction of an algorithmic risk assessment tool in Virginia.
Authors: Chengpiao Huang, Kaizheng Wang
Abstract: We develop a versatile framework for statistical learning in non-stationary environments. In each time period, our approach applies a stability principle to select a look-back window that maximizes the utilization of historical data while keeping the cumulative bias within an acceptable range relative to the stochastic error. Our theory and numerical experiments showcase the adaptivity of this approach to unknown non-stationarity. We prove regret bounds that are minimax optimal up to logarithmic factors when the population losses are strongly convex, or Lipschitz only. At the heart of our analysis lie two novel components: a measure of similarity between functions and a segmentation technique for dividing the non-stationary data sequence into quasi-stationary pieces.
Authors: Han-Jia Ye, Qi-Le Zhou, Huai-Hong Yin, De-Chuan Zhan, Wei-Lun Chao
Abstract: Pre-training is prevalent in deep learning for vision and text data, leveraging knowledge from other datasets to enhance downstream tasks. However, for tabular data, the inherent heterogeneity in attribute and label spaces across datasets complicates the learning of shareable knowledge. We propose Tabular data Pre-Training via Meta-representation (TabPTM), aiming to pre-train a general tabular model over diverse datasets. The core idea is to embed data instances into a shared feature space, where each instance is represented by its distance to a fixed number of nearest neighbors and their labels. This ''meta-representation'' transforms heterogeneous tasks into homogeneous local prediction problems, enabling the model to infer labels (or scores for each label) based on neighborhood information. As a result, the pre-trained TabPTM can be applied directly to new datasets, regardless of their diverse attributes and labels, without further fine-tuning. Extensive experiments on 101 datasets confirm TabPTM's effectiveness in both classification and regression tasks, with and without fine-tuning.
Authors: Jack Sandberg, Niklas {\AA}kerblom, Morteza Haghir Chehreghani
Abstract: We consider the combinatorial volatile Gaussian process (GP) semi-bandit problem. Each round, an agent is provided a set of available base arms and must select a subset of them to maximize the long-term cumulative reward. We study the Bayesian setting and provide novel Bayesian cumulative regret bounds for three GP-based algorithms: GP-UCB, GP-BayesUCB and GP-TS. Our bounds extend previous results for GP-UCB and GP-TS to the infinite, volatile and combinatorial setting, and to the best of our knowledge, we provide the first regret bound for GP-BayesUCB. Volatile arms encompass other widely considered bandit problems such as contextual bandits. Furthermore, we employ our framework to address the challenging real-world problem of online energy-efficient navigation, where we demonstrate its effectiveness compared to the alternatives.
Authors: Jan-Philipp von Bassewitz, Sebastian Kaltenbach, Petros Koumoutsakos
Abstract: Reliable predictions of critical phenomena, such as weather, wildfires and epidemics often rely on models described by Partial Differential Equations (PDEs). However, simulations that capture the full range of spatio-temporal scales described by such PDEs are often prohibitively expensive. Consequently, coarse-grained simulations are usually deployed that adopt various heuristics and empirical closure terms to account for the missing information. We propose a novel and systematic approach for identifying closures in under-resolved PDEs using grid-based Reinforcement Learning. This formulation incorporates inductive bias and exploits locality by deploying a central policy represented efficiently by a Fully Convolutional Network (FCN). We demonstrate the capabilities and limitations of our framework through numerical solutions of the advection equation and the Burgers' equation. Our results show accurate predictions for in- and out-of-distribution test cases as well as a significant speedup compared to resolving all scales.
Authors: Jun Wang, Wenjie Du, Yiyuan Yang, Linglong Qian, Wei Cao, Keli Zhang, Wenjia Wang, Yuxuan Liang, Qingsong Wen
Abstract: Missing values are ubiquitous in multivariate time series (MTS) data, posing significant challenges for accurate analysis and downstream applications. In recent years, deep learning-based methods have successfully handled missing data by leveraging complex temporal dependencies and learned data distributions. In this survey, we provide a comprehensive summary of deep learning approaches for multivariate time series imputation (MTSI) tasks. We propose a novel taxonomy that categorizes existing methods based on two key perspectives: imputation uncertainty and neural network architecture. Furthermore, we summarize existing MTSI toolkits with a particular emphasis on the PyPOTS Ecosystem, which provides an integrated and standardized foundation for MTSI research. Finally, we discuss key challenges and future research directions, which give insight for further MTSI research. This survey aims to serve as a valuable resource for researchers and practitioners in the field of time series analysis and missing data imputation tasks.
Authors: Eugenio Ressa, Alberto Marchisio, Maurizio Martina, Guido Masera, Muhammad Shafique
Abstract: The Continuous Learning (CL) paradigm consists of continuously evolving the parameters of the Deep Neural Network (DNN) model to progressively learn to perform new tasks without reducing the performance on previous tasks, i.e., avoiding the so-called catastrophic forgetting. However, the DNN parameter update in CL-based autonomous systems is extremely resource-hungry. The existing DNN accelerators cannot be directly employed in CL because they only support the execution of the forward propagation. Only a few prior architectures execute the backpropagation and weight update, but they lack the control and management for CL. Towards this, we design a hardware architecture, TinyCL, to perform CL on resource-constrained autonomous systems. It consists of a processing unit that executes both forward and backward propagation, and a control unit that manages memory-based CL workload. To minimize the memory accesses, the sliding window of the convolutional layer moves in a snake-like fashion. Moreover, the Multiply-and-Accumulate units can be reconfigured at runtime to execute different operations. As per our knowledge, our proposed TinyCL represents the first hardware accelerator that executes CL on autonomous systems. We synthesize the complete TinyCL architecture in a 65 nm CMOS technology node with the conventional ASIC design flow. It executes 1 epoch of training on a Conv + ReLU + Dense model on the CIFAR10 dataset in 1.76 s, while 1 training epoch of the same model using an Nvidia Tesla P100 GPU takes 103 s, thus achieving a 58x speedup, consuming 86 mW in a 4.74 mm2 die.
Authors: Dennis Grinwald, Philipp Wiesner, Shinichi Nakajima
Abstract: Statistical heterogeneity in federated learning poses two major challenges: slow global training due to conflicting gradient signals, and the need of personalization for local distributions. In this work, we tackle both challenges by leveraging recent advances in \emph{linear mode connectivity} -- identifying a linearly connected low-loss region in the parameter space of neural networks, which we call solution simplex. We propose federated learning over connected modes (\textsc{Floco}), where clients are assigned local subregions in this simplex based on their gradient signals, and together learn the shared global solution simplex. This allows personalization of the client models to fit their local distributions within the degrees of freedom in the solution simplex and homogenizes the update signals for the global simplex training. Our experiments show that \textsc{Floco} accelerates the global training process, and significantly improves the local accuracy with minimal computational overhead in cross-silo federated learning settings.
Authors: Mohamed Elrefaie, Angela Dai, Faez Ahmed
Abstract: This study introduces DrivAerNet, a large-scale high-fidelity CFD dataset of 3D industry-standard car shapes, and RegDGCNN, a dynamic graph convolutional neural network model, both aimed at aerodynamic car design through machine learning. DrivAerNet, with its 4000 detailed 3D car meshes using 0.5 million surface mesh faces and comprehensive aerodynamic performance data comprising of full 3D pressure, velocity fields, and wall-shear stresses, addresses the critical need for extensive datasets to train deep learning models in engineering applications. It is 60\% larger than the previously available largest public dataset of cars, and is the only open-source dataset that also models wheels and underbody. RegDGCNN leverages this large-scale dataset to provide high-precision drag estimates directly from 3D meshes, bypassing traditional limitations such as the need for 2D image rendering or Signed Distance Fields (SDF). By enabling fast drag estimation in seconds, RegDGCNN facilitates rapid aerodynamic assessments, offering a substantial leap towards integrating data-driven methods in automotive design. Together, DrivAerNet and RegDGCNN promise to accelerate the car design process and contribute to the development of more efficient cars. To lay the groundwork for future innovations in the field, the dataset and code used in our study are publicly accessible at https://github.com/Mohamedelrefaie/DrivAerNet.
Authors: Yiping Ji, Hemanth Saratchandran, Cameron Gordon, Zeyu Zhang, Simon Lucey
Abstract: Low-rank decomposition has emerged as a vital tool for enhancing parameter efficiency in neural network architectures, gaining traction across diverse applications in machine learning. These techniques significantly lower the number of parameters, striking a balance between compactness and performance. However, a common challenge has been the compromise between parameter efficiency and the accuracy of the model, where reduced parameters often lead to diminished accuracy compared to their full-rank counterparts. In this work, we propose a novel theoretical framework that integrates a sinusoidal function within the low-rank decomposition process. This approach not only preserves the benefits of the parameter efficiency characteristic of low-rank methods but also increases the decomposition's rank, thereby enhancing model performance. Our method proves to be a plug in enhancement for existing low-rank models, as evidenced by its successful application in Vision Transformers (ViT), Large Language Models (LLMs), Neural Radiance Fields (NeRF) and 3D shape modelling.
Authors: Yichuan Deng, Zhao Song, Jing Xiong, Chiwun Yang
Abstract: Sparse Attention is a technique that approximates standard attention computation with sub-quadratic complexity. This is achieved by selectively ignoring smaller entries in the attention matrix during the softmax function computation. Variations of this technique, such as pruning KV cache, sparsity-based fast attention, and Sparse Transformer, have been extensively utilized for efficient Large Language Models (LLMs) deployment. Despite its widespread use, a theoretical understanding of the conditions under which sparse attention performs on par with traditional attention remains elusive. This work aims to $\textbf{bridge this gap by examining the inherent sparsity of standard attention processes}$. Our theoretical framework reveals several brand-new key insights: $\bullet$ Attention is $n^{C}$-sparse, implying that considering only the largest $\Omega(n^{C})$ entries out of all $n$ entries is sufficient for sparse attention to approximate the exact attention matrix with decreasing loss. Here, $n$ represents the input length and $C \in (0, 1)$ is a constant. $\bullet$ Stable $o(\log(n))$-sparse attention, which approximates attention computation with $\log(n)$ or fewer entries, may not be feasible since the error will persist at a minimum of $O(1)$. $\bullet$ An adaptive strategy ($\alpha \cdot n^C, \alpha \in \mathbb{R}$) for the window size of efficient attention methods rather than a fixed one is guaranteed to perform more accurately and efficiently in a task for inference on flexible context lengths.
Authors: QingGuo Qi, Hongyang Chen, Minhao Cheng, Han Liu
Abstract: In recent years, there has been a surge in research on dynamic graph representation learning, primarily focusing on modeling the evolution of temporal-spatial patterns in real-world applications. However, within the domain of discrete-time dynamic graphs, the exploration of temporal edges remains underexplored. Existing approaches often rely on additional sequential models to capture dynamics, leading to high computational and memory costs, particularly for large-scale graphs. To address this limitation, we propose the Input {\bf S}napshots {\bf F}usion based {\bf Dy}namic {\bf G}raph Neural Network (SFDyG), which combines Hawkes processes with graph neural networks to capture temporal and structural patterns in dynamic graphs effectively. By fusing multiple snapshots into a single temporal graph, SFDyG decouples computational complexity from the number of snapshots, enabling efficient full-batch and mini-batch training. Experimental evaluations on eight diverse dynamic graph datasets for future link prediction tasks demonstrate that SFDyG consistently outperforms existing methods.
Authors: Kai Hu, Weichen Yu, Yining Li, Kai Chen, Tianjun Yao, Xiang Li, Wenhe Liu, Lijun Yu, Zhiqiang Shen, Matt Fredrikson
Abstract: Recent research indicates that large language models (LLMs) are susceptible to jailbreaking attacks that can generate harmful content. This paper introduces a novel token-level attack method, Adaptive Dense-to-Sparse Constrained Optimization (ADC), which has been shown to successfully jailbreak multiple open-source LLMs. Drawing inspiration from the difficulties of discrete token optimization, our method relaxes the discrete jailbreak optimization into a continuous optimization process while gradually increasing the sparsity of the optimizing vectors. This technique effectively bridges the gap between discrete and continuous space optimization. Experimental results demonstrate that our method is more effective and efficient than state-of-the-art token-level methods. On Harmbench, our approach achieves the highest attack success rate on seven out of eight LLMs compared to the latest jailbreak methods. Trigger Warning: This paper contains model behavior that can be offensive in nature.
Authors: Yuchen Zhu, Tianrong Chen, Lingkai Kong, Evangelos A. Theodorou, Molei Tao
Abstract: The generative modeling of data on manifolds is an important task, for which diffusion models in flat spaces typically need nontrivial adaptations. This article demonstrates how a technique called `trivialization' can transfer the effectiveness of diffusion models in Euclidean spaces to Lie groups. In particular, an auxiliary momentum variable was algorithmically introduced to help transport the position variable between data distribution and a fixed, easy-to-sample distribution. Normally, this would incur further difficulty for manifold data because momentum lives in a space that changes with the position. However, our trivialization technique creates a new momentum variable that stays in a simple fixed vector space. This design, together with a manifold preserving integrator, simplifies implementation and avoids inaccuracies created by approximations such as projections to tangent space and manifold, which were typically used in prior work, hence facilitating generation with high-fidelity and efficiency. The resulting method achieves state-of-the-art performance on protein and RNA torsion angle generation and sophisticated torus datasets. We also, arguably for the first time, tackle the generation of data on high-dimensional Special Orthogonal and Unitary groups, the latter essential for quantum problems. Code is available at https://github.com/yuchen-zhu-zyc/TDM.
Authors: Shuran Zheng, Xuan Qi, Rui Ray Chen, Yongchan Kwon, James Zou
Abstract: Data plays a central role in the development of modern artificial intelligence, with high-quality data emerging as a key driver of model performance. This has prompted the development of various data curation methods in recent years. However, measuring the effectiveness of these data curation techniques remains a major challenge. Traditional evaluation methods, which assess a trained model's performance on specific benchmarks, risk promoting practices that merely make the data more similar to the test data. This issue exemplifies Goodhart's law: when a measure becomes a target, it ceases to be a good measure. To address this, we propose an information-theoretic framework for evaluating data curation methods, where dataset quality is measured by its informativeness about the true model parameters using the Blackwell ordering. We compare informativeness by the Shannon mutual information of the evaluated data and the test data, and we propose a novel method for estimating the mutual information of datasets by training Bayesian models on embedded data and computing the mutual information from the model's parameter posteriors. Experiments on real-world data demonstrate that our mutual information-based evaluation assigns appropriately lower scores to data curation strategies that reduce dataset informativeness, while traditional test score-based evaluation methods may favor data curation strategies that overfit to the test set but compromise the training data's informativeness.
Authors: Haoran He, Emmanuel Bengio, Qingpeng Cai, Ling Pan
Abstract: The Generative Flow Network (GFlowNet) is a probabilistic framework in which an agent learns a stochastic policy and flow functions to sample objects with probability proportional to an unnormalized reward function. GFlowNets share a strong connection with reinforcement learning (RL) that typically aims to maximize reward. A number of recent works explored connections between GFlowNets and maximum entropy (MaxEnt) RL, which incorporates entropy regularization into the standard RL objective. However, the relationship between GFlowNets and standard RL remains largely unexplored, despite the inherent similarities in their sequential decision-making nature. While GFlowNets can discover diverse solutions through specialized flow-matching objectives, connecting them to standard RL can simplify their implementation through well-established RL principles and also improve RL's capabilities in diverse solution discovery (a critical requirement in many real-world applications), and bridging this gap can further unlock the potential of both fields. In this paper, we bridge this gap by revealing a fundamental connection between GFlowNets and one of the most basic components of RL -- policy evaluation. Surprisingly, we find that the value function obtained from evaluating a uniform policy is closely associated with the flow functions in GFlowNets. Building upon these insights, we introduce a rectified random policy evaluation (RPE) algorithm, which achieves the same reward-matching effect as GFlowNets based on simply evaluating a fixed random policy, offering a new perspective. Empirical results across extensive benchmarks demonstrate that RPE achieves competitive results compared to previous approaches, shedding light on the previously overlooked connection between (non-MaxEnt) RL and GFlowNets.
Authors: Micha{\l} Zawalski, Gracjan G\'oral, Micha{\l} Tyrolski, Emilia Wi\'snios, Franciszek Budrowski, Marek Cygan, {\L}ukasz Kuci\'nski, Piotr Mi{\l}o\'s
Abstract: Efficiently tackling combinatorial reasoning problems, particularly the notorious NP-hard tasks, remains a significant challenge for AI research. Recent efforts have sought to enhance planning by incorporating hierarchical high-level search strategies, known as subgoal methods. While promising, their performance against traditional low-level planners is inconsistent, raising questions about their application contexts. In this study, we conduct an in-depth exploration of subgoal-planning methods for combinatorial reasoning. We identify the attributes pivotal for leveraging the advantages of high-level search: hard-to-learn value functions, complex action spaces, presence of dead ends in the environment, or using data collected from diverse experts. We propose a consistent evaluation methodology to achieve meaningful comparisons between methods and reevaluate the state-of-the-art algorithms.
Authors: Jiawei Xu, Rui Yang, Shuang Qiu, Feng Luo, Meng Fang, Baoxiang Wang, Lei Han
Abstract: Learning policy from offline datasets through offline reinforcement learning (RL) holds promise for scaling data-driven decision-making while avoiding unsafe and costly online interactions. However, real-world data collected from sensors or humans often contains noise and errors, posing a significant challenge for existing offline RL methods, particularly when the real-world data is limited. Our study reveals that prior research focusing on adapting predominant offline RL methods based on temporal difference learning still falls short under data corruption when the dataset is limited. In contrast, we discover that vanilla sequence modeling methods, such as Decision Transformer, exhibit robustness against data corruption, even without specialized modifications. To unlock the full potential of sequence modeling, we propose **R**obust **D**ecision **T**ransformer (**RDT**) by incorporating three simple yet effective robust techniques: embedding dropout to improve the model's robustness against erroneous inputs, Gaussian weighted learning to mitigate the effects of corrupted labels, and iterative data correction to eliminate corrupted data from the source. Extensive experiments on MuJoCo, Kitchen, and Adroit tasks demonstrate RDT's superior performance under various data corruption scenarios compared to prior methods. Furthermore, RDT exhibits remarkable robustness in a more challenging setting that combines training-time data corruption with test-time observation perturbations. These results highlight the potential of sequence modeling for learning from noisy or corrupted offline datasets, thereby promoting the reliable application of offline RL in real-world scenarios.Our code is available at https://github.com/jiawei415/RobustDecisionTransformer.
URLs: https://github.com/jiawei415/RobustDecisionTransformer.
Authors: Juhani Kivim\"aki, Jakub Bia{\l}ek, Jukka K. Nurminen, Wojtek Kuberski
Abstract: After a machine learning model has been deployed into production, its predictive performance needs to be monitored. Ideally, such monitoring can be carried out by comparing the model's predictions against ground truth labels. For this to be possible, the ground truth labels must be available relatively soon after inference. However, there are many use cases where ground truth labels are available only after a significant delay, or in the worst case, not at all. In such cases, directly monitoring the model's predictive performance is impossible. Recently, novel methods for estimating the predictive performance of a model when ground truth is unavailable have been developed. Many of these methods leverage model confidence or other uncertainty estimates and are experimentally compared against a naive baseline method, namely Average Confidence (AC), which estimates model accuracy as the average of confidence scores for a given set of predictions. However, until now the theoretical properties of the AC method have not been properly explored. In this paper, we try to fill this gap by reviewing the AC method and show that under certain general assumptions, it is an unbiased and consistent estimator of model accuracy with many desirable properties. We also compare this baseline estimator against some more complex estimators empirically and show that in many cases the AC method is able to beat the others, although the comparative quality of the different estimators is heavily case-dependent.
Authors: Changhun Kim, Taewon Kim, Seungyeon Woo, June Yong Yang, Eunho Yang
Abstract: In real-world scenarios, tabular data often suffer from distribution shifts that threaten the performance of machine learning models. Despite its prevalence and importance, handling distribution shifts in the tabular domain remains underexplored due to the inherent challenges within the tabular data itself. In this sense, test-time adaptation (TTA) offers a promising solution by adapting models to target data without accessing source data, crucial for privacy-sensitive tabular domains. However, existing TTA methods either 1) overlook the nature of tabular distribution shifts, often involving label distribution shifts, or 2) impose architectural constraints on the model, leading to a lack of applicability. To this end, we propose AdapTable, a novel TTA framework for tabular data. AdapTable operates in two stages: 1) calibrating model predictions using a shift-aware uncertainty calibrator, and 2) adjusting these predictions to match the target label distribution with a label distribution handler. We validate the effectiveness of AdapTable through theoretical analysis and extensive experiments on various distribution shift scenarios. Our results demonstrate AdapTable's ability to handle various real-world distribution shifts, achieving up to a 16% improvement on the HELOC dataset.
Authors: Jianke Yu, Hanchen Wang, Chen Chen, Xiaoyang Wang, Lu Qin, Wenjie Zhang, Ying Zhang, Xijuan Liu
Abstract: Graph Neural Networks (GNNs) are vital in data science but are increasingly susceptible to adversarial attacks. To help researchers develop more robust GNN models, it's essential to focus on designing strong attack models as foundational benchmarks and guiding references. Among adversarial attacks, gray-box poisoning attacks are noteworthy due to their effectiveness and fewer constraints. These attacks exploit GNNs' need for retraining on updated data, thereby impacting their performance by perturbing these datasets. However, current research overlooks the real-world scenario of incomplete graphs. To address this gap, we introduce the Robust Incomplete Deep Attack Framework (RIDA). It is the first algorithm for robust gray-box poisoning attacks on incomplete graphs. The approach innovatively aggregates distant vertex information and ensures powerful data utilization. Extensive tests against 9 SOTA baselines on 3 real-world datasets demonstrate that RIDA's superiority in handling incompleteness and high attack performance on the incomplete graph.
Authors: Paul-Gauthier No\'e, Miquel Perell\'o-Nieto, Jean-Fran\c{c}ois Bonastre, Peter Flach
Abstract: Originating in game theory, Shapley values are widely used for explaining a machine learning model's prediction by quantifying the contribution of each feature's value to the prediction. This requires a scalar prediction as in binary classification, whereas a multiclass probabilistic prediction is a discrete probability distribution, living on a multidimensional simplex. In such a multiclass setting the Shapley values are typically computed separately on each class in a one-vs-rest manner, ignoring the compositional nature of the output distribution. In this paper, we introduce Shapley compositions as a well-founded way to properly explain a multiclass probabilistic prediction, using the Aitchison geometry from compositional data analysis. We prove that the Shapley composition is the unique quantity satisfying linearity, symmetry and efficiency on the Aitchison simplex, extending the corresponding axiomatic properties of the standard Shapley value. We demonstrate this proper multiclass treatment in a range of scenarios.
Authors: Yam Eitan, Yoav Gelberg, Guy Bar-Shalom, Fabrizio Frasca, Michael Bronstein, Haggai Maron
Abstract: Topological deep learning (TDL) is a rapidly growing field that seeks to leverage topological structure in data and facilitate learning from data supported on topological objects, ranging from molecules to 3D shapes. Most TDL architectures can be unified under the framework of higher-order message-passing (HOMP), which generalizes graph message-passing to higher-order domains. In the first part of the paper, we explore HOMP's expressive power from a topological perspective, demonstrating the framework's inability to capture fundamental topological and metric invariants such as diameter, orientability, planarity, and homology. In addition, we demonstrate HOMP's limitations in fully leveraging lifting and pooling methods on graphs. To the best of our knowledge, this is the first work to study the expressivity of TDL from a \emph{topological} perspective. In the second part of the paper, we develop two new classes of architectures -- multi-cellular networks (MCN) and scalable MCN (SMCN) -- which draw inspiration from expressive GNNs. MCN can reach full expressivity, but scaling it to large data objects can be computationally expansive. Designed as a more scalable alternative, SMCN still mitigates many of HOMP's expressivity limitations. Finally, we create new benchmarks for evaluating models based on their ability to learn topological properties of complexes. We then evaluate SMCN on these benchmarks and on real-world graph datasets, demonstrating improvements over both HOMP baselines and expressive graph methods, highlighting the value of expressively leveraging topological information. Code and data are available at https://github.com/yoavgelberg/SMCN.
Authors: Yucheng Sheng, Kai Huang, Le Liang, Peng Liu, Shi Jin, Geoffrey Ye Li
Abstract: In this letter, we use large language models (LLMs) to develop a high-performing and robust beam prediction method. We formulate the millimeter wave (mmWave) beam prediction problem as a time series forecasting task, where the historical observations are aggregated through cross-variable attention and then transformed into text-based representations using a trainable tokenizer. By leveraging the prompt-as-prefix (PaP) technique for contextual enrichment, our method harnesses the power of LLMs to predict future optimal beams. Simulation results demonstrate that our LLM-based approach outperforms traditional learning-based models in prediction accuracy as well as robustness, highlighting the significant potential of LLMs in enhancing wireless communication systems.
Authors: Tom Jacobs, Rebekka Burkholz
Abstract: Continuous sparsification strategies are among the most effective methods for reducing the inference costs and memory demands of large-scale neural networks. A key factor in their success is the implicit $L_1$ regularization induced by jointly learning both mask and weight variables, which has been shown experimentally to outperform explicit $L_1$ regularization. We provide a theoretical explanation for this observation by analyzing the learning dynamics, revealing that early continuous sparsification is governed by an implicit $L_2$ regularization that gradually transitions to an $L_1$ penalty over time. Leveraging this insight, we propose a method to dynamically control the strength of this implicit bias. Through an extension of the mirror flow framework, we establish convergence and optimality guarantees in the context of underdetermined linear regression. Our theoretical findings may be of independent interest, as we demonstrate how to enter the rich regime and show that the implicit bias can be controlled via a time-dependent Bregman potential. To validate these insights, we introduce PILoT, a continuous sparsification approach with novel initialization and dynamic regularization, which consistently outperforms baselines in standard experiments.
Authors: Georgy Sokolov, Maximilian Thiessen, Margarita Akhmejanova, Fabio Vitale, Francesco Orabona
Abstract: We study the problem of classifying the nodes of a given graph in the self-directed learning setup. This learning setting is a variant of online learning, where rather than an adversary determining the sequence in which nodes are presented, the learner autonomously and adaptively selects them. While self-directed learning of Euclidean halfspaces, linear functions, and general multiclass hypothesis classes was recently considered, no results previously existed specifically for self-directed node classification on graphs. In this paper, we address this problem developing efficient algorithms for it. More specifically, we focus on the case of (geodesically) convex clusters, i.e., for every two nodes sharing the same label, all nodes on every shortest path between them also share the same label. In particular, we devise an algorithm with runtime polynomial in $n$ that makes only $3(h(G)+1)^4 \ln n$ mistakes on graphs with two convex clusters, where $n$ is the total number of nodes and $h(G)$ is the Hadwiger number, i.e., the size of the largest clique minor of the graph $G$. We also show that our algorithm is robust to the case that clusters are slightly non-convex, still achieving a mistake bound logarithmic in $n$. Finally, we devise a simple and efficient algorithm for homophilic clusters, where strongly connected nodes tend to belong to the same class.
Authors: Markus Knauer, Alin Albu-Sch\"affer, Freek Stulp, Jo\~ao Silv\'erio
Abstract: The problem of generalization in learning from demonstration (LfD) has received considerable attention over the years, particularly within the context of movement primitives, where a number of approaches have emerged. Recently, two important approaches have gained recognition. While one leverages via-points to adapt skills locally by modulating demonstrated trajectories, another relies on so-called task-parameterized models that encode movements with respect to different coordinate systems, using a product of probabilities for generalization. While the former are well-suited to precise, local modulations, the latter aim at generalizing over large regions of the workspace and often involve multiple objects. Addressing the quality of generalization by leveraging both approaches simultaneously has received little attention. In this work, we propose an interactive imitation learning framework that simultaneously leverages local and global modulations of trajectory distributions. Building on the kernelized movement primitives (KMP) framework, we introduce novel mechanisms for skill modulation from direct human corrective feedback. Our approach particularly exploits the concept of via-points to incrementally and interactively 1) improve the model accuracy locally, 2) add new objects to the task during execution and 3) extend the skill into regions where demonstrations were not provided. We evaluate our method on a bearing ring-loading task using a torque-controlled, 7-DoF, DLR SARA robot.
Authors: Ziheng Cheng, Margalit Glasgow
Abstract: We study distributed adaptive algorithms with local updates (intermittent communication). Despite the great empirical success of adaptive methods in distributed training of modern machine learning models, the theoretical benefits of local updates within adaptive methods, particularly in terms of reducing communication complexity, have not been fully understood yet. In this paper, for the first time, we prove that \em Local SGD \em with momentum (\em Local \em SGDM) and \em Local \em Adam can outperform their minibatch counterparts in convex and weakly convex settings in certain regimes, respectively. Our analysis relies on a novel technique to prove contraction during local iterations, which is a crucial yet challenging step to show the advantages of local updates, under generalized smoothness assumption and gradient clipping strategy.
Authors: Byron A Jacobs, Ifthakaar Shaik, Frando Lin
Abstract: Globally, infertility rates are increasing, with 2.5\% of all births being assisted by in vitro fertilisation (IVF) in 2022. Male infertility is the cause for approximately half of these cases. The quality of sperm DNA has substantial impact on the success of IVF. The assessment of sperm DNA is traditionally done through chemical assays which render sperm cells ineligible for IVF. Many compounding factors lead to the population crisis, with fertility rates dropping globally in recent history. As such assisted reproductive technologies (ART) have been the focus of recent research efforts. Simultaneously, artificial intelligence has grown ubiquitous and is permeating more aspects of modern life. With the advent of state-of-the-art machine learning and its exceptional performance in many sectors, this work builds on these successes and proposes a novel framework for the prediction of sperm cell DNA fragmentation from images of unstained sperm. Rendering a predictive model which preserves sperm integrity and allows for optimal selection of sperm for IVF.
Authors: Xingzhou Lou, Dong Yan, Wei Shen, Yuzi Yan, Jian Xie, Junge Zhang
Abstract: Reward models (RMs) are essential for aligning large language models (LLM) with human expectations. However, existing RMs struggle to capture the stochastic and uncertain nature of human preferences and fail to assess the reliability of reward predictions. To address these challenges, we introduce the Uncertainty-aware Reward Model (URM) and its ensemble variant, URME. URM employs a probabilistic value head to capture aleatoric uncertainty by modeling the distribution of disentangled human preference attributes. URME further quantifies epistemic uncertainty by examining discrepancies among individual URMs within the ensemble, enabling identification of unreliable evaluations. Our empirical evaluations demonstrate that URM achieves strong performance on RewardBench, outperforming competitive large-scale models. Additionally, extensive experiments, including best-of-n sampling (BoN), iterative direct preference optimization (iterative DPO), and proximal policy optimization (PPO), demonstrate that URM and URME significantly enhance LLMs' generation quality. Notably, reward predictions with lower uncertainty are far more reliable, demonstrate significantly higher quality, and result in substantially improved alignment.
Authors: Jiecheng Lu, Xu Han, Yan Sun, Shihao Yang
Abstract: We propose a Weighted Autoregressive Varying gatE (WAVE) attention mechanism equipped with both Autoregressive (AR) and Moving-average (MA) components. It can adapt to various attention mechanisms, enhancing and decoupling their ability to capture long-range and local temporal patterns in time series data. In this paper, we first demonstrate that, for the time series forecasting (TSF) task, the previously overlooked decoder-only autoregressive Transformer model can achieve results comparable to the best baselines when appropriate tokenization and training methods are applied. Moreover, inspired by the ARMA model from statistics and recent advances in linear attention, we introduce the full ARMA structure into existing autoregressive attention mechanisms. By using an indirect MA weight generation method, we incorporate the MA term while maintaining the time complexity and parameter size of the underlying efficient attention models. We further explore how indirect parameter generation can produce implicit MA weights that align with the modeling requirements for local temporal impacts. Experimental results show that WAVE attention that incorporates the ARMA structure consistently improves the performance of various AR attentions on TSF tasks, achieving state-of-the-art results.
Authors: T. Konstantin Rusch, Daniela Rus
Abstract: We propose Linear Oscillatory State-Space models (LinOSS) for efficiently learning on long sequences. Inspired by cortical dynamics of biological neural networks, we base our proposed LinOSS model on a system of forced harmonic oscillators. A stable discretization, integrated over time using fast associative parallel scans, yields the proposed state-space model. We prove that LinOSS produces stable dynamics only requiring nonnegative diagonal state matrix. This is in stark contrast to many previous state-space models relying heavily on restrictive parameterizations. Moreover, we rigorously show that LinOSS is universal, i.e., it can approximate any continuous and causal operator mapping between time-varying functions, to desired accuracy. In addition, we show that an implicit-explicit discretization of LinOSS perfectly conserves the symmetry of time reversibility of the underlying dynamics. Together, these properties enable efficient modeling of long-range interactions, while ensuring stable and accurate long-horizon forecasting. Finally, our empirical results, spanning a wide range of time-series tasks from mid-range to very long-range classification and regression, as well as long-horizon forecasting, demonstrate that our proposed LinOSS model consistently outperforms state-of-the-art sequence models. Notably, LinOSS outperforms Mamba by nearly 2x and LRU by 2.5x on a sequence modeling task with sequences of length 50k.
Authors: Wen Ye, Yizhou Zhang, Wei Yang, Defu Cao, Lumingyuan Tang, Jie Cai, Yan Liu
Abstract: Time series analysis is crucial in real-world applications, yet traditional methods focus on isolated tasks only, and recent studies on time series reasoning remain limited to simple, single-step inference constrained to natural language answer. In this work, we propose a practical novel task: multi-step time series inference that demands both compositional reasoning and computation precision of time series analysis. To address such challenge, we propose a simple but effective program-aided inference agent that leverages LLMs' reasoning ability to decompose complex tasks into structured execution pipelines. By integrating in-context learning, self-correction, and program-aided execution, our proposed approach ensures accurate and interpretable results. To benchmark performance, we introduce a new dataset and a unified evaluation framework with task-specific success criteria. Experiments show that our approach outperforms standalone general purpose LLMs in both basic time series concept understanding as well as multi-step time series inference task, highlighting the importance of hybrid approaches that combine reasoning with computational precision.
Authors: Divyat Mahajan, Jannes Gladrow, Agrin Hilmkil, Cheng Zhang, Meyer Scetbon
Abstract: With the increasing acquisition of datasets over time, we now have access to precise and varied descriptions of the world, encompassing a broad range of phenomena. These datasets can be seen as observations from unknown causal generative processes, commonly described by Structural Causal Models (SCMs). Recovering SCMs from observations poses formidable challenges, and often requires us to learn a specific generative model for each dataset. In this work, we propose to learn a \emph{single} model capable of inferring the SCMs in a zero-shot manner. Rather than learning a specific SCM for each dataset, we enable the Fixed-Point Approach (FiP)~\citep{scetbon2024fip} to infer the generative SCMs conditionally on their empirical representations. As a by-product, our approach can perform zero-shot generation of new dataset samples and intervened samples. We demonstrate via experiments that our amortized procedure achieves performances on par with SoTA methods trained specifically for each dataset on both in and out-of-distribution problems. To the best of our knowledge, this is the first time that SCMs are inferred in a zero-shot manner from observations, paving the way for a paradigmatic shift toward the assimilation of causal knowledge across datasets. The code is available on Github.
Authors: Wenxuan Bao, Zhichen Zeng, Zhining Liu, Hanghang Tong, Jingrui He
Abstract: Powerful as they are, graph neural networks (GNNs) are known to be vulnerable to distribution shifts. Recently, test-time adaptation (TTA) has attracted attention due to its ability to adapt a pre-trained model to a target domain, without re-accessing the source domain. However, existing TTA algorithms are primarily designed for attribute shifts in vision tasks, where samples are independent. These methods perform poorly on graph data that experience structure shifts, where node connectivity differs between source and target graphs. We attribute this performance gap to the distinct impact of node attribute shifts versus graph structure shifts: the latter significantly degrades the quality of node representations and blurs the boundaries between different node categories. To address structure shifts in graphs, we propose Matcha, an innovative framework designed for effective and efficient adaptation to structure shifts by adjusting the htop-aggregation parameters in GNNs. To enhance the representation quality, we design a prediction-informed clustering loss to encourage the formation of distinct clusters for different node categories. Additionally, Matcha seamlessly integrates with existing TTA algorithms, allowing it to handle attribute shifts effectively while improving overall performance under combined structure and attribute shifts. We validate the effectiveness of Matcha on both synthetic and real-world datasets, demonstrating its robustness across various combinations of structure and attribute shifts. Our code is available at https://github.com/baowenxuan/Matcha .
Authors: Roey Magen, Shuning Shang, Zhiwei Xu, Spencer Frei, Wei Hu, Gal Vardi
Abstract: The phenomenon of benign overfitting, where a trained neural network perfectly fits noisy training data but still achieves near-optimal test performance, has been extensively studied in recent years for linear models and fully-connected/convolutional networks. In this work, we study benign overfitting in a single-head softmax attention model, which is the fundamental building block of Transformers. We prove that under appropriate conditions, the model exhibits benign overfitting in a classification setting already after two steps of gradient descent. Moreover, we show conditions where a minimum-norm/maximum-margin interpolator exhibits benign overfitting. We study how the overfitting behavior depends on the signal-to-noise ratio (SNR) of the data distribution, namely, the ratio between norms of signal and noise tokens, and prove that a sufficiently large SNR is both necessary and sufficient for benign overfitting.
Authors: Julien Zhou (Thoth, STATIFY), Pierre Gaillard (Thoth), Thibaud Rahier (STATIFY), Julyan Arbel (STATIFY)
Abstract: We address the online unconstrained submodular maximization problem (Online USM), in a setting with stochastic bandit feedback. In this framework, a decision-maker receives noisy rewards from a non monotone submodular function taking values in a known bounded interval. This paper proposes Double-Greedy - Explore-then-Commit (DG-ETC), adapting the Double-Greedy approach from the offline and online full-information settings. DG-ETC satisfies a $O(d\log(dT))$ problem-dependent upper bound for the $1/2$-approximate pseudo-regret, as well as a $O(dT^{2/3}\log(dT)^{1/3})$ problem-free one at the same time, outperforming existing approaches. In particular, we introduce a problem-dependent notion of hardness characterizing the transition between logarithmic and polynomial regime for the upper bounds.
Authors: Sujai Hiremath, Promit Ghosal, Kyra Gan
Abstract: Inferring causal relationships from observational data is crucial when experiments are costly or infeasible. Additive noise models (ANMs) enable unique directed acyclic graph (DAG) identification, but existing ANM methods often rely on restrictive assumptions on the data generating process, limiting their applicability to real-world settings. We propose local search in additive noise models, LoSAM, a topological ordering method for learning a unique DAG in ANMs with mixed causal mechanisms and general noise distributions. We introduce new causal substructures and criteria for identifying roots and leaves, enabling efficient top-down learning. We prove asymptotic consistency and polynomial runtime, ensuring scalability and sample efficiency. We test LoSAM on synthetic and real-world data, demonstrating state-of-the-art performance across all mixed mechanism settings.
Authors: Yulun Wu, Louie McConnell, Claudia Iriondo
Abstract: Estimating an individual's counterfactual outcomes under interventions is a challenging task for traditional causal inference and supervised learning approaches when the outcome is high-dimensional (e.g. gene expressions, facial images) and covariates are relatively limited. In this case, to predict one's outcomes under counterfactual treatments, it is crucial to leverage individual information contained in the observed outcome in addition to the covariates. Prior works using variational inference in counterfactual generative modeling have been focusing on neural adaptations and model variants within the conditional variational autoencoder formulation, which we argue is fundamentally ill-suited to the notion of counterfactual in causal inference. In this work, we present a novel variational Bayesian causal inference framework and its theoretical backings to properly handle counterfactual generative modeling tasks, through which we are able to conduct counterfactual supervision end-to-end during training without any counterfactual samples, and encourage disentangled exogenous noise abduction that aids the correct identification of causal effect in counterfactual generations. In experiments, we demonstrate the advantage of our framework compared to state-of-the-art models in counterfactual generative modeling on multiple benchmarks.
Authors: Mateusz Cedro, David Martens
Abstract: Graph Neural Networks (GNNs) are a powerful technique for machine learning on graph-structured data, yet they pose challenges in interpretability. Existing GNN explanation methods usually yield technical outputs, such as subgraphs and feature importance scores, that are difficult for non-data scientists to understand and thereby violate the purpose of explanations. Motivated by recent Explainable AI (XAI) research, we propose GraphXAIN, a method that generates natural language narratives explaining GNN predictions. GraphXAIN is a model- and explainer-agnostic method that uses Large Language Models (LLMs) to translate explanatory subgraphs and feature importance scores into coherent, story-like explanations of GNN decision-making processes. Evaluations on real-world datasets demonstrate GraphXAIN's ability to improve graph explanations. A survey of machine learning researchers and practitioners reveals that GraphXAIN enhances four explainability dimensions: understandability, satisfaction, convincingness, and suitability for communicating model predictions. When combined with another graph explainer method, GraphXAIN further improves trustworthiness, insightfulness, confidence, and usability. Notably, 95% of participants found GraphXAIN to be a valuable addition to the GNN explanation method. By incorporating natural language narratives, our approach serves both graph practitioners and non-expert users by providing clearer and more effective explanations.
Authors: Siddhant Dutta, Iago Leal de Freitas, Pedro Maciel Xavier, Claudio Miceli de Farias, David Esteban Bernal Neira
Abstract: Federated Learning (FL) is a decentralized machine learning approach that has gained attention for its potential to enable collaborative model training across clients while protecting data privacy, making it an attractive solution for the chemical industry. This work aims to provide the chemical engineering community with an accessible introduction to the discipline. Supported by a hands-on tutorial and a comprehensive collection of examples, it explores the application of FL in tasks such as manufacturing optimization, multimodal data integration, and drug discovery while addressing the unique challenges of protecting proprietary information and managing distributed datasets. The tutorial was built using key frameworks such as $\texttt{Flower}$ and $\texttt{TensorFlow Federated}$ and was designed to provide chemical engineers with the right tools to adopt FL in their specific needs. We compare the performance of FL against centralized learning across three different datasets relevant to chemical engineering applications, demonstrating that FL will often maintain or improve classification performance, particularly for complex and heterogeneous data. We conclude with an outlook on the open challenges in federated learning to be tackled and current approaches designed to remediate and improve this framework.
Authors: Zhi-Yi Chin, Mario Fritz, Pin-Yu Chen, Wei-Chen Chiu
Abstract: Text-to-image (T2I) models have shown remarkable progress, but their potential to generate harmful content remains a critical concern in the ML community. While various safety mechanisms have been developed, the field lacks systematic tools for evaluating their effectiveness against real-world misuse scenarios. In this work, we propose ICER, a novel red-teaming framework that leverages Large Language Models (LLMs) and a bandit optimization-based algorithm to generate interpretable and semantic meaningful problematic prompts by learning from past successful red-teaming attempts. Our ICER efficiently probes safety mechanisms across different T2I models without requiring internal access or additional training, making it broadly applicable to deployed systems. Through extensive experiments, we demonstrate that ICER significantly outperforms existing prompt attack methods in identifying model vulnerabilities while maintaining high semantic similarity with intended content. By uncovering that successful jailbreaking instances can systematically facilitate the discovery of new vulnerabilities, our work provides crucial insights for developing more robust safety mechanisms in T2I systems.
Authors: Xianliang Li, Jun Luo, Zhiwei Zheng, Hanxiao Wang, Li Luo, Lingkun Wen, Linlong Wu, Sheng Xu
Abstract: Momentum-based optimizers are widely adopted for training neural networks. However, the optimal selection of momentum coefficients remains elusive. This uncertainty impedes a clear understanding of the role of momentum in stochastic gradient methods. In this paper, we present a frequency domain analysis framework that interprets the momentum method as a time-variant filter for gradients, where adjustments to momentum coefficients modify the filter characteristics. Our experiments support this perspective and provide a deeper understanding of the mechanism involved. Moreover, our analysis reveals the following significant findings: high-frequency gradient components are undesired in the late stages of training; preserving the original gradient in the early stages, and gradually amplifying low-frequency gradient components during training both enhance performance. Based on these insights, we propose Frequency Stochastic Gradient Descent with Momentum (FSGDM), a heuristic optimizer that dynamically adjusts the momentum filtering characteristic with an empirically effective dynamic magnitude response. Experimental results demonstrate the superiority of FSGDM over conventional momentum optimizers.
Authors: Xuanze Chen, Jiajun Zhou, Shanqing Yu, Qi Xuan
Abstract: Graph neural networks excel at graph representation learning but struggle with heterophilous data and long-range dependencies. And graph transformers address these issues through self-attention, yet face scalability and noise challenges on large-scale graphs. To overcome these limitations, we propose GNNMoE, a universal model architecture for node classification. This architecture flexibly combines fine-grained message-passing operations with a mixture-of-experts mechanism to build feature encoding blocks. Furthermore, by incorporating soft and hard gating layers to assign the most suitable expert networks to each node, we enhance the model's expressive power and adaptability to different graph types. In addition, we introduce adaptive residual connections and an enhanced FFN module into GNNMoE, further improving the expressiveness of node representation. Extensive experimental results demonstrate that GNNMoE performs exceptionally well across various types of graph data, effectively alleviating the over-smoothing issue and global noise, enhancing model robustness and adaptability, while also ensuring computational efficiency on large-scale graphs.
Authors: Youssef Allouah, Joshua Kazdan, Rachid Guerraoui, Sanmi Koyejo
Abstract: Machine unlearning, the process of selectively removing data from trained models, is increasingly crucial for addressing privacy concerns and knowledge gaps post-deployment. Despite this importance, existing approaches are often heuristic and lack formal guarantees. In this paper, we analyze the fundamental utility, time, and space complexity trade-offs of approximate unlearning, providing rigorous certification analogous to differential privacy. For in-distribution forget data -- data similar to the retain set -- we show that a surprisingly simple and general procedure, empirical risk minimization with output perturbation, achieves tight unlearning-utility-complexity trade-offs, addressing a previous theoretical gap on the separation from unlearning "for free" via differential privacy, which inherently facilitates the removal of such data. However, such techniques fail with out-of-distribution forget data -- data significantly different from the retain set -- where unlearning time complexity can exceed that of retraining, even for a single sample. To address this, we propose a new robust and noisy gradient descent variant that provably amortizes unlearning time complexity without compromising utility.
Authors: Xin Huang, Eric M. Wolff, Paul Vernaza, Tung Phan-Minh, Hongge Chen, David S. Hayden, Mark Edmonds, Brian Pierce, Xinxin Chen, Pratik Elias Jacob, Xiaobai Chen, Chingiz Tairbekov, Pratik Agarwal, Tianshi Gao, Yuning Chai, Siddhartha Srinivasa
Abstract: We present DriveGPT, a scalable behavior model for autonomous driving. We model driving as a sequential decision-making task, and learn a transformer model to predict future agent states as tokens in an autoregressive fashion. We scale up our model parameters and training data by multiple orders of magnitude, enabling us to explore the scaling properties in terms of dataset size, model parameters, and compute. We evaluate DriveGPT across different scales in a planning task, through both quantitative metrics and qualitative examples, including closed-loop driving in complex real-world scenarios. In a separate prediction task, DriveGPT outperforms state-of-the-art baselines and exhibits improved performance by pretraining on a large-scale dataset, further validating the benefits of data scaling.
Authors: Manele Ait Habouche, Micka\"el Kerboeuf, Goulven Guillou, Jean-Philippe Babau
Abstract: Unmanned Surface Vehicles (USVs) have become critical tools for marine exploration, environmental monitoring, and autonomous navigation. Accurate estimation of wave direction is essential for improving USV navigation and ensuring operational safety, but traditional methods often suffer from high costs and limited spatial resolution. This paper proposes a machine learning-based approach leveraging LSTM (Long Short-Term Memory) networks to predict wave direction using sensor data collected from USVs. Experimental results show the capability of the LSTM model to learn temporal dependencies and provide accurate predictions, outperforming simpler baselines.
Authors: Katherine Goode, J. Derek Tucker, Daniel Ries, Heike Hofmann
Abstract: Machine learning (ML) models have shown success in applications with an objective of prediction, but the algorithmic complexity of some models makes them difficult to interpret. Methods have been proposed to provide insight into these "black-box" models, but there is little research that focuses on supervised ML when the model inputs are functional data. In this work, we consider two applications from high-consequence spaces with objectives of making predictions using functional data inputs. One application aims to classify material types to identify explosive materials given hyperspectral computed tomography scans of the materials. The other application considers the forensics science task of connecting an inkjet printed document to the source printer using color signatures extracted by Raman spectroscopy. An instinctive route to consider for analyzing these data is a data driven ML model for classification, but due to the high consequence nature of the applications, we argue it is important to appropriately account for the nature of the data in the analysis to not obscure or misrepresent patterns. As such, we propose the Variable importance Explainable Elastic Shape Analysis (VEESA) pipeline for training ML models with functional data that (1) accounts for the vertical and horizontal variability in the functional data and (2) provides an explanation in the original data space of how the model uses variability in the functional data for prediction. The pipeline makes use of elastic functional principal components analysis (efPCA) to generate uncorrelated model inputs and permutation feature importance (PFI) to identify the principal components important for prediction. The variability captured by the important principal components in visualized the original data space. We ultimately discuss ideas for natural extensions of the VEESA pipeline and challenges for future research.
Authors: Yiming Yang, Xiaoyuan Cheng, Daniel Giles, Sibo Cheng, Yi He, Xiao Xue, Boli Chen, Yukun Hu
Abstract: Variational data assimilation estimates the dynamical system states by minimizing a cost function that fits the numerical models with observational data. The widely used method, four-dimensional variational assimilation (4D-Var), has two primary challenges: (1) computationally demanding for complex nonlinear systems and (2) relying on state-observation mappings, which are often not perfectly known. Deep learning (DL) has been used as a more expressive class of efficient model approximators to address these challenges. However, integrating such models into 4D-Var remains challenging due to their inherent nonlinearities and the lack of theoretical guarantees for consistency in assimilation results. In this paper, we propose Tensor-Var to address these challenges using kernel Conditional Mean Embedding (CME). Tensor-Var improves optimization efficiency by characterizing system dynamics and state-observation mappings as linear operators, leading to a convex cost function in the feature space. Furthermore, our method provides a new perspective to incorporate CME into 4D-Var, offering theoretical guarantees of consistent assimilation results between the original and feature spaces. To improve scalability, we propose a method to learn deep features (DFs) using neural networks within the Tensor-Var framework. Experiments on chaotic systems and global weather prediction with real-time observations show that Tensor-Var outperforms conventional and DL hybrid 4D-Var baselines in accuracy while achieving efficiency comparable to the static 3D-Var method.
Authors: Maria-Florina Balcan, Anh Tuan Nguyen, Dravyansh Sharma
Abstract: Modern machine learning algorithms, especially deep learning based techniques, typically involve careful hyperparameter tuning to achieve the best performance. Despite the surge of intense interest in practical techniques like Bayesian optimization and random search based approaches to automating this laborious and compute intensive task, the fundamental learning theoretic complexity of tuning hyperparameters for deep neural networks is poorly understood. Inspired by this glaring gap, we initiate the formal study of hyperparameter tuning complexity in deep learning through a recently introduced data driven setting. We assume that we have a series of deep learning tasks, and we have to tune hyperparameters to do well on average over the distribution of tasks. A major difficulty is that the utility function as a function of the hyperparameter is very volatile and furthermore, it is given implicitly by an optimization problem over the model parameters. To tackle this challenge, we introduce a new technique to characterize the discontinuities and oscillations of the utility function on any fixed problem instance as we vary the hyperparameter; our analysis relies on subtle concepts including tools from differential/algebraic geometry and constrained optimization. This can be used to show that the learning theoretic complexity of the corresponding family of utility functions is bounded. We instantiate our results and provide sample complexity bounds for concrete applications tuning a hyperparameter that interpolates neural activation functions and setting the kernel parameter in graph neural networks.
Authors: Long Phan (Michael Pokorny), Alice Gatti (Michael Pokorny), Ziwen Han (Michael Pokorny), Nathaniel Li (Michael Pokorny), Josephina Hu (Michael Pokorny), Hugh Zhang (Michael Pokorny), Chen Bo Calvin Zhang (Michael Pokorny), Mohamed Shaaban (Michael Pokorny), John Ling (Michael Pokorny), Sean Shi (Michael Pokorny), Michael Choi (Michael Pokorny), Anish Agrawal (Michael Pokorny), Arnav Chopra (Michael Pokorny), Adam Khoja (Michael Pokorny), Ryan Kim (Michael Pokorny), Richard Ren (Michael Pokorny), Jason Hausenloy (Michael Pokorny), Oliver Zhang (Michael Pokorny), Mantas Mazeika (Michael Pokorny), Tung Nguyen (Michael Pokorny), Daron Anderson (Michael Pokorny), Imad Ali Shah (Michael Pokorny), Mikhail Doroshenko (Michael Pokorny), Alun Cennyth Stokes (Michael Pokorny), Mobeen Mahmood (Michael Pokorny), Jaeho Lee (Michael Pokorny), Oleksandr Pokutnyi (Michael Pokorny), Oleg Iskra (Michael Pokorny), Jessica P. Wang (Michael Pokorny), Robert Gerbicz (Michael Pokorny), John-Clark Levin (Michael Pokorny), Serguei Popov (Michael Pokorny), Fiona Feng (Michael Pokorny), Steven Y. Feng (Michael Pokorny), Haoran Zhao (Michael Pokorny), Michael Yu (Michael Pokorny), Varun Gangal (Michael Pokorny), Chelsea Zou (Michael Pokorny), Zihan Wang (Michael Pokorny), Mstyslav Kazakov (Michael Pokorny), Geoff Galgon (Michael Pokorny), Johannes Schmitt (Michael Pokorny), Alvaro Sanchez (Michael Pokorny), Yongki Lee (Michael Pokorny), Will Yeadon (Michael Pokorny), Scott Sauers (Michael Pokorny), Marc Roth (Michael Pokorny), Chidozie Agu (Michael Pokorny), S{\o}ren Riis (Michael Pokorny), Fabian Giska (Michael Pokorny), Saiteja Utpala (Michael Pokorny), Antrell Cheatom (Michael Pokorny), Zachary Giboney (Michael Pokorny), Gashaw M. Goshu (Michael Pokorny), Sarah-Jane Crowson (Michael Pokorny), Mohinder Maheshbhai Naiya (Michael Pokorny), Noah Burns (Michael Pokorny), Lennart Finke (Michael Pokorny), Zerui Cheng (Michael Pokorny), Hyunwoo Park (Michael Pokorny), Francesco Fournier-Facio (Michael Pokorny), Jennifer Zampese (Michael Pokorny), John Wydallis (Michael Pokorny), John B. Wydallis (Michael Pokorny), Ryan G. Hoerr (Michael Pokorny), Mark Nandor (Michael Pokorny), Tim Gehrunger (Michael Pokorny), Jiaqi Cai (Michael Pokorny), Ben McCarty (Michael Pokorny), Jungbae Nam (Michael Pokorny), Edwin Taylor (Michael Pokorny), Jun Jin (Michael Pokorny), Gautier Abou Loume (Michael Pokorny), Hangrui Cao (Michael Pokorny), Alexis C Garretson (Michael Pokorny), Damien Sileo (Michael Pokorny), Qiuyu Ren (Michael Pokorny), Doru Cojoc (Michael Pokorny), Pavel Arkhipov (Michael Pokorny), Usman Qazi (Michael Pokorny), Aras Bacho (Michael Pokorny), Lianghui Li (Michael Pokorny), Sumeet Motwani (Michael Pokorny), Christian Schroeder de Witt (Michael Pokorny), Alexei Kopylov (Michael Pokorny), Johannes Veith (Michael Pokorny), Eric Singer (Michael Pokorny), Paolo Rissone (Michael Pokorny), Jaehyeok Jin (Michael Pokorny), Jack Wei Lun Shi (Michael Pokorny), Chris G. Willcocks (Michael Pokorny), Ameya Prabhu (Michael Pokorny), Longke Tang (Michael Pokorny), Kevin Zhou (Michael Pokorny), Emily de Oliveira Santos (Michael Pokorny), Andrey Pupasov Maksimov (Michael Pokorny), Edward Vendrow (Michael Pokorny), Kengo Zenitani (Michael Pokorny), Joshua Robinson (Michael Pokorny), Aleksandar Mikov (Michael Pokorny), Julien Guillod (Michael Pokorny), Yuqi Li (Michael Pokorny), Ben Pageler (Michael Pokorny), Joshua Vendrow (Michael Pokorny), Vladyslav Kuchkin (Michael Pokorny), Pierre Marion (Michael Pokorny), Denis Efremov (Michael Pokorny), Jayson Lynch (Michael Pokorny), Kaiqu Liang (Michael Pokorny), Andrew Gritsevskiy (Michael Pokorny), Dakotah Martinez (Michael Pokorny), Nick Crispino (Michael Pokorny), Dimitri Zvonkine (Michael Pokorny), Natanael Wildner Fraga (Michael Pokorny), Saeed Soori (Michael Pokorny), Ori Press (Michael Pokorny), Henry Tang (Michael Pokorny), Julian Salazar (Michael Pokorny), Sean R. Green (Michael Pokorny), Lina Br\"ussel (Michael Pokorny), Moon Twayana (Michael Pokorny), Aymeric Dieuleveut (Michael Pokorny), T. Ryan Rogers (Michael Pokorny), Wenjin Zhang (Michael Pokorny), Ross Finocchio (Michael Pokorny), Bikun Li (Michael Pokorny), Jinzhou Yang (Michael Pokorny), Arun Rao (Michael Pokorny), Gabriel Loiseau (Michael Pokorny), Mikhail Kalinin (Michael Pokorny), Marco Lukas (Michael Pokorny), Ciprian Manolescu (Michael Pokorny), Nate Stambaugh (Michael Pokorny), Subrata Mishra (Michael Pokorny), Ariel Ghislain Kemogne Kamdoum (Michael Pokorny), Tad Hogg (Michael Pokorny), Alvin Jin (Michael Pokorny), Carlo Bosio (Michael Pokorny), Gongbo Sun (Michael Pokorny), Brian P Coppola (Michael Pokorny), Haline Heidinger (Michael Pokorny), Rafael Sayous (Michael Pokorny), Stefan Ivanov (Michael Pokorny), Joseph M Cavanagh (Michael Pokorny), Jiawei Shen (Michael Pokorny), Joseph Marvin Imperial (Michael Pokorny), Philippe Schwaller (Michael Pokorny), Shaipranesh Senthilkuma (Michael Pokorny), Andres M Bran (Michael Pokorny), Andres Algaba (Michael Pokorny), Brecht Verbeken (Michael Pokorny), Kelsey Van den Houte (Michael Pokorny), Lynn Van Der Sypt (Michael Pokorny), David Noever (Michael Pokorny), Lisa Schut (Michael Pokorny), Ilia Sucholutsky (Michael Pokorny), Evgenii Zheltonozhskii (Michael Pokorny), Qiaochu Yuan (Michael Pokorny), Derek Lim (Michael Pokorny), Richard Stanley (Michael Pokorny), Shankar Sivarajan (Michael Pokorny), Tong Yang (Michael Pokorny), John Maar (Michael Pokorny), Julian Wykowski (Michael Pokorny), Mart\'i Oller (Michael Pokorny), Jennifer Sandlin (Michael Pokorny), Anmol Sahu (Michael Pokorny), Cesare Giulio Ardito (Michael Pokorny), Yuzheng Hu (Michael Pokorny), Felipe Meneguitti Dias (Michael Pokorny), Tobias Kreiman (Michael Pokorny), Kaivalya Rawal (Michael Pokorny), Tobias Garcia Vilchis (Michael Pokorny), Yuexuan Zu (Michael Pokorny), Martin Lackner (Michael Pokorny), James Koppel (Michael Pokorny), Jeremy Nguyen (Michael Pokorny), Daniil S. Antonenko (Michael Pokorny), Steffi Chern (Michael Pokorny), Bingchen Zhao (Michael Pokorny), Pierrot Arsene (Michael Pokorny), Sergey Ivanov (Michael Pokorny), Rafa{\l} Po\'swiata (Michael Pokorny), Chenguang Wang (Michael Pokorny), Daofeng Li (Michael Pokorny), Donato Crisostomi (Michael Pokorny), Ali Dehghan (Michael Pokorny), Andrea Achilleos (Michael Pokorny), John Arnold Ambay (Michael Pokorny), Benjamin Myklebust (Michael Pokorny), Archan Sen (Michael Pokorny), David Perrella (Michael Pokorny), Nurdin Kaparov (Michael Pokorny), Mark H Inlow (Michael Pokorny), Allen Zang (Michael Pokorny), Kalyan Ramakrishnan (Michael Pokorny), Daniil Orel (Michael Pokorny), Vladislav Poritski (Michael Pokorny), Shalev Ben-David (Michael Pokorny), Zachary Berger (Michael Pokorny), Parker Whitfill (Michael Pokorny), Michael Foster (Michael Pokorny), Daniel Munro (Michael Pokorny), Linh Ho (Michael Pokorny), Dan Bar Hava (Michael Pokorny), Aleksey Kuchkin (Michael Pokorny), Robert Lauff (Michael Pokorny), David Holmes (Michael Pokorny), Frank Sommerhage (Michael Pokorny), Anji Zhang (Michael Pokorny), Richard Moat (Michael Pokorny), Keith Schneider (Michael Pokorny), Daniel Pyda (Michael Pokorny), Zakayo Kazibwe (Michael Pokorny), Mukhwinder Singh (Michael Pokorny), Don Clarke (Michael Pokorny), Dae Hyun Kim (Michael Pokorny), Sara Fish (Michael Pokorny), Veit Elser (Michael Pokorny), Victor Efren Guadarrama Vilchis (Michael Pokorny), Immo Klose (Michael Pokorny), Christoph Demian (Michael Pokorny), Ujjwala Anantheswaran (Michael Pokorny), Adam Zweiger (Michael Pokorny), Guglielmo Albani (Michael Pokorny), Jeffery Li (Michael Pokorny), Nicolas Daans (Michael Pokorny), Maksim Radionov (Michael Pokorny), V\'aclav Rozho\v{n} (Michael Pokorny), Vincent Ginis (Michael Pokorny), Ziqiao Ma (Michael Pokorny), Christian Stump (Michael Pokorny), Jacob Platnick (Michael Pokorny), Volodymyr Nevirkovets (Michael Pokorny), Luke Basler (Michael Pokorny), Marco Piccardo (Michael Pokorny), Niv Cohen (Michael Pokorny), Virendra Singh (Michael Pokorny), Josef Tkadlec (Michael Pokorny), Paul Rosu (Michael Pokorny), Alan Goldfarb (Michael Pokorny), Piotr Padlewski (Michael Pokorny), Stanislaw Barzowski (Michael Pokorny), Kyle Montgomery (Michael Pokorny), Aline Menezes (Michael Pokorny), Arkil Patel (Michael Pokorny), Zixuan Wang (Michael Pokorny), Jamie Tucker-Foltz (Michael Pokorny), Jack Stade (Michael Pokorny), Declan Grabb (Michael Pokorny), Tom Goertzen (Michael Pokorny), Fereshteh Kazemi (Michael Pokorny), Jeremiah Milbauer (Michael Pokorny), Abhishek Shukla (Michael Pokorny), Hossam Elgnainy (Michael Pokorny), Yan Carlos Leyva Labrador (Michael Pokorny), Hao He (Michael Pokorny), Ling Zhang (Michael Pokorny), Alan Givr\'e (Michael Pokorny), Hew Wolff (Michael Pokorny), G\"ozdenur Demir (Michael Pokorny), Muhammad Fayez Aziz (Michael Pokorny), Younesse Kaddar (Michael Pokorny), Ivar \"Angquist (Michael Pokorny), Yanxu Chen (Michael Pokorny), Elliott Thornley (Michael Pokorny), Robin Zhang (Michael Pokorny), Jiayi Pan (Michael Pokorny), Antonio Terpin (Michael Pokorny), Niklas Muennighoff (Michael Pokorny), Hailey Schoelkopf (Michael Pokorny), Eric Zheng (Michael Pokorny), Avishy Carmi (Michael Pokorny), Jainam Shah (Michael Pokorny), Ethan D. L. Brown (Michael Pokorny), Kelin Zhu (Michael Pokorny), Max Bartolo (Michael Pokorny), Richard Wheeler (Michael Pokorny), Andrew Ho (Michael Pokorny), Shaul Barkan (Michael Pokorny), Jiaqi Wang (Michael Pokorny), Martin Stehberger (Michael Pokorny), Egor Kretov (Michael Pokorny), Peter Bradshaw (Michael Pokorny), JP Heimonen (Michael Pokorny), Kaustubh Sridhar (Michael Pokorny), Zaki Hossain (Michael Pokorny), Ido Akov (Michael Pokorny), Yury Makarychev (Michael Pokorny), Joanna Tam (Michael Pokorny), Hieu Hoang (Michael Pokorny), David M. Cunningham (Michael Pokorny), Vladimir Goryachev (Michael Pokorny), Demosthenes Patramanis (Michael Pokorny), Michael Krause (Michael Pokorny), Andrew Redenti (Michael Pokorny), David Aldous (Michael Pokorny), Jesyin Lai (Michael Pokorny), Shannon Coleman (Michael Pokorny), Jiangnan Xu (Michael Pokorny), Sangwon Lee (Michael Pokorny), Ilias Magoulas (Michael Pokorny), Sandy Zhao (Michael Pokorny), Ning Tang (Michael Pokorny), Michael K. Cohen (Michael Pokorny), Micah Carroll (Michael Pokorny), Orr Paradise (Michael Pokorny), Jan Hendrik Kirchner (Michael Pokorny), Stefan Steinerberger (Michael Pokorny), Maksym Ovchynnikov (Michael Pokorny), Jason O. Matos (Michael Pokorny), Adithya Shenoy (Michael Pokorny), Michael Wang (Michael Pokorny), Yuzhou Nie (Michael Pokorny), Paolo Giordano (Michael Pokorny), Philipp Petersen (Michael Pokorny), Anna Sztyber-Betley (Michael Pokorny), Paolo Faraboschi (Michael Pokorny), Robin Riblet (Michael Pokorny), Jonathan Crozier (Michael Pokorny), Shiv Halasyamani (Michael Pokorny), Antonella Pinto (Michael Pokorny), Shreyas Verma (Michael Pokorny), Prashant Joshi (Michael Pokorny), Eli Meril (Michael Pokorny), Zheng-Xin Yong (Michael Pokorny), Allison Tee (Michael Pokorny), J\'er\'emy Andr\'eoletti (Michael Pokorny), Orion Weller (Michael Pokorny), Raghav Singhal (Michael Pokorny), Gang Zhang (Michael Pokorny), Alexander Ivanov (Michael Pokorny), Seri Khoury (Michael Pokorny), Nils Gustafsson (Michael Pokorny), Hamid Mostaghimi (Michael Pokorny), Kunvar Thaman (Michael Pokorny), Qijia Chen (Michael Pokorny), Tran Quoc Kh\'anh (Michael Pokorny), Jacob Loader (Michael Pokorny), Stefano Cavalleri (Michael Pokorny), Hannah Szlyk (Michael Pokorny), Zachary Brown (Michael Pokorny), Himanshu Narayan (Michael Pokorny), Jonathan Roberts (Michael Pokorny), William Alley (Michael Pokorny), Kunyang Sun (Michael Pokorny), Ryan Stendall (Michael Pokorny), Max Lamparth (Michael Pokorny), Anka Reuel (Michael Pokorny), Ting Wang (Michael Pokorny), Hanmeng Xu (Michael Pokorny), Pablo Hern\'andez-C\'amara (Michael Pokorny), Freddie Martin (Michael Pokorny), Thomas Preu (Michael Pokorny), Tomek Korbak (Michael Pokorny), Marcus Abramovitch (Michael Pokorny), Dominic Williamson (Michael Pokorny), Ida Bosio (Michael Pokorny), Ziye Chen (Michael Pokorny), Bir\'o B\'alint (Michael Pokorny), Eve J. Y. Lo (Michael Pokorny), Maria In\^es S. Nunes (Michael Pokorny), Yibo Jiang (Michael Pokorny), M Saiful Bari (Michael Pokorny), Peyman Kassani (Michael Pokorny), Zihao Wang (Michael Pokorny), Behzad Ansarinejad (Michael Pokorny), Yewen Sun (Michael Pokorny), Stephane Durand (Michael Pokorny), Guillaume Douville (Michael Pokorny), Daniel Tordera (Michael Pokorny), George Balabanian (Michael Pokorny), Earth Anderson (Michael Pokorny), Lynna Kvistad (Michael Pokorny), Alejandro Jos\'e Moyano (Michael Pokorny), Hsiaoyun Milliron (Michael Pokorny), Ahmad Sakor (Michael Pokorny), Murat Eron (Michael Pokorny), Isaac C. McAlister (Michael Pokorny), Andrew Favre D. O. (Michael Pokorny), Shailesh Shah (Michael Pokorny), Xiaoxiang Zhou (Michael Pokorny), Firuz Kamalov (Michael Pokorny), Ronald Clark (Michael Pokorny), Sherwin Abdoli (Michael Pokorny), Tim Santens (Michael Pokorny), Harrison K Wang (Michael Pokorny), Evan Chen (Michael Pokorny), Alessandro Tomasiello (Michael Pokorny), G. Bruno De Luca (Michael Pokorny), Shi-Zhuo Looi (Michael Pokorny), Vinh-Kha Le (Michael Pokorny), Noam Kolt (Michael Pokorny), Niels M\"undler (Michael Pokorny), Avi Semler (Michael Pokorny), Emma Rodman (Michael Pokorny), Jacob Drori (Michael Pokorny), Carl J Fossum (Michael Pokorny), Luk Gloor (Michael Pokorny), Milind Jagota (Michael Pokorny), Ronak Pradeep (Michael Pokorny), Honglu Fan (Michael Pokorny), Tej Shah (Michael Pokorny), Jonathan Eicher (Michael Pokorny), Michael Chen (Michael Pokorny), Kushal Thaman (Michael Pokorny), William Merrill (Michael Pokorny), Moritz Firsching (Michael Pokorny), Carter Harris (Michael Pokorny), Stefan Ciob\^ac\u{a} (Michael Pokorny), Jason Gross (Michael Pokorny), Rohan Pandey (Michael Pokorny), Ilya Gusev (Michael Pokorny), Adam Jones (Michael Pokorny), Shashank Agnihotri (Michael Pokorny), Pavel Zhelnov (Michael Pokorny), Siranut Usawasutsakorn (Michael Pokorny), Mohammadreza Mofayezi (Michael Pokorny), Alexander Piperski (Michael Pokorny), Marc Carauleanu (Michael Pokorny), David K. Zhang (Michael Pokorny), Kostiantyn Dobarskyi (Michael Pokorny), Dylan Ler (Michael Pokorny), Roman Leventov (Michael Pokorny), Ignat Soroko (Michael Pokorny), Thorben Jansen (Michael Pokorny), Scott Creighton (Michael Pokorny), Pascal Lauer (Michael Pokorny), Joshua Duersch (Michael Pokorny), Vage Taamazyan (Michael Pokorny), Dario Bezzi (Michael Pokorny), Wiktor Morak (Michael Pokorny), Wenjie Ma (Michael Pokorny), William Held (Michael Pokorny), Tran {\DJ}uc Huy (Michael Pokorny), Ruicheng Xian (Michael Pokorny), Armel Randy Zebaze (Michael Pokorny), Mohanad Mohamed (Michael Pokorny), Julian Noah Leser (Michael Pokorny), Michelle X Yuan (Michael Pokorny), Laila Yacar (Michael Pokorny), Johannes Lengler (Michael Pokorny), Katarzyna Olszewska (Michael Pokorny), Hossein Shahrtash (Michael Pokorny), Edson Oliveira (Michael Pokorny), Joseph W. Jackson (Michael Pokorny), Daniel Espinosa Gonzalez (Michael Pokorny), Andy Zou (Michael Pokorny), Muthu Chidambaram (Michael Pokorny), Timothy Manik (Michael Pokorny), Hector Haffenden (Michael Pokorny), Dashiell Stander (Michael Pokorny), Ali Dasouqi (Michael Pokorny), Alexander Shen (Michael Pokorny), Emilien Duc (Michael Pokorny), Bita Golshani (Michael Pokorny), David Stap (Michael Pokorny), Mikalai Uzhou (Michael Pokorny), Alina Borisovna Zhidkovskaya (Michael Pokorny), Lukas Lewark (Michael Pokorny), Miguel Orbegozo Rodriguez (Michael Pokorny), M\'aty\'as Vincze (Michael Pokorny), Dustin Wehr (Michael Pokorny), Colin Tang (Michael Pokorny), Shaun Phillips (Michael Pokorny), Fortuna Samuele (Michael Pokorny), Jiang Muzhen (Michael Pokorny), Fredrik Ekstr\"om (Michael Pokorny), Angela Hammon (Michael Pokorny), Oam Patel (Michael Pokorny), Faraz Farhidi (Michael Pokorny), George Medley (Michael Pokorny), Forough Mohammadzadeh (Michael Pokorny), Madellene Pe\~naflor (Michael Pokorny), Haile Kassahun (Michael Pokorny), Alena Friedrich (Michael Pokorny), Claire Sparrow (Michael Pokorny), Rayner Hernandez Perez (Michael Pokorny), Taom Sakal (Michael Pokorny), Omkar Dhamane (Michael Pokorny), Ali Khajegili Mirabadi (Michael Pokorny), Eric Hallman (Michael Pokorny), Kenchi Okutsu (Michael Pokorny), Mike Battaglia (Michael Pokorny), Mohammad Maghsoudimehrabani (Michael Pokorny), Alon Amit (Michael Pokorny), Dave Hulbert (Michael Pokorny), Roberto Pereira (Michael Pokorny), Simon Weber (Michael Pokorny), Handoko (Michael Pokorny), Anton Peristyy (Michael Pokorny), Stephen Malina (Michael Pokorny), Samuel Albanie (Michael Pokorny), Will Cai (Michael Pokorny), Mustafa Mehkary (Michael Pokorny), Rami Aly (Michael Pokorny), Frank Reidegeld (Michael Pokorny), Anna-Katharina Dick (Michael Pokorny), Cary Friday (Michael Pokorny), Jasdeep Sidhu (Michael Pokorny), Hassan Shapourian (Michael Pokorny), Wanyoung Kim (Michael Pokorny), Mariana Costa (Michael Pokorny), Hubeyb Gurdogan (Michael Pokorny), Brian Weber (Michael Pokorny), Harsh Kumar (Michael Pokorny), Tong Jiang (Michael Pokorny), Arunim Agarwal (Michael Pokorny), Chiara Ceconello (Michael Pokorny), Warren S. Vaz (Michael Pokorny), Chao Zhuang (Michael Pokorny), Haon Park (Michael Pokorny), Andrew R. Tawfeek (Michael Pokorny), Daattavya Aggarwal (Michael Pokorny), Michael Kirchhof (Michael Pokorny), Linjie Dai (Michael Pokorny), Evan Kim (Michael Pokorny), Johan Ferret (Michael Pokorny), Yuzhou Wang (Michael Pokorny), Minghao Yan (Michael Pokorny), Krzysztof Burdzy (Michael Pokorny), Lixin Zhang (Michael Pokorny), Antonio Franca (Michael Pokorny), Diana T. Pham (Michael Pokorny), Kang Yong Loh (Michael Pokorny), Joshua Robinson (Michael Pokorny), Abram Jackson (Michael Pokorny), Shreen Gul (Michael Pokorny), Gunjan Chhablani (Michael Pokorny), Zhehang Du (Michael Pokorny), Adrian Cosma (Michael Pokorny), Jesus Colino (Michael Pokorny), Colin White (Michael Pokorny), Jacob Votava (Michael Pokorny), Vladimir Vinnikov (Michael Pokorny), Ethan Delaney (Michael Pokorny), Petr Spelda (Michael Pokorny), Vit Stritecky (Michael Pokorny), Syed M. Shahid (Michael Pokorny), Jean-Christophe Mourrat (Michael Pokorny), Lavr Vetoshkin (Michael Pokorny), Koen Sponselee (Michael Pokorny), Renas Bacho (Michael Pokorny), Florencia de la Rosa (Michael Pokorny), Xiuyu Li (Michael Pokorny), Guillaume Malod (Michael Pokorny), Leon Lang (Michael Pokorny), Julien Laurendeau (Michael Pokorny), Dmitry Kazakov (Michael Pokorny), Fatimah Adesanya (Michael Pokorny), Julien Portier (Michael Pokorny), Lawrence Hollom (Michael Pokorny), Victor Souza (Michael Pokorny), Yuchen Anna Zhou (Michael Pokorny), Julien Degorre (Michael Pokorny), Yi\u{g}it Yal{\i}n (Michael Pokorny), Gbenga Daniel Obikoya (Michael Pokorny), Luca Arnaboldi (Michael Pokorny), Rai (Michael Pokorny), Filippo Bigi (Quinn), M. C. Bosc\'a (Quinn), Oleg Shumar (Quinn), Kaniuar Bacho (Quinn), Pierre Clavier (Quinn), Gabriel Recchia (Quinn), Mara Popescu (Quinn), Nikita Shulga (Quinn), Ngefor Mildred Tanwie (Quinn), Denis Peskoff (Quinn), Thomas C. H. Lux (Quinn), Ben Rank (Quinn), Colin Ni (Quinn), Matthew Brooks (Quinn), Alesia Yakimchyk (Quinn), Huanxu (Quinn), Liu (Tony), Olle H\"aggstr\"om (Tony), Emil Verkama (Tony), Hans Gundlach (Tony), Leonor Brito-Santana (Tony), Brian Amaro (Tony), Vivek Vajipey (Tony), Rynaa Grover (Tony), Yiyang Fan (Tony), Gabriel Poesia Reis e Silva (Tony), Linwei Xin (Tony), Yosi Kratish (Tony), Jakub {\L}ucki (Tony), Wen-Ding Li (Tony), Sivakanth Gopi (Tony), Andrea Caciolai (Tony), Justin Xu (Tony), Kevin Joseph Scaria (Tony), Freddie Vargus (Tony), Farzad Habibi (Tony), Long (Tony), Lian, Emanuele Rodol\`a, Jules Robins, Vincent Cheng, Tony Fruhauff, Brad Raynor, Hao Qi, Xi Jiang, Ben Segev, Jingxuan Fan, Sarah Martinson, Erik Y. Wang, Kaylie Hausknecht, Michael P. Brenner, Mao Mao, Xinyu Zhang, David Avagian, Eshawn Jessica Scipio, Alon Ragoler, Justin Tan, Blake Sims, Rebeka Plecnik, Aaron Kirtland, Omer Faruk Bodur, D. P. Shinde, Zahra Adoul, Mohamed Zekry, Ali Karakoc, Tania C. B. Santos, Samir Shamseldeen, Loukmane Karim, Anna Liakhovitskaia, Nate Resman, Nicholas Farina, Juan Carlos Gonzalez, Gabe Maayan, Sarah Hoback, Rodrigo De Oliveira Pena, Glen Sherman, Elizabeth Kelley, Hodjat Mariji, Rasoul Pouriamanesh, Wentao Wu, Sandra Mendoza, Ismail Alarab, Joshua Cole, Danyelle Ferreira, Bryan Johnson, Mohammad Safdari, Liangti Dai, Siriphan Arthornthurasuk, Alexey Pronin, Jing Fan, Angel Ramirez-Trinidad, Ashley Cartwright, Daphiny Pottmaier, Omid Taheri, David Outevsky, Stanley Stepanic, Samuel Perry, Luke Askew, Ra\'ul Adri\'an Huerta Rodr\'iguez, Ali M. R. Minissi, Sam Ali, Ricardo Lorena, Krishnamurthy Iyer, Arshad Anil Fasiludeen, Sk Md Salauddin, Murat Islam, Juan Gonzalez, Josh Ducey, Maja Somrak, Vasilios Mavroudis, Eric Vergo, Juehang Qin, Benj\'amin Borb\'as, Eric Chu, Jack Lindsey, Anil Radhakrishnan, Antoine Jallon, I. M. J. McInnis, Pawan Kumar, Laxman Prasad Goswami, Daniel Bugas, Nasser Heydari, Ferenc Jeanplong, Archimedes Apronti, Abdallah Galal, Ng Ze-An, Ankit Singh, Joan of Arc Xavier, Kanu Priya Agarwal, Mohammed Berkani, Benedito Alves de Oliveira Junior, Dmitry Malishev, Nicolas Remy, Taylor D. Hartman, Tim Tarver, Stephen Mensah, Javier Gimenez, Roselynn Grace Montecillo, Russell Campbell, Asankhaya Sharma, Khalida Meer, Xavier Alapont, Deepakkumar Patil, Rajat Maheshwari, Abdelkader Dendane, Priti Shukla, Sergei Bogdanov, S\"oren M\"oller, Muhammad Rehan Siddiqi, Prajvi Saxena, Himanshu Gupta, Innocent Enyekwe, Ragavendran P V, Zienab EL-Wasif, Aleksandr Maksapetyan, Vivien Rossbach, Chris Harjadi, Mohsen Bahaloohoreh, Song Bian, John Lai, Justine Leon Uro, Greg Bateman, Mohamed Sayed, Ahmed Menshawy, Darling Duclosel, Yashaswini Jain, Ashley Aaron, Murat Tiryakioglu, Sheeshram Siddh, Keith Krenek, Alex Hoover, Joseph McGowan, Tejal Patwardhan, Summer Yue, Alexandr Wang, Dan Hendrycks
Abstract: Benchmarks are important tools for tracking the rapid advancements in large language model (LLM) capabilities. However, benchmarks are not keeping pace in difficulty: LLMs now achieve over 90\% accuracy on popular benchmarks like MMLU, limiting informed measurement of state-of-the-art LLM capabilities. In response, we introduce Humanity's Last Exam (HLE), a multi-modal benchmark at the frontier of human knowledge, designed to be the final closed-ended academic benchmark of its kind with broad subject coverage. HLE consists of 3,000 questions across dozens of subjects, including mathematics, humanities, and the natural sciences. HLE is developed globally by subject-matter experts and consists of multiple-choice and short-answer questions suitable for automated grading. Each question has a known solution that is unambiguous and easily verifiable, but cannot be quickly answered via internet retrieval. State-of-the-art LLMs demonstrate low accuracy and calibration on HLE, highlighting a significant gap between current LLM capabilities and the expert human frontier on closed-ended academic questions. To inform research and policymaking upon a clear understanding of model capabilities, we publicly release HLE at https://lastexam.ai.
URLs: https://lastexam.ai.
Authors: Xu Chu, Hanlin Xue, Zhijie Tan, Bingce Wang, Tong Mo, Weiping Li
Abstract: The success of Large Language Models (LLMs) in various domains has led researchers to apply them to graph-related problems by converting graph data into natural language text. However, unlike graph data, natural language inherently has sequential order. We observe a counter-intuitive fact that when the order of nodes or edges in the natural language description of a graph is shuffled, despite describing the same graph, model performance fluctuates between high performance and random guessing. Additionally, due to LLMs' limited input context length, current methods typically randomly sample neighbors of target nodes as representatives of their neighborhood, which may not always be effective for accurate reasoning. To address these gaps, we introduce GraphSOS (Graph Sampling and Order Selection). This novel model framework features an Order Selector Module to ensure proper serialization order of the graph and a Subgraph Sampling Module to sample subgraphs with better structure for better reasoning. Furthermore, we propose Graph CoT obtained through distillation, and enhance LLM's reasoning and zero-shot learning capabilities for graph tasks through instruction tuning. Experiments on multiple datasets for node classification and graph question-answering demonstrate that GraphSOS improves LLMs' performance and generalization ability on graph tasks.
Authors: Yixing Jiang, Kameron C. Black, Gloria Geng, Danny Park, James Zou, Andrew Y. Ng, Jonathan H. Chen
Abstract: Recent large language models (LLMs) have demonstrated significant advancements, particularly in their ability to serve as agents thereby surpassing their traditional role as chatbots. These agents can leverage their planning and tool utilization capabilities to address tasks specified at a high level. However, a standardized dataset to benchmark the agent capabilities of LLMs in medical applications is currently lacking, making the evaluation of LLMs on complex tasks in interactive healthcare environments challenging. To address this gap, we introduce MedAgentBench, a broad evaluation suite designed to assess the agent capabilities of large language models within medical records contexts. MedAgentBench encompasses 300 patient-specific clinically-derived tasks from 10 categories written by human physicians, realistic profiles of 100 patients with over 700,000 data elements, a FHIR-compliant interactive environment, and an accompanying codebase. The environment uses the standard APIs and communication infrastructure used in modern EMR systems, so it can be easily migrated into live EMR systems. MedAgentBench presents an unsaturated agent-oriented benchmark that current state-of-the-art LLMs exhibit some ability to succeed at. The best model (Claude 3.5 Sonnet v2) achieves a success rate of 69.67%. However, there is still substantial space for improvement which gives the community a next direction to optimize. Furthermore, there is significant variation in performance across task categories. MedAgentBench establishes this and is publicly available at https://github.com/stanfordmlgroup/MedAgentBench , offering a valuable framework for model developers to track progress and drive continuous improvements in the agent capabilities of large language models within the medical domain.
Authors: Makoto Shing, Kou Misaki, Han Bao, Sho Yokoi, Takuya Akiba
Abstract: Causal language models have demonstrated remarkable capabilities, but their size poses significant challenges for deployment in resource-constrained environments. Knowledge distillation, a widely-used technique for transferring knowledge from a large teacher model to a small student model, presents a promising approach for model compression. A significant remaining issue lies in the major differences between teacher and student models, namely the substantial capacity gap, mode averaging, and mode collapse, which pose barriers during distillation. To address these issues, we introduce $\textit{Temporally Adaptive Interpolated Distillation (TAID)}$, a novel knowledge distillation approach that dynamically interpolates student and teacher distributions through an adaptive intermediate distribution, gradually shifting from the student's initial distribution towards the teacher's distribution. We provide a theoretical analysis demonstrating TAID's ability to prevent mode collapse and empirically show its effectiveness in addressing the capacity gap while balancing mode averaging and mode collapse. Our comprehensive experiments demonstrate TAID's superior performance across various model sizes and architectures in both instruction tuning and pre-training scenarios. Furthermore, we showcase TAID's practical impact by developing two state-of-the-art compact foundation models: $\texttt{TAID-LLM-1.5B}$ for language tasks and $\texttt{TAID-VLM-2B}$ for vision-language tasks. These results demonstrate TAID's effectiveness in creating high-performing and efficient models, advancing the development of more accessible AI technologies.
Authors: Pir Bakhsh Khokhar, Viviana Pentangelo, Fabio Palomba, Carmine Gravino
Abstract: Diabetes mellitus (DM) is a global health issue of significance that must be diagnosed as early as possible and managed well. This study presents a framework for diabetes prediction using Machine Learning (ML) models, complemented with eXplainable Artificial Intelligence (XAI) tools, to investigate both the predictive accuracy and interpretability of the predictions from ML models. Data Preprocessing is based on the Synthetic Minority Oversampling Technique (SMOTE) and feature scaling used on the Diabetes Binary Health Indicators dataset to deal with class imbalance and variability of clinical features. The ensemble model provided high accuracy, with a test accuracy of 92.50% and an ROC-AUC of 0.975. BMI, Age, General Health, Income, and Physical Activity were the most influential predictors obtained from the model explanations. The results of this study suggest that ML combined with XAI is a promising means of developing accurate and computationally transparent tools for use in healthcare systems.
Authors: Harshwardhan Praveen, Jacob Brown, Christopher Earls
Abstract: In this work, we present a mesh-independent, data-driven library, chebgreen, to mathematically model one-dimensional systems, possessing an associated control parameter, and whose governing partial differential equation is unknown. The proposed method learns an Empirical Green's Function for the associated, but hidden, boundary value problem, in the form of a Rational Neural Network from which we subsequently construct a bivariate representation in a Chebyshev basis. We uncover the Green's function, at an unseen control parameter value, by interpolating the left and right singular functions within a suitable library, expressed as points on a manifold of Quasimatrices, while the associated singular values are interpolated with Lagrange polynomials.
Authors: Gon\c{c}alo Paulo, Stepan Shabalin, Nora Belrose
Abstract: Sparse autoencoders (SAEs) extract human-interpretable features from deep neural networks by transforming their activations into a sparse, higher dimensional latent space, and then reconstructing the activations from these latents. Transcoders are similar to SAEs, but they are trained to reconstruct the output of a component of a deep network given its input. In this work, we compare the features found by transcoders and SAEs trained on the same model and data, finding that transcoder features are significantly more interpretable. We also propose skip transcoders, which add an affine skip connection to the transcoder architecture, and show that these achieve lower reconstruction loss with no effect on interpretability.
Authors: Haitong Ma, Tianyi Chen, Kai Wang, Na Li, Bo Dai
Abstract: Diffusion policies have achieved superior performance in imitation learning and offline reinforcement learning (RL) due to their rich expressiveness. However, the vanilla diffusion training procedure requires samples from target distribution, which is impossible in online RL since we cannot sample from the optimal policy, making training diffusion policies highly non-trivial in online RL. Backpropagating policy gradient through the diffusion process incurs huge computational costs and instability, thus being expensive and impractical. To enable efficient diffusion policy training for online RL, we propose Soft Diffusion Actor-Critic (SDAC), exploiting the viewpoint of diffusion models as noise-perturbed energy-based models. The proposed SDAC relies solely on the state-action value function as the energy functions to train diffusion policies, bypassing sampling from the optimal policy while maintaining lightweight computations. We conducted comprehensive comparisons on MuJoCo benchmarks. The empirical results show that SDAC outperforms all recent diffusion-policy online RLs on most tasks, and improves more than 120% over soft actor-critic on complex locomotion tasks such as Humanoid and Ant.
Authors: Isha Puri, Shivchander Sudalairaj, Guangxuan Xu, Kai Xu, Akash Srivastava
Abstract: Large language models (LLMs) have achieved significant performance gains via scaling up model sizes and/or data. However, recent evidence suggests diminishing returns from such approaches, motivating scaling the computation spent at inference time. Existing inference-time scaling methods, usually with reward models, cast the task as a search problem, which tends to be vulnerable to reward hacking as a consequence of approximation errors in reward models. In this paper, we instead cast inference-time scaling as a probabilistic inference task and leverage sampling-based techniques to explore the typical set of the state distribution of a state-space model with an approximate likelihood, rather than optimize for its mode directly. We propose a novel inference-time scaling approach by adapting particle-based Monte Carlo methods to this task. Our empirical evaluation demonstrates that our methods have a 4-16x better scaling rate over our deterministic search counterparts on various challenging mathematical reasoning tasks. Using our approach, we show that Qwen2.5-Math-1.5B-Instruct can surpass GPT-4o accuracy in only 4 rollouts, while Qwen2.5-Math-7B-Instruct scales to o1 level accuracy in only 32 rollouts. Our work not only presents an effective method to inference-time scaling, but also connects the rich literature in probabilistic inference with inference-time scaling of LLMs to develop more robust algorithms in future work. Code, videos, and further information available at https://probabilistic-inference-scaling.github.io.
Authors: Jiayang Meng, Tao Huang, Hong Chen, Xin Shi, Qingyu Huang, Chen Hou
Abstract: The widespread deployment of deep learning models in privacy-sensitive domains has amplified concerns regarding privacy risks, particularly those stemming from gradient leakage during training. Current privacy assessments primarily rely on post-training attack simulations. However, these methods are inherently reactive, unable to encompass all potential attack scenarios, and often based on idealized adversarial assumptions. These limitations underscore the need for proactive approaches to privacy risk assessment during the training process. To address this gap, we propose the concept of privacy tokens, which are derived directly from private gradients during training. Privacy tokens encapsulate gradient features and, when combined with data features, offer valuable insights into the extent of private information leakage from training data, enabling real-time measurement of privacy risks without relying on adversarial attack simulations. Additionally, we employ Mutual Information (MI) as a robust metric to quantify the relationship between training data and gradients, providing precise and continuous assessments of privacy leakage throughout the training process. Extensive experiments validate our framework, demonstrating the effectiveness of privacy tokens and MI in identifying and quantifying privacy risks. This proactive approach marks a significant advancement in privacy monitoring, promoting the safer deployment of deep learning models in sensitive applications.
Authors: Ruoyan Li, Dipti Ranjan Sahu, Guy Van den Broeck, Zhe Zeng
Abstract: While deep generative models~(DGMs) have demonstrated remarkable success in capturing complex data distributions, they consistently fail to learn constraints that encode domain knowledge and thus require constraint integration. Existing solutions to this challenge have primarily relied on heuristic methods and often ignore the underlying data distribution, harming the generative performance. In this work, we propose a probabilistically sound approach for enforcing the hard constraints into DGMs to generate constraint-compliant and realistic data. This is achieved by our proposed gradient estimators that allow the constrained distribution, the data distribution conditioned on constraints, to be differentiably learned. We carry out extensive experiments with various DGM model architectures over five image datasets and three scientific applications in which domain knowledge is governed by linear equality constraints. We validate that the standard DGMs almost surely generate data violating the constraints. Among all the constraint integration strategies, ours not only guarantees the satisfaction of constraints in generation but also archives superior generative performance than the other methods across every benchmark.
Authors: Xinyu Yang, Tianqi Chen, Beidi Chen
Abstract: Context-augmented generation (CAG) techniques, including RAG and ICL, require the efficient combination of multiple contexts to generate responses to user queries. Directly inputting these contexts as a sequence introduces a considerable computational burden by re-encoding the combined selection of contexts for every request. To address this, we explore the promising potential of parallel encoding to independently pre-compute and cache each context's KV states. This approach enables the direct loading of cached states during inference while accommodating more contexts through position reuse across contexts. However, due to misalignments in attention distribution, directly applying parallel encoding results in a significant performance drop. To enable effective and efficient CAG, we propose Adaptive Parallel Encoding ($\textbf{APE}$), which brings shared prefix, attention temperature, and scaling factor to align the distribution of parallel encoding with sequential encoding. Results on RAG and ICL tasks demonstrate that APE can preserve 98% and 93% sequential encoding performance using the same inputs while outperforming parallel encoding by 3.6% and 7.9%, respectively. It also scales to many-shot CAG, effectively encoding hundreds of contexts in parallel. Efficiency evaluation shows that APE can achieve an end-to-end 4.5$\times$ speedup by reducing 28$\times$ prefilling time for a 128K-length context.
Authors: Han Qi, Haochen Yang, Qiaosheng Zhang, Zhuoran Yang
Abstract: We study the problem of reinforcement learning from human feedback (RLHF), a critical problem in training large language models, from a theoretical perspective. Our main contribution is the design of novel sample-efficient RLHF algorithms based on information-directed sampling (IDS), an online decision-making principle inspired by information theory. Our algorithms maximize the sum of the value function and a mutual information term that encourages exploration of the unknown environment (which quantifies the information gained about the environment through observed human feedback data). To tackle the challenge of large state spaces and improve sample efficiency, we construct a simplified \emph{surrogate environment} and introduce a novel distance measure (named the \emph{$\ell_g$-distance}), enabling our IDS-based algorithm to achieve a Bayesian regret upper bound of order $O(H^{\frac{3}{2}}\sqrt{\log(K(\epsilon)) T})$, where $H$ is the episode length, $T$ is the number of episode and $K(\epsilon)$ is related to the covering number of the environment. Specializing to the tabular settings, this regret bound is of order $\tilde{O}(H^2\sqrt{SAT})$, where $S$ and $A$ are the numbers of states and actions. Finally, we propose an Approximate-IDS algorithm that is computationally more efficient while maintaining nearly the same sample efficiency. The design principle of this approximate algorithm is not only effective in RLHF settings but also applicable to the standard RL framework. Moreover, our work showcases the value of information theory in reinforcement learning and in the training of large language models.
Authors: Shizhou Xu, Thomas Strohmer
Abstract: How can we effectively remove or "unlearn" undesirable information, such as specific features or individual data points, from a learning outcome while minimizing utility loss and ensuring rigorous guarantees? We introduce a mathematical framework based on information-theoretic regularization to address both feature and data point unlearning. For feature unlearning, we derive a unified solution that simultaneously optimizes diverse learning objectives, including entropy, conditional entropy, KL-divergence, and the energy of conditional probability. For data point unlearning, we first propose a novel definition that serves as a practical condition for unlearning via retraining, is easy to verify, and aligns with the principles of differential privacy from an inference perspective. Then, we provide provable guarantees for our framework on data point unlearning. By combining flexibility in learning objectives with simplicity in regularization design, our approach is highly adaptable and practical for a wide range of machine learning and AI applications.
Authors: Naoki Saito, David Weber
Abstract: We propose a new method to extract discriminant and explainable features from a particular machine learning model, i.e., a combination of the scattering transform and the multiclass logistic regression. Although this model is well-known for its ability to learn various signal classes with high classification rate, it remains elusive to understand why it can generate such successful classification, mainly due to the nonlinearity of the scattering transform. In order to uncover the meaning of the scattering transform coefficients selected by the multiclass logistic regression (with the Lasso penalty), we adopt zeroth-order optimization algorithms to search an input pattern that maximizes the class probability of a class of interest given the learned model. In order to do so, it turns out that imposing sparsity and smoothness of input patterns is important. We demonstrate the effectiveness of our proposed method using a couple of synthetic time-series classification problems.
Authors: Xingrun Xing, Zheng Liu, Shitao Xiao, Boyan Gao, Yiming Liang, Wanpeng Zhang, Haokun Lin, Guoqi Li, Jiajun Zhang
Abstract: Modern large language models (LLMs) driven by scaling laws, achieve intelligence emergency in large model sizes. Recently, the increasing concerns about cloud costs, latency, and privacy make it an urgent requirement to develop compact edge language models. Distinguished from direct pretraining that bounded by the scaling law, this work proposes the pruning-aware pretraining, focusing on retaining performance of much larger optimized models. It features following characteristics: 1) Data-scalable: we introduce minimal parameter groups in LLM and continuously optimize structural pruning, extending post-training pruning methods like LLM-Pruner and SparseGPT into the pretraining phase. 2) Architecture-agnostic: the LLM architecture is auto-designed using saliency-driven pruning, which is the first time to exceed SoTA human-designed LLMs in modern pretraining. We reveal that it achieves top-quality edge language models, termed EfficientLLM, by scaling up LLM compression and extending its boundary. EfficientLLM significantly outperforms SoTA baselines with $100M \sim 1B$ parameters, such as MobileLLM, SmolLM, Qwen2.5-0.5B, OLMo-1B, Llama3.2-1B in common sense benchmarks. As the first attempt, EfficientLLM bridges the performance gap between traditional LLM compression and direct pretraining methods, and we will fully open source at https://github.com/Xingrun-Xing2/EfficientLLM.
Authors: Xu Zhang, Kaidi Xu, Ziqing Hu, Ren Wang
Abstract: Mixture of Experts (MoE) have shown remarkable success in leveraging specialized expert networks for complex machine learning tasks. However, their susceptibility to adversarial attacks presents a critical challenge for deployment in robust applications. This paper addresses the critical question of how to incorporate robustness into MoEs while maintaining high natural accuracy. We begin by analyzing the vulnerability of MoE components, finding that expert networks are notably more susceptible to adversarial attacks than the router. Based on this insight, we propose a targeted robust training technique that integrates a novel loss function to enhance the adversarial robustness of MoE, requiring only the robustification of one additional expert without compromising training or inference efficiency. Building on this, we introduce a dual-model strategy that linearly combines a standard MoE model with our robustified MoE model using a smoothing parameter. This approach allows for flexible control over the robustness-accuracy trade-off. We further provide theoretical foundations by deriving certified robustness bounds for both the single MoE and the dual-model. To push the boundaries of robustness and accuracy, we propose a novel joint training strategy JTDMoE for the dual-model. This joint training enhances both robustness and accuracy beyond what is achievable with separate models. Experimental results on CIFAR-10 and TinyImageNet datasets using ResNet18 and Vision Transformer (ViT) architectures demonstrate the effectiveness of our proposed methods.
Authors: Yun Gong, Zebang Shen, Niao He
Abstract: Potential functions in highly pertinent applications, such as deep learning in over-parameterized regime, are empirically observed to admit non-isolated minima. To understand the convergence behavior of stochastic dynamics in such landscapes, we propose to study the class of \logPLmeasure\ measures $\mu_\epsilon \propto \exp(-V/\epsilon)$, where the potential $V$ satisfies a local Polyak-{\L}ojasiewicz (P\L) inequality, and its set of local minima is provably \emph{connected}. Notably, potentials in this class can exhibit local maxima and we characterize its optimal set S to be a compact $\mathcal{C}^2$ \emph{embedding submanifold} of $\mathbb{R}^d$ without boundary. The \emph{non-contractibility} of S distinguishes our function class from the classical convex setting topologically. Moreover, the embedding structure induces a naturally defined Laplacian-Beltrami operator on S, and we show that its first non-trivial eigenvalue provides an \emph{$\epsilon$-independent} lower bound for the \Poincare\ constant in the \Poincare\ inequality of $\mu_\epsilon$. As a direct consequence, Langevin dynamics with such non-convex potential $V$ and diffusion coefficient $\epsilon$ converges to its equilibrium $\mu_\epsilon$ at a rate of $\tilde{\mathcal{O}}(1/\epsilon)$, provided $\epsilon$ is sufficiently small. Here $\tilde{\mathcal{O}}$ hides logarithmic terms.
Authors: Tai Hoang, Huy Le, Philipp Becker, Vien Anh Ngo, Gerhard Neumann
Abstract: Manipulating objects with varying geometries and deformable objects is a major challenge in robotics. Tasks such as insertion with different objects or cloth hanging require precise control and effective modelling of complex dynamics. In this work, we frame this problem through the lens of a heterogeneous graph that comprises smaller sub-graphs, such as actuators and objects, accompanied by different edge types describing their interactions. This graph representation serves as a unified structure for both rigid and deformable objects tasks, and can be extended further to tasks comprising multiple actuators. To evaluate this setup, we present a novel and challenging reinforcement learning benchmark, including rigid insertion of diverse objects, as well as rope and cloth manipulation with multiple end-effectors. These tasks present a large search space, as both the initial and target configurations are uniformly sampled in 3D space. To address this issue, we propose a novel graph-based policy model, dubbed Heterogeneous Equivariant Policy (HEPi), utilizing $SE(3)$ equivariant message passing networks as the main backbone to exploit the geometric symmetry. In addition, by modeling explicit heterogeneity, HEPi can outperform Transformer-based and non-heterogeneous equivariant policies in terms of average returns, sample efficiency, and generalization to unseen objects.
Authors: Cong Lu, Shengran Hu, Jeff Clune
Abstract: Foundation models have become general-purpose assistants, exhibiting diverse capabilities across numerous domains through training on web-scale data. It remains challenging to precisely characterize even a fraction of the full spectrum of capabilities and potential risks in any new model. Existing evaluation approaches often require significant human effort, and it is taking increasing effort to design ever harder challenges for more capable models. We introduce Automated Capability Discovery (ACD), a framework that designates one foundation model as a scientist to systematically propose open-ended tasks probing the abilities of a subject model (potentially itself). By combining frontier models with ideas from the field of open-endedness, ACD automatically and systematically uncovers both surprising capabilities and failures in the subject model. We demonstrate ACD across a range of foundation models (including the GPT, Claude, and Llama series), showing that it automatically reveals thousands of capabilities that would be challenging for any single team to uncover. We further validate our method's automated scoring with extensive human surveys, observing high agreement between model-generated and human evaluations. By leveraging foundation models' ability to both create tasks and self-evaluate, ACD is a significant step toward scalable, automated evaluation of novel AI systems. All code and evaluation logs are open-sourced at https://github.com/conglu1997/ACD.
Authors: Arvind Pillai, Dimitris Spathis, Subigya Nepal, Amanda C Collins, Daniel M Mackin, Michael V Heinz, Tess Z Griffin, Nicholas C Jacobson, Andrew Campbell
Abstract: Large language models (LLMs) show promise for health applications when combined with behavioral sensing data. Traditional approaches convert sensor data into text prompts, but this process is prone to errors, computationally expensive, and requires domain expertise. These challenges are particularly acute when processing extended time series data. While time series foundation models (TFMs) have recently emerged as powerful tools for learning representations from temporal data, bridging TFMs and LLMs remains challenging. Here, we present Time2Lang, a framework that directly maps TFM outputs to LLM representations without intermediate text conversion. Our approach first trains on synthetic data using periodicity prediction as a pretext task, followed by evaluation on mental health classification tasks. We validate Time2Lang on two longitudinal wearable and mobile sensing datasets: daily depression prediction using step count data (17,251 days from 256 participants) and flourishing classification based on conversation duration (46 participants over 10 weeks). Time2Lang maintains near constant inference times regardless of input length, unlike traditional prompting methods. The generated embeddings preserve essential time-series characteristics such as auto-correlation. Our results demonstrate that TFMs and LLMs can be effectively integrated while minimizing information loss and enabling performance transfer across these distinct modeling paradigms. To our knowledge, we are the first to integrate a TFM and an LLM for health, thus establishing a foundation for future research combining general-purpose large models for complex healthcare tasks.
Authors: Yulun Wu, Layne C. Price, Zichen Wang, Vassilis N. Ioannidis, Robert A. Barton, George Karypis
Abstract: Estimating an individual's potential outcomes under counterfactual treatments is a challenging task for traditional causal inference and supervised learning approaches when the outcome is high-dimensional (e.g. gene expressions, impulse responses, human faces) and covariates are relatively limited. In this case, to construct one's outcome under a counterfactual treatment, it is crucial to leverage individual information contained in its observed factual outcome on top of the covariates. We propose a deep variational Bayesian framework that rigorously integrates two main sources of information for outcome construction under a counterfactual treatment: one source is the individual features embedded in the high-dimensional factual outcome; the other source is the response distribution of similar subjects (subjects with the same covariates) that factually received this treatment of interest.
Authors: Zahra Ramezani, Kenan \v{S}ehi\'c, Luigi Nardi, Knut {\AA}kesson
Abstract: Cyber-physical systems (CPSs) are often complex and safety-critical, making it both challenging and crucial to ensure that the system's specifications are met. Simulation-based falsification is a practical testing technique for increasing confidence in a CPS's correctness, as it only requires that the system be simulated. Reducing the number of computationally intensive simulations needed for falsification is a key concern. In this study, we investigate Bayesian optimization (BO), a sample-efficient approach that learns a surrogate model to capture the relationship between input signal parameterization and specification evaluation. We propose two enhancements to the basic BO for improving falsification: (1) leveraging local surrogate models, and (2) utilizing the user's prior knowledge. Additionally, we address the formulation of acquisition functions for falsification by proposing and evaluating various alternatives. Our benchmark evaluation demonstrates significant improvements when using local surrogate models in BO for falsifying challenging benchmark examples. Incorporating prior knowledge is found to be especially beneficial when the simulation budget is constrained. For some benchmark problems, the choice of acquisition function noticeably impacts the number of simulations required for successful falsification.
Authors: Matthieu Cordonnier, Nicolas Keriven, Nicolas Tremblay, Samuel Vaiter
Abstract: We study the convergence of message passing graph neural networks on random graph models to their continuous counterpart as the number of nodes tends to infinity. Until now, this convergence was only known for architectures with aggregation functions in the form of normalized means, or, equivalently, of an application of classical operators like the adjacency matrix or the graph Laplacian. We extend such results to a large class of aggregation functions, that encompasses all classically used message passing graph neural networks, such as attention-based message passing, max convolutional message passing, (degree-normalized) convolutional message passing, or moment-based aggregation message passing. Under mild assumptions, we give non-asymptotic bounds with high probability to quantify this convergence. Our main result is based on the McDiarmid inequality. Interestingly, this result does not apply to the case where the aggregation is a coordinate-wise maximum. We treat this case separately and obtain a different convergence rate.
Authors: Ali Ayub, Zachary De Francesco, Patrick Holthaus, Chrystopher L. Nehaniv, Kerstin Dautenhahn
Abstract: For long-term deployment in dynamic real-world environments, assistive robots must continue to learn and adapt to their environments. Researchers have developed various computational models for continual learning (CL) that can allow robots to continually learn from limited training data, and avoid forgetting previous knowledge. While these CL models can mitigate forgetting on static, systematically collected datasets, it is unclear how human users might perceive a robot that continually learns over multiple interactions with them. In this paper, we developed a system that integrates CL models for object recognition with a Fetch mobile manipulator robot and allows human participants to directly teach and test the robot over multiple sessions. We conducted an in-person study with 60 participants that interacted with our system in 300 sessions (5 sessions per participant). We conducted a between-subject study with three different CL models to understand human perceptions of continual learning robots over multiple sessions. Our results suggest that participants' perceptions of trust, competence, and usability of a continual learning robot significantly decrease over multiple sessions if the robot forgets previously learned objects. However, the perceived task load on participants for teaching and testing the robot remains the same over multiple sessions even if the robot forgets previously learned objects. Our results also indicate that state-of-the-art CL models might perform unreliably when applied on robots interacting with human participants. Further, continual learning robots are not perceived as very trustworthy or competent by human participants, regardless of the underlying continual learning model or the session number.
Authors: Da-Wei Zhou, Yuanhan Zhang, Yan Wang, Jingyi Ning, Han-Jia Ye, De-Chuan Zhan, Ziwei Liu
Abstract: Class-Incremental Learning (CIL) or continual learning is a desired capability in the real world, which requires a learning system to adapt to new tasks without forgetting former ones. While traditional CIL methods focus on visual information to grasp core features, recent advances in Vision-Language Models (VLM) have shown promising capabilities in learning generalizable representations with the aid of textual information. However, when continually trained with new classes, VLMs often suffer from catastrophic forgetting of former knowledge. Applying VLMs to CIL poses two major challenges: 1) how to adapt the model without forgetting; and 2) how to make full use of the multi-modal information. To this end, we propose PROjectiOn Fusion (PROOF) that enables VLMs to learn without forgetting. To handle the first challenge, we propose training task-specific projections based on the frozen image/text encoders. When facing new tasks, new projections are expanded and former projections are fixed, alleviating the forgetting of old concepts. For the second challenge, we propose the fusion module to better utilize the cross-modality information. By jointly adjusting visual and textual features, the model can capture semantic information with stronger representation ability. Extensive experiments on nine benchmark datasets validate PROOF achieves state-of-the-art performance. Code is available at https://github.com/zhoudw-zdw/PROOF
Authors: Yuzhou Gu, Zhao Song, Lichen Zhang
Abstract: Quadratic programming is a ubiquitous prototype in convex programming. Many machine learning problems can be formulated as quadratic programming, including the famous Support Vector Machines (SVMs). Linear and kernel SVMs have been among the most popular models in machine learning over the past three decades, prior to the deep learning era. Generally, a quadratic program has an input size of $\Theta(n^2)$, where $n$ is the number of variables. Assuming the Strong Exponential Time Hypothesis ($\textsf{SETH}$), it is known that no $O(n^{2-o(1)})$ time algorithm exists when the quadratic objective matrix is positive semidefinite (Backurs, Indyk, and Schmidt, NeurIPS'17). However, problems such as SVMs usually admit much smaller input sizes: one is given $n$ data points, each of dimension $d$, and $d$ is oftentimes much smaller than $n$. Furthermore, the SVM program has only $O(1)$ equality linear constraints. This suggests that faster algorithms are feasible, provided the program exhibits certain structures. In this work, we design the first nearly-linear time algorithm for solving quadratic programs whenever the quadratic objective admits a low-rank factorization, and the number of linear constraints is small. Consequently, we obtain results for SVMs: * For linear SVM when the input data is $d$-dimensional, our algorithm runs in time $\widetilde O(nd^{(\omega+1)/2}\log(1/\epsilon))$ where $\omega\approx 2.37$ is the fast matrix multiplication exponent; * For Gaussian kernel SVM, when the data dimension $d = {\color{black}O(\log n)}$ and the squared dataset radius is sub-logarithmic in $n$, our algorithm runs in time $O(n^{1+o(1)}\log(1/\epsilon))$. We also prove that when the squared dataset radius is at least $\Omega(\log^2 n)$, then $\Omega(n^{2-o(1)})$ time is required. This improves upon the prior best lower bound in both the dimension $d$ and the squared dataset radius.
Authors: Ryan Sweke, Erik Recio-Armengol, Sofiene Jerbi, Elies Gil-Fuster, Bryce Fuller, Jens Eisert, Johannes Jakob Meyer
Abstract: Quantum machine learning is arguably one of the most explored applications of near-term quantum devices. Much focus has been put on notions of variational quantum machine learning where parameterized quantum circuits (PQCs) are used as learning models. These PQC models have a rich structure which suggests that they might be amenable to efficient dequantization via random Fourier features (RFF). In this work, we establish necessary and sufficient conditions under which RFF does indeed provide an efficient dequantization of variational quantum machine learning for regression. We build on these insights to make concrete suggestions for PQC architecture design, and to identify structures which are necessary for a regression problem to admit a potential quantum advantage via PQC based optimization.
Authors: Aashish Gottipati, Sami Khairy, Gabriel Mittag, Vishak Gopal, Ross Cutler
Abstract: Real-time video applications require accurate bandwidth estimation (BWE) to maintain user experience across varying network conditions. However, increasing network heterogeneity challenges general-purpose BWE algorithms, necessitating solutions that adapt to end-user environments. While widely adopted, heuristic-based methods are difficult to individualize without extensive domain expertise. Conversely, online reinforcement learning (RL) offers ease of customization but neglects prior domain expertise and suffers from sample inefficiency. Thus, we present Merlin, an imitation learning-based solution that replaces the manual parameter tuning of heuristic-based methods with data-driven updates to streamline end-user personalization. Our key insight is that transforming heuristic-based BWE algorithms into neural networks facilitates data-driven personalization. Merlin utilizes Behavioral Cloning to efficiently learn from offline telemetry logs, capturing heuristic policies without live network interactions. The cloned policy can then be seamlessly tailored to end user network conditions through online finetuning. In real intercontinental videoconferencing calls, Merlin matches our heuristic's policy with no statistically significant differences in user quality of experience (QoE). Finetuning Merlin's control policy to end-user environments enables QoE improvements of up to 7.8% compared to the heuristic policy. Lastly, our IL-based design performs competitively with current state-of-the-art online RL techniques but converges with 80% fewer videoconferencing samples, facilitating practical end-user personalization.
Authors: Wei-Fu Tseng, Kai-Chun Chen, Zi-Hong Xiao, Yen-Huan Li
Abstract: Online learning of quantum states with the logarithmic loss (LL-OLQS) is a quantum generalization of online portfolio selection (OPS), a classic open problem in online learning for over three decades. This problem also emerges in designing stochastic optimization algorithms for maximum-likelihood quantum state tomography. Recently, Jezequel et al. (arXiv:2209.13932) proposed the VB-FTRL algorithm, the first regret-optimal algorithm for OPS with moderate computational complexity. In this paper, we generalize VB-FTRL for LL-OLQS. Let $d$ denote the dimension and $T$ the number of rounds. The generalized algorithm achieves a regret rate of $O ( d^2 \log ( d + T ) )$ for LL-OLQS. Each iteration of the algorithm consists of solving a semidefinite program that can be implemented in polynomial time by, for example, cutting-plane methods. For comparison, the best-known regret rate for LL-OLQS is currently $O ( d^2 \log T )$, achieved by an exponential weight method. However, no explicit implementation is available for the exponential weight method for LL-OLQS. To facilitate the generalization, we introduce the notion of VB-convexity. VB-convexity is a sufficient condition for the volumetric barrier associated with any function to be convex and is of independent interest.
Authors: Chenyu Liu, Xinliang Zhou, Yihao Wu, Ruizhi Yang, Zhongruo Wang, Liming Zhai, Ziyu Jia, Yang Liu
Abstract: Compared to other modalities, EEG-based emotion recognition can intuitively respond to the emotional patterns in the human brain and, therefore, has become one of the most concerning tasks in the brain-computer interfaces field. Since dependencies within brain regions are closely related to emotion, a significant trend is to develop Graph Neural Networks (GNNs) for EEG-based emotion recognition. However, brain region dependencies in emotional EEG have physiological bases that distinguish GNNs in this field from those in other time series fields. Besides, there is neither a comprehensive review nor guidance for constructing GNNs in EEG-based emotion recognition. In the survey, our categorization reveals the commonalities and differences of existing approaches under a unified framework of graph construction. We analyze and categorize methods from three stages in the framework to provide clear guidance on constructing GNNs in EEG-based emotion recognition. In addition, we discuss several open challenges and future directions, such as Temporal full-connected graph and Graph condensation.
Authors: Ashish Hooda, Mihai Christodorescu, Miltiadis Allamanis, Aaron Wilson, Kassem Fawaz, Somesh Jha
Abstract: Large Language Models' success on text generation has also made them better at code generation and coding tasks. While a lot of work has demonstrated their remarkable performance on tasks such as code completion and editing, it is still unclear as to why. We help bridge this gap by exploring to what degree auto-regressive models understand the logical constructs of the underlying programs. We propose Counterfactual Analysis for Programming Concept Predicates (CACP) as a counterfactual testing framework to evaluate whether Large Code Models understand programming concepts. With only black-box access to the model, we use CACP to evaluate ten popular Large Code Models for four different programming concepts. Our findings suggest that current models lack understanding of concepts such as data flow and control flow.
Authors: Shijing Si, Yuwei Wu, Le Tang, Yugui Zhang, Jedrek Wosik, Qinliang Su
Abstract: Email continues to be a pivotal and extensively utilized communication medium within professional and commercial domains. Nonetheless, the prevalence of spam emails poses a significant challenge for users, disrupting their daily routines and diminishing productivity. Consequently, accurately identifying and filtering spam based on content has become crucial for cybersecurity. Recent advancements in natural language processing, particularly with large language models like ChatGPT, have shown remarkable performance in tasks such as question answering and text generation. However, its potential in spam identification remains underexplored. To fill in the gap, this study attempts to evaluate ChatGPT's capabilities for spam identification in both English and Chinese email datasets. We employ ChatGPT for spam email detection using in-context learning, which requires a prompt instruction with (or without) a few demonstrations. We also investigate how the number of demonstrations in the prompt affects the performance of ChatGPT. For comparison, we also implement five popular benchmark methods, including naive Bayes, support vector machines (SVM), logistic regression (LR), feedforward dense neural networks (DNN), and BERT classifiers. Through extensive experiments, the performance of ChatGPT is significantly worse than deep supervised learning methods in the large English dataset, while it presents superior performance on the low-resourced Chinese dataset. This study provides insights into the potential and limitations of ChatGPT for spam identification, highlighting its potential as a viable solution for resource-constrained language domains.
Authors: Jingyang Zhang, Jingwei Sun, Eric Yeats, Yang Ouyang, Martin Kuo, Jianyi Zhang, Hao Frank 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. Despite improved performance, existing methods (including the state-of-the-art, Min-K%) are mostly developed upon simple heuristics and lack solid, reasonable foundations. In this work, we propose a novel and theoretically motivated methodology for pre-training data detection, named Min-K%++. Specifically, we present a key insight that training samples tend to be local maxima of the modeled distribution along each input dimension through maximum likelihood training, which in turn allow us to insightfully translate the problem into identification of local maxima. Then, we design our method accordingly that works under the discrete distribution modeled by LLMs, whose core idea is to determine whether the input forms a mode or has relatively high probability under the conditional categorical distribution. Empirically, the proposed method achieves new SOTA performance across multiple settings. On the WikiMIA benchmark, Min-K%++ outperforms the runner-up by 6.2% to 10.5% in detection AUROC averaged over five models. On the more challenging MIMIR benchmark, it consistently improves upon reference-free methods while performing on par with reference-based method that requires an extra reference model.
Authors: Yuchi Liu, Lei Wang, Yuli Zou, James Zou, Liang Zheng
Abstract: Model calibration aims to align confidence with prediction correctness. The Cross-Entropy (CE) loss is widely used for calibrator training, which enforces the model to increase confidence on the ground truth class. However, we find the CE loss has intrinsic limitations. For example, for a narrow misclassification (e.g., a test sample is wrongly classified and its softmax score on the ground truth class is 0.4), a calibrator trained by the CE loss often produces high confidence on the wrongly predicted class, which is undesirable. In this paper, we propose a new post-hoc calibration objective derived from the aim of calibration. Intuitively, the proposed objective function asks that the calibrator decrease model confidence on wrongly predicted samples and increase confidence on correctly predicted samples. Because a sample itself has insufficient ability to indicate correctness, we use its transformed versions (e.g., rotated, greyscaled, and color-jittered) during calibrator training. Trained on an in-distribution validation set and tested with isolated, individual test samples, our method achieves competitive calibration performance on both in-distribution and out-of-distribution test sets compared with the state of the art. Further, our analysis points out the difference between our method and commonly used objectives such as CE loss and Mean Square Error (MSE) loss, where the latters sometimes deviates from the calibration aim.
Authors: Denise Moussa, Germans Hirsch, Christian Riess
Abstract: Audio recordings may provide important evidence in criminal investigations. One such case is the forensic association of a recorded audio to its recording location. For example, a voice message may be the only investigative cue to narrow down the candidate sites for a crime. Up to now, several works provide supervised classification tools for closed-set recording environment identification under relatively clean recording conditions. However, in forensic investigations, the candidate locations are case-specific. Thus, supervised learning techniques are not applicable without retraining a classifier on a sufficient amount of training samples for each case and respective candidate set. In addition, a forensic tool has to deal with audio material from uncontrolled sources with variable properties and quality. In this work, we therefore attempt a major step towards practical forensic application scenarios. We propose a representation learning framework called EnvId, short for environment identification. EnvId avoids case-specific retraining by modeling the task as a few-shot classification problem. We demonstrate that EnvId can handle forensically challenging material. It provides good quality predictions even under unseen signal degradations, out-of-distribution reverberation characteristics or recording position mismatches.
Authors: Camilo Ram\'irez, Jorge F. Silva, Ferhat Tamssaouet, Tom\'as Rojas, Marcos E. Orchard
Abstract: The ability to detect when a system undergoes an incipient fault is of paramount importance in preventing a critical failure. Classic methods for fault detection (including model-based and data-driven approaches) rely on thresholding error statistics or simple input-residual dependencies but face difficulties with non-linear or non-Gaussian systems. Behavioral methods (e.g., those relying on digital twins) address these difficulties but still face challenges when faulty data is scarce, decision guarantees are required, or working with already-deployed models is required. In this work, we propose an information-driven fault detection method based on a novel concept drift detector, addressing these challenges. The method is tailored to identifying drifts in input-output relationships of additive noise models (i.e., model drifts) and is based on a distribution-free mutual information (MI) estimator. Our scheme does not require prior faulty examples and can be applied distribution-free over a large class of system models. Our core contributions are twofold. First, we demonstrate the connection between fault detection, model drift detection, and testing independence between two random variables. Second, we prove several theoretical properties of the proposed MI-based fault detection scheme: (i) strong consistency, (ii) exponentially fast detection of the non-faulty case, and (iii) control of both significance levels and power of the test. To conclude, we validate our theory with synthetic data and the benchmark dataset N-CMAPSS of aircraft turbofan engines. These empirical results support the usefulness of our methodology in many practical and realistic settings, and the theoretical results show performance guarantees that other methods cannot offer.
Authors: Rahul Parhi, Pakshal Bohra, Ayoub El Biari, Mehrsa Pourya, Michael Unser
Abstract: We consider a large class of shallow neural networks with randomly initialized parameters and rectified linear unit activation functions. We prove that these random neural networks are well-defined non-Gaussian processes. As a by-product, we demonstrate that these networks are solutions to stochastic differential equations driven by impulsive white noise (combinations of random Dirac measures). These processes are parameterized by the law of the weights and biases as well as the density of activation thresholds in each bounded region of the input domain. We prove that these processes are isotropic and wide-sense self-similar with Hurst exponent 3/2. We also derive a remarkably simple closed-form expression for their autocovariance function. Our results are fundamentally different from prior work in that we consider a non-asymptotic viewpoint: The number of neurons in each bounded region of the input domain (i.e., the width) is itself a random variable with a Poisson law with mean proportional to the density parameter. Finally, we show that, under suitable hypotheses, as the expected width tends to infinity, these processes can converge in law not only to Gaussian processes, but also to non-Gaussian processes depending on the law of the weights. Our asymptotic results provide a new take on several classical results (wide networks converge to Gaussian processes) as well as some new ones (wide networks can converge to non-Gaussian processes).
Authors: Patrick Chao, Yan Sun, Edgar Dobriban, Hamed Hassani
Abstract: Recent progress in large language models enables the creation of realistic machine-generated content. Watermarking is a promising approach to distinguish machine-generated text from human text, embedding statistical signals in the output that are ideally undetectable to humans. We propose a watermarking framework that encodes such signals through an error correcting code. Our method, termed robust binary code (RBC) watermark, introduces no distortion compared to the original probability distribution, and no noticeable degradation in quality. We evaluate our watermark on base and instruction fine-tuned models and find our watermark is robust to edits, deletions, and translations. We provide an information-theoretic perspective on watermarking, a powerful statistical test for detection and for generating p-values, and theoretical guarantees. Our empirical findings suggest our watermark is fast, powerful, and robust, comparing favorably to the state-of-the-art.
Authors: Sungjun Lim, Jeyoon Yeom, Sooyon Kim, Hoyoon Byun, Jinho Kang, Yohan Jung, Jiyoung Jung, Kyungwoo Song
Abstract: Bayesian neural networks (BNNs) estimate the posterior distribution of model parameters and utilize posterior samples for Bayesian Model Aver- aging (BMA) in prediction. However, despite the crucial role of flatness in the loss landscape in improving the generalization of neural networks, its impact on BMA has been largely overlooked. In this work, we explore how posterior flatness influences BMA generalization and empirically demonstrate that (1) most approximate Bayesian inference methods fail to yield a flat posterior and (2) BMA predictions, without considering posterior flatness, are less effective at improving generalization. To address this, we propose Flat Posterior-aware Bayesian Model Averaging (FP-BMA), a novel training objective that explicitly encourages flat posteriors in a principled Bayesian manner. We also introduce a Flat Posterior-aware Bayesian Transfer Learning scheme that enhances generalization in downstream tasks. Empirically, we show that FP-BMA successfully captures flat posteriors, improving generalization performance.
Authors: Anton Xue, Avishree Khare, Rajeev Alur, Surbhi Goel, Eric Wong
Abstract: We study how to subvert large language models (LLMs) from following prompt-specified rules. We first formalize rule-following as inference in propositional Horn logic, a mathematical system in which rules have the form "if $P$ and $Q$, then $R$" for some propositions $P$, $Q$, and $R$. Next, we prove that although small transformers can faithfully follow such rules, maliciously crafted prompts can still mislead both theoretical constructions and models learned from data. Furthermore, we demonstrate that popular attack algorithms on LLMs find adversarial prompts and induce attention patterns that align with our theory. Our novel logic-based framework provides a foundation for studying LLMs in rule-based settings, enabling a formal analysis of tasks like logical reasoning and jailbreak attacks.
Authors: Dan Friedman, Abhishek Panigrahi, Danqi Chen
Abstract: What kind of internal mechanisms might Transformers use to conduct fluid, natural-sounding conversations? Prior work has illustrated by construction how Transformers can solve various synthetic tasks, such as sorting a list or recognizing formal languages, but it remains unclear how to extend this approach to a conversational setting. In this work, we propose using ELIZA, a classic rule-based chatbot, as a setting for formal, mechanistic analysis of Transformer-based chatbots. ELIZA allows us to formally model key aspects of conversation, including local pattern matching and long-term dialogue state tracking. We first present a theoretical construction of a Transformer that implements the ELIZA chatbot. Building on prior constructions, particularly those for simulating finite-state automata, we show how simpler mechanisms can be composed and extended to produce more sophisticated behavior. Next, we conduct a set of empirical analyses of Transformers trained on synthetically generated ELIZA conversations. Our analysis illustrates the kinds of mechanisms these models tend to prefer--for example, models favor an induction head mechanism over a more precise, position-based copying mechanism; and using intermediate generations to simulate recurrent data structures, akin to an implicit scratchpad or Chain-of-Thought. Overall, by drawing an explicit connection between neural chatbots and interpretable, symbolic mechanisms, our results provide a new framework for the mechanistic analysis of conversational agents.
Authors: Siqi Lu, Junlin Guo, James R Zimmer-Dauphinee, Jordan M Nieusma, Xiao Wang, Parker VanValkenburgh, Steven A Wernke, Yuankai Huo
Abstract: Artificial Intelligence (AI) technologies have profoundly transformed the field of remote sensing, revolutionizing data collection, processing, and analysis. Traditionally reliant on manual interpretation and task-specific models, remote sensing research has been significantly enhanced by the advent of foundation models-large-scale, pre-trained AI models capable of performing a wide array of tasks with unprecedented accuracy and efficiency. This paper provides a comprehensive survey of foundation models in the remote sensing domain. We categorize these models based on their architectures, pre-training datasets, and methodologies. Through detailed performance comparisons, we highlight emerging trends and the significant advancements achieved by those foundation models. Additionally, we discuss technical challenges, practical implications, and future research directions, addressing the need for high-quality data, computational resources, and improved model generalization. Our research also finds that pre-training methods, particularly self-supervised learning techniques like contrastive learning and masked autoencoders, remarkably enhance the performance and robustness of foundation models. This survey aims to serve as a resource for researchers and practitioners by providing a panorama of advances and promising pathways for continued development and application of foundation models in remote sensing.
Authors: Frank Qiu, Joshua Michalenko, Lilian K. Casias, Cameron J. Radosevich, Jon Slater, Eric A. Shields
Abstract: Point-source transient events (PSTEs) - optical events that are both extremely fast and extremely small - pose several challenges to an imaging system. Due to their speed, accurately characterizing such events often requires detectors with very high frame rates. Due to their size, accurately detecting such events requires maintaining coverage over an extended field-of-view, often through the use of imaging focal plane arrays (FPA) with a global shutter readout. Traditional imaging systems that meet these requirements are costly in terms of price, size, weight, power consumption, and data bandwidth, and there is a need for cheaper solutions with adequate temporal and spatial coverage. To address these issues, we develop a novel compressed sensing algorithm adapted to the rolling shutter readout of an imaging system. This approach enables reconstruction of a PSTE signature at the sampling rate of the rolling shutter, offering a 1-2 order of magnitude temporal speedup and a proportional reduction in data bandwidth. We present empirical results demonstrating accurate recovery of PSTEs using measurements that are spatially undersampled by a factor of 25, and our simulations show that, relative to other compressed sensing algorithms, our algorithm is both faster and yields higher quality reconstructions. We also present theoretical results characterizing our algorithm and corroborating simulations. The potential impact of our work includes the development of much faster, cheaper sensor solutions for PSTE detection and characterization.
Authors: Zeyi Liao, Lingbo Mo, Chejian Xu, Mintong Kang, Jiawei Zhang, Chaowei Xiao, Yuan Tian, Bo Li, Huan Sun
Abstract: Generalist web agents have demonstrated remarkable potential in autonomously completing a wide range of tasks on real websites, significantly boosting human productivity. However, web tasks, such as booking flights, usually involve users' PII, which may be exposed to potential privacy risks if web agents accidentally interact with compromised websites, a scenario that remains largely unexplored in the literature. In this work, we narrow this gap by conducting the first study on the privacy risks of generalist web agents in adversarial environments. First, we present a realistic threat model for attacks on the website, where we consider two adversarial targets: stealing users' specific PII or the entire user request. Then, we propose a novel attack method, termed Environmental Injection Attack (EIA). EIA injects malicious content designed to adapt well to environments where the agents operate and our work instantiates EIA specifically for privacy scenarios in web environments. We collect 177 action steps that involve diverse PII categories on realistic websites from the Mind2Web, and conduct experiments using one of the most capable generalist web agent frameworks to date. The results demonstrate that EIA achieves up to 70% ASR in stealing specific PII and 16% ASR for full user request. Additionally, by accessing the stealthiness and experimenting with a defensive system prompt, we indicate that EIA is hard to detect and mitigate. Notably, attacks that are not well adapted for a webpage can be detected via human inspection, leading to our discussion about the trade-off between security and autonomy. However, extra attackers' efforts can make EIA seamlessly adapted, rendering such supervision ineffective. Thus, we further discuss the defenses at the pre- and post-deployment stages of the websites without relying on human supervision and call for more advanced defense strategies.
Authors: Jun Yuan, Guohao Cai, Zhenhua Dong
Abstract: Multi-task ranking models have become essential for modern real-world recommendation systems. While most recommendation researches focus on designing sophisticated models for specific scenarios, achieving performance improvement for multi-task ranking models across various scenarios still remains a significant challenge. Training all tasks naively can result in inconsistent learning, highlighting the need for the development of multi-task optimization (MTO) methods to tackle this challenge. Conventional methods assume that the optimal joint gradient on shared parameters leads to optimal parameter updates. However, the actual update on model parameters may deviates significantly from gradients when using momentum based optimizers such as Adam, and we design and execute statistical experiments to support the observation. In this paper, we propose a novel Parameter Update Balancing algorithm for multi-task optimization, denoted as PUB. In contrast to traditional MTO method which are based on gradient level tasks fusion or loss level tasks fusion, PUB is the first work to optimize multiple tasks through parameter update balancing. Comprehensive experiments on benchmark multi-task ranking datasets demonstrate that PUB consistently improves several multi-task backbones and achieves state-of-the-art performance. Additionally, experiments on benchmark computer vision datasets show the great potential of PUB in various multi-task learning scenarios. Furthermore, we deployed our method for an industrial evaluation on the real-world commercial platform, HUAWEI AppGallery, where PUB significantly enhances the online multi-task ranking model, efficiently managing the primary traffic of a crucial channel.
Authors: Elizabeth Teng, Ugur Demir, Zoheyr Doctor, Philipp M. Srivastava, Shamal Lalvani, Vicky Kalogera, Aggelos Katsaggelos, Jeff J. Andrews, Simone S. Bavera, Max M. Briel, Seth Gossage, Konstantinos Kovlakas, Matthias U. Kruckow, Kyle Akira Rocha, Meng Sun, Zepei Xing, Emmanouil Zapartas
Abstract: Knowledge about the internal physical structure of stars is crucial to understanding their evolution. The novel binary population synthesis code POSYDON includes a module for interpolating the stellar and binary properties of any system at the end of binary MESA evolution based on a pre-computed set of models. In this work, we present a new emulation method for predicting stellar profiles, i.e., the internal stellar structure along the radial axis, using machine learning techniques. We use principal component analysis for dimensionality reduction and fully-connected feed-forward neural networks for making predictions. We find accuracy to be comparable to that of nearest neighbor approximation, with a strong advantage in terms of memory and storage efficiency. By providing a versatile framework for modeling stellar internal structure, the emulation method presented here will enable faster simulations of higher physical fidelity, offering a foundation for a wide range of large-scale population studies of stellar and binary evolution.
Authors: Hongkuan Zhou, Lavdim Halilaj, Sebastian Monka, Stefan Schmid, Yuqicheng Zhu, Bo Xiong, Steffen Staab
Abstract: Despite the remarkable success of deep neural networks (DNNs) in computer vision, they fail to remain high-performing when facing distribution shifts between training and testing data. In this paper, we propose Knowledge-Guided Visual representation learning (KGV) - a distribution-based learning approach leveraging multi-modal prior knowledge - to improve generalization under distribution shift. It integrates knowledge from two distinct modalities: 1) a knowledge graph (KG) with hierarchical and association relationships; and 2) generated synthetic images of visual elements semantically represented in the KG. The respective embeddings are generated from the given modalities in a common latent space, i.e., visual embeddings from original and synthetic images as well as knowledge graph embeddings (KGEs). These embeddings are aligned via a novel variant of translation-based KGE methods, where the node and relation embeddings of the KG are modeled as Gaussian distributions and translations, respectively. We claim that incorporating multi-model prior knowledge enables more regularized learning of image representations. Thus, the models are able to better generalize across different data distributions. We evaluate KGV on different image classification tasks with major or minor distribution shifts, namely road sign classification across datasets from Germany, China, and Russia, image classification with the mini-ImageNet dataset and its variants, as well as the DVM-CAR dataset. The results demonstrate that KGV consistently exhibits higher accuracy and data efficiency across all experiments.
Authors: Yunhan Zhao, Xiang Zheng, Lin Luo, Yige Li, Xingjun Ma, Yu-Gang Jiang
Abstract: In this paper, we focus on black-box defense for VLMs against jailbreak attacks. Existing black-box defense methods are either unimodal or bimodal. Unimodal methods enhance either the vision or language module of the VLM, while bimodal methods robustify the model through text-image representation realignment. However, these methods suffer from two limitations: 1) they fail to fully exploit the cross-modal information, or 2) they degrade the model performance on benign inputs. To address these limitations, we propose a novel blue-team method BlueSuffix that defends target VLMs against jailbreak attacks without compromising its performance under black-box setting. BlueSuffix includes three key components: 1) a visual purifier against jailbreak images, 2) a textual purifier against jailbreak texts, and 3) a blue-team suffix generator using reinforcement fine-tuning for enhancing cross-modal robustness. We empirically show on four VLMs (LLaVA, MiniGPT-4, InstructionBLIP, and Gemini) and four safety benchmarks (Harmful Instruction, AdvBench, MM-SafetyBench, and RedTeam-2K) that BlueSuffix outperforms the baseline defenses by a significant margin. Our BlueSuffix opens up a promising direction for defending VLMs against jailbreak attacks. Code is available at https://github.com/Vinsonzyh/BlueSuffix.
Authors: Seong-Hyun Hong, Hyun-Sung Kim, Zian Jang, Deunsol Yoon, Hyungseok Song, Byung-Jun Lee
Abstract: Recent advancements in neural combinatorial optimization (NCO) methods have shown promising results in generating near-optimal solutions without the need for expert-crafted heuristics. However, high performance of these approaches often rely on problem-specific human-expertise-based search after generating candidate solutions, limiting their applicability to commonly solved CO problems such as Traveling Salesman Problem (TSP). In this paper, we present IC/DC, an unsupervised CO framework that directly trains a diffusion model from scratch. We train our model in a self-supervised way to minimize the cost of the solution while adhering to the problem-specific constraints. IC/DC is specialized in addressing CO problems involving two distinct sets of items, and it does not need problem-specific search processes to generate valid solutions. IC/DC employs a novel architecture capable of capturing the intricate relationships between items, and thereby enabling effective optimization in challenging CO scenarios. IC/DC achieves state-of-the-art performance relative to existing NCO methods on the Parallel Machine Scheduling Problem (PMSP) and Asymmetric Traveling Salesman Problem (ATSP).
Authors: Guixian Xu, Jinglai Li, Junqi Tang
Abstract: Equivariant Imaging (EI) regularization has become the de-facto technique for unsupervised training of deep imaging networks, without any need of ground-truth data. Observing that the EI-based unsupervised training paradigm currently has significant computational redundancy leading to inefficiency in high-dimensional applications, we propose a sketched EI regularization which leverages the randomized sketching techniques for acceleration. We then extend our sketched EI regularization to develop an accelerated deep internal learning framework, Sketched Equivariant Deep Image Prior (Sk-EI-DIP), which can be efficiently applied for single-image and task-adapted reconstruction. Additionally, for network adaptation tasks, we propose a parameter-efficient approach for accelerating both EI-DIP and Sk-EI-DIP via optimizing only the normalization layers. Our numerical study on X-ray CT and multi-coil MRI image reconstruction tasks demonstrate that our approach can achieve significant computational acceleration over standard EI-based counterpart in single-input setting and network adaptation at test time.
Authors: Egor Sevriugov, Ivan Oseledets
Abstract: Non-autoregressive language models generate all tokens simultaneously, offering potential speed advantages over traditional autoregressive models, but they face challenges in modeling the complex dependencies inherent in text data. In this work, we investigate a conditional flow matching approach for text generation. We represent tokens as one-hot vectors in a \(V\)-dimensional simplex and utilize geodesics under the Kullback-Leibler (KL) divergence, which correspond to linear interpolation in logit space. We provide a theoretical justification that maximizing the conditional likelihood \(P_{\theta}(x_1 \mid x_t, t)\) yields the exact flow matching velocity under logit interpolation. To address the suboptimal performance of basic inference, we propose a novel empirical sampling scheme that iteratively samples from the conditional distribution and introduces additional noise, significantly improving results despite lacking full theoretical underpinnings. Furthermore, we propose a hybrid inference method that combines the basic approach with the sampling scheme. This method demonstrates superior performance on both conditional and unconditional text generation experiments compared to previous SOTA method for discrete flow matching.
Authors: Yi Tang, Peng Sun, Zhenglin Cheng, Tao Lin
Abstract: Recent studies indicate that the denoising process in deep generative diffusion models implicitly learns and memorizes semantic information from the data distribution. These findings suggest that capturing more complex data distributions requires larger neural networks, leading to a substantial increase in computational demands, which in turn become the primary bottleneck in both training and inference of diffusion models. To this end, we introduce GMem: A Modular Approach for Ultra-Efficient Generative Models. Our approach GMem decouples the memory capacity from model and implements it as a separate, immutable memory set that preserves the essential semantic information in the data. The results are significant: GMem enhances both training, sampling efficiency, and diversity generation. This design on one hand reduces the reliance on network for memorize complex data distribution and thus enhancing both training and sampling efficiency. On ImageNet at $256 \times 256$ resolution, GMem achieves a $50\times$ training speedup compared to SiT, reaching FID $=7.66$ in fewer than $28$ epochs ($\sim 4$ hours training time), while SiT requires $1400$ epochs. Without classifier-free guidance, GMem achieves state-of-the-art (SoTA) performance FID $=1.53$ in $160$ epochs with only $\sim 20$ hours of training, outperforming LightningDiT which requires $800$ epochs and $\sim 95$ hours to attain FID $=2.17$.
Authors: Giordano Cicchetti, Eleonora Grassucci, Luigi Sigillo, Danilo Comminiello
Abstract: Human perception integrates multiple modalities, such as vision, hearing, and language, into a unified understanding of the surrounding reality. While recent multimodal models have achieved significant progress by aligning pairs of modalities via contrastive learning, their solutions are unsuitable when scaling to multiple modalities. These models typically align each modality to a designated anchor without ensuring the alignment of all modalities with each other, leading to suboptimal performance in tasks requiring a joint understanding of multiple modalities. In this paper, we structurally rethink the pairwise conventional approach to multimodal learning and we present the novel Gramian Representation Alignment Measure (GRAM), which overcomes the above-mentioned limitations. GRAM learns and then aligns $n$ modalities directly in the higher-dimensional space in which modality embeddings lie by minimizing the Gramian volume of the $k$-dimensional parallelotope spanned by the modality vectors, ensuring the geometric alignment of all modalities simultaneously. GRAM can replace cosine similarity in any downstream method, holding for 2 to $n$ modalities and providing more meaningful alignment with respect to previous similarity measures. The novel GRAM-based contrastive loss function enhances the alignment of multimodal models in the higher-dimensional embedding space, leading to new state-of-the-art performance in downstream tasks such as video-audio-text retrieval and audio-video classification. The project page, the code, and the pretrained models are available at https://ispamm.github.io/GRAM/.
Authors: Hourun Li, Yifan Wang, Zhiping Xiao, Jia Yang, Changling Zhou, Ming Zhang, Wei Ju
Abstract: Recommender systems are widely used in various real-world applications, but they often encounter the persistent challenge of the user cold-start problem. Cross-domain recommendation (CDR), which leverages user interactions from one domain to improve prediction performance in another, has emerged as a promising solution. However, users with similar preferences in the source domain may exhibit different interests in the target domain. Therefore, directly transferring embeddings may introduce irrelevant source-domain collaborative information. In this paper, we propose a novel graph-based disentangled contrastive learning framework to capture fine-grained user intent and filter out irrelevant collaborative information, thereby avoiding negative transfer. Specifically, for each domain, we use a multi-channel graph encoder to capture diverse user intents. We then construct the affinity graph in the embedding space and perform multi-step random walks to capture high-order user similarity relationships. Treating one domain as the target, we propose a disentangled intent-wise contrastive learning approach, guided by user similarity, to refine the bridging of user intents across domains. Extensive experiments on four benchmark CDR datasets demonstrate that DisCo consistently outperforms existing state-of-the-art baselines, thereby validating the effectiveness of both DisCo and its components.
Authors: Hyogon Ryu, NaHyeon Park, Hyunjung Shim
Abstract: Despite the widespread use of text-to-image diffusion models across various tasks, their computational and memory demands limit practical applications. To mitigate this issue, quantization of diffusion models has been explored. It reduces memory usage and computational costs by compressing weights and activations into lower-bit formats. However, existing methods often struggle to preserve both image quality and text-image alignment, particularly in lower-bit($<$ 8bits) quantization. In this paper, we analyze the challenges associated with quantizing text-to-image diffusion models from a distributional perspective. Our analysis reveals that activation outliers play a crucial role in determining image quality. Additionally, we identify distinctive patterns in cross-attention scores, which significantly affects text-image alignment. To address these challenges, we propose Distribution-aware Group Quantization (DGQ), a method that identifies and adaptively handles pixel-wise and channel-wise outliers to preserve image quality. Furthermore, DGQ applies prompt-specific logarithmic quantization scales to maintain text-image alignment. Our method demonstrates remarkable performance on datasets such as MS-COCO and PartiPrompts. We are the first to successfully achieve low-bit quantization of text-to-image diffusion models without requiring additional fine-tuning of weight quantization parameters. Code is available at https://github.com/ugonfor/DGQ.
Authors: Ruilin Luo, Zhuofan Zheng, Yifan Wang, Yiyao Yu, Xinzhe Ni, Zicheng Lin, Jin Zeng, Yujiu Yang
Abstract: Chain-of-Thought (CoT) reasoning is widely used to enhance the mathematical reasoning capabilities of large language models (LLMs). The introduction of process supervision for CoT trajectories has sparked discussions on improving test-time scaling, thereby unlocking the System 2-style thinking capabilities of these models. However, in multimodal mathematical reasoning, the scarcity of high-quality CoT training data has hindered existing models from achieving both deliberate reasoning and fine-grained verification. In this work, we propose a novel framework that introduces System 2-style thinking to multimodal mathematical reasoning. We introduce a three-module CoT data synthesis process that integrates CoT distillation, trajectory-format rewriting, and format unification. This process generates MMathCoT-1M, a high-quality CoT reasoning instruction fine-tuning dataset. Furthermore, we implement a dual-view trajectory labeling automation that targets both visual grounding fidelity and deductive chain validity, resulting in the DualMath-1.1M dataset. The URSA-8B model, trained on MMathCoT-1M, achieves new state-of-the-art (SOTA) performance among similarly sized multimodal LLMs on six popular reasoning benchmarks. Training URSA-8B further on the DualMath-1.1M dataset yields URSA-RM-8B, a verifier that enhances URSA-8B's test-time performance and surpasses strong closed-source multimodal MLLMs like GPT-4o. The model weights, training data, and code have been open-sourced: https://github.com/URSA-MATH/URSA-MATH.
Authors: Zixuan Ke, Yifei Ming, Xuan-Phi Nguyen, Caiming Xiong, Shafiq Joty
Abstract: Domain-adaptive post-training of large language models (LLMs) has emerged as a promising approach for specialized domains such as medicine and finance. However, significant challenges remain in identifying optimal adaptation criteria and training strategies across varying data and model configurations. To address these challenges, we introduce FINDAP, a systematic and fine-grained investigation into domain adaptive post-training of LLMs for the finance domain. Our approach consists of four key components: FinCap, which defines the core capabilities required for the target domain; FinRec, an effective training recipe that jointly optimizes continual pre-training and instruction-following, along with a novel preference data distillation method leveraging process signals from a generative reward model; FinTrain, a curated set of training datasets supporting FinRec; and FinEval, a comprehensive evaluation suite aligned with FinCap. The resulting model, Llama-Fin, achieves state-of-the-art performance across a wide range of financial tasks. Our analysis also highlights how each post-training stage contributes to distinct capabilities, uncovering specific challenges and effective solutions, providing valuable insights for domain adaptation of LLMs.
Authors: Jingyi Wang, Haowei Wang, Nai-Yuan Chiang, Cosmin G. Petra
Abstract: Expected improvement (EI) is one of the most widely used acquisition functions in Bayesian optimization (BO). Despite its proven success in applications for decades, important open questions remain on the theoretical convergence behaviors and rates for EI. In this paper, we contribute to the convergence theory of EI in three novel and critical areas. First, we consider objective functions that fit under the Gaussian process (GP) prior assumption, whereas existing works mostly focus on functions in the reproducing kernel Hilbert space (RKHS). Second, we establish for the first time the asymptotic error bound and its corresponding rate for GP-EI with noisy observations under the GP prior assumption. Third, by investigating the exploration and exploitation properties of the non-convex EI function, we establish improved error bounds of GP-EI for both the noise-free and noisy cases.
Authors: Samiran Dey, Christopher R. S. Banerji, Partha Basuchowdhuri, Sanjoy K. Saha, Deepak Parashar, Tapabrata Chakraborti
Abstract: Emerging research has highlighted that artificial intelligence based multimodal fusion of digital pathology and transcriptomic features can improve cancer diagnosis (grading/subtyping) and prognosis (survival risk) prediction. However, such direct fusion for joint decision is impractical in real clinical settings, where histopathology is still the gold standard for diagnosis and transcriptomic tests are rarely requested, at least in the public healthcare system. With our novel diffusion based crossmodal generative AI model PathGen, we show that genomic expressions synthesized from digital histopathology jointly predicts cancer grading and patient survival risk with high accuracy (state-of-the-art performance), certainty (through conformal coverage guarantee) and interpretability (through distributed attention maps). PathGen code is available for open use by the research community through GitHub at https://github.com/Samiran-Dey/PathGen.
Authors: Thibault de Surrel, Fabien Lotte, Sylvain Chevallier, Florian Yger
Abstract: Circular and non-flat data distributions are prevalent across diverse domains of data science, yet their specific geometric structures often remain underutilized in machine learning frameworks. A principled approach to accounting for the underlying geometry of such data is pivotal, particularly when extending statistical models, like the pervasive Gaussian distribution. In this work, we tackle those issue by focusing on the manifold of symmetric positive definite matrices, a key focus in information geometry. We introduced a non-isotropic wrapped Gaussian by leveraging the exponential map, we derive theoretical properties of this distribution and propose a maximum likelihood framework for parameter estimation. Furthermore, we reinterpret established classifiers on SPD through a probabilistic lens and introduce new classifiers based on the wrapped Gaussian model. Experiments on synthetic and real-world datasets demonstrate the robustness and flexibility of this geometry-aware distribution, underscoring its potential to advance manifold-based data analysis. This work lays the groundwork for extending classical machine learning and statistical methods to more complex and structured data.
Authors: Yuwei Yin, Giuseppe Carenini
Abstract: Large language models (LLMs) achieve remarkable performance on challenging benchmarks that are often structured as multiple-choice question-answering (QA) tasks. Zero-shot Chain-of-Thought (CoT) prompting enhances reasoning in LLMs but provides only vague and generic guidance ("think step by step"). This paper introduces ARR, an intuitive and effective zero-shot prompting method that explicitly incorporates three key steps in QA solving: analyzing the intent of the question, retrieving relevant information, and reasoning step by step. Comprehensive experiments across diverse and challenging QA tasks demonstrate that ARR consistently improves the Baseline (without ARR prompting) and outperforms CoT. Ablation and case studies further validate the positive contributions of each component: analyzing, retrieving, and reasoning. Notably, intent analysis plays a vital role in ARR. Additionally, extensive evaluations across various model sizes, LLM series, and generation settings solidify the effectiveness, robustness, and generalizability of ARR.
Authors: Zhenwei Wu, Jinxiong Lu, Yuxiao Chen, Yunxin Liu, Yueting Zhuang, Luhui Hu
Abstract: Humanoid robotics presents significant challenges in artificial intelligence, requiring precise coordination and control of high-degree-of-freedom systems. Designing effective reward functions for deep reinforcement learning (DRL) in this domain remains a critical bottleneck, demanding extensive manual effort, domain expertise, and iterative refinement. To overcome these challenges, we introduce STRIDE, a novel framework built on agentic engineering to automate reward design, DRL training, and feedback optimization for humanoid robot locomotion tasks. By combining the structured principles of agentic engineering with large language models (LLMs) for code-writing, zero-shot generation, and in-context optimization, STRIDE generates, evaluates, and iteratively refines reward functions without relying on task-specific prompts or templates. Across diverse environments featuring humanoid robot morphologies, STRIDE outperforms the state-of-the-art reward design framework EUREKA, achieving an average improvement of round 250% in efficiency and task performance. Using STRIDE-generated rewards, simulated humanoid robots achieve sprint-level locomotion across complex terrains, highlighting its ability to advance DRL workflows and humanoid robotics research.
Authors: Yijia Xiao, Runhui Wang, Luyang Kong, Davor Golac, Wei Wang
Abstract: The increasing complexity of computer science research projects demands more effective tools for deploying code repositories. Large Language Models (LLMs), such as Anthropic Claude and Meta Llama, have demonstrated significant advancements across various fields of computer science research, including the automation of diverse software engineering tasks. To evaluate the effectiveness of LLMs in handling complex code development tasks of research projects, particularly for NLP/CV/AI/ML/DM topics, we introduce CSR-Bench, a benchmark for Computer Science Research projects. This benchmark assesses LLMs from various aspects including accuracy, efficiency, and deployment script quality, aiming to explore their potential in conducting computer science research autonomously. We also introduce a novel framework, CSR-Agents, that utilizes multiple LLM agents to automate the deployment of GitHub code repositories of computer science research projects. Specifically, by checking instructions from markdown files and interpreting repository structures, the model generates and iteratively improves bash commands that set up the experimental environments and deploy the code to conduct research tasks. Preliminary results from CSR-Bench indicate that LLM agents can significantly enhance the workflow of repository deployment, thereby boosting developer productivity and improving the management of developmental workflows.
Authors: Yuchen He, Chihao Zhang
Abstract: We study the problem of sampling from a $d$-dimensional distribution with density $p(x)\propto e^{-f(x)}$, which does not necessarily satisfy good isoperimetric conditions. Specifically, we show that for any $L,M$ satisfying $LM\ge d\ge 5$, $\epsilon\in \left(0,\frac{1}{32}\right)$, and any algorithm with query accesses to the value of $f(x)$ and $\nabla f(x)$, there exists an $L$-log-smooth distribution with second moment at most $M$ such that the algorithm requires $\left(\frac{LM}{d\epsilon}\right)^{\Omega(d)}$ queries to compute a sample whose distribution is within $\epsilon$ in total variation distance to the target distribution. We complement the lower bound with an algorithm requiring $\left(\frac{LM}{d\epsilon}\right)^{\mathcal O(d)}$ queries, thereby characterizing the tight (up to the constant in the exponent) query complexity for sampling from the family of non-log-concave distributions. Our results are in sharp contrast with the recent work of Huang et al. (COLT'24), where an algorithm with quasi-polynomial query complexity was proposed for sampling from a non-log-concave distribution when $M=\mathtt{poly}(d)$. Their algorithm works under the stronger condition that all distributions along the trajectory of the Ornstein-Uhlenbeck process, starting from the target distribution, are $\mathcal O(1)$-log-smooth. We investigate this condition and prove that it is strictly stronger than requiring the target distribution to be $\mathcal O(1)$-log-smooth. Additionally, we study this condition in the context of mixtures of Gaussians. Finally, we place our results within the broader theme of ``sampling versus optimization'', as studied in Ma et al. (PNAS'19). We show that for a wide range of parameters, sampling is strictly easier than optimization by a super-exponential factor in the dimension $d$.
Authors: Linxiao Gong, Hao Yang, Gaoyun Fang, Bobo Ju, Juncen Guo, Xiaoguang Zhu, Yan Wang, Xiping Hu, Peng Sun, Azzedine Boukerche
Abstract: The explosive growth of video data has driven the development of distributed video analytics in cloud-edge-terminal collaborative (CETC) systems, enabling efficient video processing, real-time inference, and privacy-preserving analysis. Among multiple advantages, CETC systems can distribute video processing tasks and enable adaptive analytics across cloud, edge, and terminal devices, leading to breakthroughs in video surveillance, autonomous driving, and smart cities. In this survey, we first analyze fundamental architectural components, including hierarchical, distributed, and hybrid frameworks, alongside edge computing platforms and resource management mechanisms. Building upon these foundations, edge-centric approaches emphasize on-device processing, edge-assisted offloading, and edge intelligence, while cloud-centric methods leverage powerful computational capabilities for complex video understanding and model training. Our investigation also covers hybrid video analytics incorporating adaptive task offloading and resource-aware scheduling techniques that optimize performance across the entire system. Beyond conventional approaches, recent advances in large language models and multimodal integration reveal both opportunities and challenges in platform scalability, data protection, and system reliability. Future directions also encompass explainable systems, efficient processing mechanisms, and advanced video analytics, offering valuable insights for researchers and practitioners in this dynamic field.
Authors: Chenao Li, Shuo Yan, Enyan Dai
Abstract: Enzyme-catalyzed protein cleavage is essential for many biological functions. Accurate prediction of cleavage sites can facilitate various applications such as drug development, enzyme design, and a deeper understanding of biological mechanisms. However, most existing models are restricted to an individual enzyme, which neglects shared knowledge of enzymes and fails generalize to novel enzymes. Thus, we introduce a unified protein cleavage site predictor named UniZyme, which can generalize across diverse enzymes. To enhance the enzyme encoding for the protein cleavage site prediction, UniZyme employs a novel biochemically-informed model architecture along with active-site knowledge of proteolytic enzymes. Extensive experiments demonstrate that UniZyme achieves high accuracy in predicting cleavage sites across a range of proteolytic enzymes, including unseen enzymes. The code is available in https://anonymous.4open.science/r/UniZyme-4A67.
Authors: Leonardo Berti, Bardh Prenkaj, Paola Velardi
Abstract: Financial markets are complex systems characterized by high statistical noise, nonlinearity, and constant evolution. Thus, modeling them is extremely hard. We address the task of generating realistic and responsive Limit Order Book (LOB) market simulations, which are fundamental for calibrating and testing trading strategies, performing market impact experiments, and generating synthetic market data. Previous works lack realism, usefulness, and responsiveness of the generated simulations. To bridge this gap, we propose a novel TRAnsformer-based Denoising Diffusion Probabilistic Engine for LOB Simulations (TRADES). TRADES generates realistic order flows conditioned on the state of the market, leveraging a transformer-based architecture that captures the temporal and spatial characteristics of high-frequency market data. There is a notable absence of quantitative metrics for evaluating generative market simulation models in the literature. To tackle this problem, we adapt the predictive score, a metric measured as an MAE, by training a stock price predictive model on synthetic data and testing it on real data. We compare TRADES with previous works on two stocks, reporting an x3.27 and x3.47 improvement over SoTA according to the predictive score, demonstrating that we generate useful synthetic market data for financial downstream tasks. We assess TRADES's market simulation realism and responsiveness, showing that it effectively learns the conditional data distribution and successfully reacts to an experimental agent, giving sprout to possible calibrations and evaluations of trading strategies and market impact experiments. We developed DeepMarket, the first open-source Python framework for market simulation with deep learning. Our repository includes a synthetic LOB dataset composed of TRADES's generates simulations. We release the code at github.com/LeonardoBerti00/DeepMarket.
Authors: Atharva Mehta, Shivam Chauhan, Amirbek Djanibekov, Atharva Kulkarni, Gus Xia, Monojit Choudhury
Abstract: The advent of Music-Language Models has greatly enhanced the automatic music generation capability of AI systems, but they are also limited in their coverage of the musical genres and cultures of the world. We present a study of the datasets and research papers for music generation and quantify the bias and under-representation of genres. We find that only 5.7% of the total hours of existing music datasets come from non-Western genres, which naturally leads to disparate performance of the models across genres. We then investigate the efficacy of Parameter-Efficient Fine-Tuning (PEFT) techniques in mitigating this bias. Our experiments with two popular models -- MusicGen and Mustango, for two underrepresented non-Western music traditions -- Hindustani Classical and Turkish Makam music, highlight the promises as well as the non-triviality of cross-genre adaptation of music through small datasets, implying the need for more equitable baseline music-language models that are designed for cross-cultural transfer learning.
Authors: Sixiao Zheng, Zimian Peng, Yanpeng Zhou, Yi Zhu, Hang Xu, Xiangru Huang, Yanwei Fu
Abstract: Recent image-to-video generation methods have demonstrated success in enabling control over one or two visual elements, such as camera trajectory or object motion. However, these methods are unable to offer control over multiple visual elements due to limitations in data and network efficacy. In this paper, we introduce VidCRAFT3, a novel framework for precise image-to-video generation that enables control over camera motion, object motion, and lighting direction simultaneously. To better decouple control over each visual element, we propose the Spatial Triple-Attention Transformer, which integrates lighting direction, text, and image in a symmetric way. Since most real-world video datasets lack lighting annotations, we construct a high-quality synthetic video dataset, the VideoLightingDirection (VLD) dataset. This dataset includes lighting direction annotations and objects of diverse appearance, enabling VidCRAFT3 to effectively handle strong light transmission and reflection effects. Additionally, we propose a three-stage training strategy that eliminates the need for training data annotated with multiple visual elements (camera motion, object motion, and lighting direction) simultaneously. Extensive experiments on benchmark datasets demonstrate the efficacy of VidCRAFT3 in producing high-quality video content, surpassing existing state-of-the-art methods in terms of control granularity and visual coherence. All code and data will be publicly available.