Authors: Yukun Chen, Zihuan Qiu, Fanman Meng, Hongliang Li, Linfeng Xu, Qingbo Wu
Abstract: Unlike traditional Multimodal Class-Incremental Learning (MCIL) methods that focus only on vision and text, this paper explores MCIL across vision, audio and text modalities, addressing challenges in integrating complementary information and mitigating catastrophic forgetting. To tackle these issues, we propose an MCIL method based on multimodal pre-trained models. Firstly, a Multimodal Incremental Feature Extractor (MIFE) based on Mixture-of-Experts (MoE) structure is introduced to achieve effective incremental fine-tuning for AudioCLIP. Secondly, to enhance feature discriminability and generalization, we propose an Adaptive Audio-Visual Fusion Module (AAVFM) that includes a masking threshold mechanism and a dynamic feature fusion mechanism, along with a strategy to enhance text diversity. Thirdly, a novel multimodal class-incremental contrastive training loss is proposed to optimize cross-modal alignment in MCIL. Finally, two MCIL-specific evaluation metrics are introduced for comprehensive assessment. Extensive experiments on three multimodal datasets validate the effectiveness of our method.
Authors: Wei Li, Mengcheng Lan, Jiaxing Xu, Yiping Ke
Abstract: Graphs are essential for modeling complex interactions across domains such as social networks, biology, and recommendation systems. Traditional Graph Neural Networks, particularly Message Passing Neural Networks (MPNNs), rely heavily on supervised learning, limiting their generalization and applicability in label-scarce scenarios. Recent self-supervised approaches still require labeled fine-tuning, limiting their effectiveness in zero-shot scenarios. Meanwhile, Large Language Models (LLMs) excel in natural language tasks but face significant challenges when applied to graphs, including preserving reasoning abilities, managing extensive token lengths from rich node attributes, and being limited to textual-attributed graphs (TAGs) and a single level task. To overcome these limitations, we propose the Node-Oriented Conceptualization LLM (NOCL), a novel framework that leverages two core techniques: 1) node description, which converts heterogeneous node attributes into structured natural language, extending LLM from TAGs to non-TAGs; 2) node concept, which encodes node descriptions into compact semantic embeddings using pretrained language models, significantly reducing token lengths by up to 93.9% compared to directly using node descriptions. Additionally, our NOCL employs graph representation descriptors to unify graph tasks at various levels into a shared, language-based query format, paving a new direction for Graph Foundation Models. Experimental results validate NOCL's competitive supervised performance relative to traditional MPNNs and hybrid LLM-MPNN methods and demonstrate superior generalization in zero-shot settings.
Authors: Uijun Jung, Deokho Jang, Sungchul Kim, Jungho Kim
Abstract: Optical properties of thin film are greatly influenced by the thickness of each layer. Accurately predicting these thicknesses and their corresponding optical properties is important in the optical inverse design of thin films. However, traditional inverse design methods usually demand extensive numerical simulations and optimization procedures, which are time-consuming. In this paper, we utilize deep learning for the inverse design of the transmission spectra of SiO2/TiO2 multilayer thin films. We implement a tandem neural network (TNN), which can solve the one-to-many mapping problem that greatly degrades the performance of deep-learning-based inverse designs. In general, the TNN has been implemented by a back-to-back connection of an inverse neural network and a pre-trained forward neural network, both of which have been implemented based on multilayer perceptron (MLP) algorithms. In this paper, we propose to use not only MLP, but also convolutional neural network (CNN) or long short-term memory (LSTM) algorithms in the configuration of the TNN. We show that an LSTM-LSTM-based TNN yields the highest accuracy but takes the longest training time among nine configurations of TNNs. We also find that a CNN-LSTM-based TNN will be an optimal solution in terms of accuracy and speed because it could integrate the strengths of the CNN and LSTM algorithms.
Authors: Shangpin Peng, Weinong Wang, Zhuotao Tian, Senqiao Yang, Xing Wu, Haotian Xu, Chengquan Zhang, Takashi Isobe, Baotian Hu, Min Zhang
Abstract: Direct Preference Optimization (DPO) has become a cornerstone of reinforcement learning from human feedback (RLHF) due to its simplicity and efficiency. However, existing DPO-based approaches typically treat all preference pairs uniformly, ignoring critical variations in their inherent quality and learning utility, leading to suboptimal data utilization and performance. To address this challenge, we propose Omni-DPO, a dual-perspective optimization framework that jointly accounts for (1) the inherent quality of each preference pair and (2) the model's evolving performance on those pairs. By adaptively weighting samples according to both data quality and the model's learning dynamics during training, Omni-DPO enables more effective training data utilization and achieves better performance. Experimental results on various models and benchmarks demonstrate the superiority and generalization capabilities of Omni-DPO. On textual understanding tasks, Gemma-2-9b-it finetuned with Omni-DPO beats the leading LLM, Claude 3 Opus, by a significant margin of 6.7 points on the Arena-Hard benchmark. On mathematical reasoning tasks, Omni-DPO consistently outperforms the baseline methods across all benchmarks, providing strong empirical evidence for the effectiveness and robustness of our approach. Code and models will be available at https://github.com/pspdada/Omni-DPO.
Authors: Brendan Leigh Ross, No\"el Vouitsis, Atiyeh Ashari Ghomi, Rasa Hosseinzadeh, Ji Xin, Zhaoyan Liu, Yi Sui, Shiyi Hou, Kin Kwan Leung, Gabriel Loaiza-Ganem, Jesse C. Cresswell
Abstract: Although large language models (LLMs) are becoming increasingly capable of solving challenging real-world tasks, accurately quantifying their uncertainty remains a critical open problem, which limits their applicability in high-stakes domains. This challenge is further compounded by the closed-source, black-box nature of many state-of-the-art LLMs. Moreover, LLM-based systems can be highly sensitive to the prompts that bind them together, which often require significant manual tuning (i.e., prompt engineering). In this work, we address these challenges by viewing LLM-based systems through a Bayesian lens. We interpret prompts as textual parameters in a statistical model, allowing us to use a small training dataset to perform Bayesian inference over these prompts. This novel perspective enables principled uncertainty quantification over both the model's textual parameters and its downstream predictions, while also incorporating prior beliefs about these parameters expressed in free-form text. To perform Bayesian inference, a difficult problem even for well-studied data modalities, we introduce Metropolis-Hastings through LLM Proposals (MHLP), a novel Markov chain Monte Carlo (MCMC) algorithm that combines prompt optimization techniques with standard MCMC methods. MHLP is a turnkey modification to existing LLM pipelines, including those that rely exclusively on closed-source models. Empirically, we demonstrate that our method yields improvements in both predictive accuracy and uncertainty quantification (UQ) on a range of LLM benchmarks and UQ tasks. More broadly, our work demonstrates a viable path for incorporating methods from the rich Bayesian literature into the era of LLMs, paving the way for more reliable and calibrated LLM-based systems.
Authors: Dane Williamson, Yangfeng Ji, Matthew Dwyer
Abstract: Out-of-distribution (OOD) detection is a critical task in machine learning, particularly for safety-critical applications where unexpected inputs must be reliably flagged. While hierarchical variational autoencoders (HVAEs) offer improved representational capacity over traditional VAEs, their performance is highly sensitive to how latent dimensions are distributed across layers. Existing approaches often allocate latent capacity arbitrarily, leading to ineffective representations or posterior collapse. In this work, we introduce a theoretically grounded framework for optimizing latent dimension allocation in HVAEs, drawing on principles from information theory to formalize the trade-off between information loss and representational attenuation. We prove the existence of an optimal allocation ratio $r^{\ast}$ under a fixed latent budget, and empirically show that tuning this ratio consistently improves OOD detection performance across datasets and architectures. Our approach outperforms baseline HVAE configurations and provides practical guidance for principled latent structure design, leading to more robust OOD detection with deep generative models.
Authors: Bingshan Hu, Zheng He, Danica J. Sutherland
Abstract: We consider a kernelized bandit problem with a compact arm set ${X} \subset \mathbb{R}^d $ and a fixed but unknown reward function $f^*$ with a finite norm in some Reproducing Kernel Hilbert Space (RKHS). We propose a class of computationally efficient kernelized bandit algorithms, which we call GP-Generic, based on a novel concept: exploration distributions. This class of algorithms includes Upper Confidence Bound-based approaches as a special case, but also allows for a variety of randomized algorithms. With careful choice of exploration distribution, our proposed generic algorithm realizes a wide range of concrete algorithms that achieve $\tilde{O}(\gamma_T\sqrt{T})$ regret bounds, where $\gamma_T$ characterizes the RKHS complexity. This matches known results for UCB- and Thompson Sampling-based algorithms; we also show that in practice, randomization can yield better practical results.
Authors: Md. Faizul Islam Ansari
Abstract: This research implements an advanced unsupervised clustering system for MNIST handwritten digits through two-phase deep autoencoder architecture. A deep neural autoencoder requires a training process during phase one to develop minimal yet interpretive representations of images by minimizing reconstruction errors. During the second phase we unify the reconstruction error with a KMeans clustering loss for learned latent embeddings through a joint distance-based objective. Our model contains three elements which include batch normalization combined with dropout and weight decay for achieving generalized and stable results. The framework achieves superior clustering performance during extensive tests which used intrinsic measurements including Silhouette Score and Davies-Bouldin Index coupled with extrinsic metrics NMI and ARI when processing image features. The research uses t-SNE visualization to present learned embeddings that show distinct clusters for digits. Our approach reaches an optimal combination between data reconstruction accuracy and cluster separation purity when adding the benefit of understandable results and scalable implementations. The approach creates a dependable base that helps deploy unsupervised representation learning in different large-scale image clustering applications.
Authors: Ahmed Elbakary, Chaouki Ben Issaid, Mehdi Bennis
Abstract: We present a novel federated multi-task learning method that leverages cross-client similarity to enable personalized learning for each client. To avoid transmitting the entire model to the parameter server, we propose a communication-efficient scheme that introduces a feature anchor, a compact vector representation that summarizes the features learned from the client's local classes. This feature anchor is shared with the server to account for local clients' distribution. In addition, the clients share the classification heads, a lightweight linear layer, and perform a graph-based regularization to enable collaboration among clients. By modeling collaboration between clients as a dynamic graph and continuously updating and refining this graph, we can account for any drift from the clients. To ensure beneficial knowledge transfer and prevent negative collaboration, we leverage a community detection-based approach that partitions this dynamic graph into homogeneous communities, maximizing the sum of task similarities, represented as the graph edges' weights, within each community. This mechanism restricts collaboration to highly similar clients within their formed communities, ensuring positive interaction and preserving personalization. Extensive experiments on two heterogeneous datasets demonstrate that our method significantly outperforms state-of-the-art baselines. Furthermore, we show that our method exhibits superior computation and communication efficiency and promotes fairness across clients.
Authors: Nadav Torem, Tamar Sde-Chen, Yoav Y. Schechner
Abstract: Most natural objects have inherent complexity and variability. While some simple objects can be modeled from first principles, many real-world phenomena, such as cloud formation, require computationally expensive simulations that limit scalability. This work focuses on a class of physically meaningful, nonnegative objects that are computationally tractable but costly to simulate. To dramatically reduce computational costs, we propose nonnegative diffusion (NnD). This is a learned generative model using score based diffusion. It adapts annealed Langevin dynamics to enforce, by design, non-negativity throughout iterative scene generation and analysis (inference). NnD trains on high-quality physically simulated objects. Once trained, it can be used for generation and inference. We demonstrate generation of 3D volumetric clouds, comprising inherently nonnegative microphysical fields. Our generated clouds are consistent with cloud physics trends. They are effectively not distinguished as non-physical by expert perception.
Authors: Maryam Khalid, Akane Sano
Abstract: Graph-based Active Learning (AL) leverages the structure of graphs to efficiently prioritize label queries, reducing labeling costs and user burden in applications like health monitoring, human behavior analysis, and sensor networks. By identifying strategically positioned nodes, graph AL minimizes data collection demands while maintaining model performance, making it a valuable tool for dynamic environments. Despite its potential, existing graph AL methods are often evaluated on static graph datasets and primarily focus on prediction accuracy, neglecting user-centric considerations such as sampling diversity, query fairness, and adaptability to dynamic settings. To bridge this gap, we introduce GRAIL, a novel benchmarking framework designed to evaluate graph AL strategies in dynamic, real-world environments. GRAIL introduces novel metrics to assess sustained effectiveness, diversity, and user burden, enabling a comprehensive evaluation of AL methods under varying conditions. Extensive experiments on datasets featuring dynamic, real-life human sensor data reveal trade-offs between prediction performance and user burden, highlighting limitations in existing AL strategies. GRAIL demonstrates the importance of balancing node importance, query diversity, and network topology, providing an evaluation mechanism for graph AL solutions in dynamic environments.
Authors: Xinyi Hu, Aldo Pacchiano
Abstract: We study the decentralized multi-player multi-armed bandits (MMAB) problem under a no-sensing setting, where each player receives only their own reward and obtains no information about collisions. Each arm has an unknown capacity, and if the number of players pulling an arm exceeds its capacity, all players involved receive zero reward. This setting generalizes the classical unit-capacity model and introduces new challenges in coordination and capacity discovery under severe feedback limitations. We propose A-CAPELLA (Algorithm for Capacity-Aware Parallel Elimination for Learning and Allocation), a decentralized algorithm that achieves logarithmic regret in this generalized regime. Our main contribution is a collaborative hypothesis testing protocol that enables synchronized successive elimination and capacity estimation through carefully structured collision patterns. This represents a provably efficient learning result in decentralized no-sensing MMAB with unknown arm capacities.
Authors: Arnaud Fickinger, Abderrahim Bendahi, Stuart Russell
Abstract: Reinforcement-learning agents often struggle when deployed from simulation to the real-world. A dominant strategy for reducing the sim-to-real gap is domain randomization (DR) which trains the policy across many simulators produced by sampling dynamics parameters, but standard DR ignores offline data already available from the real system. We study offline domain randomization (ODR), which first fits a distribution over simulator parameters to an offline dataset. While a growing body of empirical work reports substantial gains with algorithms such as DROPO, the theoretical foundations of ODR remain largely unexplored. In this work, we (i) formalize ODR as a maximum-likelihood estimation over a parametric simulator family, (ii) prove consistency of this estimator under mild regularity and identifiability conditions, showing it converges to the true dynamics as the dataset grows, (iii) derive gap bounds demonstrating ODRs sim-to-real error is up to an O(M) factor tighter than uniform DR in the finite-simulator case (and analogous gains in the continuous setting), and (iv) introduce E-DROPO, a new version of DROPO which adds an entropy bonus to prevent variance collapse, yielding broader randomization and more robust zero-shot transfer in practice.
Authors: Daniel Lawson, Adriana Hugessen, Charlotte Cloutier, Glen Berseth, Khimya Khetarpal
Abstract: Behavioral cloning (BC) methods trained with supervised learning (SL) are an effective way to learn policies from human demonstrations in domains like robotics. Goal-conditioning these policies enables a single generalist policy to capture diverse behaviors contained within an offline dataset. While goal-conditioned behavior cloning (GCBC) methods can perform well on in-distribution training tasks, they do not necessarily generalize zero-shot to tasks that require conditioning on novel state-goal pairs, i.e. combinatorial generalization. In part, this limitation can be attributed to a lack of temporal consistency in the state representation learned by BC; if temporally related states are encoded to similar latent representations, then the out-of-distribution gap for novel state-goal pairs would be reduced. Hence, encouraging this temporal consistency in the representation space should facilitate combinatorial generalization. Successor representations, which encode the distribution of future states visited from the current state, nicely encapsulate this property. However, previous methods for learning successor representations have relied on contrastive samples, temporal-difference (TD) learning, or both. In this work, we propose a simple yet effective representation learning objective, $\text{BYOL-}\gamma$ augmented GCBC, which is not only able to theoretically approximate the successor representation in the finite MDP case without contrastive samples or TD learning, but also, results in competitive empirical performance across a suite of challenging tasks requiring combinatorial generalization.
Authors: Mohammad Taufeeque, Aaron David Tucker, Adam Gleave, Adri\`a Garriga-Alonso
Abstract: We partially reverse-engineer a convolutional recurrent neural network (RNN) trained to play the puzzle game Sokoban with model-free reinforcement learning. Prior work found that this network solves more levels with more test-time compute. Our analysis reveals several mechanisms analogous to components of classic bidirectional search. For each square, the RNN represents its plan in the activations of channels associated with specific directions. These state-action activations are analogous to a value function - their magnitudes determine when to backtrack and which plan branch survives pruning. Specialized kernels extend these activations (containing plan and value) forward and backward to create paths, forming a transition model. The algorithm is also unlike classical search in some ways. State representation is not unified; instead, the network considers each box separately. Each layer has its own plan representation and value function, increasing search depth. Far from being inscrutable, the mechanisms leveraging test-time compute learned in this network by model-free training can be understood in familiar terms.
Authors: Andrei V. Konstantinov, Vlada A. Efremenko, Lev V. Utkin
Abstract: Survival analysis is a fundamental tool for modeling time-to-event data in healthcare, engineering, and finance, where censored observations pose significant challenges. While traditional methods like the Beran estimator offer nonparametric solutions, they often struggle with the complex data structures and heavy censoring. This paper introduces three novel survival models, iSurvM (the imprecise Survival model based on Mean likelihood functions), iSurvQ (the imprecise Survival model based on the Quantiles of likelihood functions), and iSurvJ (the imprecise Survival model based on the Joint learning), that combine imprecise probability theory with attention mechanisms to handle censored data without parametric assumptions. The first idea behind the models is to represent censored observations by interval-valued probability distributions for each instance over time intervals between events moments. The second idea is to employ the kernel-based Nadaraya-Watson regression with trainable attention weights for computing the imprecise probability distribution over time intervals for the entire dataset. The third idea is to consider three decision strategies for training, which correspond to the proposed three models. Experiments on synthetic and real datasets demonstrate that the proposed models, especially iSurvJ, consistently outperform the Beran estimator from the accuracy and computational complexity points of view. Codes implementing the proposed models are publicly available.
Authors: Yaowen Zhang, Libera Fresiello, Peter H. Veltink, Dirk W. Donker, Ying Wang
Abstract: Heart failure (HF) poses a significant global health challenge, with early detection offering opportunities for improved outcomes. Abnormalities in heart rate (HR), particularly during daily activities, may serve as early indicators of HF risk. However, existing HR monitoring tools for HF detection are limited by their reliability on population-based averages. The estimation of individualized HR serves as a dynamic digital twin, enabling precise tracking of cardiac health biomarkers. Current HR estimation methods, categorized into physiologically-driven and purely data-driven models, struggle with efficiency and interpretability. This study introduces a novel physiological-model-based neural network (PMB-NN) framework for HR estimation based on oxygen uptake (VO2) data during daily physical activities. The framework was trained and tested on individual datasets from 12 participants engaged in activities including resting, cycling, and running. By embedding physiological constraints, which were derived from our proposed simplified human movement physiological model (PM), into the neural network training process, the PMB-NN model adheres to human physiological principles while achieving high estimation accuracy, with a median R$^2$ score of 0.8 and an RMSE of 8.3 bpm. Comparative statistical analysis demonstrates that the PMB-NN achieves performance on par with the benchmark neural network model while significantly outperforming traditional physiological model (p=0.002). In addition, our PMB-NN is adept at identifying personalized parameters of the PM, enabling the PM to generate reasonable HR estimation. The proposed framework with a precise VO2 estimation system derived from body movements enables the future possibilities of personalized and real-time cardiac monitoring during daily life physical activities.
Authors: Tejaswi Kasarla, Max van Spengler, Pascal Mettes
Abstract: Out-of-distribution recognition forms an important and well-studied problem in deep learning, with the goal to filter out samples that do not belong to the distribution on which a network has been trained. The conclusion of this paper is simple: a good hierarchical hyperbolic embedding is preferred for discriminating in- and out-of-distribution samples. We introduce Balanced Hyperbolic Learning. We outline a hyperbolic class embedding algorithm that jointly optimizes for hierarchical distortion and balancing between shallow and wide subhierarchies. We then use the class embeddings as hyperbolic prototypes for classification on in-distribution data. We outline how to generalize existing out-of-distribution scoring functions to operate with hyperbolic prototypes. Empirical evaluations across 13 datasets and 13 scoring functions show that our hyperbolic embeddings outperform existing out-of-distribution approaches when trained on the same data with the same backbones. We also show that our hyperbolic embeddings outperform other hyperbolic approaches, beat state-of-the-art contrastive methods, and natively enable hierarchical out-of-distribution generalization.
Authors: Minoh Jeong, Seonho Kim, Alfred Hero
Abstract: Deterministic embeddings learned by contrastive learning (CL) methods such as SimCLR and SupCon achieve state-of-the-art performance but lack a principled mechanism for uncertainty quantification. We propose Variational Contrastive Learning (VCL), a decoder-free framework that maximizes the evidence lower bound (ELBO) by interpreting the InfoNCE loss as a surrogate reconstruction term and adding a KL divergence regularizer to a uniform prior on the unit hypersphere. We model the approximate posterior $q_\theta(z|x)$ as a projected normal distribution, enabling the sampling of probabilistic embeddings. Our two instantiations--VSimCLR and VSupCon--replace deterministic embeddings with samples from $q_\theta(z|x)$ and incorporate a normalized KL term into the loss. Experiments on multiple benchmarks demonstrate that VCL mitigates dimensional collapse, enhances mutual information with class labels, and matches or outperforms deterministic baselines in classification accuracy, all the while providing meaningful uncertainty estimates through the posterior model. VCL thus equips contrastive learning with a probabilistic foundation, serving as a new basis for contrastive approaches.
Authors: Gilad Landau, Miran \"Ozdogan, Gereon Elvers, Francesco Mantegna, Pratik Somaiya, Dulhan Jayalath, Luisa Kurth, Teyun Kwon, Brendan Shillingford, Greg Farquhar, Minqi Jiang, Karim Jerbi, Hamza Abdelhedi, Yorguin Mantilla Ramos, Caglar Gulcehre, Mark Woolrich, Natalie Voets, Oiwi Parker Jones
Abstract: The advance of speech decoding from non-invasive brain data holds the potential for profound societal impact. Among its most promising applications is the restoration of communication to paralysed individuals affected by speech deficits such as dysarthria, without the need for high-risk surgical interventions. The ultimate aim of the 2025 PNPL competition is to produce the conditions for an "ImageNet moment" or breakthrough in non-invasive neural decoding, by harnessing the collective power of the machine learning community. To facilitate this vision we present the largest within-subject MEG dataset recorded to date (LibriBrain) together with a user-friendly Python library (pnpl) for easy data access and integration with deep learning frameworks. For the competition we define two foundational tasks (i.e. Speech Detection and Phoneme Classification from brain data), complete with standardised data splits and evaluation metrics, illustrative benchmark models, online tutorial code, a community discussion board, and public leaderboard for submissions. To promote accessibility and participation the competition features a Standard track that emphasises algorithmic innovation, as well as an Extended track that is expected to reward larger-scale computing, accelerating progress toward a non-invasive brain-computer interface for speech.
Authors: Zahra Shahrooei, Ali Baheri
Abstract: Deep off-policy actor-critic algorithms have emerged as the leading framework for reinforcement learning in continuous control domains. However, most of these algorithms suffer from poor sample efficiency, especially in environments with sparse rewards. In this paper, we take a step towards addressing this issue by providing a principled directed exploration strategy. We propose Wasserstein Barycenter Soft Actor-Critic (WBSAC) algorithm, which benefits from a pessimistic actor for temporal difference learning and an optimistic actor to promote exploration. This is achieved by using the Wasserstein barycenter of the pessimistic and optimistic policies as the exploration policy and adjusting the degree of exploration throughout the learning process. We compare WBSAC with state-of-the-art off-policy actor-critic algorithms and show that WBSAC is more sample-efficient on MuJoCo continuous control tasks.
Authors: Defang Chen, Zhenyu Zhou, Can Wang, Siwei Lyu
Abstract: Diffusion-based generative models employ stochastic differential equations (SDEs) and their equivalent probability flow ordinary differential equations (ODEs) to establish a smooth transformation between complex high-dimensional data distributions and tractable prior distributions. In this paper, we reveal a striking geometric regularity in the deterministic sampling dynamics: each simulated sampling trajectory lies within an extremely low-dimensional subspace, and all trajectories exhibit an almost identical ''boomerang'' shape, regardless of the model architecture, applied conditions, or generated content. We characterize several intriguing properties of these trajectories, particularly under closed-form solutions based on kernel-estimated data modeling. We also demonstrate a practical application of the discovered trajectory regularity by proposing a dynamic programming-based scheme to better align the sampling time schedule with the underlying trajectory structure. This simple strategy requires minimal modification to existing ODE-based numerical solvers, incurs negligible computational overhead, and achieves superior image generation performance, especially in regions with only $5 \sim 10$ function evaluations.
Authors: Mowafaq Salem Alzboon, Mohammad Al-Batah, Muhyeeddin Alqaraleh, Ahmad Abuashour, Ahmad Fuad Bader
Abstract: In many nations, diabetes is becoming a significant health problem, and early identification and control are crucial. Using machine learning algorithms to predict diabetes has yielded encouraging results. Using the Pima Indians Diabetes dataset, this study attempts to evaluate the efficacy of several machine-learning methods for diabetes prediction. The collection includes information on 768 patients, such as their ages, BMIs, and glucose levels. The techniques assessed are Logistic Regression, Decision Tree, Random Forest, k-Nearest Neighbors, Naive Bayes, Support Vector Machine, Gradient Boosting, and Neural Network. The findings indicate that the Neural Network algorithm performed the best, with an accuracy of 78.57 percent, followed by the Random Forest method, with an accuracy of 76.30 percent. The study implies that machine learning algorithms can aid diabetes prediction and be an efficient early detection tool.
Authors: Mohammad Subhi Al-Batah, Mowafaq Salem Alzboon, Muhyeeddin Alqaraleh
Abstract: This study conducts an empirical examination of MLP networks investigated through a rigorous methodical experimentation process involving three diverse datasets: TinyFace, Heart Disease, and Iris. Study Overview: The study includes three key methods: a) a baseline training using the default settings for the Multi-Layer Perceptron (MLP), b) feature selection using Genetic Algorithm (GA) based refinement c) Principal Component Analysis (PCA) based dimension reduction. The results show important information on how such techniques affect performance. While PCA had showed benefits in low-dimensional and noise-free datasets GA consistently increased accuracy in complex datasets by accurately identifying critical features. Comparison reveals that feature selection and dimensionality reduction play interdependent roles in enhancing MLP performance. The study contributes to the literature on feature engineering and neural network parameter optimization, offering practical guidelines for a wide range of machine learning tasks
Authors: Yuhui Ding, Thomas Hofmann
Abstract: Equivariant diffusion models have achieved impressive performance in 3D molecule generation. These models incorporate Euclidean symmetries of 3D molecules by utilizing an SE(3)-equivariant denoising network. However, specialized equivariant architectures limit the scalability and efficiency of diffusion models. In this paper, we propose an approach that relaxes such equivariance constraints. Specifically, our approach learns a sample-dependent SO(3) transformation for each molecule to construct an aligned latent space. A non-equivariant diffusion model is then trained over the aligned representations. Experimental results demonstrate that our approach performs significantly better than previously reported non-equivariant models. It yields sample quality comparable to state-of-the-art equivariant diffusion models and offers improved training and sampling efficiency. Our code is available at https://github.com/skeletondyh/RADM
Authors: Mohammad Subhi Al-Batah, Muhyeeddin Alqaraleh, Mowafaq Salem Alzboon
Abstract: Oral cancer presents a formidable challenge in oncology, necessitating early diagnosis and accurate prognosis to enhance patient survival rates. Recent advancements in machine learning and data mining have revolutionized traditional diagnostic methodologies, providing sophisticated and automated tools for differentiating between benign and malignant oral lesions. This study presents a comprehensive review of cutting-edge data mining methodologies, including Neural Networks, K-Nearest Neighbors (KNN), Support Vector Machines (SVM), and ensemble learning techniques, specifically applied to the diagnosis and prognosis of oral cancer. Through a rigorous comparative analysis, our findings reveal that Neural Networks surpass other models, achieving an impressive classification accuracy of 93,6 % in predicting oral cancer. Furthermore, we underscore the potential benefits of integrating feature selection and dimensionality reduction techniques to enhance model performance. These insights underscore the significant promise of advanced data mining techniques in bolstering early detection, optimizing treatment strategies, and ultimately improving patient outcomes in the realm of oral oncology.
Authors: Tina Behrouzi, Sana Tonekaboni, Rahul G. Krishnan, Anna Goldenberg
Abstract: Real-world observational data often contain existing or emerging heterogeneous subpopulations that deviate from global patterns. The majority of models tend to overlook these underrepresented groups, leading to inaccurate or even harmful predictions. Existing solutions often rely on detecting these samples as Out-of-domain (OOD) rather than adapting the model to new emerging patterns. We introduce DynaSubVAE, a Dynamic Subgrouping Variational Autoencoder framework that jointly performs representation learning and adaptive OOD detection. Unlike conventional approaches, DynaSubVAE evolves with the data by dynamically updating its latent structure to capture new trends. It leverages a novel non-parametric clustering mechanism, inspired by Gaussian Mixture Models, to discover and model latent subgroups based on embedding similarity. Extensive experiments show that DynaSubVAE achieves competitive performance in both near-OOD and far-OOD detection, and excels in class-OOD scenarios where an entire class is missing during training. We further illustrate that our dynamic subgrouping mechanism outperforms standalone clustering methods such as GMM and KMeans++ in terms of both OOD accuracy and regret precision.
Authors: Jing Liu, Toshiaki Koike-Akino, Ye Wang, Hassan Mansour, Matthew Brand
Abstract: To address the enormous size of Large Language Models (LLMs), model compression methods, such as quantization and pruning, are often deployed, especially on edge devices. In this work, we focus on layer-wise post-training quantization and pruning. Drawing connections between activation-aware weight pruning and sparse approximation problems, and motivated by the success of Iterative Hard Thresholding (IHT), we propose a unified method for Activation-aware Weight pruning and quantization via Projected gradient descent (AWP). Our experiments demonstrate that AWP outperforms state-of-the-art LLM pruning and quantization methods. Theoretical convergence guarantees of the proposed method for pruning are also provided.
Authors: Sajjad Karimi, Amit J. Shah, Gari D. Clifford, Reza Sameni
Abstract: Simultaneous electrocardiography (ECG) and phonocardiogram (PCG) provide a comprehensive, multimodal perspective on cardiac function by capturing the heart's electrical and mechanical activities, respectively. However, the distinct and overlapping information content of these signals, as well as their potential for mutual reconstruction and biomarker extraction, remains incompletely understood, especially under varying physiological conditions and across individuals. In this study, we systematically investigate the common and exclusive characteristics of ECG and PCG using the EPHNOGRAM dataset of simultaneous ECG-PCG recordings during rest and exercise. We employ a suite of linear and nonlinear machine learning models, including non-causal LSTM networks, to reconstruct each modality from the other and analyze the influence of causality, physiological state, and cross-subject variability. Our results demonstrate that nonlinear models, particularly non-causal LSTM, provide superior reconstruction performance, with reconstructing ECG from PCG proving more tractable than the reverse. Exercise and cross-subject scenarios present significant challenges, but envelope-based modeling that utilizes instantaneous amplitude features substantially improves cross-subject generalizability for cross-modal learning. Furthermore, we demonstrate that clinically relevant ECG biomarkers, such as fiducial points and QT intervals, can be estimated from PCG in cross-subject settings. These findings advance our understanding of the relationship between electromechanical cardiac modalities, in terms of both waveform characteristics and the timing of cardiac events, with potential applications in novel multimodal cardiac monitoring technologies.
Authors: Chen-Chia Chang, Wan-Hsuan Lin, Yikang Shen, Yiran Chen, Xin Zhang
Abstract: Automation of analog topology design is crucial due to customized requirements of modern applications with heavily manual engineering efforts. The state-of-the-art work applies a sequence-to-sequence approach and supervised finetuning on language models to generate topologies given user specifications. However, its circuit formulation is inefficient due to O(|V |2) token length and suffers from low precision sensitivity to numeric inputs. In this work, we introduce LaMAGIC2, a succinct float-input canonical formulation with identifier (SFCI) for language model-based analog topology generation. SFCI addresses these challenges by improving component-type recognition through identifier-based representations, reducing token length complexity to O(|V |), and enhancing numeric precision sensitivity for better performance under tight tolerances. Our experiments demonstrate that LaMAGIC2 achieves 34% higher success rates under a tight tolerance of 0.01 and 10X lower MSEs compared to a prior method. LaMAGIC2 also exhibits better transferability for circuits with more vertices with up to 58.5% improvement. These advancements establish LaMAGIC2 as a robust framework for analog topology generation.
Authors: Yuji Kawamata, Kaoru Kamijo, Maki Kihira, Akihiro Toyoda, Tomoru Nakayama, Akira Imakura, Tetsuya Sakurai, Yukihiko Okada
Abstract: In recent years, the growing need to leverage sensitive data across institutions has led to increased attention on federated learning (FL), a decentralized machine learning paradigm that enables model training without sharing raw data. However, existing FL-based clustering methods, known as federated clustering, typically assume simple data partitioning scenarios such as horizontal or vertical splits, and cannot handle more complex distributed structures. This study proposes data collaboration clustering (DC-Clustering), a novel federated clustering method that supports clustering over complex data partitioning scenarios where horizontal and vertical splits coexist. In DC-Clustering, each institution shares only intermediate representations instead of raw data, ensuring privacy preservation while enabling collaborative clustering. The method allows flexible selection between k-means and spectral clustering, and achieves final results with a single round of communication with the central server. We conducted extensive experiments using synthetic and open benchmark datasets. The results show that our method achieves clustering performance comparable to centralized clustering where all data are pooled. DC-Clustering addresses an important gap in current FL research by enabling effective knowledge discovery from distributed heterogeneous data. Its practical properties -- privacy preservation, communication efficiency, and flexibility -- make it a promising tool for privacy-sensitive domains such as healthcare and finance.
Authors: Atsutoshi Kumagai, Tomoharu Iwata, Taishi Nishiyama, Yasutoshi Ida, Yasuhiro Fujiwara
Abstract: We propose a meta-learning method for learning from multiple noisy annotators. In many applications such as crowdsourcing services, labels for supervised learning are given by multiple annotators. Since the annotators have different skills or biases, given labels can be noisy. To learn accurate classifiers, existing methods require many noisy annotated data. However, sufficient data might be unavailable in practice. To overcome the lack of data, the proposed method uses labeled data obtained in different but related tasks. The proposed method embeds each example in tasks to a latent space by using a neural network and constructs a probabilistic model for learning a task-specific classifier while estimating annotators' abilities on the latent space. This neural network is meta-learned to improve the expected test classification performance when the classifier is adapted to a given small amount of annotated data. This classifier adaptation is performed by maximizing the posterior probability via the expectation-maximization (EM) algorithm. Since each step in the EM algorithm is easily computed as a closed-form and is differentiable, the proposed method can efficiently backpropagate the loss through the EM algorithm to meta-learn the neural network. We show the effectiveness of our method with real-world datasets with synthetic noise and real-world crowdsourcing datasets.
Authors: Ryota Ueda, Takami Sato, Ken Kobayashi, Kazuhide Nakata
Abstract: Semidefinite programming (SDP) relaxation has emerged as a promising approach for neural network verification, offering tighter bounds than other convex relaxation methods for deep neural networks (DNNs) with ReLU activations. However, we identify a critical limitation in the SDP relaxation when applied to deep networks: interior-point vanishing, which leads to the loss of strict feasibility -- a crucial condition for the numerical stability and optimality of SDP. Through rigorous theoretical and empirical analysis, we demonstrate that as the depth of DNNs increases, the strict feasibility is likely to be lost, creating a fundamental barrier to scaling SDP-based verification. To address the interior-point vanishing, we design and investigate five solutions to enhance the feasibility conditions of the verification problem. Our methods can successfully solve 88% of the problems that could not be solved by existing methods, accounting for 41% of the total. Our analysis also reveals that the valid constraints for the lower and upper bounds for each ReLU unit are traditionally inherited from prior work without solid reasons, but are actually not only unbeneficial but also even harmful to the problem's feasibility. This work provides valuable insights into the fundamental challenges of SDP-based DNN verification and offers practical solutions to improve its applicability to deeper neural networks, contributing to the development of more reliable and secure systems with DNNs.
Authors: Jiajin Liu, Dongzhe Fan, Jiacheng Shen, Chuanhao Ji, Daochen Zha, Qiaoyu Tan
Abstract: Multimodal Large Language Models (MLLMs) have demonstrated remarkable capabilities in representing and understanding diverse modalities. However, they typically focus on modality alignment in a pairwise manner while overlooking structural relationships across data points. Integrating multimodality with structured graph information (i.e., multimodal graphs, MMGs) is essential for real-world applications such as social networks, healthcare, and recommendation systems. Existing MMG learning methods fall into three paradigms based on how they leverage MLLMs: Encoder, Aligner, and Predictor. MLLM-as-Encoder focuses on enhancing graph neural networks (GNNs) via multimodal feature fusion; MLLM-as-Aligner aligns multimodal attributes in language or hidden space to enable LLM-based graph reasoning; MLLM-as-Predictor treats MLLMs as standalone reasoners with in-context learning or fine-tuning. Despite their advances, the MMG field lacks a unified benchmark to fairly evaluate across these approaches, making it unclear what progress has been made. To bridge this gap, we present Graph-MLLM, a comprehensive benchmark for multimodal graph learning by systematically evaluating these three paradigms across six datasets with different domains. Through extensive experiments, we observe that jointly considering the visual and textual attributes of the nodes benefits graph learning, even when using pre-trained text-to-image alignment models (e.g., CLIP) as encoders. We also find that converting visual attributes into textual descriptions further improves performance compared to directly using visual inputs. Moreover, we observe that fine-tuning MLLMs on specific MMGs can achieve state-of-the-art results in most scenarios, even without explicit graph structure information. We hope that our open-sourced library will facilitate rapid, equitable evaluation and inspire further innovative research in this field.
Authors: Mo\"ise Blanchard, Vineet Goyal
Abstract: We study the impact of sharing exploration in multi-armed bandits in a grouped setting where a set of groups have overlapping feasible action sets [Baek and Farias '24]. In this grouped bandit setting, groups share reward observations, and the objective is to minimize the collaborative regret, defined as the maximum regret across groups. This naturally captures applications in which one aims to balance the exploration burden between groups or populations -- it is known that standard algorithms can lead to significantly imbalanced exploration cost between groups. We address this problem by introducing an algorithm Col-UCB that dynamically coordinates exploration across groups. We show that Col-UCB achieves both optimal minimax and instance-dependent collaborative regret up to logarithmic factors. These bounds are adaptive to the structure of shared action sets between groups, providing insights into when collaboration yields significant benefits over each group learning their best action independently.
Authors: Luc Raszewski, Christine De Kock
Abstract: Malicious sockpuppet detection on Wikipedia is critical to preserving access to reliable information on the internet and preventing the spread of disinformation. Prior machine learning approaches rely on stylistic and meta-data features, but do not prioritise adaptability to author-specific behaviours. As a result, they struggle to effectively model the behaviour of specific sockpuppet-groups, especially when text data is limited. To address this, we propose the application of meta-learning, a machine learning technique designed to improve performance in data-scarce settings by training models across multiple tasks. Meta-learning optimises a model for rapid adaptation to the writing style of a new sockpuppet-group. Our results show that meta-learning significantly enhances the precision of predictions compared to pre-trained models, marking an advancement in combating sockpuppetry on open editing platforms. We release a new dataset of sockpuppet investigations to foster future research in both sockpuppetry and meta-learning fields.
Authors: Paul Janson, Benjamin Therien, Quentin Anthony, Xiaolong Huang, Abhinav Moudgil, Eugene Belilovsky
Abstract: Learned optimizers have been an active research topic over the past decade, with increasing progress toward practical, general-purpose optimizers that can serve as drop-in replacements for widely used methods like Adam. However, recent advances -- such as VeLO, which was meta-trained for 4000 TPU-months -- remain largely inaccessible to the broader community, in part due to their reliance on JAX and the absence of user-friendly packages for applying the optimizers after meta-training. To address this gap, we introduce PyLO, a PyTorch-based library that brings learned optimizers to the broader machine learning community through familiar, widely adopted workflows. Unlike prior work focused on synthetic or convex tasks, our emphasis is on applying learned optimization to real-world large-scale pre-training tasks. Our release includes a CUDA-accelerated version of the small_fc_lopt learned optimizer architecture from (Metz et al., 2022a), delivering substantial speedups -- from 39.36 to 205.59 samples/sec throughput for training ViT B/16 with batch size 32. PyLO also allows us to easily combine learned optimizers with existing optimization tools such as learning rate schedules and weight decay. When doing so, we find that learned optimizers can substantially benefit. Our code is available at https://github.com/Belilovsky-Lab/pylo
Authors: Aaryam Sharma
Abstract: Air pollution has become a significant health risk in developing countries. While governments routinely publish air-quality index (AQI) data to track pollution, these values fail to capture the local reality, as sensors are often very sparse. In this paper, we address this gap by predicting AQI in 1 km^2 neighborhoods, using the example of AirDelhi dataset. Using Spatio-temporal GNNs we surpass existing works by 71.654 MSE a 79% reduction, even on unseen coordinates. New insights about AQI such as the existence of strong repetitive short-term patterns and changing spatial relations are also discovered. The code is available on GitHub.
Authors: Wanqiao Xu, Allen Nie, Ruijie Zheng, Aditya Modi, Adith Swaminathan, Ching-An Cheng
Abstract: Interactively learning from observation and language feedback is an increasingly studied area driven by the emergence of large language model (LLM) agents. While impressive empirical demonstrations have been shown, so far a principled framing of these decision problems remains lacking. In this paper, we formalize the Learning from Language Feedback (LLF) problem, assert sufficient assumptions to enable learning despite latent rewards, and introduce $\textit{transfer eluder dimension}$ as a complexity measure to characterize the hardness of LLF problems. We show that transfer eluder dimension captures the intuition that information in the feedback changes the learning complexity of the LLF problem. We demonstrate cases where learning from rich language feedback can be exponentially faster than learning from reward. We develop a no-regret algorithm, called $\texttt{HELiX}$, that provably solves LLF problems through sequential interactions, with performance guarantees that scale with the transfer eluder dimension of the problem. Across several empirical domains, we show that $\texttt{HELiX}$ performs well even when repeatedly prompting LLMs does not work reliably. Our contributions mark a first step towards designing principled interactive learning algorithms from generic language feedback.
Authors: Yanlong Chen, Mattia Orlandi, Pierangelo Maria Rapa, Simone Benatti, Luca Benini, Yawei Li
Abstract: Physiological signals are often corrupted by motion artifacts, baseline drift, and other low-SNR disturbances, which pose significant challenges for analysis. Additionally, these signals exhibit strong non-stationarity, with sharp peaks and abrupt changes that evolve continuously, making them difficult to represent using traditional time-domain or filtering methods. To address these issues, a novel wavelet-based approach for physiological signal analysis is presented, aiming to capture multi-scale time-frequency features in various physiological signals. Leveraging this technique, two large-scale pretrained models specific to EMG and ECG are introduced for the first time, achieving superior performance and setting new baselines in downstream tasks. Additionally, a unified multi-modal framework is constructed by integrating pretrained EEG model, where each modality is guided through its dedicated branch and fused via learnable weighted fusion. This design effectively addresses challenges such as low signal-to-noise ratio, high inter-subject variability, and device mismatch, outperforming existing methods on multi-modal tasks. The proposed wavelet-based architecture lays a solid foundation for analysis of diverse physiological signals, while the multi-modal design points to next-generation physiological signal processing with potential impact on wearable health monitoring, clinical diagnostics, and broader biomedical applications.
Authors: Binyao Guo, Zihan Lin, QiZhi He
Abstract: This study presents an end-to-end learning framework for data-driven modeling of path-dependent inelastic materials using neural operators. The framework is built on the premise that irreversible evolution of material responses, governed by hidden dynamics, can be inferred from observable data. We develop the History-Aware Neural Operator (HANO), an autoregressive model that predicts path-dependent material responses from short segments of recent strain-stress history without relying on hidden state variables, thereby overcoming self-consistency issues commonly encountered in recurrent neural network (RNN)-based models. Built on a Fourier-based neural operator backbone, HANO enables discretization-invariant learning. To enhance its ability to capture both global loading patterns and critical local path dependencies, we embed a hierarchical self-attention mechanism that facilitates multiscale feature extraction. Beyond ensuring self-consistency, HANO mitigates sensitivity to initial hidden states, a commonly overlooked issue that can lead to instability in recurrent models when applied to generalized loading paths. By modeling stress-strain evolution as a continuous operator rather than relying on fixed input-output mappings, HANO naturally accommodates varying path discretizations and exhibits robust performance under complex conditions, including irregular sampling, multi-cycle loading, noisy data, and pre-stressed states. We evaluate HANO on two benchmark problems: elastoplasticity with hardening and progressive anisotropic damage in brittle solids. Results show that HANO consistently outperforms baseline models in predictive accuracy, generalization, and robustness. With its demonstrated capabilities, HANO provides an effective data-driven surrogate for simulating inelastic materials and is well-suited for integration with classical numerical solvers.
Authors: Yu-Yang Qian, Yuan-Ze Xu, Zhen-Yu Zhang, Peng Zhao, Zhi-Hua Zhou
Abstract: Many real-world applications collect data in a streaming environment, where learning tasks are encountered sequentially. This necessitates continual learning (CL) to update models online, enabling adaptation to new tasks while preserving past knowledge to prevent catastrophic forgetting. Nowadays, with the flourish of large pre-trained models (LPMs), efficiency has become increasingly critical for CL, due to their substantial computational demands and growing parameter sizes. In this paper, we introduce TreeLoRA (K-D Tree of Low-Rank Adapters), a novel approach that constructs layer-wise adapters by leveraging hierarchical gradient similarity to enable efficient CL, particularly for LPMs. To reduce the computational burden of task similarity estimation, we employ bandit techniques to develop an algorithm based on lower confidence bounds to efficiently explore the task structure. Furthermore, we use sparse gradient updates to facilitate parameter optimization, making the approach better suited for LPMs. Theoretical analysis is provided to justify the rationale behind our approach, and experiments on both vision transformers (ViTs) and large language models (LLMs) demonstrate the effectiveness and efficiency of our approach across various domains, including vision and natural language processing tasks.
Authors: Penguin Huang, Chhavi Yadav, Ruihan Wu, Kamalika Chaudhuri
Abstract: Large language models (LLMs) are increasingly fine-tuned on domain-specific datasets to support applications in fields such as healthcare, finance, and law. These fine-tuning datasets often have sensitive and confidential dataset-level properties -- such as patient demographics or disease prevalence -- that are not intended to be revealed. While prior work has studied property inference attacks on discriminative models (e.g., image classification models) and generative models (e.g., GANs for image data), it remains unclear if such attacks transfer to LLMs. In this work, we introduce PropInfer, a benchmark task for evaluating property inference in LLMs under two fine-tuning paradigms: question-answering and chat-completion. Built on the ChatDoctor dataset, our benchmark includes a range of property types and task configurations. We further propose two tailored attacks: a prompt-based generation attack and a shadow-model attack leveraging word frequency signals. Empirical evaluations across multiple pretrained LLMs show the success of our attacks, revealing a previously unrecognized vulnerability in LLMs.
Authors: Jikai Jin, Vasilis Syrgkanis, Sham Kakade, Hanlin Zhang
Abstract: Faithful evaluation of language model capabilities is crucial for deriving actionable insights that can inform model development. However, rigorous causal evaluations in this domain face significant methodological challenges, including complex confounding effects and prohibitive computational costs associated with extensive retraining. To tackle these challenges, we propose a causal representation learning framework wherein observed benchmark performance is modeled as a linear transformation of a few latent capability factors. Crucially, these latent factors are identified as causally interrelated after appropriately controlling for the base model as a common confounder. Applying this approach to a comprehensive dataset encompassing over 1500 models evaluated across six benchmarks from the Open LLM Leaderboard, we identify a concise three-node linear causal structure that reliably explains the observed performance variations. Further interpretation of this causal structure provides substantial scientific insights beyond simple numerical rankings: specifically, we reveal a clear causal direction starting from general problem-solving capabilities, advancing through instruction-following proficiency, and culminating in mathematical reasoning ability. Our results underscore the essential role of carefully controlling base model variations during evaluation, a step critical to accurately uncovering the underlying causal relationships among latent model capabilities.
Authors: Yuhang Chen, Zhen Tan, Tianlong Chen
Abstract: Reward Models (RMs), vital for large model alignment, are underexplored for complex embodied tasks like Embodied Question Answering (EQA) where nuanced evaluation of agents' spatial, temporal, and logical understanding is critical yet not considered by generic approaches. We introduce EQA-RM, a novel generative multimodal reward model specifically architected for EQA, trained via our innovative Contrastive Group Relative Policy Optimization (C-GRPO) strategy to learn fine-grained behavioral distinctions. The generative nature of EQA-RM provides interpretable, structured reward feedback (beyond simple scalars), uniquely enabling test-time scaling to dynamically adjust evaluation granularity, from concise scores to detailed critiques of reasoning and grounding, at inference without retraining. Concurrently, we introduce EQARewardBench, a new benchmark built on OpenEQA for standardized EQA reward model assessment. Demonstrating high sample efficiency, EQA-RM (fine-tuning Qwen2-VL-2B-Instruct) achieves 61.9\% accuracy on EQA-RM-Bench with only 700 samples, outperforming strong proprietary baselines, including Gemini-2.5-Flash, GPT-4o, Claude-3.5-Haiku, and open-sourced state-of-the-art models such as RoVRM and VisualPRM. The code and dataset can be found here https://github.com/UNITES-Lab/EQA-RM.
Authors: Tzu-Heng Huang, Harit Vishwakarma, Frederic Sala
Abstract: Large language models (LLMs) are widely used to evaluate the quality of LLM generations and responses, but this leads to significant challenges: high API costs, uncertain reliability, inflexible pipelines, and inherent biases. To address these, we introduce PAJAMA (Program-As-a-Judge for Automated Model Assessment), a new alternative that uses LLMs to synthesize executable judging programs instead of directly scoring responses. These synthesized programs can be stored and run locally, costing orders of magnitude less while providing interpretable, and auditable judging logic that can be easily adapted. Program-based judges mitigate biases, improving judgment consistency by 15.83% and reducing biased responses by 23.7% on average compared to a Qwen2.5-14B-based LLM-as-a-judge. When program judgments are distilled into a model, PAJAMA outperforms LLM-as-a-judge on the challenging CHAT-HARD subset of RewardBench, outperforming metrics by 2.19% on Prometheus and 8.67% on the JudgeLM dataset, all at three orders of magnitude lower cost.
Authors: Bryan Shaddy, Brianna Binder, Agnimitra Dasgupta, Haitong Qin, James Haley, Angel Farguell, Kyle Hilburn, Derek V. Mallia, Adam Kochanski, Jan Mandel, Assad Oberai
Abstract: Increasing wildfire occurrence has spurred growing interest in wildfire spread prediction. However, even the most complex wildfire models diverge from observed progression during multi-day simulations, motivating need for data assimilation. A useful approach to assimilating measurement data into complex coupled atmosphere-wildfire models is to estimate wildfire progression from measurements and use this progression to develop a matching atmospheric state. In this study, an approach is developed for estimating fire progression from VIIRS active fire measurements, GOES-derived ignition times, and terrain height data. A conditional Generative Adversarial Network is trained with simulations of historic wildfires from the atmosphere-wildfire model WRF-SFIRE, thus allowing incorporation of WRF-SFIRE physics into estimates. Fire progression is succinctly represented by fire arrival time, and measurements for training are obtained by applying an approximate observation operator to WRF-SFIRE solutions, eliminating need for satellite data during training. The model is trained on tuples of fire arrival times, measurements, and terrain, and once trained leverages measurements of real fires and corresponding terrain data to generate samples of fire arrival times. The approach is validated on five Pacific US wildfires, with results compared against high-resolution perimeters measured via aircraft, finding an average Sorensen-Dice coefficient of 0.81. The influence of terrain height on the arrival time inference is also evaluated and it is observed that terrain has minimal influence when the inference is conditioned on satellite measurements.
Authors: Ching Chang, Jeehyun Hwang, Yidan Shi, Haixin Wang, Wen-Chih Peng, Tien-Fu Chen, Wei Wang
Abstract: Time series data in real-world applications such as healthcare, climate modeling, and finance are often irregular, multimodal, and messy, with varying sampling rates, asynchronous modalities, and pervasive missingness. However, existing benchmarks typically assume clean, regularly sampled, unimodal data, creating a significant gap between research and real-world deployment. We introduce Time-IMM, a dataset specifically designed to capture cause-driven irregularity in multimodal multivariate time series. Time-IMM represents nine distinct types of time series irregularity, categorized into trigger-based, constraint-based, and artifact-based mechanisms. Complementing the dataset, we introduce IMM-TSF, a benchmark library for forecasting on irregular multimodal time series, enabling asynchronous integration and realistic evaluation. IMM-TSF includes specialized fusion modules, including a timestamp-to-text fusion module and a multimodality fusion module, which support both recency-aware averaging and attention-based integration strategies. Empirical results demonstrate that explicitly modeling multimodality on irregular time series data leads to substantial gains in forecasting performance. Time-IMM and IMM-TSF provide a foundation for advancing time series analysis under real-world conditions. The dataset is publicly available at https://www.kaggle.com/datasets/blacksnail789521/time-imm/data, and the benchmark library can be accessed at https://anonymous.4open.science/r/IMMTSF_NeurIPS2025.
URLs: https://www.kaggle.com/datasets/blacksnail789521/time-imm/data,, https://anonymous.4open.science/r/IMMTSF_NeurIPS2025.
Authors: Weiying Zhao, Aleksei Unagaev, Natalia Efremova
Abstract: Soil organic carbon (SOC) monitoring often relies on selecting representative field sampling locations based on environmental covariates. We propose a novel hybrid methodology that integrates spectral clustering - an unsupervised machine learning technique with conditioned Latin hypercube sampling (cLHS) to enhance the representativeness of SOC sampling. In our approach, spectral clustering partitions the study area into $K$ homogeneous zones using multivariate covariate data, and cLHS is then applied within each zone to select sampling locations that collectively capture the full diversity of environmental conditions. This hybrid spectral-cLHS method ensures that even minor but important environmental clusters are sampled, addressing a key limitation of vanilla cLHS which can overlook such areas. We demonstrate on a real SOC mapping dataset that spectral-cLHS provides more uniform coverage of covariate feature space and spatial heterogeneity than standard cLHS. This improved sampling design has the potential to yield more accurate SOC predictions by providing better-balanced training data for machine learning models.
Authors: Nart Gashi, Panagiotis Kakosimos, George Papafotiou
Abstract: Kolmogorov-Arnold Networks (KANs) are emerging as a powerful framework for interpretable and efficient system identification in dynamic systems. By leveraging the Kolmogorov-Arnold representation theorem, KANs enable function approximation through learnable activation functions, offering improved scalability, accuracy, and interpretability compared to traditional neural networks. This paper investigates the application of KANs to model and analyze the dynamics of a buck converter system, focusing on state-space parameter estimation along with discovering the system equations. Using simulation data, the methodology involves approximating state derivatives with KANs, constructing interpretable state-space representations, and validating these models through numerical experiments. The results demonstrate the ability of KANs to accurately identify system dynamics, verify model consistency, and detect parameter changes, providing valuable insights into their applicability for system identification in modern industrial systems.
Authors: Zhaode Wang, Jingbang Yang, Xinyu Qian, Shiwen Xing, Xiaotang Jiang, Chengfei Lv, Shengyu Zhang
Abstract: Large language models (LLMs) have demonstrated exceptional performance across a variety of tasks. However, their substantial scale leads to significant computational resource consumption during inference, resulting in high costs. Consequently, edge device inference presents a promising solution. The primary challenges of edge inference include memory usage and inference speed. This paper introduces MNN-LLM, a framework specifically designed to accelerate the deployment of large language models on mobile devices. MNN-LLM addresses the runtime characteristics of LLMs through model quantization and DRAM-Flash hybrid storage, effectively reducing memory usage. It rearranges weights and inputs based on mobile CPU instruction sets and GPU characteristics while employing strategies such as multicore load balancing, mixed-precision floating-point operations, and geometric computations to enhance performance. Notably, MNN-LLM achieves up to a 8.6x speed increase compared to current mainstream LLM-specific frameworks.
Authors: Fran\c{c}ois Cornet, Grigory Bartosh, Mikkel N. Schmidt, Christian A. Naesseth
Abstract: We introduce Equivariant Neural Diffusion (END), a novel diffusion model for molecule generation in 3D that is equivariant to Euclidean transformations. Compared to current state-of-the-art equivariant diffusion models, the key innovation in END lies in its learnable forward process for enhanced generative modelling. Rather than pre-specified, the forward process is parameterized through a time- and data-dependent transformation that is equivariant to rigid transformations. Through a series of experiments on standard molecule generation benchmarks, we demonstrate the competitive performance of END compared to several strong baselines for both unconditional and conditional generation.
Authors: Vasilis Michalakopoulos, Christoforos Menos-Aikateriniadis, Elissaios Sarmas, Antonis Zakynthinos, Pavlos S. Georgilakis, Dimitris Askounis
Abstract: This study investigates the performance of machine learning models in forecasting electricity Day-Ahead Market (DAM) prices using short historical training windows, with a focus on detecting seasonal trends and price spikes. We evaluate four models, namely LSTM with Feed Forward Error Correction (FFEC), XGBoost, LightGBM, and CatBoost, across three European energy markets (Greece, Belgium, Ireland) using feature sets derived from ENTSO-E forecast data. Training window lengths range from 7 to 90 days, allowing assessment of model adaptability under constrained data availability. Results indicate that LightGBM consistently achieves the highest forecasting accuracy and robustness, particularly with 45 and 60 day training windows, which balance temporal relevance and learning depth. Furthermore, LightGBM demonstrates superior detection of seasonal effects and peak price events compared to LSTM and other boosting models. These findings suggest that short-window training approaches, combined with boosting methods, can effectively support DAM forecasting in volatile, data-scarce environments.
Authors: Pascal Plettenberg, Andr\'e Alcalde, Bernhard Sick, Josephine M. Thomas
Abstract: The design and optimization of Printed Circuit Board (PCB) schematics is crucial for the development of high-quality electronic devices. Thereby, an important task is to optimize drafts by adding components that improve the robustness and reliability of the circuit, e.g., pull-up resistors or decoupling capacitors. Since there is a shortage of skilled engineers and manual optimizations are very time-consuming, these best practices are often neglected. However, this typically leads to higher costs for troubleshooting in later development stages as well as shortened product life cycles, resulting in an increased amount of electronic waste that is difficult to recycle. Here, we present an approach for automating the addition of new components into PCB schematics by representing them as bipartite graphs and utilizing a node pair prediction model based on Graph Neural Networks (GNNs). We apply our approach to three highly relevant PCB design optimization tasks and compare the performance of several popular GNN architectures on real-world datasets labeled by human experts. We show that GNNs can solve these problems with high accuracy and demonstrate that our approach offers the potential to automate PCB design optimizations in a time- and cost-efficient manner.
Authors: Antonio Ferrara, Francesco Cozzi, Alan Perotti, Andr\'e Panisson, Francesco Bonchi
Abstract: Determining whether an algorithmic decision-making system discriminates against a specific demographic typically involves comparing a single point estimate of a fairness metric against a predefined threshold. This practice is statistically brittle: it ignores sampling error and treats small demographic subgroups the same as large ones. The problem intensifies in intersectional analyses, where multiple sensitive attributes are considered jointly, giving rise to a larger number of smaller groups. As these groups become more granular, the data representing them becomes too sparse for reliable estimation, and fairness metrics yield excessively wide confidence intervals, precluding meaningful conclusions about potential unfair treatments. In this paper, we introduce a unified, size-adaptive, hypothesis-testing framework that turns fairness assessment into an evidence-based statistical decision. Our contribution is twofold. (i) For sufficiently large subgroups, we prove a Central-Limit result for the statistical parity difference, leading to analytic confidence intervals and a Wald test whose type-I (false positive) error is guaranteed at level $\alpha$. (ii) For the long tail of small intersectional groups, we derive a fully Bayesian Dirichlet-multinomial estimator; Monte-Carlo credible intervals are calibrated for any sample size and naturally converge to Wald intervals as more data becomes available. We validate our approach empirically on benchmark datasets, demonstrating how our tests provide interpretable, statistically rigorous decisions under varying degrees of data availability and intersectionality.
Authors: Yu-Jie Zhang, Peng Zhao, Masashi Sugiyama
Abstract: Non-stationary online learning has drawn much attention in recent years. Despite considerable progress, dynamic regret minimization has primarily focused on convex functions, leaving the functions with stronger curvature (e.g., squared or logistic loss) underexplored. In this work, we address this gap by showing that the regret can be substantially improved by leveraging the concept of mixability, a property that generalizes exp-concavity to effectively capture loss curvature. Let $d$ denote the dimensionality and $P_T$ the path length of comparators that reflects the environmental non-stationarity. We demonstrate that an exponential-weight method with fixed-share updates achieves an $\mathcal{O}(d T^{1/3} P_T^{2/3} \log T)$ dynamic regret for mixable losses, improving upon the best-known $\mathcal{O}(d^{10/3} T^{1/3} P_T^{2/3} \log T)$ result (Baby and Wang, 2021) in $d$. More importantly, this improvement arises from a simple yet powerful analytical framework that exploits the mixability, which avoids the Karush-Kuhn-Tucker-based analysis required by existing work.
Authors: Reza Karbasi, Masoud Rahimi, Abdol-Hossein Vahabie, Hadi Moradi
Abstract: This paper addresses the persistent challenge of accurately digitizing paper-based electrocardiogram (ECG) recordings, with a particular focus on robustly handling single leads compromised by signal overlaps-a common yet under-addressed issue in existing methodologies. We propose a two-stage pipeline designed to overcome this limitation. The first stage employs a U-Net based segmentation network, trained on a dataset enriched with overlapping signals and fortified with custom data augmentations, to accurately isolate the primary ECG trace. The subsequent stage converts this refined binary mask into a time-series signal using established digitization techniques, enhanced by an adaptive grid detection module for improved versatility across different ECG formats and scales. Our experimental results demonstrate the efficacy of our approach. The U-Net architecture achieves an IoU of 0.87 for the fine-grained segmentation task. Crucially, our proposed digitization method yields superior performance compared to a well-established baseline technique across both non-overlapping and challenging overlapping ECG samples. For non-overlapping signals, our method achieved a Mean Squared Error (MSE) of 0.0010 and a Pearson Correlation Coefficient (rho) of 0.9644, compared to 0.0015 and 0.9366, respectively, for the baseline. On samples with signal overlap, our method achieved an MSE of 0.0029 and a rho of 0.9641, significantly improving upon the baseline's 0.0178 and 0.8676. This work demonstrates an effective strategy to significantly enhance digitization accuracy, especially in the presence of signal overlaps, thereby laying a strong foundation for the reliable conversion of analog ECG records into analyzable digital data for contemporary research and clinical applications. The implementation is publicly available at this GitHub repository: https://github.com/masoudrahimi39/ECG-code.
Authors: Thu Ha Phi, Alexandre Hippert-Ferrer, Florent Bouchard, Arnaud Breloy
Abstract: This paper addresses the problem of learning an undirected graph from data gathered at each nodes. Within the graph signal processing framework, the topology of such graph can be linked to the support of the conditional correlation matrix of the data. The corresponding graph learning problem then scales to the squares of the number of variables (nodes), which is usually problematic at large dimension. To tackle this issue, we propose a graph learning framework that leverages a low-rank factorization of the conditional correlation matrix. In order to solve for the resulting optimization problems, we derive tools required to apply Riemannian optimization techniques for this particular structure. The proposal is then particularized to a low-rank constrained counterpart of the GLasso algorithm, i.e., the penalized maximum likelihood estimation of a Gaussian graphical model. Experiments on synthetic and real data evidence that a very efficient dimension-versus-performance trade-off can be achieved with this approach.
Authors: Yucheng Yang, Tianyi Zhou, Qiang He, Lei Han, Mykola Pechenizkiy, Meng Fang
Abstract: Unsupervised reinforcement learning (URL) aims to learn general skills for unseen downstream tasks. Mutual Information Skill Learning (MISL) addresses URL by maximizing the mutual information between states and skills but lacks sufficient theoretical analysis, e.g., how well its learned skills can initialize a downstream task's policy. Our new theoretical analysis in this paper shows that the diversity and separability of learned skills are fundamentally critical to downstream task adaptation but MISL does not necessarily guarantee these properties. To complement MISL, we propose a novel disentanglement metric LSEPIN. Moreover, we build an information-geometric connection between LSEPIN and downstream task adaptation cost. For better geometric properties, we investigate a new strategy that replaces the KL divergence in information geometry with Wasserstein distance. We extend the geometric analysis to it, which leads to a novel skill-learning objective WSEP. It is theoretically justified to be helpful to downstream task adaptation and it is capable of discovering more initial policies for downstream tasks than MISL. We finally propose another Wasserstein distance-based algorithm PWSEP that can theoretically discover all optimal initial policies.
Authors: Yucong Luo, Yitong Zhou, Mingyue Cheng, Jiahao Wang, Daoyu Wang, Tingyue Pan, Jintao Zhang
Abstract: To advance time series forecasting (TSF), various methods have been proposed to improve prediction accuracy, evolving from statistical techniques to data-driven deep learning architectures. Despite their effectiveness, most existing methods still adhere to a fast thinking paradigm-relying on extracting historical patterns and mapping them to future values as their core modeling philosophy, lacking an explicit thinking process that incorporates intermediate time series reasoning. Meanwhile, emerging slow-thinking LLMs (e.g., OpenAI-o1) have shown remarkable multi-step reasoning capabilities, offering an alternative way to overcome these issues. However, prompt engineering alone presents several limitations - including high computational cost, privacy risks, and limited capacity for in-depth domain-specific time series reasoning. To address these limitations, a more promising approach is to train LLMs to develop slow thinking capabilities and acquire strong time series reasoning skills. For this purpose, we propose Time-R1, a two-stage reinforcement fine-tuning framework designed to enhance multi-step reasoning ability of LLMs for time series forecasting. Specifically, the first stage conducts supervised fine-tuning for warmup adaptation, while the second stage employs reinforcement learning to improve the model's generalization ability. Particularly, we design a fine-grained multi-objective reward specifically for time series forecasting, and then introduce GRIP (group-based relative importance for policy optimization), which leverages non-uniform sampling to further encourage and optimize the model's exploration of effective reasoning paths. Experiments demonstrate that Time-R1 significantly improves forecast performance across diverse datasets.
Authors: Alexander Lobashev, Dmitry Guskov, Maria Larchenko, Mikhail Tamm
Abstract: This paper presents a novel method for analyzing the latent space geometry of generative models, including statistical physics models and diffusion models, by reconstructing the Fisher information metric. The method approximates the posterior distribution of latent variables given generated samples and uses this to learn the log-partition function, which defines the Fisher metric for exponential families. Theoretical convergence guarantees are provided, and the method is validated on the Ising and TASEP models, outperforming existing baselines in reconstructing thermodynamic quantities. Applied to diffusion models, the method reveals a fractal structure of phase transitions in the latent space, characterized by abrupt changes in the Fisher metric. We demonstrate that while geodesic interpolations are approximately linear within individual phases, this linearity breaks down at phase boundaries, where the diffusion model exhibits a divergent Lipschitz constant with respect to the latent space. These findings provide new insights into the complex structure of diffusion model latent spaces and their connection to phenomena like phase transitions. Our source code is available at https://github.com/alobashev/hessian-geometry-of-diffusion-models.
URLs: https://github.com/alobashev/hessian-geometry-of-diffusion-models.
Authors: Lang Yin, Debangshu Banerjee, Gagandeep Singh
Abstract: Chain of Thought (CoT) has been applied to various large language models (LLMs) and proven to be effective in improving the quality of outputs. In recent studies, transformers are proven to have absolute upper bounds in terms of expressive power, and consequently, they cannot solve many computationally difficult problems. However, empowered by CoT, transformers are proven to be able to solve some difficult problems effectively, such as the $k$-parity problem. Nevertheless, those works rely on two imperative assumptions: (1) identical training and testing distribution, and (2) corruption-free training data with correct reasoning steps. However, in the real world, these assumptions do not always hold. Although the risks of data shifts have caught attention, our work is the first to rigorously study the exact harm caused by such shifts to the best of our knowledge. Focusing on the $k$-parity problem, in this work we investigate the joint impact of two types of data shifts: the distribution shifts and data poisoning, on the quality of trained models obtained by a well-established CoT decomposition. In addition to revealing a surprising phenomenon that CoT leads to worse performance on learning parity than directly generating the prediction, our technical results also give a rigorous and comprehensive explanation of the mechanistic reasons of such impact.
Authors: Igor Urbanik, Pawe{\l} Gajewski
Abstract: Continual learning poses a fundamental challenge for neural systems, which often suffer from catastrophic forgetting when exposed to sequential tasks. Self-Organizing Maps (SOMs), despite their interpretability and efficiency, are not immune to this issue. In this paper, we introduce Saturation Self-Organizing Maps (SatSOM)-an extension of SOMs designed to improve knowledge retention in continual learning scenarios. SatSOM incorporates a novel saturation mechanism that gradually reduces the learning rate and neighborhood radius of neurons as they accumulate information. This effectively freezes well-trained neurons and redirects learning to underutilized areas of the map.
Authors: Floris Holstege, Shauli Ravfogel, Bram Wouters
Abstract: Modern neural networks often encode unwanted concepts alongside task-relevant information, leading to fairness and interpretability concerns. Existing post-hoc approaches can remove undesired concepts but often degrade useful signals. We introduce SPLICE-Simultaneous Projection for LInear concept removal and Covariance prEservation-which eliminates sensitive concepts from representations while exactly preserving their covariance with a target label. SPLICE achieves this via an oblique projection that "splices out" the unwanted direction yet protects important label correlations. Theoretically, it is the unique solution that removes linear concept predictability and maintains target covariance with minimal embedding distortion. Empirically, SPLICE outperforms baselines on benchmarks such as Bias in Bios and Winobias, removing protected attributes while minimally damaging main-task information.
Authors: Marco Spinaci, Marek Polewczyk, Maximilian Schambach, Sam Thelin
Abstract: Tabular in-context learning (ICL) has recently achieved state-of-the-art (SOTA) performance on several tabular prediction tasks. Previously restricted to classification problems on small tables, recent advances such as TabPFN and TabICL have extended its use to larger datasets. While being architecturally efficient and well-adapted to tabular data structures, current table-native ICL architectures, being trained exclusively on synthetic data, do not fully leverage the rich semantics and world knowledge contained in real-world tabular data. On another end of this spectrum, tabular ICL models based on pretrained large language models such as TabuLa-8B integrate deep semantic understanding and world knowledge but are only able to make use of a small amount of context due to inherent architectural limitations. With the aim to combine the best of both these worlds, we introduce ConTextTab, integrating semantic understanding and alignment into a table-native ICL framework. By employing specialized embeddings for different data modalities and by training on large-scale real-world tabular data, our model is competitive with SOTA across a broad set of benchmarks while setting a new standard on the semantically rich CARTE benchmark.
Authors: Sai Prasanna Teja Reddy Bogireddy, Abrar Majeedi, Viswanatha Reddy Gajjala, Zhuoyan Xu, Siddhant Rai, Vaishnav Potlapalli
Abstract: Automated question answering (QA) over electronic health records (EHRs) can bridge critical information gaps for clinicians and patients, yet it demands both precise evidence retrieval and faithful answer generation under limited supervision. In this work, we present Neural, the runner-up in the BioNLP 2025 ArchEHR-QA shared task on evidence-grounded clinical QA. Our proposed method decouples the task into (1) sentence-level evidence identification and (2) answer synthesis with explicit citations. For each stage, we automatically explore the prompt space with DSPy's MIPROv2 optimizer, jointly tuning instructions and few-shot demonstrations on the development set. A self-consistency voting scheme further improves evidence recall without sacrificing precision. On the hidden test set, our method attains an overall score of 51.5, placing second stage while outperforming standard zero-shot and few-shot prompting by over 20 and 10 points, respectively. These results indicate that data-driven prompt optimization is a cost-effective alternative to model fine-tuning for high-stakes clinical QA, advancing the reliability of AI assistants in healthcare.
Authors: Ferran Alet, Ilan Price, Andrew El-Kadi, Dominic Masters, Stratis Markou, Tom R. Andersson, Jacklynn Stott, Remi Lam, Matthew Willson, Alvaro Sanchez-Gonzalez, Peter Battaglia
Abstract: Machine learning (ML)-based weather models have rapidly risen to prominence due to their greater accuracy and speed than traditional forecasts based on numerical weather prediction (NWP), recently outperforming traditional ensembles in global probabilistic weather forecasting. This paper presents FGN, a simple, scalable and flexible modeling approach which significantly outperforms the current state-of-the-art models. FGN generates ensembles via learned model-perturbations with an ensemble of appropriately constrained models. It is trained directly to minimize the continuous rank probability score (CRPS) of per-location forecasts. It produces state-of-the-art ensemble forecasts as measured by a range of deterministic and probabilistic metrics, makes skillful ensemble tropical cyclone track predictions, and captures joint spatial structure despite being trained only on marginals.
Authors: Yufei Tao
Abstract: In monotone classification, the input is a multi-set $P$ of points in $\mathbb{R}^d$, each associated with a hidden label from $\{-1, 1\}$. The goal is to identify a monotone function $h$, which acts as a classifier, mapping from $\mathbb{R}^d$ to $\{-1, 1\}$ with a small {\em error}, measured as the number of points $p \in P$ whose labels differ from the function values $h(p)$. The cost of an algorithm is defined as the number of points having their labels revealed. This article presents the first study on the lowest cost required to find a monotone classifier whose error is at most $(1 + \epsilon) \cdot k^*$ where $\epsilon \ge 0$ and $k^*$ is the minimum error achieved by an optimal monotone classifier -- in other words, the error is allowed to exceed the optimal by at most a relative factor. Nearly matching upper and lower bounds are presented for the full range of $\epsilon$. All previous work on the problem can only achieve an error higher than the optimal by an absolute factor.
Authors: Benjamin Hoover, Zhaoyang Shi, Krishnakumar Balasubramanian, Dmitry Krotov, Parikshit Ram
Abstract: We propose a novel energy function for Dense Associative Memory (DenseAM) networks, the log-sum-ReLU (LSR), inspired by optimal kernel density estimation. Unlike the common log-sum-exponential (LSE) function, LSR is based on the Epanechnikov kernel and enables exact memory retrieval with exponential capacity without requiring exponential separation functions. Moreover, it introduces abundant additional \emph{emergent} local minima while preserving perfect pattern recovery -- a characteristic previously unseen in DenseAM literature. Empirical results show that LSR energy has significantly more local minima (memories) that have comparable log-likelihood to LSE-based models. Analysis of LSR's emergent memories on image datasets reveals a degree of creativity and novelty, hinting at this method's potential for both large-scale memory storage and generative tasks.
Authors: Alex McKenzie, Urja Pawar, Phil Blandfort, William Bankes, David Krueger, Ekdeep Singh Lubana, Dmitrii Krasheninnikov
Abstract: Monitoring is an important aspect of safely deploying Large Language Models (LLMs). This paper examines activation probes for detecting "high-stakes" interactions -- where the text indicates that the interaction might lead to significant harm -- as a critical, yet underexplored, target for such monitoring. We evaluate several probe architectures trained on synthetic data, and find them to exhibit robust generalization to diverse, out-of-distribution, real-world data. Probes' performance is comparable to that of prompted or finetuned medium-sized LLM monitors, while offering computational savings of six orders-of-magnitude. Our experiments also highlight the potential of building resource-aware hierarchical monitoring systems, where probes serve as an efficient initial filter and flag cases for more expensive downstream analysis. We release our novel synthetic dataset and codebase to encourage further study.
Authors: Ravishka Rathnasuriya, Tingxi Li, Zexin Xu, Zihe Song, Mirazul Haque, Simin Chen, Wei Yang
Abstract: Deep Learning Systems (DLSs) are increasingly deployed in real-time applications, including those in resourceconstrained environments such as mobile and IoT devices. To address efficiency challenges, Dynamic Deep Learning Systems (DDLSs) adapt inference computation based on input complexity, reducing overhead. While this dynamic behavior improves efficiency, such behavior introduces new attack surfaces. In particular, efficiency adversarial attacks exploit these dynamic mechanisms to degrade system performance. This paper systematically explores efficiency robustness of DDLSs, presenting the first comprehensive taxonomy of efficiency attacks. We categorize these attacks based on three dynamic behaviors: (i) attacks on dynamic computations per inference, (ii) attacks on dynamic inference iterations, and (iii) attacks on dynamic output production for downstream tasks. Through an in-depth evaluation, we analyze adversarial strategies that target DDLSs efficiency and identify key challenges in securing these systems. In addition, we investigate existing defense mechanisms, demonstrating their limitations against increasingly popular efficiency attacks and the necessity for novel mitigation strategies to secure future adaptive DDLSs.
Authors: Nudrat Fariha, Md Nazmuddin Moin Khan, Md Iqbal Hossain, Syed Ali Reza, Joy Chakra Bortty, Kazi Sharmin Sultana, Md Shadidur Islam Jawad, Saniah Safat, Md Abdul Ahad, Maksuda Begum
Abstract: The rise of digital payments has accelerated the need for intelligent and scalable systems to detect fraud. This research presents an end-to-end, feature-rich machine learning framework for detecting credit card transaction anomalies and fraud using real-world data. The study begins by merging transactional, cardholder, merchant, and merchant category datasets from a relational database to create a unified analytical view. Through the feature engineering process, we extract behavioural signals such as average spending, deviation from historical patterns, transaction timing irregularities, and category frequency metrics. These features are enriched with temporal markers such as hour, day of week, and weekend indicators to expose all latent patterns that indicate fraudulent behaviours. Exploratory data analysis reveals contextual transaction trends across all the dataset features. Using the transactional data, we train and evaluate a range of unsupervised models: Isolation Forest, One Class SVM, and a deep autoencoder trained to reconstruct normal behavior. These models flag the top 1% of reconstruction errors as outliers. PCA visualizations illustrate each models ability to separate anomalies into a two-dimensional latent space. We further segment the transaction landscape using K-Means clustering and DBSCAN to identify dense clusters of normal activity and isolate sparse, suspicious regions.
Authors: Pierre-Fran\c{c}ois Massiani, Alexander von Rohr, Lukas Haverbeck, Sebastian Trimpe
Abstract: Despite the many recent advances in reinforcement learning (RL), the question of learning policies that robustly satisfy state constraints under unknown disturbances remains open. In this paper, we offer a new perspective on achieving robust safety by analyzing the interplay between two well-established techniques in model-free RL: entropy regularization, and constraints penalization. We reveal empirically that entropy regularization in constrained RL inherently biases learning toward maximizing the number of future viable actions, thereby promoting constraints satisfaction robust to action noise. Furthermore, we show that by relaxing strict safety constraints through penalties, the constrained RL problem can be approximated arbitrarily closely by an unconstrained one and thus solved using standard model-free RL. This reformulation preserves both safety and optimality while empirically improving resilience to disturbances. Our results indicate that the connection between entropy regularization and robustness is a promising avenue for further empirical and theoretical investigation, as it enables robust safety in RL through simple reward shaping.
Authors: Lucas Gnecco-Heredia, Benjamin Negrevergne, Yann Chevaleyre
Abstract: Finite mixtures of classifiers (a.k.a. randomized ensembles) have been proposed as a way to improve robustness against adversarial attacks. However, existing attacks have been shown to not suit this kind of classifier. In this paper, we discuss the problem of attacking a mixture in a principled way and introduce two desirable properties of attacks based on a geometrical analysis of the problem (effectiveness and maximality). We then show that existing attacks do not meet both of these properties. Finally, we introduce a new attack called {\em lattice climber attack} with theoretical guarantees in the binary linear setting, and demonstrate its performance by conducting experiments on synthetic and real datasets.
Authors: Subham Sekhar Sahoo, Justin Deschenaux, Aaron Gokaslan, Guanghan Wang, Justin Chiu, Volodymyr Kuleshov
Abstract: Uniform-state discrete diffusion models hold the promise of fast text generation due to their inherent ability to self-correct. However, they are typically outperformed by autoregressive models and masked diffusion models. In this work, we narrow this performance gap by leveraging a key insight: Uniform-state diffusion processes naturally emerge from an underlying Gaussian diffusion. Our method, Duo, transfers powerful techniques from Gaussian diffusion to improve both training and sampling. First, we introduce a curriculum learning strategy guided by the Gaussian process, doubling training speed by reducing variance. Models trained with curriculum learning surpass autoregressive models in zero-shot perplexity on 3 of 7 benchmarks. Second, we present Discrete Consistency Distillation, which adapts consistency distillation from the continuous to the discrete setting. This algorithm unlocks few-step generation in diffusion language models by accelerating sampling by two orders of magnitude. We provide the code and model checkpoints on the project page: http://s-sahoo.github.io/duo
Authors: Jari Kolehmainen, Nikolay Blagoev, John Donaghy, O\u{g}uzhan Ersoy, Christopher Nies
Abstract: Training large language models is generally done via optimization methods on clusters containing tens of thousands of accelerators, communicating over a high-bandwidth interconnect. Scaling up these clusters is expensive and can become impractical, imposing limits on the size of models that can be trained. Several recent studies have proposed training methods that are less communication intensive, avoiding the need for a highly connected compute cluster. These state-of-the-art low communication training methods still employ a synchronization step for model parameters, which, when performed over all model replicas, can become costly on a low-bandwidth network. In this work, we propose a novel optimization method, NoLoCo, that does not explicitly synchronize all model parameters during training and, as a result, does not require any collective communication. NoLoCo implicitly synchronizes model weights via a novel variant of the Nesterov momentum optimizer by partially averaging model weights with a randomly selected other one. We provide both a theoretical convergence analysis for our proposed optimizer as well as empirical results from language model training. We benchmark NoLoCo on a wide range of accelerator counts and model sizes, between 125M to 6.8B parameters. Our method requires significantly less communication overhead than fully sharded data parallel training or even widely used low communication training method, DiLoCo. The synchronization step itself is estimated to be one magnitude faster than the all-reduce used in DiLoCo for few hundred accelerators training over the internet. We also do not have any global blocking communication that reduces accelerator idling time. Compared to DiLoCo, we also observe up to $4\%$ faster convergence rate with wide range of model sizes and accelerator counts.
Authors: Yuchen Ma, Dennis Frauen, Emil Javurek, Stefan Feuerriegel
Abstract: Prior-data fitted networks (PFNs) have recently been proposed as a promising way to train tabular foundation models. PFNs are transformers that are pre-trained on synthetic data generated from a prespecified prior distribution and that enable Bayesian inference through in-context learning. In this paper, we introduce CausalFM, a comprehensive framework for training PFN-based foundation models in various causal inference settings. First, we formalize the construction of Bayesian priors for causal inference based on structural causal models (SCMs) in a principled way and derive necessary criteria for the validity of such priors. Building on this, we propose a novel family of prior distributions using causality-inspired Bayesian neural networks that enable CausalFM to perform Bayesian causal inference in various settings, including back-door, front-door, and instrumental variable adjustment. Finally, we instantiate CausalFM and explicitly train a foundation model for estimating conditional average treatment effects (CATEs) using back-door adjustment. We show that CausalFM performs competitively for CATE estimation using various synthetic and semi-synthetic benchmarks. In sum, our framework can be used as a general recipe to train foundation models for various causal inference settings. In contrast to the current state-of-the-art in causal inference, CausalFM offers a novel paradigm with the potential to fundamentally change how practitioners perform causal inference in medicine, economics, and other disciplines.
Authors: Morris Yau, Sharut Gupta, Valerie Engelmayer, Kazuki Irie, Stefanie Jegelka, Jacob Andreas
Abstract: Modern neural sequence models are designed to meet the dual mandate of parallelizable training and fast sequential inference. Recent developments have given rise to various models, such as Gated Linear Attention (GLA) and Mamba, that achieve such ``sequential-parallel duality.'' This raises a natural question: can we characterize the full class of neural sequence models that support near-constant-time parallel evaluation and linear-time, constant-space sequential inference? We begin by describing a broad class of such models -- state space models -- as those whose state updates can be computed using the classic parallel prefix scan algorithm with a custom associative aggregation operator. We then define a more general class, Prefix-Scannable Models (PSMs), by relaxing the state aggregation operator to allow arbitrary (potentially non-associative) functions such as softmax attention. This generalization unifies many existing architectures, including element-wise RNNs (e.g., Mamba) and linear transformers (e.g., GLA, Mamba2, mLSTM), while also introducing new models with softmax-like operators that achieve O(1) amortized compute per token and log(N) memory for sequence length N. We empirically evaluate such models on illustrative small-scale language modeling and canonical synthetic tasks, including state tracking and associative recall. Empirically, we find that PSMs retain the expressivity of transformer-based architectures while matching the inference efficiency of state space models -- in some cases exhibiting better length generalization than either.
Authors: Adam Karvonen, Samuel Marks
Abstract: Large language models (LLMs) are increasingly deployed in high-stakes hiring applications, making decisions that directly impact people's careers and livelihoods. While prior studies suggest simple anti-bias prompts can eliminate demographic biases in controlled evaluations, we find these mitigations fail when realistic contextual details are introduced. We address these failures through internal bias mitigation: by identifying and neutralizing sensitive attribute directions within model activations, we achieve robust bias reduction across all tested scenarios. Across leading commercial (GPT-4o, Claude 4 Sonnet, Gemini 2.5 Flash) and open-source models (Gemma-2 27B, Gemma-3, Mistral-24B), we find that adding realistic context such as company names, culture descriptions from public careers pages, and selective hiring constraints (e.g.,``only accept candidates in the top 10\%") induces significant racial and gender biases (up to 12\% differences in interview rates). When these biases emerge, they consistently favor Black over White candidates and female over male candidates across all tested models and scenarios. Moreover, models can infer demographics and become biased from subtle cues like college affiliations, with these biases remaining invisible even when inspecting the model's chain-of-thought reasoning. To address these limitations, our internal bias mitigation identifies race and gender-correlated directions and applies affine concept editing at inference time. Despite using directions from a simple synthetic dataset, the intervention generalizes robustly, consistently reducing bias to very low levels (typically under 1\%, always below 2.5\%) while largely maintaining model performance. Our findings suggest that practitioners deploying LLMs for hiring should adopt more realistic evaluation methodologies and consider internal mitigation strategies for equitable outcomes.
Authors: Thanathai Lertpetchpun, Tiantian Feng, Dani Byrd, Shrikanth Narayanan
Abstract: Speech emotion recognition (SER) in naturalistic conditions presents a significant challenge for the speech processing community. Challenges include disagreement in labeling among annotators and imbalanced data distributions. This paper presents a reproducible framework that achieves superior (top 1) performance in the Emotion Recognition in Naturalistic Conditions Challenge (IS25-SER Challenge) - Task 2, evaluated on the MSP-Podcast dataset. Our system is designed to tackle the aforementioned challenges through multimodal learning, multi-task learning, and imbalanced data handling. Specifically, our best system is trained by adding text embeddings, predicting gender, and including ``Other'' (O) and ``No Agreement'' (X) samples in the training set. Our system's results secured both first and second places in the IS25-SER Challenge, and the top performance was achieved by a simple two-system ensemble.
Authors: Adam Zweiger, Jyothish Pari, Han Guo, Ekin Aky\"urek, Yoon Kim, Pulkit Agrawal
Abstract: Large language models (LLMs) are powerful but static; they lack mechanisms to adapt their weights in response to new tasks, knowledge, or examples. We introduce Self-Adapting LLMs (SEAL), a framework that enables LLMs to self-adapt by generating their own finetuning data and update directives. Given a new input, the model produces a self-edit-a generation that may restructure the information in different ways, specify optimization hyperparameters, or invoke tools for data augmentation and gradient-based updates. Through supervised finetuning (SFT), these self-edits result in persistent weight updates, enabling lasting adaptation. To train the model to produce effective self-edits, we use a reinforcement learning loop with the downstream performance of the updated model as the reward signal. Unlike prior approaches that rely on separate adaptation modules or auxiliary networks, SEAL directly uses the model's own generation to control its adaptation process. Experiments on knowledge incorporation and few-shot generalization show that SEAL is a promising step toward language models capable of self-directed adaptation. Our website and code is available at https://jyopari.github.io/posts/seal.
Authors: Evelyn Ma, Duo Zhou, Peizhi Niu, Huiting Zhou, Huan Zhang, Olgica Milenkovic, S. Rasoul Etesami
Abstract: Unlearning in large language models (LLMs) is becoming increasingly important due to regulatory compliance, copyright protection, and privacy concerns. However, a key challenge in LLM unlearning is unintended forgetting, where the removal of specific data inadvertently impairs the utility of the model and its retention of valuable, desired information. While prior work has primarily focused on architectural innovations, the influence of data-level factors on unlearning performance remains underexplored. As a result, existing methods often suffer from degraded retention when forgetting high-impact data. To address this, we propose GUARD-a novel framework for Guided Unlearning And Retention via Data attribution. At its core, GUARD introduces a lightweight proxy data attribution metric tailored for LLM unlearning, which quantifies the "alignment" between the forget and retain sets while remaining computationally efficient. Building on this, we design a novel unlearning objective that assigns adaptive, nonuniform unlearning weights to samples, inversely proportional to their proxy attribution scores. Through such a reallocation of unlearning power, GUARD mitigates unintended losses in retention. We provide rigorous theoretical guarantees that GUARD significantly enhances retention while maintaining forgetting metrics comparable to prior methods. Extensive experiments on the TOFU benchmark across multiple LLM architectures demonstrate that GUARD substantially improves utility preservation while ensuring effective unlearning. Notably, GUARD reduces utility sacrifice on the Retain Set by up to 194.92% in terms of Truth Ratio when forgetting 10% of the training data.
Authors: Boaz Lavon, Shahar Katz, Lior Wolf
Abstract: We present a novel approach to neural code generation that incorporates real-time execution signals into the language model generation process. While large language models (LLMs) have demonstrated impressive code generation capabilities, they typically do not utilize execution feedback during inference, a critical signal that human programmers regularly leverage. Our method, Execution-Guided Classifier-Free Guidance (EG-CFG), dynamically incorporates execution signals as the model generates code, providing line-by-line feedback that guides the generation process toward executable solutions. EG-CFG employs a multi-stage process: first, we conduct beam search to sample candidate program completions for each line; second, we extract execution signals by executing these candidates against test cases; and finally, we incorporate these signals into the prompt during generation. By maintaining consistent signals across tokens within the same line and refreshing signals at line boundaries, our approach provides coherent guidance while preserving syntactic structure. Moreover, the method naturally supports native parallelism at the task level in which multiple agents operate in parallel, exploring diverse reasoning paths and collectively generating a broad set of candidate solutions. Our experiments across diverse coding tasks demonstrate that EG-CFG significantly improves code generation performance compared to standard approaches, achieving state-of-the-art results across various levels of complexity, from foundational problems to challenging competitive programming tasks. Our code is available at: https://github.com/boazlavon/eg_cfg
Authors: Xing Han L\`u, Gaurav Kamath, Marius Mosbach, Siva Reddy
Abstract: Recent advancements in Large Language Models (LLMs) and multimodal counterparts have spurred significant interest in developing web agents -- AI systems capable of autonomously navigating and completing tasks within web environments. While holding tremendous promise for automating complex web interactions, current approaches face substantial challenges due to the fundamental mismatch between human-designed interfaces and LLM capabilities. Current methods struggle with the inherent complexity of web inputs, whether processing massive DOM trees, relying on screenshots augmented with additional information, or bypassing the user interface entirely through API interactions. This position paper advocates for a paradigm shift in web agent research: rather than forcing web agents to adapt to interfaces designed for humans, we should develop a new interaction paradigm specifically optimized for agentic capabilities. To this end, we introduce the concept of an Agentic Web Interface (AWI), an interface specifically designed for agents to navigate a website. We establish six guiding principles for AWI design, emphasizing safety, efficiency, and standardization, to account for the interests of all primary stakeholders. This reframing aims to overcome fundamental limitations of existing interfaces, paving the way for more efficient, reliable, and transparent web agent design, which will be a collaborative effort involving the broader ML community.
Authors: Aayush Karan, Kulin Shah, Sitan Chen
Abstract: There has been a flurry of activity around using pretrained diffusion models as informed data priors for solving inverse problems, and more generally around steering these models using reward models. Training-free methods like diffusion posterior sampling (DPS) and its many variants have offered flexible heuristic algorithms for these tasks, but when the reward is not informative enough, e.g., in hard inverse problems with low signal-to-noise ratio, these techniques veer off the data manifold, failing to produce realistic outputs. In this work, we devise a simple wrapper, ReGuidance, for boosting both the sample realism and reward achieved by these methods. Given a candidate solution $\hat{x}$ produced by an algorithm of the user's choice, we propose inverting the solution by running the unconditional probability flow ODE in reverse starting from $\hat{x}$, and then using the resulting latent as an initialization for DPS. We evaluate our wrapper on hard inverse problems like large box in-painting and super-resolution with high upscaling. Whereas state-of-the-art baselines visibly fail, we find that applying our wrapper on top of these baselines significantly boosts sample quality and measurement consistency. We complement these findings with theory proving that on certain multimodal data distributions, ReGuidance simultaneously boosts the reward and brings the candidate solution closer to the data manifold. To our knowledge, this constitutes the first rigorous algorithmic guarantee for DPS.
Authors: Zhaiming Shen, Alexander Hsu, Rongjie Lai, Wenjing Liao
Abstract: While in-context learning (ICL) has achieved remarkable success in natural language and vision domains, its theoretical understanding--particularly in the context of structured geometric data--remains unexplored. In this work, we initiate a theoretical study of ICL for regression of H\"older functions on manifolds. By establishing a novel connection between the attention mechanism and classical kernel methods, we derive generalization error bounds in terms of the prompt length and the number of training tasks. When a sufficient number of training tasks are observed, transformers give rise to the minimax regression rate of H\"older functions on manifolds, which scales exponentially with the intrinsic dimension of the manifold, rather than the ambient space dimension. Our result also characterizes how the generalization error scales with the number of training tasks, shedding light on the complexity of transformers as in-context algorithm learners. Our findings provide foundational insights into the role of geometry in ICL and novels tools to study ICL of nonlinear models.
Authors: Houyi Li, Wenzhen Zheng, Qiufeng Wang, Zhenyu Ding, Haoying Wang, Zili Wang, Shijie Xuyang, Ning Ding, Shuigeng Zhou, Xiangyu Zhang, Daxin Jiang
Abstract: Training Large Language Models (LLMs) is prohibitively expensive, creating a critical scaling gap where insights from small-scale experiments often fail to transfer to resource-intensive production systems, thereby hindering efficient innovation. To bridge this, we introduce Farseer, a novel and refined scaling law offering enhanced predictive accuracy across scales. By systematically constructing a model loss surface $L(N,D)$, Farseer achieves a significantly better fit to empirical data than prior laws (e.g., Chinchilla's law). Our methodology yields accurate, robust, and highly generalizable predictions, demonstrating excellent extrapolation capabilities, improving upon Chinchilla's law by reducing extrapolation error by 433\%. This allows for the reliable evaluation of competing training strategies across all $(N,D)$ settings, enabling conclusions from small-scale ablation studies to be confidently extrapolated to predict large-scale performance. Furthermore, Farseer provides new insights into optimal compute allocation, better reflecting the nuanced demands of modern LLM training. To validate our approach, we trained an extensive suite of approximately 1,000 LLMs across diverse scales and configurations, consuming roughly 3 million NVIDIA H100 GPU hours. We are comprehensively open-sourcing all models, data, results, and logs at https://github.com/Farseer-Scaling-Law/Farseer to foster further research.
Authors: Julius Berner, Miguel Liu-Schiaffini, Jean Kossaifi, Valentin Duruisseaux, Boris Bonev, Kamyar Azizzadenesheli, Anima Anandkumar
Abstract: A wide range of scientific problems, such as those described by continuous-time dynamical systems and partial differential equations (PDEs), are naturally formulated on function spaces. While function spaces are typically infinite-dimensional, deep learning has predominantly advanced through applications in computer vision and natural language processing that focus on mappings between finite-dimensional spaces. Such fundamental disparities in the nature of the data have limited neural networks from achieving a comparable level of success in scientific applications as seen in other fields. Neural operators are a principled way to generalize neural networks to mappings between function spaces, offering a pathway to replicate deep learning's transformative impact on scientific problems. For instance, neural operators can learn solution operators for entire classes of PDEs, e.g., physical systems with different boundary conditions, coefficient functions, and geometries. A key factor in deep learning's success has been the careful engineering of neural architectures through extensive empirical testing. Translating these neural architectures into neural operators allows operator learning to enjoy these same empirical optimizations. However, prior neural operator architectures have often been introduced as standalone models, not directly derived as extensions of existing neural network architectures. In this paper, we identify and distill the key principles for constructing practical implementations of mappings between infinite-dimensional function spaces. Using these principles, we propose a recipe for converting several popular neural architectures into neural operators with minimal modifications. This paper aims to guide practitioners through this process and details the steps to make neural operators work in practice. Our code can be found at https://github.com/neuraloperator/NNs-to-NOs
Authors: Sebastian Sanokowski, Lukas Gruber, Christoph Bartmann, Sepp Hochreiter, Sebastian Lehner
Abstract: Diffusion bridges are a promising class of deep-learning methods for sampling from unnormalized distributions. Recent works show that the Log Variance (LV) loss consistently outperforms the reverse Kullback-Leibler (rKL) loss when using the reparametrization trick to compute rKL-gradients. While the on-policy LV loss yields identical gradients to the rKL loss when combined with the log-derivative trick for diffusion samplers with non-learnable forward processes, this equivalence does not hold for diffusion bridges or when diffusion coefficients are learned. Based on this insight we argue that for diffusion bridges the LV loss does not represent an optimization objective that can be motivated like the rKL loss via the data processing inequality. Our analysis shows that employing the rKL loss with the log-derivative trick (rKL-LD) does not only avoid these conceptual problems but also consistently outperforms the LV loss. Experimental results with different types of diffusion bridges on challenging benchmarks show that samplers trained with the rKL-LD loss achieve better performance. From a practical perspective we find that rKL-LD requires significantly less hyperparameter optimization and yields more stable training behavior.
Authors: Jason Yang, Ariane Mora, Shengchao Liu, Bruce J. Wittmann, Anima Anandkumar, Frances H. Arnold, Yisong Yue
Abstract: Enzymes are important proteins that catalyze chemical reactions. In recent years, machine learning methods have emerged to predict enzyme function from sequence; however, there are no standardized benchmarks to evaluate these methods. We introduce CARE, a benchmark and dataset suite for the Classification And Retrieval of Enzymes (CARE). CARE centers on two tasks: (1) classification of a protein sequence by its enzyme commission (EC) number and (2) retrieval of an EC number given a chemical reaction. For each task, we design train-test splits to evaluate different kinds of out-of-distribution generalization that are relevant to real use cases. For the classification task, we provide baselines for state-of-the-art methods. Because the retrieval task has not been previously formalized, we propose a method called Contrastive Reaction-EnzymE Pretraining (CREEP) as one of the first baselines for this task and compare it to the recent method, CLIPZyme. CARE is available at https://github.com/jsunn-y/CARE/.
Authors: Shangshang Wang, Julian Asilis, \"Omer Faruk Akg\"ul, Enes Burak Bilgin, Ollie Liu, Deqing Fu, Willie Neiswanger
Abstract: How cost-effectively can we elicit strong reasoning in language models by leveraging their underlying representations? We answer this question with Resa, a family of 1.5B reasoning models trained via a novel and efficient sparse autoencoder tuning (SAE-Tuning) procedure. This method first trains an SAE to capture reasoning abilities from a source model, and then uses the trained SAE to guide a standard supervised fine-tuning process to elicit such abilities in a target model, all using verified question-answer data without any reasoning traces. Notably, when applied to certain base models before further RL post-training, SAE-Tuning retains >97% of its RL-trained counterpart's reasoning performance while reducing training costs by >2000x to roughly \$1 and training time by >450x to around 20 minutes. Furthermore, when applied to lightly RL-trained models (e.g., within 1 hour on 2 GPUs), it enables reasoning performance such as 43.33% Pass@1 on AIME24 and 90% Pass@1 on AMC23 for only around \$1 additional cost. Surprisingly, the reasoning abilities extracted via SAEs are potentially both generalizable and modular. Generality means abilities extracted from one dataset still elevate performance on a larger and overlapping corpus. Modularity means abilities extracted from Qwen or Qwen-Math can be attached to the R1-Distill model at test time, without any retraining, and yield comparable gains. Extensive ablations validate these findings and all artifacts are fully open-sourced.
Authors: Jie Qin, Wei Yang, Yan Su, Yiran Zhu, Weizhen Li, Yunyue Pan, Chengchang Pan, Honggang Qi
Abstract: Current HER2 assessment models for breast cancer predominantly analyze H&E or IHC images in isolation,despite clinical reliance on their synergistic interpretation. However, concurrent acquisition of both modalities is often hindered by workflow complexity and cost constraints. We propose an adaptive bimodal framework enabling flexible single-/dual-modality HER2 prediction through three innovations: 1) A dynamic branch selector that activates either single-modality reconstruction or dual-modality joint inference based on input completeness; 2) A bidirectional cross-modal GAN performing context-aware feature-space reconstruction of missing modalities; 3) A hybrid training protocol integrating adversarial learning and multi-task optimization. This architecture elevates single-modality H&E prediction accuracy from 71.44% to 94.25% while achieving 95.09% dual-modality accuracy, maintaining 90.28% reliability with sole IHC inputs. The framework's "dual-preferred, single-compatible" design delivers near-bimodal performance without requiring synchronized acquisition, particularly benefiting resource-limited settings through IHC infrastructure cost reduction. Experimental validation confirms 22.81%/12.90% accuracy improvements over H&E/IHC baselines respectively, with cross-modal reconstruction enhancing F1-scores to 0.9609 (HE to IHC) and 0.9251 (IHC to HE). By dynamically routing inputs through reconstruction-enhanced or native fusion pathways, the system mitigates performance degradation from missing data while preserving computational efficiency (78.55% parameter reduction in lightweight variant). This elastic architecture demonstrates significant potential for democratizing precise HER2 assessment across diverse healthcare settings.
Authors: Von Ralph Dane Marquez Herbuela, Yukie Nagai
Abstract: Human emotional expression emerges through coordinated vocal, facial, and gestural signals. While speech face alignment is well established, the broader dynamics linking emotionally expressive speech to regional facial and hand motion remains critical for gaining a deeper insight into how emotional and behavior cues are communicated in real interactions. Further modulating the coordination is the structure of conversational exchange like sequential turn taking, which creates stable temporal windows for multimodal synchrony, and simultaneous speech, often indicative of high arousal moments, disrupts this alignment and impacts emotional clarity. Understanding these dynamics enhances realtime emotion detection by improving the accuracy of timing and synchrony across modalities in both human interactions and AI systems. This study examines multimodal emotion coupling using region specific motion capture from dyadic interactions in the IEMOCAP corpus. Speech features included low level prosody, MFCCs, and model derived arousal, valence, and categorical emotions (Happy, Sad, Angry, Neutral), aligned with 3D facial and hand marker displacements. Expressive activeness was quantified through framewise displacement magnitudes, and speech to gesture prediction mapped speech features to facial and hand movements. Nonoverlapping speech consistently elicited greater activeness particularly in the lower face and mouth. Sadness showed increased expressivity during nonoverlap, while anger suppressed gestures during overlaps. Predictive mapping revealed highest accuracy for prosody and MFCCs in articulatory regions while arousal and valence had lower and more context sensitive correlations. Notably, hand speech synchrony was enhanced under low arousal and overlapping speech, but not for valence.
Authors: Chuqiao Zhang, Sarath Chandra Dantu, Debarghya Mitra, Dalia Chakrabarty
Abstract: Identification of critical residues of a protein is actively pursued, since such residues are essential for protein function. We present three ways of recognising critical residues of an example protein, the evolution of which is tracked via molecular dynamical simulations. Our methods are based on learning a Random Geometric Graph (RGG) variable, where the state variable of each of 156 residues, is attached to a node of this graph, with the RGG learnt using the matrix of correlations between state variables of each residue-pair. Given the categorical nature of the state variable, correlation between a residue pair is computed using Cramer's V. We advance an organic thresholding to learn an RGG, and compare results against extant thresholding techniques, when parametrising criticality as the nodal degree in the learnt RGG. Secondly, we develop a criticality measure by ranking the computed differences between the posterior probability of the full graph variable defined on all 156 residues, and that of the graph with all but one residue omitted. A third parametrisation of criticality informs on the dynamical variation of nodal degrees as the protein evolves during the simulation. Finally, we compare results obtained with the three distinct criticality parameters, against experimentally-ascertained critical residues.
Authors: Tian Lan, Yang-Hao Zhou, Zi-Ao Ma, Fanshu Sun, Rui-Qing Sun, Junyu Luo, Rong-Cheng Tu, Heyan Huang, Chen Xu, Zhijing Wu, Xian-Ling Mao
Abstract: Recent advances in deep learning have significantly enhanced generative AI capabilities across text, images, and audio. However, automatically evaluating the quality of these generated outputs presents ongoing challenges. Although numerous automatic evaluation methods exist, current research lacks a systematic framework that comprehensively organizes these methods across text, visual, and audio modalities. To address this issue, we present a comprehensive review and a unified taxonomy of automatic evaluation methods for generated content across all three modalities; We identify five fundamental paradigms that characterize existing evaluation approaches across these domains. Our analysis begins by examining evaluation methods for text generation, where techniques are most mature. We then extend this framework to image and audio generation, demonstrating its broad applicability. Finally, we discuss promising directions for future research in cross-modal evaluation methodologies.
Authors: Yanwen Huang, Lok Ming Lui, Gary P. T. Choi
Abstract: Density-equalizing map (DEM) serves as a powerful technique for creating shape deformations with the area changes reflecting an underlying density function. In recent decades, DEM has found widespread applications in fields such as data visualization, geometry processing, and medical imaging. Traditional approaches to DEM primarily rely on iterative numerical solvers for diffusion equations or optimization-based methods that minimize handcrafted energy functionals. However, these conventional techniques often face several challenges: they may suffer from limited accuracy, produce overlapping artifacts in extreme cases, and require substantial algorithmic redesign when extended from 2D to 3D, due to the derivative-dependent nature of their energy formulations. In this work, we propose a novel learning-based density-equalizing mapping framework (LDEM) using deep neural networks. Specifically, we introduce a loss function that enforces density uniformity and geometric regularity, and utilize a hierarchical approach to predict the transformations at both the coarse and dense levels. Our method demonstrates superior density-equalizing and bijectivity properties compared to prior methods for a wide range of simple and complex density distributions, and can be easily applied to surface remeshing with different effects. Also, it generalizes seamlessly from 2D to 3D domains without structural changes to the model architecture or loss formulation. Altogether, our work opens up new possibilities for scalable and robust computation of density-equalizing maps for practical applications.
Authors: Olga Ovcharenko, Florian Barkmann, Philip Toma, Imant Daunhawer, Julia Vogt, Sebastian Schelter, Valentina Boeva
Abstract: Self-supervised learning (SSL) has proven to be a powerful approach for extracting biologically meaningful representations from single-cell data. To advance our understanding of SSL methods applied to single-cell data, we present scSSL-Bench, a comprehensive benchmark that evaluates nineteen SSL methods. Our evaluation spans nine datasets and focuses on three common downstream tasks: batch correction, cell type annotation, and missing modality prediction. Furthermore, we systematically assess various data augmentation strategies. Our analysis reveals task-specific trade-offs: the specialized single-cell frameworks, scVI, CLAIRE, and the finetuned scGPT excel at uni-modal batch correction, while generic SSL methods, such as VICReg and SimCLR, demonstrate superior performance in cell typing and multi-modal data integration. Random masking emerges as the most effective augmentation technique across all tasks, surpassing domain-specific augmentations. Notably, our results indicate the need for a specialized single-cell multi-modal data integration framework. scSSL-Bench provides a standardized evaluation platform and concrete recommendations for applying SSL to single-cell analysis, advancing the convergence of deep learning and single-cell genomics.
Authors: Giannis Daras, Adrian Rodriguez-Munoz, Adam Klivans, Antonio Torralba, Constantinos Daskalakis
Abstract: We show how to use low-quality, synthetic, and out-of-distribution images to improve the quality of a diffusion model. Typically, diffusion models are trained on curated datasets that emerge from highly filtered data pools from the Web and other sources. We show that there is immense value in the lower-quality images that are often discarded. We present Ambient Diffusion Omni, a simple, principled framework to train diffusion models that can extract signal from all available images during training. Our framework exploits two properties of natural images -- spectral power law decay and locality. We first validate our framework by successfully training diffusion models with images synthetically corrupted by Gaussian blur, JPEG compression, and motion blur. We then use our framework to achieve state-of-the-art ImageNet FID, and we show significant improvements in both image quality and diversity for text-to-image generative modeling. The core insight is that noise dampens the initial skew between the desired high-quality distribution and the mixed distribution we actually observe. We provide rigorous theoretical justification for our approach by analyzing the trade-off between learning from biased data versus limited unbiased data across diffusion times.
Authors: Malvern Madondo, Yuan Shao, Yingzi Liu, Jun Zhou, Xiaofeng Yang, Zhen Tian
Abstract: Anatomical changes during intensity-modulated proton therapy (IMPT) for head-and-neck cancer (HNC) can shift Bragg peaks, risking tumor underdosing and organ-at-risk overdosing. As a result, treatment replanning is often required to maintain clinically acceptable treatment quality. However, current manual replanning processes are resource-intensive and time-consuming. We propose a patient-specific deep reinforcement learning (DRL) framework for automated IMPT replanning, with a reward-shaping mechanism based on a $150$-point plan quality score addressing competing clinical objectives. We formulate the planning process as an RL problem where agents learn control policies to adjust optimization priorities, maximizing plan quality. Unlike population-based approaches, our framework trains personalized agents for each patient using their planning CT (Computed Tomography) and augmented anatomies simulating anatomical changes (tumor progression and regression). This patient-specific approach leverages anatomical similarities throughout treatment, enabling effective plan adaptation. We implemented two DRL algorithms, Deep Q-Network and Proximal Policy Optimization, using dose-volume histograms (DVHs) as state representations and a $22$-dimensional action space of priority adjustments. Evaluation on five HNC patients using actual replanning CT data showed both DRL agents improved initial plan scores from $120.63 \pm 21.40$ to $139.78 \pm 6.84$ (DQN) and $142.74 \pm 5.16$ (PPO), surpassing manual replans generated by a human planner ($137.20 \pm 5.58$). Clinical validation confirms that improvements translate to better tumor coverage and OAR sparing across diverse anatomical changes. This work demonstrates DRL's potential in addressing geometric and dosimetric complexities of adaptive proton therapy, offering efficient offline adaptation solutions and advancing online adaptive proton therapy.
Authors: Christian Reichenb\"acher, Philipp Rank, Jochen Hipp, Oliver Bringmann
Abstract: This paper presents the first application of Gaussian Mixture Copula Models to the statistical modeling of driving scenarios for the safety validation of automated driving systems. Knowledge of the joint probability distribution of scenario parameters is essential for scenario-based safety assessment, where risk quantification depends on the likelihood of concrete parameter combinations. Gaussian Mixture Copula Models bring together the multimodal expressivity of Gaussian Mixture Models and the flexibility of copulas, enabling separate modeling of marginal distributions and dependencies. We benchmark Gaussian Mixture Copula Models against previously proposed approaches - Gaussian Mixture Models and Gaussian Copula Models - using real-world driving data drawn from scenarios defined in United Nations Regulation No. 157. Our evaluation across 18 million scenario instances demonstrates that Gaussian Mixture Copula Models provide a better fit to the data in terms of both likelihood and Sinkhorn distance. These results suggest that Gaussian Mixture Copula Models are a compelling foundation for future scenario-based validation frameworks.
Authors: Seyed Amir Hossein Saberi, Amir Najafi, Abolfazl Motahari, Babak H. khalaj
Abstract: In this paper, we establish sample complexity bounds for learning high-dimensional simplices in $\mathbb{R}^K$ from noisy data. Specifically, we consider $n$ i.i.d. samples uniformly drawn from an unknown simplex in $\mathbb{R}^K$, each corrupted by additive Gaussian noise of unknown variance. We prove an algorithm exists that, with high probability, outputs a simplex within $\ell_2$ or total variation (TV) distance at most $\varepsilon$ from the true simplex, provided $n \ge (K^2/\varepsilon^2) e^{\mathcal{O}(K/\mathrm{SNR}^2)}$, where $\mathrm{SNR}$ is the signal-to-noise ratio. Extending our prior work~\citep{saberi2023sample}, we derive new information-theoretic lower bounds, showing that simplex estimation within TV distance $\varepsilon$ requires at least $n \ge \Omega(K^3 \sigma^2/\varepsilon^2 + K/\varepsilon)$ samples, where $\sigma^2$ denotes the noise variance. In the noiseless scenario, our lower bound $n \ge \Omega(K/\varepsilon)$ matches known upper bounds up to constant factors. We resolve an open question by demonstrating that when $\mathrm{SNR} \ge \Omega(K^{1/2})$, noisy-case complexity aligns with the noiseless case. Our analysis leverages sample compression techniques (Ashtiani et al., 2018) and introduces a novel Fourier-based method for recovering distributions from noisy observations, potentially applicable beyond simplex learning.
Authors: Abiodun Ganiyu, Pranshav Gajjar, Vijay K Shah
Abstract: The advent of Open Radio Access Networks (O-RAN) has transformed the telecommunications industry by promoting interoperability, vendor diversity, and rapid innovation. However, its disaggregated architecture introduces complex testing challenges, particularly in validating multi-vendor components against O-RAN ALLIANCE and 3GPP specifications. Existing frameworks, such as those provided by Open Testing and Integration Centres (OTICs), rely heavily on manual processes, are fragmented and prone to human error, leading to inconsistency and scalability issues. To address these limitations, we present AI5GTest -- an AI-powered, specification-aware testing framework designed to automate the validation of O-RAN components. AI5GTest leverages a cooperative Large Language Models (LLM) framework consisting of Gen-LLM, Val-LLM, and Debug-LLM. Gen-LLM automatically generates expected procedural flows for test cases based on 3GPP and O-RAN specifications, while Val-LLM cross-references signaling messages against these flows to validate compliance and detect deviations. If anomalies arise, Debug-LLM performs root cause analysis, providing insight to the failure cause. To enhance transparency and trustworthiness, AI5GTest incorporates a human-in-the-loop mechanism, where the Gen-LLM presents top-k relevant official specifications to the tester for approval before proceeding with validation. Evaluated using a range of test cases obtained from O-RAN TIFG and WG5-IOT test specifications, AI5GTest demonstrates a significant reduction in overall test execution time compared to traditional manual methods, while maintaining high validation accuracy.
Authors: Natanael Lucena, F\'abio S. da Silva, Ricardo Rios
Abstract: This paper presents a comparison of the performance of Convolutional Neural Networks (CNNs) and Vision Transformers (ViTs) in the task of multi-classifying images containing lesions of psoriasis and diseases similar to it. Models pre-trained on ImageNet were adapted to a specific data set. Both achieved high predictive metrics, but the ViTs stood out for their superior performance with smaller models. Dual Attention Vision Transformer-Base (DaViT-B) obtained the best results, with an f1-score of 96.4%, and is recommended as the most efficient architecture for automated psoriasis detection. This article reinforces the potential of ViTs for medical image classification tasks.
Authors: Xiyao Wang, Zhengyuan Yang, Chao Feng, Yongyuan Liang, Yuhang Zhou, Xiaoyu Liu, Ziyi Zang, Ming Li, Chung-Ching Lin, Kevin Lin, Linjie Li, Furong Huang, Lijuan Wang
Abstract: Reinforcement learning (RL) has shown great effectiveness for fine-tuning large language models (LLMs) using tasks that are challenging yet easily verifiable, such as math reasoning or code generation. However, extending this success to visual perception in vision-language models (VLMs) has been impeded by the scarcity of vision-centric tasks that are simultaneously challenging and unambiguously verifiable. To this end, we introduce ViCrit (Visual Caption Hallucination Critic), an RL proxy task that trains VLMs to localize a subtle, synthetic visual hallucination injected into paragraphs of human-written image captions. Starting from a 200-word captions, we inject a single, subtle visual description error-altering a few words on objects, attributes, counts, or spatial relations-and task the model to pinpoint the corrupted span given the image and the modified caption. This formulation preserves the full perceptual difficulty while providing a binary, exact-match reward that is easy to compute and unambiguous. Models trained with the ViCrit Task exhibit substantial gains across a variety of VL benchmarks. Crucially, the improvements transfer beyond natural-image training data to abstract image reasoning and visual math, showing promises of learning to perceive rather than barely memorizing seen objects. To facilitate evaluation, we further introduce ViCrit-Bench, a category-balanced diagnostic benchmark that systematically probes perception errors across diverse image domains and error types. Together, our results demonstrate that fine-grained hallucination criticism is an effective and generalizable objective for enhancing visual perception in VLMs.
Authors: Yuke Xie, Herv\'e Chauris, Nicolas Desassis
Abstract: Diffusion models have recently shown promise as powerful generative priors for inverse problems. However, conventional applications require solving the full reverse diffusion process and operating on noisy intermediate states, which poses challenges for physics-constrained computational seismic imaging. In particular, such instability is pronounced in non-linear solvers like those used in Full Waveform Inversion (FWI), where wave propagation through noisy velocity fields can lead to numerical artifacts and poor inversion quality. In this work, we propose a simple yet effective framework that directly integrates a pretrained Denoising Diffusion Probabilistic Model (DDPM) as a score-based generative diffusion prior into FWI through a score rematching strategy. Unlike traditional diffusion approaches, our method avoids the reverse diffusion sampling and needs fewer iterations. We operate the image inversion entirely in the clean image space, eliminating the need to operate through noisy velocity models. The generative diffusion prior can be introduced as a simple regularization term in the standard FWI update rule, requiring minimal modification to existing FWI pipelines. This promotes stable wave propagation and can improve convergence behavior and inversion quality. Numerical experiments suggest that the proposed method offers enhanced fidelity and robustness compared to conventional and GAN-based FWI approaches, while remaining practical and computationally efficient for seismic imaging and other inverse problem tasks.
Authors: Zhecheng Liu, Jeff D. Eldredge
Abstract: A linear flow control strategy designed for weak disturbances may not remain effective in sequences of strong disturbances due to nonlinear interactions, but it is sensible to leverage it for developing a better strategy. In the present study, we propose a transformer-based reinforcement learning (RL) framework to learn an effective control strategy for regulating aerodynamic lift in gust sequences via pitch control. The transformer addresses the challenge of partial observability from limited surface pressure sensors. We demonstrate that the training can be accelerated with two techniques -- pretraining with an expert policy (here, linear control) and task-level transfer learning (here, extending a policy trained on isolated gusts to multiple gusts). We show that the learned strategy outperforms the best proportional control, with the performance gap widening as the number of gusts increases. The control strategy learned in an environment with a small number of successive gusts is shown to effectively generalize to an environment with an arbitrarily long sequence of gusts. We investigate the pivot configuration and show that quarter-chord pitching control can achieve superior lift regulation with substantially less control effort compared to mid-chord pitching control. Through a decomposition of the lift, we attribute this advantage to the dominant added-mass contribution accessible via quarter-chord pitching. The success on multiple configurations shows the generalizability of the proposed transformer-based RL framework, which offers a promising approach to solve more computationally demanding flow control problems when combined with the proposed acceleration techniques.
Authors: Bidyarthi Paul, SM Musfiqur Rahman, Dipta Biswas, Md. Ziaul Hasan, Md. Zahid Hossain
Abstract: Research on understanding emotions in written language continues to expand, especially for understudied languages with distinctive regional expressions and cultural features, such as Bangla. This study examines emotion analysis using 22,698 social media comments from the EmoNoBa dataset. For language analysis, we employ machine learning models: Linear SVM, KNN, and Random Forest with n-gram data from a TF-IDF vectorizer. We additionally investigated how PCA affects the reduction of dimensionality. Moreover, we utilized a BiLSTM model and AdaBoost to improve decision trees. To make our machine learning models easier to understand, we used LIME to explain the predictions of the AdaBoost classifier, which uses decision trees. With the goal of advancing sentiment analysis in languages with limited resources, our work examines various techniques to find efficient techniques for emotion identification in Bangla.
Authors: Elizabeth Demers, Victor Xiaoqi Wang, Kean Wu
Abstract: Human capital (HC) is increasingly important to corporate value creation. Unlike other assets, however, HC is not currently subject to well-defined measurement or disclosure rules. We use a machine learning algorithm (word2vec) trained on a confirmed set of HC disclosures to develop a comprehensive list of HC-related keywords classified into five subcategories (DEI; health and safety; labor relations and culture; compensation and benefits; and demographics and other) that capture the multidimensional nature of HC management. We share our lexicon, corporate HC disclosures, and the Python code used to develop the lexicon, and we provide detailed examples of using our data and code, including for fine-tuning a BERT model. Researchers can use our HC lexicon (or modify the code to capture another construct of interest) with their samples of corporate communications to address pertinent HC questions. We close with a discussion of future research opportunities related to HC management and disclosure.
Authors: Panagiotis Theodoropoulos, Augustinos D. Saravanos, Evangelos A. Theodorou, Guan-Horng Liu
Abstract: Understanding complex systems by inferring trajectories from sparse sample snapshots is a fundamental challenge in a wide range of domains, e.g., single-cell biology, meteorology, and economics. Despite advancements in Bridge and Flow matching frameworks, current methodologies rely on pairwise interpolation between adjacent snapshots. This hinders their ability to capture long-range temporal dependencies and potentially affects the coherence of the inferred trajectories. To address these issues, we introduce \textbf{Momentum Multi-Marginal Schr\"odinger Bridge Matching (3MSBM)}, a novel matching framework that learns smooth measure-valued splines for stochastic systems that satisfy multiple positional constraints. This is achieved by lifting the dynamics to phase space and generalizing stochastic bridges to be conditioned on several points, forming a multi-marginal conditional stochastic optimal control problem. The underlying dynamics are then learned by minimizing a variational objective, having fixed the path induced by the multi-marginal conditional bridge. As a matching approach, 3MSBM learns transport maps that preserve intermediate marginals throughout training, significantly improving convergence and scalability. Extensive experimentation in a series of real-world applications validates the superior performance of 3MSBM compared to existing methods in capturing complex dynamics with temporal dependencies, opening new avenues for training matching frameworks in multi-marginal settings.
Authors: Mohammad Jalali, Haoyu Lei, Amin Gohari, Farzan Farnia
Abstract: Diffusion models have demonstrated remarkable success in high-fidelity image synthesis and prompt-guided generative modeling. However, ensuring adequate diversity in generated samples of prompt-guided diffusion models remains a challenge, particularly when the prompts span a broad semantic spectrum and the diversity of generated data needs to be evaluated in a prompt-aware fashion across semantically similar prompts. Recent methods have introduced guidance via diversity measures to encourage more varied generations. In this work, we extend the diversity measure-based approaches by proposing the Scalable Prompt-Aware R\'eny Kernel Entropy Diversity Guidance (SPARKE) method for prompt-aware diversity guidance. SPARKE utilizes conditional entropy for diversity guidance, which dynamically conditions diversity measurement on similar prompts and enables prompt-aware diversity control. While the entropy-based guidance approach enhances prompt-aware diversity, its reliance on the matrix-based entropy scores poses computational challenges in large-scale generation settings. To address this, we focus on the special case of Conditional latent RKE Score Guidance, reducing entropy computation and gradient-based optimization complexity from the $O(n^3)$ of general entropy measures to $O(n)$. The reduced computational complexity allows for diversity-guided sampling over potentially thousands of generation rounds on different prompts. We numerically test the SPARKE method on several text-to-image diffusion models, demonstrating that the proposed method improves the prompt-aware diversity of the generated data without incurring significant computational costs. We release our code on the project page: https://mjalali.github.io/SPARKE
Authors: Souvik Ghosh, Sayak Roy
Abstract: The Su-Schrieffer-Heeger (SSH) model serves as a canonical example of a one-dimensional topological insulator, yet its behavior under more complex, realistic conditions remains a fertile ground for research. This paper presents a comprehensive computational investigation into generalized SSH models, exploring the interplay between topology, quasi-periodic disorder, non-Hermiticity, and time-dependent driving. Using exact diagonalization and specialized numerical solvers, we map the system's phase space through its spectral properties and localization characteristics, quantified by the Inverse Participation Ratio (IPR). We demonstrate that while the standard SSH model exhibits topologically protected edge states, these are destroyed by a localization transition induced by strong Aubry-Andr\'e-Harper (AAH) modulation. Further, we employ unsupervised machine learning (PCA) to autonomously classify the system's phases, revealing that strong localization can obscure underlying topological signatures. Extending the model beyond Hermiticity, we uncover the non-Hermitian skin effect, a dramatic localization of all bulk states at a boundary. Finally, we apply a periodic Floquet drive to a topologically trivial chain, successfully engineering a Floquet topological insulator characterized by the emergence of anomalous edge states at the boundaries of the quasi-energy zone. These findings collectively provide a multi-faceted view of the rich phenomena hosted in generalized 1D topological systems.
Authors: Andrei Paleyes, Radzim Sendyka, Diana Robinson, Christian Cabrera, Neil D. Lawrence
Abstract: Code generation is one of the most active areas of application of Large Language Models (LLMs). While LLMs lower barriers to writing code and accelerate development process, the overall quality of generated programs depends on the quality of given prompts. Specifically, functionality and quality of generated code can be sensitive to user's background and familiarity with software development. It is therefore important to quantify LLM's sensitivity to variations in the input. To this end we propose a synthetic evaluation pipeline for code generation with LLMs, as well as a systematic persona-based evaluation approach to expose qualitative differences of LLM responses dependent on prospective user background. Both proposed methods are completely independent from specific programming tasks and LLMs, and thus are widely applicable. We provide experimental evidence illustrating utility of our methods and share our code for the benefit of the community.
Authors: Parsa Rahimi, Sebastien Marcel
Abstract: In this paper, we propose ScoreMix, a novel yet simple data augmentation strategy leveraging the score compositional properties of diffusion models to enhance discriminator performance, particularly under scenarios with limited labeled data. By convexly mixing the scores from different class-conditioned trajectories during diffusion sampling, we generate challenging synthetic samples that significantly improve discriminative capabilities in all studied benchmarks. We systematically investigate class-selection strategies for mixing and discover that greater performance gains arise when combining classes distant in the discriminator's embedding space, rather than close in the generator's condition space. Moreover, we empirically show that, under standard metrics, the correlation between the generator's learned condition space and the discriminator's embedding space is minimal. Our approach achieves notable performance improvements without extensive parameter searches, demonstrating practical advantages for training discriminative models while effectively mitigating problems regarding collections of large datasets. Paper website: https://parsa-ra.github.io/scoremix
Authors: Yeonwoo Jang, Shariqah Hossain, Ashwin Sreevatsa, Diogo Cruz
Abstract: In this work, we show that some machine unlearning methods may fail when subjected to straightforward prompt attacks. We systematically evaluate eight unlearning techniques across three model families, and employ output-based, logit-based, and probe analysis to determine to what extent supposedly unlearned knowledge can be retrieved. While methods like RMU and TAR demonstrate robust unlearning, ELM remains vulnerable to specific prompt attacks (e.g., Hindi filler text in original prompt recovering 57.3% accuracy). Our logit analysis also confirms that unlearned models are generally not hiding knowledge by modifying the way the answer is formatted, as the correlation between output and logit accuracy is strong. These results challenge prevailing assumptions about unlearning effectiveness and highlight the need for evaluation frameworks that can reliably distinguish between true knowledge removal and superficial output suppression. We also publicly make available our evaluation framework to easily evaluate prompting techniques to retrieve unlearning knowledge.
Authors: Andrea Yaoyun Cui, Pengfei Yu
Abstract: Language models are essentially probability distributions over token sequences. Auto-regressive models generate sentences by iteratively computing and sampling from the distribution of the next token. This iterative sampling introduces stochasticity, leading to the assumption that language models make probabilistic decisions, similar to sampling from unknown distributions. Building on this assumption, prior research has used simulated Gibbs sampling, inspired by experiments designed to elicit human priors, to infer the priors of language models. In this paper, we revisit a critical question: Do language models possess Bayesian brains? Our findings show that under certain conditions, language models can exhibit near-deterministic decision-making, such as producing maximum likelihood estimations, even with a non-zero sampling temperature. This challenges the sampling assumption and undermines previous methods for eliciting human-like priors. Furthermore, we demonstrate that without proper scrutiny, a system with deterministic behavior undergoing simulated Gibbs sampling can converge to a "false prior." To address this, we propose a straightforward approach to distinguish between stochastic and deterministic decision patterns in Gibbs sampling, helping to prevent the inference of misleading language model priors. We experiment on a variety of large language models to identify their decision patterns under various circumstances. Our results provide key insights in understanding decision making of large language models.
Authors: Shiyu Jiang, Xuyin Liu, Zitong Jerry Wang
Abstract: Genomic language models (gLMs) show potential for generating novel, functional DNA sequences for synthetic biology, but doing so requires them to learn not just evolutionary plausibility, but also sequence-to-function relationships. We introduce a set of prediction tasks called Nullsettes, which assesses a model's ability to predict loss-of-function mutations created by translocating key control elements in synthetic expression cassettes. Across 12 state-of-the-art models, we find that mutation effect prediction performance strongly correlates with the predicted likelihood of the nonmutant. Furthermore, the range of likelihood values predictive of strong model performance is highly dependent on sequence length. Our work highlights the importance of considering both sequence likelihood and sequence length when using gLMs for mutation effect prediction.
Authors: Jun Qi, Chao-Han Yang, Pin-Yu Chen, Min-Hsiu Hsieh
Abstract: Variational Quantum Circuits (VQCs) offer a novel pathway for quantum machine learning, yet their practical application is hindered by inherent limitations such as constrained linear expressivity, optimization challenges, and acute sensitivity to quantum hardware noise. This work introduces VQC-MLPNet, a scalable and robust hybrid quantum-classical architecture designed to overcome these obstacles. By innovatively employing quantum circuits to dynamically generate parameters for classical Multi-Layer Perceptrons (MLPs) via amplitude encoding and parameterized quantum operations, VQC-MLPNet substantially expands representation capabilities and augments training stability. We provide rigorous theoretical guarantees via statistical learning techniques and Neural Tangent Kernel analysis, explicitly deriving upper bounds on approximation, uniform deviation, and optimization errors. These theoretical insights demonstrate exponential improvements in representation capacity relative to quantum circuit depth and the number of qubits, providing clear computational advantages over standalone quantum circuits and existing hybrid quantum architectures. Our theoretical claims are empirically corroborated through extensive experiments, including classifying semiconductor quantum-dot charge states and predicting genomic transcription factor binding sites, demonstrating resilient performance even under realistic IBM quantum noise simulations. This research establishes a theoretically sound and practically robust framework, advancing the frontiers of quantum-enhanced learning for unconventional computing paradigms in the Noisy Intermediate-Scale Quantum era and beyond.
Authors: Zige Wang, Qi Zhu, Fei Mi, Minghui Xu, Ruochun Jin, Wenjing Yang
Abstract: Gradient-based data influence approximation has been leveraged to select useful data samples in the supervised fine-tuning of large language models. However, the computation of gradients throughout the fine-tuning process requires too many resources to be feasible in practice. In this paper, we propose an efficient gradient-based data selection framework with clustering and a modified Upper Confidence Bound (UCB) algorithm. Based on the intuition that data samples with similar gradient features will have similar influences, we first perform clustering on the training data pool. Then, we frame the inter-cluster data selection as a constrained computing budget allocation problem and consider it a multi-armed bandit problem. A modified UCB algorithm is leveraged to solve this problem. Specifically, during the iterative sampling process, historical data influence information is recorded to directly estimate the distributions of each cluster, and a cold start is adopted to balance exploration and exploitation. Experimental results on various benchmarks show that our proposed framework, ClusterUCB, can achieve comparable results to the original gradient-based data selection methods while greatly reducing computing consumption.
Authors: Mo\"ise Blanchard, Samory Kpotufe
Abstract: There has been much recent interest in understanding the continuum from adversarial to stochastic settings in online learning, with various frameworks including smoothed settings proposed to bridge this gap. We consider the more general and flexible framework of distributionally constrained adversaries in which instances are drawn from distributions chosen by an adversary within some constrained distribution class [RST11]. Compared to smoothed analysis, we consider general distributional classes which allows for a fine-grained understanding of learning settings between fully stochastic and fully adversarial for which a learner can achieve non-trivial regret. We give a characterization for which distribution classes are learnable in this context against both oblivious and adaptive adversaries, providing insights into the types of interplay between the function class and distributional constraints on adversaries that enable learnability. In particular, our results recover and generalize learnability for known smoothed settings. Further, we show that for several natural function classes including linear classifiers, learning can be achieved without any prior knowledge of the distribution class -- in other words, a learner can simultaneously compete against any constrained adversary within learnable distribution classes.
Authors: Jasper Yao
Abstract: We establish fundamental computational complexity barriers to verifying AI safety as system capabilities scale. Our main results show that for AI systems with expressiveness EXP$(m)$ above a critical threshold $\tau$, safety verification requires exponential time and is coNP-complete. We formalize the Capability-Risk Scaling (CRS) dynamic, which demonstrates how increasing AI capability drives societal safety requirements toward perfection, creating an inescapable tension with verification complexity. Through four core theorems, we prove that (1) verification complexity grows exponentially with system expressiveness, (2) safe policies comprise at most a $2^{-2^m}$ fraction of the policy space, (3) no finite set of alignment techniques can provide universal coverage, and (4) robust safety properties form measure-zero sets for neural networks. These results characterize an "intractability gap" where practical safety requirements fall within the region of computational intractability. We conclude by presenting a strategic trilemma: AI development must either constrain system complexity to maintain verifiable safety, accept unverifiable risks while scaling capabilities, or develop fundamentally new safety paradigms beyond verification. Our work provides the first systematic complexity-theoretic analysis of AI alignment and establishes rigorous bounds that any safety approach must confront. A formal verification of the core theorems in Lean4 is currently in progress.
Authors: Oleg Berngardt, Ivan Lavygin
Abstract: The paper presents a method for automatic constructing a classifier for processed data obtained by decameter coherent scatter radars. Method is based only on the radar data obtained, the results of automatic modeling of radio wave propagation in the ionosphere, and mathematical criteria for estimating the quality of the models. The final classifier is the model trained at data obtained by 12 radars of the SuperDARN and SECIRA networks over two years for each radar. The number of the model coefficients is 2669. For the classification, the model uses both the calculated parameters of radio wave propagation in the model ionosphere and the parameters directly measured by the radar. Calibration of radiowave elevation measurements at each radar was made using meteor trail scattered signals. The analysis showed that the optimal number of classes in the data is 37, of which 25 are frequently observed. The analysis made it possible to choose 14 classes from them, which are confidently separated in other variants of model training. A preliminary interpretation of 10 of them was carried out. The dynamics of observation of various classes and their dependence on the geographical latitude of radars at different levels of solar and geomagnetic activity were presented, it was shown that it does not contradict with known physical mechanisms. The analysis showed that the most important parameters to identify the classes are the shape of the signal ray-tracing trajectory in its second half, the ray-traced scattering height and the Doppler velocity measured by the radar.
Authors: Cheng Wang, Siqi Chen, Donghua Mi, Yang Chen, Yudong Zhang, Yinsheng Li
Abstract: Recent advances in medical imaging have established deep learning-based segmentation as the predominant approach, though it typically requires large amounts of manually annotated data. However, obtaining annotations for intracranial hemorrhage (ICH) remains particularly challenging due to the tedious and costly labeling process. Semi-supervised learning (SSL) has emerged as a promising solution to address the scarcity of labeled data, especially in volumetric medical image segmentation. Unlike conventional SSL methods that primarily focus on high-confidence pseudo-labels or consistency regularization, we propose SWDL-Net, a novel SSL framework that exploits the complementary advantages of Laplacian pyramid and deep convolutional upsampling. The Laplacian pyramid excels at edge sharpening, while deep convolutions enhance detail precision through flexible feature mapping. Our framework achieves superior segmentation of lesion details and boundaries through a difference learning mechanism that effectively integrates these complementary approaches. Extensive experiments on a 271-case ICH dataset and public benchmarks demonstrate that SWDL-Net outperforms current state-of-the-art methods in scenarios with only 2% labeled data. Additional evaluations on the publicly available Brain Hemorrhage Segmentation Dataset (BHSD) with 5% labeled data further confirm the superiority of our approach. Code and data have been released at https://github.com/SIAT-CT-LAB/SWDL.
Authors: Cameron Angliss, Jiaxun Cui, Jiaheng Hu, Arrasy Rahman, Peter Stone
Abstract: Developing AI agents that can robustly adapt to dramatically different strategic landscapes without retraining is a central challenge for multi-agent learning. Pok\'emon Video Game Championships (VGC) is a domain with an extraordinarily large space of possible team configurations of approximately $10^{139}$ - far larger than those of Dota or Starcraft. The highly discrete, combinatorial nature of team building in Pok\'emon VGC causes optimal strategies to shift dramatically depending on both the team being piloted and the opponent's team, making generalization uniquely challenging. To advance research on this problem, we introduce VGC-Bench: a benchmark that provides critical infrastructure, standardizes evaluation protocols, and supplies human-play datasets and a range of baselines - from large-language-model agents and behavior cloning to reinforcement learning and empirical game-theoretic methods such as self-play, fictitious play, and double oracle. In the restricted setting where an agent is trained and evaluated on a single-team configuration, our methods are able to win against a professional VGC competitor. We extensively evaluated all baseline methods over progressively larger team sets and find that even the best-performing algorithm in the single-team setting struggles at scaling up as team size grows. Thus, policy generalization across diverse team strategies remains an open challenge for the community. Our code is open sourced at https://github.com/cameronangliss/VGC-Bench.
Authors: Sadia Kamal, Tim Oates, Joy Wan
Abstract: Skin carcinoma is the most prevalent form of cancer globally, accounting for over $8 billion in annual healthcare expenditures. In clinical settings, physicians document patient visits using detailed SOAP (Subjective, Objective, Assessment, and Plan) notes. However, manually generating these notes is labor-intensive and contributes to clinician burnout. In this work, we propose a weakly supervised multimodal framework to generate clinically structured SOAP notes from limited inputs, including lesion images and sparse clinical text. Our approach reduces reliance on manual annotations, enabling scalable, clinically grounded documentation while alleviating clinician burden and reducing the need for large annotated data. Our method achieves performance comparable to GPT-4o, Claude, and DeepSeek Janus Pro across key clinical relevance metrics. To evaluate clinical quality, we introduce two novel metrics MedConceptEval and Clinical Coherence Score (CCS) which assess semantic alignment with expert medical concepts and input features, respectively.
Authors: Philipp B\"ar, Moe T. Wynn, Sander J. J. Leemans
Abstract: This technical report provides proofs for the claims in the paper "A Full Picture in Conformance Checking: Efficiently Summarizing All Optimal Alignments".
Authors: Yanhui Li, Dongxia Wang, Zhu Sun, Haonan Zhang, Huizhong Guo
Abstract: Recently, Graph Neural Networks (GNNs) have become the dominant approach for Knowledge Graph-aware Recommender Systems (KGRSs) due to their proven effectiveness. Building upon GNN-based KGRSs, Self-Supervised Learning (SSL) has been incorporated to address the sparity issue, leading to longer training time. However, through extensive experiments, we reveal that: (1)compared to other KGRSs, the existing GNN-based KGRSs fail to keep their superior performance under sparse interactions even with SSL. (2) More complex models tend to perform worse in sparse interaction scenarios and complex mechanisms, like attention mechanism, can be detrimental as they often increase learning difficulty. Inspired by these findings, we propose LightKG, a simple yet powerful GNN-based KGRS to address sparsity issues. LightKG includes a simplified GNN layer that encodes directed relations as scalar pairs rather than dense embeddings and employs a linear aggregation framework, greatly reducing the complexity of GNNs. Additionally, LightKG incorporates an efficient contrastive layer to implement SSL. It directly minimizes the node similarity in original graph, avoiding the time-consuming subgraph generation and comparison required in previous SSL methods. Experiments on four benchmark datasets show that LightKG outperforms 12 competitive KGRSs in both sparse and dense scenarios while significantly reducing training time. Specifically, it surpasses the best baselines by an average of 5.8\% in recommendation accuracy and saves 84.3\% of training time compared to KGRSs with SSL. Our code is available at https://github.com/1371149/LightKG.
Authors: Che Wang, Jeroen van Baar, Chaitanya Mitash, Shuai Li, Dylan Randle, Weiyao Wang, Sumedh Sontakke, Kostas E. Bekris, Kapil Katyal
Abstract: This work demonstrates how autonomously learning aspects of robotic operation from sparsely-labeled, real-world data of deployed, engineered solutions at industrial scale can provide with solutions that achieve improved performance. Specifically, it focuses on multi-suction robot picking and performs a comprehensive study on the application of multi-modal visual encoders for predicting the success of candidate robotic picks. Picking diverse items from unstructured piles is an important and challenging task for robot manipulation in real-world settings, such as warehouses. Methods for picking from clutter must work for an open set of items while simultaneously meeting latency constraints to achieve high throughput. The demonstrated approach utilizes multiple input modalities, such as RGB, depth and semantic segmentation, to estimate the quality of candidate multi-suction picks. The strategy is trained from real-world item picking data, with a combination of multimodal pretrain and finetune. The manuscript provides comprehensive experimental evaluation performed over a large item-picking dataset, an item-picking dataset targeted to include partial occlusions, and a package-picking dataset, which focuses on containers, such as boxes and envelopes, instead of unpackaged items. The evaluation measures performance for different item configurations, pick scenes, and object types. Ablations help to understand the effects of in-domain pretraining, the impact of different modalities and the importance of finetuning. These ablations reveal both the importance of training over multiple modalities but also the ability of models to learn during pretraining the relationship between modalities so that during finetuning and inference, only a subset of them can be used as input.
Authors: Laziz U. Abdullaev, Maksim Tkachenko, Tan M. Nguyen
Abstract: The self-attention mechanism, a cornerstone of Transformer-based state-of-the-art deep learning architectures, is largely heuristic-driven and fundamentally challenging to interpret. Establishing a robust theoretical foundation to explain its remarkable success and limitations has therefore become an increasingly prominent focus in recent research. Some notable directions have explored understanding self-attention through the lens of image denoising and nonparametric regression. While promising, existing frameworks still lack a deeper mechanistic interpretation of various architectural components that enhance self-attention, both in its original formulation and subsequent variants. In this work, we aim to advance this understanding by developing a unifying image processing framework, capable of explaining not only the self-attention computation itself but also the role of components such as positional encoding and residual connections, including numerous later variants. We also pinpoint potential distinctions between the two concepts building upon our framework, and make effort to close this gap. We introduce two independent architectural modifications within transformers. While our primary objective is interpretability, we empirically observe that image processing-inspired modifications can also lead to notably improved accuracy and robustness against data contamination and adversaries across language and vision tasks as well as better long sequence understanding.
Authors: Yuhua Jiang, Yuwen Xiong, Yufeng Yuan, Chao Xin, Wenyuan Xu, Yu Yue, Qianchuan Zhao, Lin Yan
Abstract: Large Language Models (LLMs) have demonstrated impressive capabilities in complex reasoning tasks, yet they still struggle to reliably verify the correctness of their own outputs. Existing solutions to this verification challenge often depend on separate verifier models or require multi-stage self-correction training pipelines, which limit scalability. In this paper, we propose Policy as Generative Verifier (PAG), a simple and effective framework that empowers LLMs to self-correct by alternating between policy and verifier roles within a unified multi-turn reinforcement learning (RL) paradigm. Distinct from prior approaches that always generate a second attempt regardless of model confidence, PAG introduces a selective revision mechanism: the model revises its answer only when its own generative verification step detects an error. This verify-then-revise workflow not only alleviates model collapse but also jointly enhances both reasoning and verification abilities. Extensive experiments across diverse reasoning benchmarks highlight PAG's dual advancements: as a policy, it enhances direct generation and self-correction accuracy; as a verifier, its self-verification outperforms self-consistency.
Authors: Boris Sedlak, Alireza Furutanpey, Zihang Wang, V\'ictor Casamayor Pujol, Schahram Dustdar
Abstract: Edge computing breaks with traditional autoscaling due to strict resource constraints, thus, motivating more flexible scaling behaviors using multiple elasticity dimensions. This work introduces an agent-based autoscaling framework that dynamically adjusts both hardware resources and internal service configurations to maximize requirements fulfillment in constrained environments. We compare four types of scaling agents: Active Inference, Deep Q Network, Analysis of Structural Knowledge, and Deep Active Inference, using two real-world processing services running in parallel: YOLOv8 for visual recognition and OpenCV for QR code detection. Results show all agents achieve acceptable SLO performance with varying convergence patterns. While the Deep Q Network benefits from pre-training, the structural analysis converges quickly, and the deep active inference agent combines theoretical foundations with practical scalability advantages. Our findings provide evidence for the viability of multi-dimensional agent-based autoscaling for edge environments and encourage future work in this research direction.
Authors: Florian Handke, F\'elix Koulischer, Gabriel Raya, Luca Ambrogioni
Abstract: It is well known that semantic and structural features of the generated images emerge at different times during the reverse dynamics of diffusion, a phenomenon that has been connected to physical phase transitions in magnets and other materials. In this paper, we introduce a general information-theoretic approach to measure when these class-semantic "decisions" are made during the generative process. By using an online formula for the optimal Bayesian classifier, we estimate the conditional entropy of the class label given the noisy state. We then determine the time intervals corresponding to the highest information transfer between noisy states and class labels using the time derivative of the conditional entropy. We demonstrate our method on one-dimensional Gaussian mixture models and on DDPM models trained on the CIFAR10 dataset. As expected, we find that the semantic information transfer is highest in the intermediate stages of diffusion while vanishing during the final stages. However, we found sizable differences between the entropy rate profiles of different classes, suggesting that different "semantic decisions" are located at different intermediate times.
Authors: Guowei Zhong, Ruohong Huan, Mingzhen Wu, Ronghua Liang, Peng Chen
Abstract: Recent advancements in Multimodal Emotion Recognition (MER) face challenges in addressing both modality missing and Out-Of-Distribution (OOD) data simultaneously. Existing methods often rely on specific models or introduce excessive parameters, which limits their practicality. To address these issues, we propose a novel robust MER framework, Causal Inference Distiller (CIDer), and introduce a new task, Random Modality Feature Missing (RMFM), to generalize the definition of modality missing. CIDer integrates two key components: a Model-Specific Self-Distillation (MSSD) module and a Model-Agnostic Causal Inference (MACI) module. MSSD enhances robustness under the RMFM task through a weight-sharing self-distillation approach applied across low-level features, attention maps, and high-level representations. Additionally, a Word-level Self-aligned Attention Module (WSAM) reduces computational complexity, while a Multimodal Composite Transformer (MCT) facilitates efficient multimodal fusion. To tackle OOD challenges, MACI employs a tailored causal graph to mitigate label and language biases using a Multimodal Causal Module (MCM) and fine-grained counterfactual texts. Notably, MACI can independently enhance OOD generalization with minimal additional parameters. Furthermore, we also introduce the new repartitioned MER OOD datasets. Experimental results demonstrate that CIDer achieves robust performance in both RMFM and OOD scenarios, with fewer parameters and faster training compared to state-of-the-art methods. The implementation of this work is publicly accessible at https://github.com/gw-zhong/CIDer.
Authors: Sarker Miraz Mahfuz, Thomas Slawig
Abstract: We used precomputed steady states obtained by a spin-up for a global marine ecosystem model as training data to build a mapping from the small number of biogeochemical model parameters onto the three-dimensional converged steady annual cycle. The mapping was performed by a conditional variational autoencoder (CVAE) with mass correction. Applied for test data, we show that the prediction obtained by the CVAE already gives a reasonable good approximation of the steady states obtained by a regular spin-up. However, the predictions do not reach the same level of annual periodicity as those obtained in the original spin-up data. Thus, we took the predictions as initial values for a spin-up. We could show that the number of necessary iterations, corresponding to model years, to reach a prescribed stopping criterion in the spin-up could be significantly reduced compared to the use of the originally uniform, constant initial value. The amount of reduction depends on the applied stopping criterion, measuring the periodicity of the solution. The savings in needed iterations and, thus, computing time for the spin-up ranges from 50 to 95\%, depending on the stopping criterion for the spin-up. We compared these results with the use of the mean of the training data as an initial value. We found that this also accelerates the spin-up, but only by a much lower factor.
Authors: Shuaiqi Sun, Congde Yuan, Haoqiang Yang, Mengzhuo Guo, Guiying Wei, Jiangbo Tian
Abstract: In digital gaming, long-term user lifetime value (LTV) prediction is essential for monetization strategy, yet presents major challenges due to delayed payment behavior, sparse early user data, and the presence of high-value outliers. While existing models typically rely on either short-cycle observations or strong distributional assumptions, such approaches often underestimate long-term value or suffer from poor robustness. To address these issues, we propose SHort-cycle auxiliary with Order-preserving REgression (SHORE), a novel LTV prediction framework that integrates short-horizon predictions (e.g., LTV-15 and LTV-30) as auxiliary tasks to enhance long-cycle targets (e.g., LTV-60). SHORE also introduces a hybrid loss function combining order-preserving multi-class classification and a dynamic Huber loss to mitigate the influence of zero-inflation and outlier payment behavior. Extensive offline and online experiments on real-world datasets demonstrate that SHORE significantly outperforms existing baselines, achieving a 47.91\% relative reduction in prediction error in online deployment. These results highlight SHORE's practical effectiveness and robustness in industrial-scale LTV prediction for digital games.
Authors: Surya Jasper, Minh Luu, Evan Pan, Aakash Tyagi, Michael Quinn, Jiang Hu, David Kebo Houngninou
Abstract: Hardware complexity continues to strain verification resources, motivating the adoption of machine learning (ML) methods to improve debug efficiency. However, ML-assisted debugging critically depends on diverse and scalable bug datasets, which existing manual or automated bug insertion methods fail to reliably produce. We introduce BugGen, a first of its kind, fully autonomous, multi-agent pipeline leveraging Large Language Models (LLMs) to systematically generate, insert, and validate realistic functional bugs in RTL. BugGen partitions modules, selects mutation targets via a closed-loop agentic architecture, and employs iterative refinement and rollback mechanisms to ensure syntactic correctness and functional detectability. Evaluated across five OpenTitan IP blocks, BugGen produced 500 unique bugs with 94% functional accuracy and achieved a throughput of 17.7 validated bugs per hour-over five times faster than typical manual expert insertion. Additionally, BugGen identified 104 previously undetected bugs in OpenTitan regressions, highlighting its utility in exposing verification coverage gaps. Compared against Certitude, BugGen demonstrated over twice the syntactic accuracy, deeper exposure of testbench blind spots, and more functionally meaningful and complex bug scenarios. Furthermore, when these BugGen-generated datasets were employed to train ML-based failure triage models, we achieved high classification accuracy (88.1%-93.2%) across different IP blocks, confirming the practical utility and realism of generated bugs. BugGen thus provides a scalable solution for generating high-quality bug datasets, significantly enhancing verification efficiency and ML-assisted debugging.
Authors: Junhua Lin (University of Edinburgh), Marc Juarez (University of Edinburgh)
Abstract: We present a novel attack specifically designed against Tree-Ring, a watermarking technique for diffusion models known for its high imperceptibility and robustness against removal attacks. Unlike previous removal attacks, which rely on strong assumptions about attacker capabilities, our attack only requires access to the variational autoencoder that was used to train the target diffusion model, a component that is often publicly available. By leveraging this variational autoencoder, the attacker can approximate the model's intermediate latent space, enabling more effective surrogate-based attacks. Our evaluation shows that this approach leads to a dramatic reduction in the AUC of Tree-Ring detector's ROC and PR curves, decreasing from 0.993 to 0.153 and from 0.994 to 0.385, respectively, while maintaining high image quality. Notably, our attacks outperform existing methods that assume full access to the diffusion model. These findings highlight the risk of reusing public autoencoders to train diffusion models -- a threat not considered by current industry practices. Furthermore, the results suggest that the Tree-Ring detector's precision, a metric that has been overlooked by previous evaluations, falls short of the requirements for real-world deployment.
Authors: Hongyu Yao, Zijin Hong, Hao Chen, Yuanchen Bei, Zhiqing Li, Qijie Shen, Zuobin Ying, Huan Gong, Feiran Huang
Abstract: Graph-based multi-task learning at billion-scale presents a significant challenge, as different tasks correspond to distinct billion-scale graphs. Traditional multi-task learning methods often neglect these graph structures, relying solely on individual user and item embeddings. However, disregarding graph structures overlooks substantial potential for improving performance. In this paper, we introduce the Macro Graph of Expert (MGOE) framework, the first approach capable of leveraging macro graph embeddings to capture task-specific macro features while modeling the correlations between task-specific experts. Specifically, we propose the concept of a Macro Graph Bottom, which, for the first time, enables multi-task learning models to incorporate graph information effectively. We design the Macro Prediction Tower to dynamically integrate macro knowledge across tasks. MGOE has been deployed at scale, powering multi-task learning for the homepage of a leading billion-scale recommender system. Extensive offline experiments conducted on three public benchmark datasets demonstrate its superiority over state-of-the-art multi-task learning methods, establishing MGOE as a breakthrough in multi-task graph-based recommendation. Furthermore, online A/B tests confirm the superiority of MGOE in billion-scale recommender systems.
Authors: Tony Bonnaire, Giovanni Catania, Aur\'elien Decelle, Beatriz Seoane
Abstract: We investigate the phase diagram and memory retrieval capabilities of bipartite energy-based neural networks, namely Restricted Boltzmann Machines (RBMs), as a function of the prior distribution imposed on their hidden units - including binary, multi-state, and ReLU-like activations. Drawing connections to the Hopfield model and employing analytical tools from statistical physics of disordered systems, we explore how the architectural choices and activation functions shape the thermodynamic properties of these models. Our analysis reveals that standard RBMs with binary hidden nodes and extensive connectivity suffer from reduced critical capacity, limiting their effectiveness as associative memories. To address this, we examine several modifications, such as introducing local biases and adopting richer hidden unit priors. These adjustments restore ordered retrieval phases and markedly improve recall performance, even at finite temperatures. Our theoretical findings, supported by finite-size Monte Carlo simulations, highlight the importance of hidden unit design in enhancing the expressive power of RBMs.
Authors: Kyohei Atarashi, Satoshi Oyama, Hiromi Arai, Hisashi Kashima
Abstract: Controlling the output probabilities of softmax-based models is a common problem in modern machine learning. Although the $\mathrm{Softmax}$ function provides soft control via its temperature parameter, it lacks the ability to enforce hard constraints, such as box constraints, on output probabilities, which can be critical in certain applications requiring reliable and trustworthy models. In this work, we propose the box-constrained softmax ($\mathrm{BCSoftmax}$) function, a novel generalization of the $\mathrm{Softmax}$ function that explicitly enforces lower and upper bounds on output probabilities. While $\mathrm{BCSoftmax}$ is formulated as the solution to a box-constrained optimization problem, we develop an exact and efficient computation algorithm for $\mathrm{BCSoftmax}$. As a key application, we introduce two post-hoc calibration methods based on $\mathrm{BCSoftmax}$. The proposed methods mitigate underconfidence and overconfidence in predictive models by learning the lower and upper bounds of the output probabilities or logits after model training, thereby enhancing reliability in downstream decision-making tasks. We demonstrate the effectiveness of our methods experimentally using the TinyImageNet, CIFAR-100, and 20NewsGroups datasets, achieving improvements in calibration metrics.
Authors: Henrik Sebastian Steude, Alexander Diedrich, Ingo Pill, Lukas Moddemann, Daniel Vranje\v{s}, Oliver Niggemann
Abstract: Diagnostic processes for complex cyber-physical systems often require extensive prior knowledge in the form of detailed system models or comprehensive training data. However, obtaining such information poses a significant challenge. To address this issue, we present a new diagnostic approach that operates with minimal prior knowledge, requiring only a basic understanding of subsystem relationships and data from nominal operations. Our method combines a neural network-based symptom generator, which employs subsystem-level anomaly detection, with a new graph diagnosis algorithm that leverages minimal causal relationship information between subsystems-information that is typically available in practice. Our experiments with fully controllable simulated datasets show that our method includes the true causal component in its diagnosis set for 82 p.c. of all cases while effectively reducing the search space in 73 p.c. of the scenarios. Additional tests on the real-world Secure Water Treatment dataset showcase the approach's potential for practical scenarios. Our results thus highlight our approach's potential for practical applications with large and complex cyber-physical systems where limited prior knowledge is available.
Authors: Stefano Longari, Paolo Cerracchio, Michele Carminati, Stefano Zanero
Abstract: The security of modern vehicles has become increasingly important, with the controller area network (CAN) bus serving as a critical communication backbone for various Electronic Control Units (ECUs). The absence of robust security measures in CAN, coupled with the increasing connectivity of vehicles, makes them susceptible to cyberattacks. While intrusion detection systems (IDSs) have been developed to counter such threats, they are not foolproof. Adversarial attacks, particularly evasion attacks, can manipulate inputs to bypass detection by IDSs. This paper extends our previous work by investigating the feasibility and impact of gradient-based adversarial attacks performed with different degrees of knowledge against automotive IDSs. We consider three scenarios: white-box (attacker with full system knowledge), grey-box (partial system knowledge), and the more realistic black-box (no knowledge of the IDS' internal workings or data). We evaluate the effectiveness of the proposed attacks against state-of-the-art IDSs on two publicly available datasets. Additionally, we study effect of the adversarial perturbation on the attack impact and evaluate real-time feasibility by precomputing evasive payloads for timed injection based on bus traffic. Our results demonstrate that, besides attacks being challenging due to the automotive domain constraints, their effectiveness is strongly dependent on the dataset quality, the target IDS, and the attacker's degree of knowledge.
Authors: Sergio Burdisso, Esa\'u Villatoro-Tello, Petr Motlicek
Abstract: The advancement of conversational AI systems relies on the availability of high-quality, flexible, and reproducible synthetic dialogues for training, evaluation, and benchmarking. SDialog is a modular, extensible Python toolkit designed to address the challenges of synthetic dialogue generation and analysis. By leveraging instruction-tuned Large Language Models (LLMs), SDialog provides abstractions for personas, orchestration, and scenario management, enabling the creation of realistic, diverse, and controllable conversational data for research and development. SDialog supports workflows such as multi-agent simulation and scenario-driven generation, and represents a step forward in the standardization of tools and frameworks for synthetic data generation, a crucial advancement for ensuring reproducibility in today's fast-evolving research landscape.
Authors: Wei Chen, Giacomo Dimarco, Lorenzo Pareschi
Abstract: The high dimensionality of kinetic equations with stochastic parameters poses major computational challenges for uncertainty quantification (UQ). Traditional Monte Carlo (MC) sampling methods, while widely used, suffer from slow convergence and high variance, which become increasingly severe as the dimensionality of the parameter space grows. To accelerate MC sampling, we adopt a multiscale control variates strategy that leverages low-fidelity solutions from simplified kinetic models to reduce variance. To further improve sampling efficiency and preserve the underlying physics, we introduce surrogate models based on structure and asymptotic preserving neural networks (SAPNNs). These deep neural networks are specifically designed to satisfy key physical properties, including positivity, conservation laws, entropy dissipation, and asymptotic limits. By training the SAPNNs on low-fidelity models and enriching them with selected high-fidelity samples from the full Boltzmann equation, our method achieves significant variance reduction while maintaining physical consistency and asymptotic accuracy. The proposed methodology enables efficient large-scale prediction in kinetic UQ and is validated across both homogeneous and nonhomogeneous multiscale regimes. Numerical results demonstrate improved accuracy and computational efficiency compared to standard MC techniques.
Authors: Rogier C. van Dalen, Shucong Zhang, Titouan Parcollet, Sourav Bhattacharya
Abstract: Speech recognisers usually perform optimally only in a specific environment and need to be adapted to work well in another. For adaptation to a new speaker, there is often too little data for fine-tuning to be robust, and that data is usually unlabelled. This paper proposes a combination of approaches to make adaptation to a single minute of data robust. First, instead of estimating the adaptation parameters with cross-entropy on a single error-prone hypothesis or "pseudo-label", this paper proposes a novel loss function, the conditional entropy over complete hypotheses. Using multiple hypotheses makes adaptation more robust to errors in the initial recognition. Second, a "speaker code" characterises a speaker in a vector short enough that it requires little data to estimate. On a far-field noise-augmented version of Common Voice, the proposed scheme yields a 20% relative improvement in word error rate on one minute of adaptation data, increasing on 10 minutes to 29%.
Authors: Tim Whittaker, Alejandro Di Luca
Abstract: Understanding the plausible upper bounds of extreme weather events is essential for risk assessment in a warming climate. Existing methods, based on large ensembles of physics-based models, are often computationally expensive or lack the fidelity needed to simulate rare, high-impact extremes. Here, we present a novel framework that leverages a differentiable hybrid climate model, NeuralGCM, to optimize initial conditions and generate physically consistent worst-case heatwave trajectories. Applied to the 2021 Pacific Northwest heatwave, our method produces temperature anomalies up to 3.7 $^\circ$C above the most extreme member of a 75-member ensemble. These trajectories feature intensified atmospheric blocking and amplified Rossby wave patterns--hallmarks of severe heat events. Our results demonstrate that differentiable climate models can efficiently explore the upper tails of event likelihoods, providing a powerful new approach for constructing targeted storylines of extreme weather under climate change.
Authors: Maxime Haddouche, Otmane Sakhi
Abstract: Off-policy learning serves as the primary framework for learning optimal policies from logged interactions collected under a static behavior policy. In this work, we investigate the more practical and flexible setting of adaptive off-policy learning, where policies are iteratively refined and re-deployed to collect higher-quality data. Building on the success of PAC-Bayesian learning with Logarithmic Smoothing (LS) in static settings, we extend this framework to the adaptive scenario using tools from online PAC-Bayesian theory. Furthermore, we demonstrate that a principled adjustment to the LS estimator naturally accommodates multiple rounds of deployment and yields faster convergence rates under mild conditions. Our method matches the performance of leading offline approaches in static settings, and significantly outperforms them when intermediate policy deployments are allowed. Empirical evaluations across diverse scenarios highlight both the advantages of adaptive data collection and the strength of the PAC-Bayesian formulation.
Authors: Sakhi Otmane, Gilotte Alexandre, Rohde David
Abstract: We address the problem of A/B testing, a widely used protocol for evaluating the potential improvement achieved by a new decision system compared to a baseline. This protocol segments the population into two subgroups, each exposed to a version of the system and estimates the improvement as the difference between the measured effects. In this work, we demonstrate that the commonly used difference-in-means estimator, while unbiased, can be improved. We introduce a family of unbiased off-policy estimators that achieves lower variance than the standard approach. Among this family, we identify the estimator with the lowest variance. The resulting estimator is simple, and offers substantial variance reduction when the two tested systems exhibit similarities. Our theoretical analysis and experimental results validate the effectiveness and practicality of the proposed method.
Authors: Thanh Thi Nguyen, Campbell Wilson, Janis Dalins
Abstract: Detecting life-threatening language is essential for safeguarding individuals in distress, promoting mental health and well-being, and preventing potential harm and loss of life. This paper presents an effective approach to identifying life-threatening texts using large language models (LLMs) and compares them with traditional methods such as bag of words, word embedding, topic modeling, and Bidirectional Encoder Representations from Transformers. We fine-tune three open-source LLMs including Gemma, Mistral, and Llama-2 using their 7B parameter variants on different datasets, which are constructed with class balance, imbalance, and extreme imbalance scenarios. Experimental results demonstrate a strong performance of LLMs against traditional methods. More specifically, Mistral and Llama-2 models are top performers in both balanced and imbalanced data scenarios while Gemma is slightly behind. We employ the upsampling technique to deal with the imbalanced data scenarios and demonstrate that while this method benefits traditional approaches, it does not have as much impact on LLMs. This study demonstrates a great potential of LLMs for real-world life-threatening language detection problems.
Authors: Ali Waqas, Sinem Coleri
Abstract: Sensor-based local inference at IoT devices faces severe computational limitations, often requiring data transmission over noisy wireless channels for server-side processing. To address this, split-network Deep Neural Network (DNN) based Joint Source-Channel Coding (JSCC) schemes are used to extract and transmit relevant features instead of raw data. However, most existing methods rely on fixed network splits and static configurations, lacking adaptability to varying computational budgets and channel conditions. In this paper, we propose a novel SNR- and computation-adaptive distributed CNN framework for wireless image classification across IoT devices and edge servers. We introduce a learning-assisted intelligent Genetic Algorithm (LAIGA) that efficiently explores the CNN hyperparameter space to optimize network configuration under given FLOPs constraints and given SNR. LAIGA intelligently discards the infeasible network configurations that exceed computational budget at IoT device. It also benefits from the Random Forests based learning assistance to avoid a thorough exploration of hyperparameter space and to induce application specific bias in candidate optimal configurations. Experimental results demonstrate that the proposed framework outperforms fixed-split architectures and existing SNR-adaptive methods, especially under low SNR and limited computational resources. We achieve a 10\% increase in classification accuracy as compared to existing JSCC based SNR-adaptive multilayer framework at an SNR as low as -10dB across a range of available computational budget (1M to 70M FLOPs) at IoT device.
Authors: Melika Ayoughi, Mina Ghadimi Atigh, Mohammad Mahdi Derakhshani, Cees G. M. Snoek, Pascal Mettes, Paul Groth
Abstract: Continual learning has traditionally focused on classifying either instances or classes, but real-world applications, such as robotics and self-driving cars, require models to handle both simultaneously. To mirror real-life scenarios, we introduce the task of continual learning of instances and classes, at the same time. This task challenges models to adapt to multiple levels of granularity over time, which requires balancing fine-grained instance recognition with coarse-grained class generalization. In this paper, we identify that classes and instances naturally form a hierarchical structure. To model these hierarchical relationships, we propose HyperCLIC, a continual learning algorithm that leverages hyperbolic space, which is uniquely suited for hierarchical data due to its ability to represent tree-like structures with low distortion and compact embeddings. Our framework incorporates hyperbolic classification and distillation objectives, enabling the continual embedding of hierarchical relations. To evaluate performance across multiple granularities, we introduce continual hierarchical metrics. We validate our approach on EgoObjects, the only dataset that captures the complexity of hierarchical object recognition in dynamic real-world environments. Empirical results show that HyperCLIC operates effectively at multiple granularities with improved hierarchical generalization.
Authors: Ye Yu, Yaoning Yu, Haohan Wang
Abstract: Large reasoning models (LRMs) such as Claude 3.7 Sonnet and OpenAI o1 achieve strong performance on mathematical benchmarks using lengthy chain-of-thought (CoT) reasoning, but the resulting traces are often unnecessarily verbose. This inflates token usage and cost, limiting deployment in latency-sensitive or API-constrained settings. We introduce PREMISE (PRompt-based Efficient Mathematical Inference with Strategic Evaluation), a prompt-only framework that reduces reasoning overhead without modifying model weights. PREMISE combines trace-level diagnostics with gradient-inspired prompt optimization to minimize redundant computation while preserving answer accuracy. The approach jointly optimizes brevity and correctness through a multi-objective textual search that balances token length and answer validity. Unlike prior work, PREMISE runs in a single-pass black-box interface, so it can be applied directly to commercial LLMs. On GSM8K, SVAMP, and Math500 we match or exceed baseline accuracy ($96\%\rightarrow96\%$ with Claude, $91\%\rightarrow92\%$ with Gemini) while reducing reasoning tokens by up to $87.5\%$ and cutting dollar cost by $69$--$82\%$. These results show that prompt-level optimization is a practical and scalable path to efficient LRM inference without compromising reasoning quality.
Authors: Xiaozhe Li, Jixuan Chen, Xinyu Fang, Shengyuan Ding, Haodong Duan, Qingwen Liu, Kai Chen
Abstract: Large Language Models (LLMs) have shown remarkable capabilities in solving diverse tasks. However, their proficiency in iteratively optimizing complex solutions through learning from previous feedback remains insufficiently explored. To bridge this gap, we present OPT-BENCH, a comprehensive benchmark designed to evaluate LLM agents on large-scale search space optimization problems. OPT-BENCH includes 20 real-world machine learning tasks sourced from Kaggle and 10 classical NP problems, offering a diverse and challenging environment for assessing LLM agents on iterative reasoning and solution refinement. To enable rigorous evaluation, we introduce OPT-Agent, an end-to-end optimization framework that emulates human reasoning when tackling complex problems by generating, validating, and iteratively improving solutions through leveraging historical feedback. Through extensive experiments on 9 state-of-the-art LLMs from 6 model families, we analyze the effects of optimization iterations, temperature settings, and model architectures on solution quality and convergence. Our results demonstrate that incorporating historical context significantly enhances optimization performance across both ML and NP tasks. All datasets, code, and evaluation tools are open-sourced to promote further research in advancing LLM-driven optimization and iterative reasoning. Project page: \href{https://github.com/OliverLeeXZ/OPT-BENCH}{https://github.com/OliverLeeXZ/OPT-BENCH}.
URLs: https://github.com/OliverLeeXZ/OPT-BENCH, https://github.com/OliverLeeXZ/OPT-BENCH
Authors: Jian Li, Wan Han, Ning Lin, Yu-Liang Zhan, Ruizhi Chengze, Haining Wang, Yi Zhang, Hongsheng Liu, Zidong Wang, Fan Yu, Hao Sun
Abstract: Understanding and reasoning about dynamics governed by physical laws through visual observation, akin to human capabilities in the real world, poses significant challenges. Currently, object-centric dynamic simulation methods, which emulate human behavior, have achieved notable progress but overlook two critical aspects: 1) the integration of physical knowledge into models. Humans gain physical insights by observing the world and apply this knowledge to accurately reason about various dynamic scenarios; 2) the validation of model adaptability across diverse scenarios. Real-world dynamics, especially those involving fluids and objects, demand models that not only capture object interactions but also simulate fluid flow characteristics. To address these gaps, we introduce SlotPi, a slot-based physics-informed object-centric reasoning model. SlotPi integrates a physical module based on Hamiltonian principles with a spatio-temporal prediction module for dynamic forecasting. Our experiments highlight the model's strengths in tasks such as prediction and Visual Question Answering (VQA) on benchmark and fluid datasets. Furthermore, we have created a real-world dataset encompassing object interactions, fluid dynamics, and fluid-object interactions, on which we validated our model's capabilities. The model's robust performance across all datasets underscores its strong adaptability, laying a foundation for developing more advanced world models.
Authors: Qingyan Wei, Yaojie Zhang, Zhiyuan Liu, Dongrui Liu, Linfeng Zhang
Abstract: Diffusion-based language models (dLLMs) have emerged as a promising alternative to traditional autoregressive LLMs by enabling parallel token generation and significantly reducing inference latency. However, existing sampling strategies for dLLMs, such as confidence-based or semi-autoregressive decoding, often suffer from static behavior, leading to suboptimal efficiency and limited flexibility. In this paper, we propose SlowFast Sampling, a novel dynamic sampling strategy that adaptively alternates between exploratory and accelerated decoding stages. Our method is guided by three golden principles: certainty principle, convergence principle, and positional principle, which govern when and where tokens can be confidently and efficiently decoded. We further integrate our strategy with dLLM-Cache to reduce redundant computation. Extensive experiments across benchmarks and models show that SlowFast Sampling achieves up to 15.63$\times$ speedup on LLaDA with minimal accuracy drop, and up to 34.22$\times$ when combined with caching. Notably, our approach outperforms strong autoregressive baselines like LLaMA3 8B in throughput, demonstrating that well-designed sampling can unlock the full potential of dLLMs for fast and high-quality generation.
Authors: Adel Chehade, Edoardo Ragusa, Paolo Gastaldo, Rodolfo Zunino
Abstract: In this paper, we present a practical deep learning (DL) approach for energy-efficient traffic classification (TC) on resource-limited microcontrollers, which are widely used in IoT-based smart systems and communication networks. Our objective is to balance accuracy, computational efficiency, and real-world deployability. To that end, we develop a lightweight 1D-CNN, optimized via hardware-aware neural architecture search (HW-NAS), which achieves 96.59% accuracy on the ISCX VPN-NonVPN dataset with only 88.26K parameters, a 20.12K maximum tensor size, and 10.08M floating-point operations (FLOPs). Moreover, it generalizes across various TC tasks, with accuracies ranging from 94% to 99%. To enable deployment, the model is quantized to INT8, suffering only a marginal 1-2% accuracy drop relative to its Float32 counterpart. We evaluate real-world inference performance on two microcontrollers: the high-performance STM32F746G-DISCO and the cost-sensitive Nucleo-F401RE. The deployed model achieves inference latencies of 31.43ms and 115.40ms, with energy consumption of 7.86 mJ and 29.10 mJ per inference, respectively. These results demonstrate the feasibility of on-device encrypted traffic analysis, paving the way for scalable, low-power IoT security solutions.
Authors: Rui Zhang, Qi Meng, Han Wan, Yang Liu, Zhi-Ming Ma, Hao Sun
Abstract: High-fidelity and efficient simulation of fluid dynamics drive progress in various scientific and engineering applications. Traditional computational fluid dynamics methods offer strong interpretability and guaranteed convergence, but rely on fine spatial and temporal meshes, incurring prohibitive computational costs. Physics-informed neural networks (PINNs) and neural operators aim to accelerate PDE solvers using deep learning techniques. However, PINNs require extensive retraining and careful tuning, and purely data-driven operators demand large labeled datasets. Hybrid physics-aware methods embed numerical discretizations into network architectures or loss functions, but achieve marginal speed gains and become unstable when balancing coarse priors against high-fidelity measurements. To this end, we introduce OmniFluids, a unified physics pre-trained operator learning framework that integrates physics-only pre-training, coarse-grid operator distillation, and few-shot fine-tuning, which enables fast inference and accurate prediction under limited or zero data supervision. For architectural design, the key components of OmniFluids include a mixture of operators, a multi-frame decoder, and factorized Fourier layers, which enable efficient and scalable modeling of diverse physical tasks while maintaining seamless integration with physics-based supervision. Across a broad range of two- and three-dimensional benchmarks, OmniFluids significantly outperforms state-of-the-art AI-driven methods in flow field reconstruction and turbulence statistics accuracy, delivering 10-100x speedups compared to classical solvers, and accurately recovers unknown physical parameters from sparse, noisy data. This work establishes a new paradigm for efficient and generalizable surrogate modeling in complex fluid systems under limited data availability.
Authors: Ziv Scully, Alexander Terenin
Abstract: The Gittins index is a tool that optimally solves a variety of decision-making problems involving uncertainty, including multi-armed bandit problems, minimizing mean latency in queues, and search problems like the Pandora's box model. However, despite the above examples and later extensions thereof, the space of problems that the Gittins index can solve perfectly optimally is limited, and its definition is rather subtle compared to those of other multi-armed bandit algorithms. As a result, the Gittins index is often regarded as being primarily a concept of theoretical importance, rather than a practical tool for solving decision-making problems. The aim of this tutorial is to demonstrate that the Gittins index can be fruitfully applied to practical problems. We start by giving an example-driven introduction to the Gittins index, then walk through several examples of problems it solves - some optimally, some suboptimally but still with excellent performance. Two practical highlights in the latter category are applying the Gittins index to Bayesian optimization, and applying the Gittins index to minimizing tail latency in queues.
Authors: Guanjin Wang, Xiangxue Zhao, Shapour Azarm, Balakumar Balachandran
Abstract: An alternative data-driven modeling approach has been proposed and employed to gain fundamental insights into robot motion interaction with granular terrain at certain length scales. The approach is based on an integration of dimension reduction (Sequentially Truncated Higher-Order Singular Value Decomposition), surrogate modeling (Gaussian Process), and data assimilation techniques (Reduced Order Particle Filter). This approach can be used online and is based on offline data, obtained from the offline collection of high-fidelity simulation data and a set of sparse experimental data. The results have shown that orders of magnitude reduction in computational time can be obtained from the proposed data-driven modeling approach compared with physics-based high-fidelity simulations. With only simulation data as input, the data-driven prediction technique can generate predictions that have comparable accuracy as simulations. With both simulation data and sparse physical experimental measurement as input, the data-driven approach with its embedded data assimilation techniques has the potential in outperforming only high-fidelity simulations for the long-horizon predictions. In addition, it is demonstrated that the data-driven modeling approach can also reproduce the scaling relationship recovered by physics-based simulations for maximum resistive forces, which may indicate its general predictability beyond a case-by-case basis. The results are expected to help robot navigation and exploration in unknown and complex terrains during both online and offline phases.
Authors: Pratik Worah
Abstract: We design an efficient algorithm that outputs a linear classifier for identifying homogeneous subsets (equivalently subcohorts) from large inhomogeneous datasets. Our theoretical contribution is a rounding technique, similar to that of Goemans and Williamson (1994), that approximates the optimal solution of the underlying optimization problem within a factor of $0.82$. As an application, we use our algorithm to design a simple test that can identify homogeneous subcohorts of patients, that are mainly comprised of metastatic cases, from the RNA microarray dataset for breast cancer by Curtis et al. (2012). Furthermore, we also use the test output by the algorithm to systematically identify subcohorts of patients in which statistically significant changes in methylation levels of tumor suppressor genes co-occur with statistically significant changes in nuclear receptor expression. Identifying such homogeneous subcohorts of patients can be useful for the discovery of disease pathways and therapeutics, specific to the subcohort.
Authors: Yixiao Huang, Hanlin Zhu, Tianyu Guo, Jiantao Jiao, Somayeh Sojoudi, Michael I. Jordan, Stuart Russell, Song Mei
Abstract: Large language models (LLMs) can acquire new knowledge through fine-tuning, but this process exhibits a puzzling duality: models can generalize remarkably from new facts, yet are also prone to hallucinating incorrect information. However, the reasons for this phenomenon remain poorly understood. In this work, we argue that both behaviors stem from a single mechanism known as out-of-context reasoning (OCR): the ability to deduce implications by associating concepts, even those without a causal link. Our experiments across five prominent LLMs confirm that OCR indeed drives both generalization and hallucination, depending on whether the associated concepts are causally related. To build a rigorous theoretical understanding of this phenomenon, we then formalize OCR as a synthetic factual recall task. We empirically show that a one-layer single-head attention-only transformer with factorized output and value matrices can learn to solve this task, while a model with combined weights cannot, highlighting the crucial role of matrix factorization. Our theoretical analysis shows that the OCR capability can be attributed to the implicit bias of gradient descent, which favors solutions that minimize the nuclear norm of the combined output-value matrix. This mathematical structure explains why the model learns to associate facts and implications with high sample efficiency, regardless of whether the correlation is causal or merely spurious. Ultimately, our work provides a theoretical foundation for understanding the OCR phenomenon, offering a new lens for analyzing and mitigating undesirable behaviors from knowledge injection.
Authors: Dimitri Meunier, Antoine Moulin, Jakub Wornbard, Vladimir R. Kostic, Arthur Gretton
Abstract: We address the problem of causal effect estimation in the presence of hidden confounders, using nonparametric instrumental variable (IV) regression. A leading strategy employs spectral features - that is, learned features spanning the top eigensubspaces of the operator linking treatments to instruments. We derive a generalization error bound for a two-stage least squares estimator based on spectral features, and gain insights into the method's performance and failure modes. We show that performance depends on two key factors, leading to a clear taxonomy of outcomes. In a good scenario, the approach is optimal. This occurs with strong spectral alignment, meaning the structural function is well-represented by the top eigenfunctions of the conditional operator, coupled with this operator's slow eigenvalue decay, indicating a strong instrument. Performance degrades in a bad scenario: spectral alignment remains strong, but rapid eigenvalue decay (indicating a weaker instrument) demands significantly more samples for effective feature learning. Finally, in the ugly scenario, weak spectral alignment causes the method to fail, regardless of the eigenvalues' characteristics. Our synthetic experiments empirically validate this taxonomy.
Authors: Emmanuel J. Cand\`es, Andrew Ilyas, Tijana Zrnic
Abstract: Obtaining high-quality labeled datasets is often costly, requiring either extensive human annotation or expensive experiments. We propose a method that supplements such "expert" labels with AI predictions from pre-trained models to construct labeled datasets more cost-effectively. Our approach results in probably approximately correct labels: with high probability, the overall labeling error is small. This solution enables rigorous yet efficient dataset curation using modern AI models. We demonstrate the benefits of the methodology through text annotation with large language models, image labeling with pre-trained vision models, and protein folding analysis with AlphaFold.
Authors: Jiancheng Huang, Gengwei Zhang, Zequn Jie, Siyu Jiao, Yinlong Qian, Ling Chen, Yunchao Wei, Lin Ma
Abstract: Text-to-video generation has significantly enriched content creation and holds the potential to evolve into powerful world simulators. However, modeling the vast spatiotemporal space remains computationally demanding, particularly when employing Transformers, which incur quadratic complexity in sequence processing and thus limit practical applications. Recent advancements in linear-time sequence modeling, particularly the Mamba architecture, offer a more efficient alternative. Nevertheless, its plain design limits its direct applicability to multi-modal and spatiotemporal video generation tasks. To address these challenges, we introduce M4V, a Multi-Modal Mamba framework for text-to-video generation. Specifically, we propose a multi-modal diffusion Mamba (MM-DiM) block that enables seamless integration of multi-modal information and spatiotemporal modeling through a multi-modal token re-composition design. As a result, the Mamba blocks in M4V reduce FLOPs by 45% compared to the attention-based alternative when generating videos at 768$\times$1280 resolution. Additionally, to mitigate the visual quality degradation in long-context autoregressive generation processes, we introduce a reward learning strategy that further enhances per-frame visual realism. Extensive experiments on text-to-video benchmarks demonstrate M4V's ability to produce high-quality videos while significantly lowering computational costs. Code and models will be publicly available at https://huangjch526.github.io/M4V_project.
Authors: Or Shafran, Atticus Geiger, Mor Geva
Abstract: A central goal for mechanistic interpretability has been to identify the right units of analysis in large language models (LLMs) that causally explain their outputs. While early work focused on individual neurons, evidence that neurons often encode multiple concepts has motivated a shift toward analyzing directions in activation space. A key question is how to find directions that capture interpretable features in an unsupervised manner. Current methods rely on dictionary learning with sparse autoencoders (SAEs), commonly trained over residual stream activations to learn directions from scratch. However, SAEs often struggle in causal evaluations and lack intrinsic interpretability, as their learning is not explicitly tied to the computations of the model. Here, we tackle these limitations by directly decomposing MLP activations with semi-nonnegative matrix factorization (SNMF), such that the learned features are (a) sparse linear combinations of co-activated neurons, and (b) mapped to their activating inputs, making them directly interpretable. Experiments on Llama 3.1, Gemma 2 and GPT-2 show that SNMF derived features outperform SAEs and a strong supervised baseline (difference-in-means) on causal steering, while aligning with human-interpretable concepts. Further analysis reveals that specific neuron combinations are reused across semantically-related features, exposing a hierarchical structure in the MLP's activation space. Together, these results position SNMF as a simple and effective tool for identifying interpretable features and dissecting concept representations in LLMs.
Authors: Fabio Demaria
Abstract: Feature selection is a critical step in high-dimensional classification tasks, particularly under challenging conditions of double imbalance, namely settings characterized by both class imbalance in the response variable and dimensional asymmetry in the data $(n \gg p)$. In such scenarios, traditional feature selection methods applied to Random Forests (RF) often yield unstable or misleading importance rankings. This paper proposes a novel thresholding scheme for feature selection based on minimal depth, which exploits the tree topology to assess variable relevance. Extensive experiments on simulated and real-world datasets demonstrate that the proposed approach produces more parsimonious and accurate subsets of variables compared to conventional minimal depth-based selection. The method provides a practical and interpretable solution for variable selection in RF under double imbalance conditions.
Authors: Leigang Qu, Feng Cheng, Ziyan Yang, Qi Zhao, Shanchuan Lin, Yichun Shi, Yicong Li, Wenjie Wang, Tat-Seng Chua, Lu Jiang
Abstract: In-context image editing aims to modify images based on a contextual sequence comprising text and previously generated images. Existing methods typically depend on task-specific pipelines and expert models (e.g., segmentation and inpainting) to curate training data. In this work, we explore whether an in-context image editing model can be learned directly from videos. We introduce a scalable approach to annotate videos as interleaved multimodal sequences. To effectively learn from this data, we design a block-causal diffusion transformer trained on three proxy tasks: next-image prediction, current segmentation prediction, and next-segmentation prediction. Additionally, we propose a novel multi-turn image editing benchmark to advance research in this area. Extensive experiments demonstrate that our model exhibits strong in-context image editing capabilities and achieves state-of-the-art results on two multi-turn image editing benchmarks. Despite being trained exclusively on videos, our model also shows promising abilities in multi-concept composition, story generation, and chain-of-editing applications.
Authors: Jingxuan Ding, Laura Zichi, Matteo Carli, Menghang Wang, Albert Musaelian, Yu Xie, Boris Kozinsky
Abstract: Understanding and controlling the atomistic-level reactions governing the formation of the solid-electrolyte interphase (SEI) is crucial for the viability of next-generation solid state batteries. However, challenges persist due to difficulties in experimentally characterizing buried interfaces and limits in simulation speed and accuracy. We conduct large-scale explicit reactive simulations with quantum accuracy for a symmetric battery cell, {\symcell}, enabled by active learning and deep equivariant neural network interatomic potentials. To automatically characterize the coupled reactions and interdiffusion at the interface, we formulate and use unsupervised classification techniques based on clustering in the space of local atomic environments. Our analysis reveals the formation of a previously unreported crystalline disordered phase, Li$_2$S$_{0.72}$P$_{0.14}$Cl$_{0.14}$, in the SEI, that evaded previous predictions based purely on thermodynamics, underscoring the importance of explicit modeling of full reaction and transport kinetics. Our simulations agree with and explain experimental observations of the SEI formations and elucidate the Li creep mechanisms, critical to dendrite initiation, characterized by significant Li motion along the interface. Our approach is to crease a digital twin from first principles, without adjustable parameters fitted to experiment. As such, it offers capabilities to gain insights into atomistic dynamics governing complex heterogeneous processes in solid-state synthesis and electrochemistry.
Authors: Rulin Shao, Shuyue Stella Li, Rui Xin, Scott Geng, Yiping Wang, Sewoong Oh, Simon Shaolei Du, Nathan Lambert, Sewon Min, Ranjay Krishna, Yulia Tsvetkov, Hannaneh Hajishirzi, Pang Wei Koh, Luke Zettlemoyer
Abstract: We show that reinforcement learning with verifiable rewards (RLVR) can elicit strong mathematical reasoning in certain models even with spurious rewards that have little, no, or even negative correlation with the correct answer. For example, RLVR improves MATH-500 performance for Qwen2.5-Math-7B in absolute points by 21.4% (random reward), 13.8% (format reward), 24.1% (incorrect label), 26.0% (1-shot RL), and 27.1% (majority voting) -- nearly matching the 29.1% gained with ground truth rewards. However, the spurious rewards that work for Qwen often fail to yield gains with other model families like Llama3 or OLMo2. In particular, we find code reasoning -- thinking in code without actual code execution -- to be a distinctive Qwen2.5-Math behavior that becomes significantly more frequent after RLVR, from 65% to over 90%, even with spurious rewards. Overall, we hypothesize that, given the lack of useful reward signal, RLVR must somehow be surfacing useful reasoning representations learned during pretraining, although the exact mechanism remains a topic for future work. We suggest that future RLVR research should possibly be validated on diverse models rather than a single de facto choice, as we show that it is easy to get significant performance gains on Qwen models even with completely spurious reward signals.
Authors: Mozhi Zhang, Howe Tissue, Lu Wang, Xipeng Qiu
Abstract: We introduce~\textsc{Domain2Vec}, a novel approach that decomposes any dataset into a linear combination of several \emph{meta-domains}, a new concept designed to capture the key underlying features of datasets. \textsc{Domain2Vec} maintains a vocabulary of meta-domains and uses a classifier to decompose any given dataset into a domain vector that corresponds to a distribution over this vocabulary. These domain vectors enable the identification of the optimal data mixture for language model (LM) pretraining in a training-free manner under the \emph{\textbf{D}istribution \textbf{A}lignment \textbf{A}ssumption} (DA$^{2}$), which suggests that when the data distributions of the training set and the validation set are better aligned, a lower validation loss is achieved. Moreover, \textsc{Domain2vec} can be seamlessly integrated into previous works to model the relationship between domain vectors and LM performance, greatly enhancing the efficiency and scalability of previous methods. Extensive experiments demonstrate that \textsc{Domain2Vec} helps find the data mixture that enhances downstream task performance with minimal computational overhead. Specifically, \textsc{Domain2Vec} achieves the same validation loss on Pile-CC using only $51.5\%$ of the computation required when training on the original mixture of The Pile dataset. Under equivalent compute budget, \textsc{Domain2Vec} improves downstream performance by an average of $2.83\%$.
Authors: Kangwei Liu, Siyuan Cheng, Bozhong Tian, Xiaozhuan Liang, Yuyang Yin, Meng Han, Ningyu Zhang, Bryan Hooi, Xi Chen, Shumin Deng
Abstract: Large language models (LLMs) have been increasingly applied to automated harmful content detection tasks, assisting moderators in identifying policy violations and improving the overall efficiency and accuracy of content review. However, existing resources for harmful content detection are predominantly focused on English, with Chinese datasets remaining scarce and often limited in scope. We present a comprehensive, professionally annotated benchmark for Chinese content harm detection, which covers six representative categories and is constructed entirely from real-world data. Our annotation process further yields a knowledge rule base that provides explicit expert knowledge to assist LLMs in Chinese harmful content detection. In addition, we propose a knowledge-augmented baseline that integrates both human-annotated knowledge rules and implicit knowledge from large language models, enabling smaller models to achieve performance comparable to state-of-the-art LLMs. Code and data are available at https://github.com/zjunlp/ChineseHarm-bench.
Authors: Yuanhui Huang, Weiliang Chen, Wenzhao Zheng, Yueqi Duan, Jie Zhou, Jiwen Lu
Abstract: Autoregressive visual generation has garnered increasing attention due to its scalability and compatibility with other modalities compared with diffusion models. Most existing methods construct visual sequences as spatial patches for autoregressive generation. However, image patches are inherently parallel, contradicting the causal nature of autoregressive modeling. To address this, we propose a Spectral AutoRegressive (SpectralAR) visual generation framework, which realizes causality for visual sequences from the spectral perspective. Specifically, we first transform an image into ordered spectral tokens with Nested Spectral Tokenization, representing lower to higher frequency components. We then perform autoregressive generation in a coarse-to-fine manner with the sequences of spectral tokens. By considering different levels of detail in images, our SpectralAR achieves both sequence causality and token efficiency without bells and whistles. We conduct extensive experiments on ImageNet-1K for image reconstruction and autoregressive generation, and SpectralAR achieves 3.02 gFID with only 64 tokens and 310M parameters. Project page: https://huang-yh.github.io/spectralar/.
Authors: He Ye, Rojas Kevin, Tao Molei
Abstract: We study masked discrete diffusion models with classifier-free guidance (CFG). Assuming no score error nor discretization error, we derive an explicit solution to the guided reverse dynamics, so that how guidance influences the sampling behavior can be precisely characterized. When the full data distribution is a mixture over classes and the goal is to sample from a specific class, guidance amplifies class-specific regions while suppresses regions shared with other classes. This effect depends on the guidance strength $w$ and induces distinct covariance structures in the sampled distribution. Notably, we observe quantitatively different behaviors in $1$D and $2$D. We also show that for large $w$, the decay rate of the total variation ($\mathrm{TV}$) along the reverse dynamics is double-exponential in $w$ for both $1$D and $2$D. These findings highlight the role of guidance, not just in shaping the output distribution, but also in controlling the dynamics of the sampling trajectory. Our theoretical analysis is supported by experiments that illustrate the geometric effects of guidance and its impact on convergence.
Authors: Yixin Ou, Yujie Luo, Jingsheng Zheng, Lanning Wei, Shuofei Qiao, Jintian Zhang, Da Zheng, Huajun Chen, Ningyu Zhang
Abstract: Large Language Model (LLM) agents have shown great potential in addressing real-world data science problems. LLM-driven data science agents promise to automate the entire machine learning pipeline, yet their real-world effectiveness remains limited. Existing frameworks depend on rigid, pre-defined workflows and inflexible coding strategies; consequently, they excel only on relatively simple, classical problems and fail to capture the empirical expertise that human practitioners bring to complex, innovative tasks. In this work, we introduce AutoMind, an adaptive, knowledgeable LLM-agent framework that overcomes these deficiencies through three key advances: (1) a curated expert knowledge base that grounds the agent in domain expert knowledge, (2) an agentic knowledgeable tree search algorithm that strategically explores possible solutions, and (3) a self-adaptive coding strategy that dynamically tailors code generation to task complexity. Evaluations on two automated data science benchmarks demonstrate that AutoMind delivers superior performance versus state-of-the-art baselines. Additional analyses confirm favorable effectiveness, efficiency, and qualitative solution quality, highlighting AutoMind as an efficient and robust step toward fully automated data science.
Authors: Donghoon Ahn, Jiwon Kang, Sanghyun Lee, Minjae Kim, Jaewon Min, Wooseok Jang, Saungwu Lee, Sayak Paul, Susung Hong, Seungryong Kim
Abstract: Recent guidance methods in diffusion models steer reverse sampling by perturbing the model to construct an implicit weak model and guide generation away from it. Among these approaches, attention perturbation has demonstrated strong empirical performance in unconditional scenarios where classifier-free guidance is not applicable. However, existing attention perturbation methods lack principled approaches for determining where perturbations should be applied, particularly in Diffusion Transformer (DiT) architectures where quality-relevant computations are distributed across layers. In this paper, we investigate the granularity of attention perturbations, ranging from the layer level down to individual attention heads, and discover that specific heads govern distinct visual concepts such as structure, style, and texture quality. Building on this insight, we propose "HeadHunter", a systematic framework for iteratively selecting attention heads that align with user-centric objectives, enabling fine-grained control over generation quality and visual attributes. In addition, we introduce SoftPAG, which linearly interpolates each selected head's attention map toward an identity matrix, providing a continuous knob to tune perturbation strength and suppress artifacts. Our approach not only mitigates the oversmoothing issues of existing layer-level perturbation but also enables targeted manipulation of specific visual styles through compositional head selection. We validate our method on modern large-scale DiT-based text-to-image models including Stable Diffusion 3 and FLUX.1, demonstrating superior performance in both general quality enhancement and style-specific guidance. Our work provides the first head-level analysis of attention perturbation in diffusion models, uncovering interpretable specialization within attention layers and enabling practical design of effective perturbation strategies.
Authors: Chara Podimata, Aleksandrs Slivkins
Abstract: Lipschitz bandits is a prominent version of multi-armed bandits that studies large, structured action spaces such as the $[0,1]$ interval, where similar actions are guaranteed to have similar rewards. A central theme here is the adaptive discretization of the action space, which gradually ``zooms in'' on the more promising regions thereof. The goal is to take advantage of ``nicer'' problem instances, while retaining near-optimal worst-case performance. While the stochastic version of the problem is well-understood, the general version with adversarial rewards is not. We provide the first algorithm (\emph{Adversarial Zooming}) for adaptive discretization in the adversarial version, and derive instance-dependent regret bounds. In particular, we recover the worst-case optimal regret bound for the adversarial version, and the instance-dependent regret bound for the stochastic version. We apply our algorithm to several fundamental applications -- including dynamic pricing and auction reserve tuning -- all under adversarial reward models. While these domains often violate Lipschitzness, our analysis only requires a weaker version thereof, allowing for meaningful regret bounds without additional smoothness assumptions. Notably, we extend our results to multi-product dynamic pricing with non-smooth reward structures, a setting which does not even satisfy one-sided Lipschitzness.
Authors: Valentino Delle Rose, Alexander Kozachinskiy, Crist\'obal Rojas, Mircea Petrache, Pablo Barcel\'o
Abstract: The Weisfeiler--Lehman (WL) test is a fundamental iterative algorithm for checking isomorphism of graphs. It has also been observed that it underlies the design of several graph neural network architectures, whose capabilities and performance can be understood in terms of the expressive power of this test. Motivated by recent developments in machine learning applications to datasets involving three-dimensional objects, we study when the WL test is {\em complete} for clouds of euclidean points represented by complete distance graphs, i.e., when it can distinguish, up to isometry, any arbitrary such cloud. %arbitrary clouds of euclidean points represented by complete distance graphs. % How many dimensions of the Weisfeiler--Lehman test is enough to distinguish any two non-isometric point clouds in $d$-dimensional Euclidean space, assuming that these point clouds are given as complete graphs labeled by distances between the points? This question is important for understanding, which architectures of graph neural networks are capable of fully exploiting the spacial structure of a point cloud. Our main result states that the $(d-1)$-dimensional WL test is complete for point clouds in $d$-dimensional Euclidean space, for any $d\ge 2$, and that only three iterations of the test suffice. We also observe that the $d$-dimensional WL test only requires one iteration to achieve completeness. Our paper thus provides complete understanding of the 3-dimensional case: it was shown in previous works that 1-WL is not complete in $\mathbb{R}^3$, and we show that 2-WL is complete there. We also strengthen the lower bound for 1-WL by showing that it is unable to recognize planar point clouds in $\mathbb{R}^3$. Finally, we show that 2-WL is not complete in $\mathbb{R}^6$, leaving as an open question, whether it is complete in $\mathbb{R}^{d}$ for $d = 4,5$.
Authors: Liu Ziyin, Hongchao Li, Masahito Ueda
Abstract: The stochastic gradient descent (SGD) algorithm is the algorithm we use to train neural networks. However, it remains poorly understood how the SGD navigates the highly nonlinear and degenerate loss landscape of a neural network. In this work, we show that the minibatch noise of SGD regularizes the solution towards a noise-balanced solution whenever the loss function contains a rescaling parameter symmetry. Because the difference between a simple diffusion process and SGD dynamics is the most significant when symmetries are present, our theory implies that the loss function symmetries constitute an essential probe of how SGD works. We then apply this result to derive the stationary distribution of stochastic gradient flow for a diagonal linear network with arbitrary depth and width. The stationary distribution exhibits complicated nonlinear phenomena such as phase transitions, broken ergodicity, and fluctuation inversion. These phenomena are shown to exist uniquely in deep networks, implying a fundamental difference between deep and shallow models.
Authors: Samuel E. Otto, Nicholas Zolman, J. Nathan Kutz, Steven L. Brunton
Abstract: Symmetry is present throughout nature and continues to play an increasingly central role in physics and machine learning. Fundamental symmetries, such as Poincar\'{e} invariance, allow physical laws discovered in laboratories on Earth to be extrapolated to the farthest reaches of the universe. Symmetry is essential to achieving this extrapolatory power in machine learning applications. For example, translation invariance in image classification allows models with fewer parameters, such as convolutional neural networks, to be trained on smaller data sets and achieve state-of-the-art performance. In this paper, we provide a unifying theoretical and methodological framework for incorporating symmetry into machine learning models in three ways: 1. enforcing known symmetry when training a model; 2. discovering unknown symmetries of a given model or data set; and 3. promoting symmetry during training by learning a model that breaks symmetries within a user-specified group of candidates when there is sufficient evidence in the data. We show that these tasks can be cast within a common mathematical framework whose central object is the Lie derivative associated with fiber-linear Lie group actions on vector bundles. We extend and unify several existing results by showing that enforcing and discovering symmetry are linear-algebraic tasks that are dual with respect to the bilinear structure of the Lie derivative. We also propose a novel way to promote symmetry by introducing a class of convex regularization functions based on the Lie derivative and nuclear norm relaxation to penalize symmetry breaking during training of machine learning models. We explain how these ideas can be applied to a wide range of machine learning models including basis function regression, dynamical systems discovery, neural networks, and neural operators acting on fields.
Authors: Charles-\'Etienne Joseph, Benjamin Th\'erien, Abhinav Moudgil, Boris Knyazev, Eugene Belilovsky
Abstract: Communication-efficient variants of SGD, specifically local SGD, have received a great deal of interest in recent years. These approaches compute multiple gradient steps locally on each worker, before averaging model parameters, helping relieve the critical communication bottleneck in distributed deep learning training. Although many variants of these approaches have been proposed, they can sometimes lag behind state-of-the-art adaptive optimizers for deep learning. In this work, we investigate if the recent progress in the emerging area of learned optimizers can potentially close this gap in homogeneous data and homogeneous device settings while remaining communication-efficient. Specifically, we meta-learn how to perform global updates given an update from local SGD iterations. Our results demonstrate that learned optimizers can substantially outperform local SGD and its sophisticated variants while maintaining their communication efficiency. Our learned optimizers can even generalize to unseen and much larger datasets and architectures, including ImageNet and ViTs, and to unseen modalities such as language modeling. We therefore show the potential of learned optimizers for improving communication-efficient distributed learning.
Authors: Shyam Nuggehalli, Jifan Zhang, Lalit Jain, Robert Nowak
Abstract: Class imbalance severely impacts machine learning performance on minority classes in real-world applications. While various solutions exist, active learning offers a fundamental fix by strategically collecting balanced, informative labeled examples from abundant unlabeled data. We introduce DIRECT, an algorithm that identifies class separation boundaries and selects the most uncertain nearby examples for annotation. By reducing the problem to one-dimensional active learning, DIRECT leverages established theory to handle batch labeling and label noise -- another common challenge in data annotation that particularly affects active learning methods. Our work presents the first comprehensive study of active learning under both class imbalance and label noise. Extensive experiments on imbalanced datasets show DIRECT reduces annotation costs by over 60\% compared to state-of-the-art active learning methods and over 80\% versus random sampling, while maintaining robustness to label noise.
Authors: Longkun Guo, Chaoqi Jia, Kewen Liao, Zhigang Lu, Minhui Xue
Abstract: Center-based clustering has attracted significant research interest from both theory and practice. In many practical applications, input data often contain background knowledge that can be used to improve clustering results. In this work, we build on widely adopted $k$-center clustering and model its input background knowledge as must-link (ML) and cannot-link (CL) constraint sets. However, most clustering problems including $k$-center are inherently $\mathcal{NP}$-hard, while the more complex constrained variants are known to suffer severer approximation and computation barriers that significantly limit their applicability. By employing a suite of techniques including reverse dominating sets, linear programming (LP) integral polyhedron, and LP duality, we arrive at the first efficient approximation algorithm for constrained $k$-center with the best possible ratio of 2. We also construct competitive baseline algorithms and empirically evaluate our approximation algorithm against them on a variety of real datasets. The results validate our theoretical findings and demonstrate the great advantages of our algorithm in terms of clustering cost, clustering quality, and running time.
Authors: Linxi Zhao, Yihe Deng, Weitong Zhang, Quanquan Gu
Abstract: The advancement of Large Vision-Language Models (LVLMs) has increasingly highlighted the critical issue of their tendency to hallucinate non-existing objects in the images. To address this issue, previous works focused on using specially curated datasets or powerful LLMs to rectify the outputs of LVLMs. However, these approaches require either costly training or fine-tuning, or API access to proprietary LLMs for post-generation correction. In response to these limitations, we propose Mitigating hallucinAtion via image-gRounded guIdaNcE (MARINE), a framework that is both training-free and API-free. MARINE effectively and efficiently reduces object hallucinations during inference by introducing image-grounded guidance to LVLMs. This is achieved by leveraging open-source vision models to extract object-level information, thereby enhancing the precision of LVLM-generated content. Our framework's flexibility further allows for the integration of multiple vision models, enabling more reliable and robust object-level guidance. Through comprehensive evaluations across 5 popular LVLMs with diverse evaluation metrics and benchmarks, we demonstrate the effectiveness of MARINE, which even outperforms existing fine-tuning-based methods. Remarkably, it reduces hallucinations consistently in GPT-4V-assisted evaluation while maintaining the detailedness of LVLMs' generations. We release our code at https://github.com/Linxi-ZHAO/MARINE.
Authors: Xavier Mart\'inez Lua\~na, Rebeca P. D\'iaz Redondo, Manuel Fern\'andez Veiga
Abstract: Federated Learning (FL) is an interesting strategy that enables the collaborative training of an AI model among different data owners without revealing their private datasets. Even so, FL has some privacy vulnerabilities that have been tried to be overcome by applying some techniques like Differential Privacy (DP), Homomorphic Encryption (HE), or Secure Multi-Party Computation (SMPC). However, these techniques have some important drawbacks that might narrow their range of application: problems to work with non-linear functions and to operate large matrix multiplications and high communication and computational costs to manage semi-honest nodes. In this context, we propose a solution to guarantee privacy in FL schemes that simultaneously solves the previously mentioned problems. Our proposal is based on the Berrut Approximated Coded Computing, a technique from the Coded Distributed Computing paradigm, adapted to a Secret Sharing configuration, to provide input privacy to FL in a scalable way. It can be applied for computing non-linear functions and treats the special case of distributed matrix multiplication, a key primitive at the core of many automated learning tasks. Because of these characteristics, it could be applied in a wide range of FL scenarios, since it is independent of the machine learning models or aggregation algorithms used in the FL scheme. We provide analysis of the achieved privacy and complexity of our solution and, due to the extensive numerical results performed, a good trade-off between privacy and precision can be observed.
Authors: Jinseong Park, Seungyun Lee, Woojin Jeong, Yujin Choi, Jaewook Lee
Abstract: Time series generation is widely used in real-world applications such as simulation, data augmentation, and hypothesis testing. Recently, diffusion models have emerged as the de facto approach to time series generation, enabling diverse synthesis scenarios. However, the fixed standard-Gaussian diffusion prior may be ill-suited for general time series data, such as temporal order and fixed points. In this paper, we propose TimeBridge, a framework that flexibly synthesizes time series data by using diffusion bridges to learn paths between a chosen prior and the data distribution. We then explore several prior designs tailored to time series synthesis. Our framework covers (i) data- and time-dependent priors for unconditional generation and (ii) scale-preserving priors for conditional generation. Experiments show that our framework with data-driven priors outperforms standard diffusion models on time series generation.
Authors: Hongyang Lei, Xiaolong Cheng, Qi Qin, Dan Wang, Kun Fan, Huazhen Huang, Qingqing Gu, Yetao Wu, Zhonglin Jiang, Yong Chen, Luo Ji
Abstract: Current multimodal learning strategies primarily optimize in the original token space. Such a framework is easy to incorporate with the backbone of pretrained language model, but might result in modality collapse. To alleviate such issues, we leverage the joint embedding predictive architecture (JEPA) on the multimodal tasks, which converts the input embedding into the output embedding space by a predictor and then conducts the cross-modal alignment on the latent space. We implement this predictor by a Multi-Gate Mixture of Experts (MMoE) and name the framework as M3-JEPA, accordingly. The gating function disentangles the modality-specific and shared information and derives information-theoretic optimality. The framework is implemented with both contrastive and regularization loss, and solved by alternative gradient descent (AGD) between different multimodal tasks. By thoroughly designed experiments, we show that M3-JEPA can obtain state-of-the-art performance on different modalities and tasks, generalize to unseen datasets and domains, and is computationally efficient in both training and inference. Our observation suggests that M3-JEPA might become a new basis to self-supervised learning in the open world.
Authors: Sourav Chatterjee, Timothy Sudijono
Abstract: We show that feedforward neural networks with ReLU activation generalize on low complexity data, suitably defined. Given i.i.d.~data generated from a simple programming language, the minimum description length (MDL) feedforward neural network which interpolates the data generalizes with high probability. We define this simple programming language, along with a notion of description length of such networks. We provide several examples on basic computational tasks, such as checking primality of a natural number. For primality testing, our theorem shows the following and more. Suppose that we draw an i.i.d.~sample of $n$ numbers uniformly at random from $1$ to $N$. For each number $x_i$, let $y_i = 1$ if $x_i$ is a prime and $0$ if it is not. Then, the interpolating MDL network accurately answers, with probability $1- O((\ln N)/n)$, whether a newly drawn number between $1$ and $N$ is a prime or not. Note that the network is not designed to detect primes; minimum description learning discovers a network which does so. Extensions to noisy data are also discussed, suggesting that MDL neural network interpolators can demonstrate tempered overfitting.
Authors: Yilin Wang, Yifei Yu, Kong Sun, Peixuan Lei, Yuxuan Zhang, Enrico Zio, Aiguo Xia, Yuanxiang Li
Abstract: In industry, the reliability of rotating machinery is critical for production efficiency and safety. Current methods of Prognostics and Health Management (PHM) often rely on task-specific models, which face significant challenges in handling diverse datasets with varying signal characteristics, fault modes and operating conditions. Inspired by advancements in generative pretrained models, we propose RmGPT, a unified model for diagnosis and prognosis tasks. RmGPT introduces a novel generative token-based framework, incorporating Signal Tokens, Prompt Tokens, Time-Frequency Task Tokens and Fault Tokens to handle heterogeneous data within a unified model architecture. We leverage self-supervised learning for robust feature extraction and introduce a next signal token prediction pretraining strategy, alongside efficient prompt learning for task-specific adaptation. Extensive experiments demonstrate that RmGPT significantly outperforms state-of-the-art algorithms, achieving near-perfect accuracy in diagnosis tasks and exceptionally low errors in prognosis tasks. Notably, RmGPT excels in few-shot learning scenarios, achieving 82\% accuracy in 16-class one-shot experiments, highlighting its adaptability and robustness. This work establishes RmGPT as a powerful PHM foundation model for rotating machinery, advancing the scalability and generalizability of PHM solutions. \textbf{Code is available at: https://github.com/Pandalin98/RmGPT.
Authors: Etienne Boursier, Nicolas Flammarion
Abstract: Understanding generalization of overparametrized neural networks remains a fundamental challenge in machine learning. Most of the literature mostly studies generalization from an interpolation point of view, taking convergence of parameters towards a global minimum of the training loss for granted. While overparametrized architectures indeed interpolated the data for typical classification tasks, this interpolation paradigm does not seem valid anymore for more complex tasks such as in-context learning or diffusion. Instead for such tasks, it has been empirically observed that the trained models goes from global minima to spurious local minima of the training loss as the number of training samples becomes larger than some level we call optimization threshold. While the former yields a poor generalization to the true population loss, the latter was observed to actually correspond to the minimiser of this true loss. This paper explores theoretically this phenomenon in the context of two-layer ReLU networks. We demonstrate that, despite overparametrization, networks often converge toward simpler solutions rather than interpolating the training data, which can lead to a drastic improvement on the test loss with respect to interpolating solutions. Our analysis relies on the so called early alignment phase, during which neurons align towards specific directions. This directional alignment, which occurs in the early stage of training, leads to a simplicity bias, wherein the network approximates the ground truth model without converging to the global minimum of the training loss. Our results suggest that this bias, resulting in an optimization threshold from which interpolation is not reached anymore, is beneficial and enhances the generalization of trained models.
Authors: Thomas Rupf, Marco Bagatella, Nico G\"urtler, Jonas Frey, Georg Martius
Abstract: Zero-shot imitation learning algorithms hold the promise of reproducing unseen behavior from as little as a single demonstration at test time. Existing practical approaches view the expert demonstration as a sequence of goals, enabling imitation with a high-level goal selector, and a low-level goal-conditioned policy. However, this framework can suffer from myopic behavior: the agent's immediate actions towards achieving individual goals may undermine long-term objectives. We introduce a novel method that mitigates this issue by directly optimizing the occupancy matching objective that is intrinsic to imitation learning. We propose to lift a goal-conditioned value function to a distance between occupancies, which are in turn approximated via a learned world model. The resulting method can learn from offline, suboptimal data, and is capable of non-myopic, zero-shot imitation, as we demonstrate in complex, continuous benchmarks. The code is available at https://github.com/martius-lab/zilot.
Authors: Khurram Yamin, Shantanu Gupta, Gaurav R. Ghosal, Zachary C. Lipton, Bryan Wilder
Abstract: In this work, we investigate the causal reasoning abilities of large language models (LLMs) through the representative problem of inferring causal relationships from narratives. We find that even state-of-the-art language models rely on unreliable shortcuts, both in terms of the narrative presentation and their parametric knowledge. For example, LLMs tend to determine causal relationships based on the topological ordering of events (i.e., earlier events cause later ones), resulting in lower performance whenever events are not narrated in their exact causal order. Similarly, we demonstrate that LLMs struggle with long-term causal reasoning and often fail when the narratives are long and contain many events. Additionally, we show LLMs appear to rely heavily on their parametric knowledge at the expense of reasoning over the provided narrative. This degrades their abilities whenever the narrative opposes parametric knowledge. We extensively validate these failure modes through carefully controlled synthetic experiments, as well as evaluations on real-world narratives. Finally, we observe that explicitly generating a causal graph generally improves performance while naive chain-of-thought is ineffective. Collectively, our results distill precise failure modes of current state-of-the-art models and can pave the way for future techniques to enhance causal reasoning in LLMs.
Authors: Julien Nicolas, C\'esar Sabater, Mohamed Maouche, Sonia Ben Mokhtar, Mark Coates
Abstract: The randomized power method has gained significant interest due to its simplicity and efficient handling of large-scale spectral analysis and recommendation tasks. However, its application to large datasets containing personal information (e.g., web interactions, search history, personal tastes) raises critical privacy problems. This paper addresses these issues by proposing enhanced privacy-preserving variants of the method. First, we propose a variant that reduces the amount of the noise required in current techniques to achieve Differential Privacy (DP). More precisely, we refine the privacy analysis so that the Gaussian noise variance no longer grows linearly with the target rank, achieving the same DP guarantees with strictly less noise. Second, we adapt our method to a decentralized framework in which data is distributed among multiple users. The decentralized protocol strengthens privacy guarantees with no accuracy penalty and a low computational and communication overhead. Our results include the provision of tighter convergence bounds for both the centralized and decentralized versions, and an empirical comparison with previous work using real recommendation datasets.
Authors: Daniel de Vassimon Manela, Linying Yang, Robin J. Evans
Abstract: Ensuring robust model performance in diverse real-world scenarios requires addressing generalizability across domains with covariate shifts. However, no formal procedure exists for statistically evaluating generalizability in machine learning algorithms. Existing predictive metrics like mean squared error (MSE) help to quantify the relative performance between models, but do not directly answer whether a model can or cannot generalize. To address this gap in the domain of causal inference, we propose a systematic framework for statistically evaluating the generalizability of high-dimensional causal inference models. Our approach uses the frugal parameterization to flexibly simulate from fully and semi-synthetic causal benchmarks, offering a comprehensive evaluation for both mean and distributional regression methods. Grounded in real-world data, our method ensures more realistic evaluations, which is often missing in current work relying on simplified datasets. Furthermore, using simulations and statistical testing, our framework is robust and avoids over-reliance on conventional metrics, providing statistical safeguards for decision making.
Authors: Joseph Pollock, Igor Shilov, Euodia Dodd, Yves-Alexandre de Montjoye
Abstract: Membership inference attacks (MIAs) are widely used to empirically assess privacy risks in machine learning models, both providing model-level vulnerability metrics and identifying the most vulnerable training samples. State-of-the-art methods, however, require training hundreds of shadow models with the same architecture as the target model. This makes the computational cost of assessing the privacy of models prohibitive for many practical applications, particularly when used iteratively as part of the model development process and for large models. We propose a novel approach for identifying the training samples most vulnerable to membership inference attacks by analyzing artifacts naturally available during the training process. Our method, Loss Trace Interquartile Range (LT-IQR), analyzes per-sample loss trajectories collected during model training to identify high-risk samples without requiring any additional model training. Through experiments on standard benchmarks, we demonstrate that LT-IQR achieves 92% precision@k=1% in identifying the samples most vulnerable to state-of-the-art MIAs. This result holds across datasets and model architectures with LT-IQR outperforming both traditional vulnerability metrics, such as loss, and lightweight MIAs using few shadow models. We also show LT-IQR to accurately identify points vulnerable to multiple MIA methods and perform ablation studies. We believe LT-IQR enables model developers to identify vulnerable training samples, for free, as part of the model development process. Our results emphasize the potential of artifact-based methods to efficiently evaluate privacy risks.
Authors: Peter St. John, Dejun Lin, Polina Binder, Malcolm Greaves, Vega Shah, John St. John, Adrian Lange, Patrick Hsu, Rajesh Illango, Arvind Ramanathan, Anima Anandkumar, David H Brookes, Akosua Busia, Abhishaike Mahajan, Stephen Malina, Neha Prasad, Sam Sinai, Lindsay Edwards, Thomas Gaudelet, Cristian Regep, Martin Steinegger, Burkhard Rost, Alexander Brace, Kyle Hippe, Luca Naef, Keisuke Kamata, George Armstrong, Kevin Boyd, Zhonglin Cao, Han-Yi Chou, Simon Chu, Allan dos Santos Costa, Sajad Darabi, Eric Dawson, Kieran Didi, Cong Fu, Mario Geiger, Michelle Gill, Darren J Hsu, Gagan Kaushik, Maria Korshunova, Steven Kothen-Hill, Youhan Lee, Meng Liu, Micha Livne, Zachary McClure, Jonathan Mitchell, Alireza Moradzadeh, Ohad Mosafi, Youssef Nashed, Saee Paliwal, Yuxing Peng, Sara Rabhi, Farhad Ramezanghorbani, Danny Reidenbach, Camir Ricketts, Brian C Roland, Kushal Shah, Tyler Shimko, Hassan Sirelkhatim, Savitha Srinivasan, Abraham C Stern, Dorota Toczydlowska, Srimukh Prasad Veccham, Niccol\`o Alberto Elia Venanzi, Anton Vorontsov, Jared Wilber, Isabel Wilkinson, Wei Jing Wong, Eva Xue, Cory Ye, Xin Yu, Yang Zhang, Guoqing Zhou, Becca Zandstein, Alejandro Chac\`on, Prashant Sohani, Maximilian Stadler, Christian Hundt, Feiwen Zhu, Christian Dallago, Bruno Trentini, Emine Kucukbenli, Saee Paliwal, Timur Rvachov, Eddie Calleja, Johnny Israeli, Harry Clifford, Risto Haukioja, Nicholas Haemel, Kyle Tretina, Neha Tadimeti, Anthony B Costa
Abstract: Artificial Intelligence models encoding biology and chemistry are opening new routes to high-throughput and high-quality in-silico drug development. However, their training increasingly relies on computational scale, with recent protein language models (pLM) training on hundreds of graphical processing units (GPUs). We introduce the BioNeMo Framework to facilitate the training of computational biology and chemistry AI models across hundreds of GPUs. Its modular design allows the integration of individual components, such as data loaders, into existing workflows and is open to community contributions. We detail technical features of the BioNeMo Framework through use cases such as pLM pre-training and fine-tuning. On 256 NVIDIA A100s, BioNeMo Framework trains a three billion parameter BERT-based pLM on over one trillion tokens in 4.2 days. The BioNeMo Framework is open-source and free for everyone to use.
Authors: Erica Coppolillo, Federico Cinus, Marco Minici, Francesco Bonchi, Giuseppe Manco
Abstract: Large Language Models (LLMs) demonstrate significant persuasive capabilities in one-on-one interactions, but their influence within social networks, where interconnected users and complex opinion dynamics pose unique challenges, remains underexplored. This paper addresses the research question: \emph{Can LLMs generate meaningful content that maximizes user engagement on social networks?} To answer this, we propose a pipeline using reinforcement learning with simulated feedback, where the network's response to LLM-generated content (i.e., the reward) is simulated through a formal engagement model. This approach bypasses the temporal cost and complexity of live experiments, enabling an efficient feedback loop between the LLM and the network under study. It also allows to control over endogenous factors such as the LLM's position within the social network and the distribution of opinions on a given topic. Our approach is adaptive to the opinion distribution of the underlying network and agnostic to the specifics of the engagement model, which is embedded as a plug-and-play component. Such flexibility makes it suitable for more complex engagement tasks and interventions in computational social science. Using our framework, we analyze the performance of LLMs in generating social engagement under different conditions, showcasing their full potential in this task. The experimental code is publicly available at https://github.com/mminici/Engagement-Driven-Content-Generation.
URLs: https://github.com/mminici/Engagement-Driven-Content-Generation.
Authors: Charita Dellaporta, Patrick O'Hara, Theodoros Damoulas
Abstract: Decision making under uncertainty is challenging as the data-generating process (DGP) is often unknown. Bayesian inference proceeds by estimating the DGP through posterior beliefs on the model's parameters. However, minimising the expected risk under these beliefs can lead to suboptimal decisions due to model uncertainty or limited, noisy observations. To address this, we introduce Distributionally Robust Optimisation with Bayesian Ambiguity Sets (DRO-BAS) which hedges against model uncertainty by optimising the worst-case risk over a posterior-informed ambiguity set. We provide two such sets, based on posterior expectations (DRO-BAS(PE)) or posterior predictives (DRO-BAS(PP)) and prove that both admit, under conditions, strong dual formulations leading to efficient single-stage stochastic programs which are solved with a sample average approximation. For DRO-BAS(PE) this covers all conjugate exponential family members while for DRO-BAS(PP) this is shown under conditions on the predictive's moment generating function. Our DRO-BAS formulations outperform existing Bayesian DRO on the Newsvendor problem and achieve faster solve times with comparable robustness on the Portfolio problem.
Authors: Etienne Lasalle (OCKHAM), R\'emi Vaudaine (OCKHAM), Titouan Vayer (OCKHAM), Pierre Borgnat (Phys-ENS), R\'emi Gribonval (OCKHAM), Paulo Gon\c{c}alves (OCKHAM), M\`arton Karsai (CEU)
Abstract: Clustering the nodes of a graph is a cornerstone of graph analysis and has been extensively studied. However, some popular methods are not suitable for very large graphs: e.g., spectral clustering requires the computation of the spectral decomposition of the Laplacian matrix, which is not applicable for large graphs with a large number of communities. This work introduces PASCO, an overlay that accelerates clustering algorithms. Our method consists of three steps: 1-We compute several independent small graphs representing the input graph by applying an efficient and structure-preserving coarsening algorithm. 2-A clustering algorithm is run in parallel onto each small graph and provides several partitions of the initial graph. 3-These partitions are aligned and combined with an optimal transport method to output the final partition. The PASCO framework is based on two key contributions: a novel global algorithm structure designed to enable parallelization and a fast, empirically validated graph coarsening algorithm that preserves structural properties. We demonstrate the strong performance of 1 PASCO in terms of computational efficiency, structural preservation, and output partition quality, evaluated on both synthetic and real-world graph datasets.
Authors: Jiaming Yu, Le Liang, Chongtao Guo, Ziyang Guo, Shi Jin, Geoffrey Ye Li
Abstract: This paper investigates the use of multi-agent reinforcement learning (MARL) to address distributed channel access in wireless local area networks. In particular, we consider the challenging yet more practical case where the agents heterogeneously adopt value-based or policy-based reinforcement learning algorithms to train the model. We propose a heterogeneous MARL training framework, named QPMIX, which adopts a centralized training with distributed execution paradigm to enable heterogeneous agents to collaborate. Moreover, we theoretically prove the convergence of the proposed heterogeneous MARL method when using the linear value function approximation. Our method maximizes the network throughput and ensures fairness among stations, therefore, enhancing the overall network performance. Simulation results demonstrate that the proposed QPMIX algorithm improves throughput, mean delay, delay jitter, and collision rates compared with conventional carrier-sense multiple access with collision avoidance (CSMA/CA) mechanism in the saturated traffic scenario. Furthermore, the QPMIX algorithm is robust in unsaturated and delay-sensitive traffic scenarios. It coexists well with the conventional CSMA/CA mechanism and promotes cooperation among heterogeneous agents.
Authors: Seyed Mahdi B. Azad, Zahra Padar, Gabriel Kalweit, Joschka Boedecker
Abstract: In this paper, we propose a novel method for learning reward functions directly from offline demonstrations. Unlike traditional inverse reinforcement learning (IRL), our approach decouples the reward function from the learner's policy, eliminating the adversarial interaction typically required between the two. This results in a more stable and efficient training process. Our reward function, called \textit{SR-Reward}, leverages successor representation (SR) to encode a state based on expected future states' visitation under the demonstration policy and transition dynamics. By utilizing the Bellman equation, SR-Reward can be learned concurrently with most reinforcement learning (RL) algorithms without altering the existing training pipeline. We also introduce a negative sampling strategy to mitigate overestimation errors by reducing rewards for out-of-distribution data, thereby enhancing robustness. This strategy inherently introduces a conservative bias into RL algorithms that employ the learned reward. We evaluate our method on the D4RL benchmark, achieving competitive results compared to offline RL algorithms with access to true rewards and imitation learning (IL) techniques like behavioral cloning. Moreover, our ablation studies on data size and quality reveal the advantages and limitations of SR-Reward as a proxy for true rewards.
Authors: Chuqin Geng, Xiaojie Xu, Anqi Xing, Ziyu Zhao, Xujie Si
Abstract: As deep neural networks continue to excel across various domains, their black-box nature has raised concerns about transparency and trust. In particular, interpretability has become increasingly essential for applications that demand high safety and knowledge rigor, such as drug discovery, autonomous driving, and genomics. However, progress in understanding even the simplest deep neural networks - such as fully connected networks - has been limited, despite their role as foundational elements in state-of-the-art models like ResNet and Transformer. In this paper, we address this challenge by introducing NeuroLogic, a novel approach for decoding interpretable logic rules from neural networks. NeuroLogic leverages neural activation patterns to capture the model's critical decision-making processes, translating them into logical rules represented by hidden predicates. Thanks to its flexible design in the grounding phase, NeuroLogic can be adapted to a wide range of neural networks. For simple fully connected neural networks, hidden predicates can be grounded in certain split patterns of original input features to derive decision-tree-like rules. For large, complex vision neural networks, NeuroLogic grounds hidden predicates into high-level visual concepts that are understandable to humans. Our empirical study demonstrates that NeuroLogic can extract global and interpretable rules from state-of-the-art models such as ResNet, a task at which existing work struggles. We believe NeuroLogic can help pave the way for understanding the black-box nature of neural networks.
Authors: Yarden Tzach, Yuval Meir, Ronit D. Gross, Ofek Tevet, Ella Koresh, Ido Kanter
Abstract: Pruning the parameters and structure of neural networks reduces the computational complexity, energy consumption, and latency during inference. Recently, a novel underlying mechanism for successful deep learning (DL) was presented based on a method that quantitatively measures the single filter performance in each layer of a DL architecture, and a new comprehensive mechanism of how deep learning works was presented. This statistical mechanics inspired viewpoint enables to reveal the macroscopic behavior of the entire network from the microscopic performance of each filter and their cooperative behavior. Herein, we demonstrate how this understanding paves the path to high quenched dilution of the convolutional layers of deep architectures without affecting their overall accuracy using applied filter cluster connections (AFCC). AFCC is exemplified on VGG-11 and EfficientNet-B0 architectures trained on CIFAR-100, and its high pruning outperforms other techniques using the same pruning magnitude. Additionally, this technique is broadened to single nodal performance and highly pruning of fully connected layers, suggesting a possible implementation to considerably reduce the complexity of over-parameterized AI tasks.
Authors: Xinshuai Dong, Ignavier Ng, Boyang Sun, Haoyue Dai, Guang-Yuan Hao, Shunxing Fan, Peter Spirtes, Yumou Qiu, Kun Zhang
Abstract: Recent advances have shown that statistical tests for the rank of cross-covariance matrices play an important role in causal discovery. These rank tests include partial correlation tests as special cases and provide further graphical information about latent variables. Existing rank tests typically assume that all the continuous variables can be perfectly measured, and yet, in practice many variables can only be measured after discretization. For example, in psychometric studies, the continuous level of certain personality dimensions of a person can only be measured after being discretized into order-preserving options such as disagree, neutral, and agree. Motivated by this, we propose Mixed data Permutation-based Rank Test (MPRT), which properly controls the statistical errors even when some or all variables are discretized. Theoretically, we establish the exchangeability and estimate the asymptotic null distribution by permutations; as a consequence, MPRT can effectively control the Type I error in the presence of discretization while previous methods cannot. Empirically, our method is validated by extensive experiments on synthetic data and real-world data to demonstrate its effectiveness as well as applicability in causal discovery.
Authors: Gil Goldshlager, Jiang Hu, Lin Lin
Abstract: Due to the ever growing amounts of data leveraged for machine learning and scientific computing, it is increasingly important to develop algorithms that sample only a small portion of the data at a time. In the case of linear least-squares, the randomized block Kaczmarz method (RBK) is an appealing example of such an algorithm, but its convergence is only understood under sampling distributions that require potentially prohibitively expensive preprocessing steps. To address this limitation, we analyze RBK when the data is sampled uniformly, showing that its iterates converge in a Monte Carlo sense to a $\textit{weighted}$ least-squares solution. Unfortunately, for general problems the condition number of the weight matrix and the variance of the iterates can become arbitrarily large. We control these issues by incorporating regularization into the RBK iterations, yielding the regularized algorithm ReBlocK. Numerical experiments including examples arising from natural gradient optimization demonstrate that ReBlocK can outperform both RBK and minibatch stochastic gradient descent for inconsistent problems with rapidly decaying singular values.
Authors: Keanu Sisouk, Julie Delon, Julien Tierny
Abstract: This paper serves as a user's guide to sampling strategies for sliced optimal transport. We provide reminders and additional regularity results on the Sliced Wasserstein distance. We detail the construction methods, generation time complexity, theoretical guarantees, and conditions for each strategy. Additionally, we provide insights into their suitability for sliced optimal transport in theory. Extensive experiments on both simulated and real-world data offer a representative comparison of the strategies, culminating in practical recommendations for their best usage.
Authors: Hanlin Yu, Arto Klami, Aapo Hyv\"arinen, Anna Korba, Omar Chehab
Abstract: Density ratio estimation in high dimensions can be reframed as integrating a certain quantity, the time score, over probability paths which interpolate between the two densities. In practice, the time score has to be estimated based on samples from the two densities. However, existing methods for this problem remain computationally expensive and can yield inaccurate estimates. Inspired by recent advances in generative modeling, we introduce a novel framework for time score estimation, based on a conditioning variable. Choosing the conditioning variable judiciously enables a closed-form objective function. We demonstrate that, compared to previous approaches, our approach results in faster learning of the time score and competitive or better estimation accuracies of the density ratio on challenging tasks. Furthermore, we establish theoretical guarantees on the error of the estimated density ratio.
Authors: Sunny Sanyal, Hayden Prairie, Rudrajit Das, Ali Kavis, Sujay Sanghavi
Abstract: Fine-tuning a pre-trained model on a downstream task often degrades its original capabilities, a phenomenon known as "catastrophic forgetting". This is especially an issue when one does not have access to the data and recipe used to develop the pre-trained model. Under this constraint, most existing methods for mitigating forgetting are inapplicable. To address this challenge, we propose a sample weighting scheme for the fine-tuning data solely based on the pre-trained model's losses. Specifically, we upweight the easy samples on which the pre-trained model's loss is low and vice versa to limit the drift from the pre-trained model. Our approach is orthogonal and yet complementary to existing methods; while such methods mostly operate on parameter or gradient space, we concentrate on the sample space. We theoretically analyze the impact of fine-tuning with our method in a linear setting, showing that it stalls learning in a certain subspace which inhibits overfitting to the target task. We empirically demonstrate the efficacy of our method on both language and vision tasks. As an example, when fine-tuning Gemma 2 2B on MetaMathQA, our method results in only a $0.8\%$ drop in accuracy on GSM8K (another math dataset) compared to standard fine-tuning, while preserving $5.4\%$ more accuracy on the pre-training datasets. Our code is publicly available at https://github.com/sanyalsunny111/FLOW_finetuning .
Authors: Shashwat Goel, Joschka Struber, Ilze Amanda Auzina, Karuna K Chandra, Ponnurangam Kumaraguru, Douwe Kiela, Ameya Prabhu, Matthias Bethge, Jonas Geiping
Abstract: As Language Model (LM) capabilities advance, evaluating and supervising them at scale is getting harder for humans. There is hope that other language models can automate both these tasks, which we refer to as ''AI Oversight''. We study how model similarity affects both aspects of AI oversight by proposing Chance Adjusted Probabilistic Agreement (CAPA): a metric for LM similarity based on overlap in model mistakes. Using CAPA, we first show that LLM-as-a-judge scores favor models similar to the judge, generalizing recent self-preference results. Then, we study training on LM annotations, and find complementary knowledge between the weak supervisor and strong student model plays a crucial role in gains from ''weak-to-strong generalization''. As model capabilities increase, it becomes harder to find their mistakes, and we might defer more to AI oversight. However, we observe a concerning trend -- model mistakes are becoming more similar with increasing capabilities, pointing to risks from correlated failures. Our work underscores the importance of reporting and correcting for model similarity, especially in the emerging paradigm of AI oversight.
Authors: Yeseul Cho, Baekrok Shin, Changmin Kang, Chulhee Yun
Abstract: Recent advances in deep learning rely heavily on massive datasets, leading to substantial storage and training costs. Dataset pruning aims to alleviate this demand by discarding redundant examples. However, many existing methods require training a model with a full dataset over a large number of epochs before being able to prune the dataset, which ironically makes the pruning process more expensive than just training the model on the entire dataset. To overcome this limitation, we introduce a Difficulty and Uncertainty-Aware Lightweight (DUAL) score, which aims to identify important samples from the early training stage by considering both example difficulty and prediction uncertainty. To address a catastrophic accuracy drop at an extreme pruning, we further propose a ratio-adaptive sampling using Beta distribution. Experiments on various datasets and learning scenarios such as image classification with label noise and image corruption, and model architecture generalization demonstrate the superiority of our method over previous state-of-the-art (SOTA) approaches. Specifically, on ImageNet-1k, our method reduces the time cost for pruning to 66% compared to previous methods while achieving a SOTA, specifically 60% test accuracy at a 90% pruning ratio. On CIFAR datasets, the time cost is reduced to just 15% while maintaining SOTA performance.
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. Our code is publicly available at http://github.com/microsoft/implicit_languagemodels .
Authors: Fanxu Meng, Pingzhi Tang, Xiaojuan Tang, Zengwei Yao, Xing Sun, Muhan Zhang
Abstract: In this paper, we present TransMLA, a framework that seamlessly converts any GQA-based pre-trained model into an MLA-based model. Our approach enables direct compatibility with DeepSeek's codebase, allowing these models to fully leverage DeepSeek-specific optimizations such as vLLM and SGlang. By compressing 93% of the KV cache in LLaMA-2-7B, TransMLA achieves a 10.6x inference speedup at an 8K context length while preserving meaningful output quality. Additionally, the model requires only 6 billion tokens for fine-tuning to regain performance on par with the original across multiple benchmarks. TransMLA offers a practical solution for migrating GQA-based models to the MLA structure. When combined with DeepSeek's advanced features, such as FP8 quantization and Multi-Token Prediction, even greater inference acceleration can be realized.
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: Richard J. Preen, Jim Smith
Abstract: Machine learning models can inadvertently expose confidential properties of their training data, making them vulnerable to membership inference attacks (MIA). While numerous evaluation methods exist, many require computationally expensive processes, such as training multiple shadow models. This article presents two new complementary approaches for efficiently identifying vulnerable tree-based models: an ante-hoc analysis of hyperparameter choices and a post-hoc examination of trained model structure. While these new methods cannot certify whether a model is safe from MIA, they provide practitioners with a means to significantly reduce the number of models that need to undergo expensive MIA assessment through a hierarchical filtering approach. More specifically, it is shown that the rank order of disclosure risk for different hyperparameter combinations remains consistent across datasets, enabling the development of simple, human-interpretable rules for identifying relatively high-risk models before training. While this ante-hoc analysis cannot determine absolute safety since this also depends on the specific dataset, it allows the elimination of unnecessarily risky configurations during hyperparameter tuning. Additionally, computationally inexpensive structural metrics serve as indicators of MIA vulnerability, providing a second filtering stage to identify risky models after training but before conducting expensive attacks. Empirical results show that hyperparameter-based risk prediction rules can achieve high accuracy in predicting the most at risk combinations of hyperparameters across different tree-based model types, while requiring no model training. Moreover, target model accuracy is not seen to correlate with privacy risk, suggesting opportunities to optimise model configurations for both performance and privacy.
Authors: Zixiao Huang, Lifeng Guo, Wenhao Li, Junjie Sheng, Chuyun Shen, Haosheng Chen, Bo Jin, Changhong Lu, Xiangfeng Wang
Abstract: Graph combinatorial optimization (GCO) problems are central to domains like logistics and bioinformatics. While traditional solvers dominate, large language models (LLMs) offer new possibilities for structured reasoning, yet struggle with complex GCO tasks requiring rigorous combinatorial analysis and multi-step deduction, often producing hallucinated steps. We first formalize the Optimal Thoughts Design (OTD) problem, which provides a structured guidance for producing high-quality intermediate reasoning steps. Building on this formulation, we introduce GraphThought, a novel framework that generates effective reasoning sequences through either heuristic-guided forward search or solver-aligned backward reasoning. By fine-tuning LLMs on these structured thought sequences, we develop Llama-GT, an 8B-parameter model that achieves state-of-the-art performance on the GraphArena benchmark, outperforming significantly larger models like DeepSeek-V3. Our results demonstrate that when scaffolded with structured reasoning priors, principled thought generation can significantly enhance LLM performance on GCO tasks without requiring increased model scale.
Authors: Phaphontee Yamchote, Saw Nay Htet Win, Chainarong Amornbunchornvej, Thanapon Noraset
Abstract: Feature interaction is crucial in predictive machine learning models, as it captures the relationships between features that influence model performance. In this work, we focus on pairwise interactions and investigate their importance in constructing feature graphs for Graph Neural Networks (GNNs). We leverage existing GNN models and tools to explore the relationship between feature graph structures and their effectiveness in modeling interactions. Through experiments on synthesized datasets, we uncover that edges between interacting features are important for enabling GNNs to model feature interactions effectively. We also observe that including non-interaction edges can act as noise, degrading model performance. Furthermore, we provide theoretical support for sparse feature graph selection using the Minimum Description Length (MDL) principle. We prove that feature graphs retaining only necessary interaction edges yield a more efficient and interpretable representation than complete graphs, aligning with Occam's Razor. Our findings offer both theoretical insights and practical guidelines for designing feature graphs that improve the performance and interpretability of GNN models.
Authors: Coleman Hooper, Sehoon Kim, Suhong Moon, Kerem Dilmen, Monishwaran Maheswaran, Nicholas Lee, Michael W. Mahoney, Sophia Shao, Kurt Keutzer, Amir Gholami
Abstract: Test-time compute scaling has emerged as a new axis along which to improve model accuracy, where additional computation is used at inference time to allow the model to think longer for more challenging problems. One promising approach for test-time compute scaling is search against a process reward model, where a model generates multiple potential candidates at each step of the search, and these partial trajectories are then scored by a separate reward model in order to guide the search process. The diversity of trajectories in the tree search process affects the accuracy of the search, since increasing diversity promotes more exploration. However, this diversity comes at a cost, as divergent trajectories have less KV sharing, which means they consume more memory and slow down the search process. Previous search methods either do not perform sufficient exploration, or else explore diverse trajectories but have high latency. We address this challenge by proposing Efficient Tree Search (ETS), which promotes KV sharing by pruning redundant trajectories while maintaining necessary diverse trajectories. ETS incorporates a linear programming cost model to promote KV cache sharing by penalizing the number of nodes retained, while incorporating a semantic coverage term into the cost model to ensure that we retain trajectories which are semantically different. We demonstrate how ETS can achieve 1.8$\times$ reduction in average KV cache size during the search process, leading to 1.4$\times$ increased throughput relative to prior state-of-the-art methods, with minimal accuracy degradation and without requiring any custom kernel implementation. Code is available at: https://github.com/SqueezeAILab/ETS.
Authors: Yongli Yan, Linglong Dai
Abstract: Neural network-based decoding methods have shown promise in enhancing error correction performance, but traditional approaches struggle with the challenges posed by punctured codes. In particular, these methods fail to address the complexities of variable code rates and the need for protocol compatibility. This paper presents a unified Long Short-Term Memory (LSTM)-based decoding architecture specifically designed to overcome these challenges. The proposed method unifies punctured convolutional and Turbo codes. A puncture embedding mechanism integrates puncturing patterns directly into the network, enabling seamless adaptation to varying code rates, while balanced bit error rate training ensures robustness across different code lengths, rates, and channels, maintaining protocol flexibility. Extensive simulations in Additive White Gaussian Noise and Rayleigh fading channels demonstrate that the proposed approach outperforms conventional decoding techniques, providing significant improvements in decoding accuracy and robustness. These results underscore the potential of LSTM-based decoding as a promising solution for next-generation artificial intelligence powered communication systems.
Authors: Metod Jazbec, Eliot Wong-Toi, Guoxuan Xia, Dan Zhang, Eric Nalisnick, Stephan Mandt
Abstract: Diffusion models have recently driven significant breakthroughs in generative modeling. While state-of-the-art models produce high-quality samples on average, individual samples can still be low quality. Detecting such samples without human inspection remains a challenging task. To address this, we propose a Bayesian framework for estimating generative uncertainty of synthetic samples. We outline how to make Bayesian inference practical for large, modern generative models and introduce a new semantic likelihood (evaluated in the latent space of a feature extractor) to address the challenges posed by high-dimensional sample spaces. Through our experiments, we demonstrate that the proposed generative uncertainty effectively identifies poor-quality samples and significantly outperforms existing uncertainty-based methods. Notably, our Bayesian framework can be applied post-hoc to any pretrained diffusion or flow matching model (via the Laplace approximation), and we propose simple yet effective techniques to minimize its computational overhead during sampling.
Authors: Jing Wang, Yu-Jie Zhang, Peng Zhao, Zhi-Hua Zhou
Abstract: We study the stochastic linear bandits with heavy-tailed noise. Two principled strategies for handling heavy-tailed noise, truncation and median-of-means, have been introduced to heavy-tailed bandits. Nonetheless, these methods rely on specific noise assumptions or bandit structures, limiting their applicability to general settings. The recent work [Huang et al.2024] develops a soft truncation method via the adaptive Huber regression to address these limitations. However, their method suffers undesired computational costs: it requires storing all historical data and performing a full pass over these data at each round. In this paper, we propose a \emph{one-pass} algorithm based on the online mirror descent framework. Our method updates using only current data at each round, reducing the per-round computational cost from $\mathcal{O}(t \log T)$ to $\mathcal{O}(1)$ with respect to current round $t$ and the time horizon $T$, and achieves a near-optimal and variance-aware regret of order $\widetilde{\mathcal{O}}\big(d T^{\frac{1-\epsilon}{2(1+\epsilon)}} \sqrt{\sum_{t=1}^T \nu_t^2} + d T^{\frac{1-\epsilon}{2(1+\epsilon)}}\big)$ where $d$ is the dimension and $\nu_t^{1+\epsilon}$ is the $(1+\epsilon)$-th central moment of reward at round $t$.
Authors: Eden Saig, Nir Rosenfeld
Abstract: When a prediction algorithm serves a collection of users, disparities in prediction quality are likely to emerge. If users respond to accurate predictions by increasing engagement, inviting friends, or adopting trends, repeated learning creates a feedback loop that shapes both the model and the population of its users. In this work, we introduce evolutionary prediction games, a framework grounded in evolutionary game theory which models such feedback loops as natural-selection processes among groups of users. Our theoretical analysis reveals a gap between idealized and real-world learning settings: In idealized settings with unlimited data and computational power, repeated learning creates competition and promotes competitive exclusion across a broad class of behavioral dynamics. However, under realistic constraints such as finite data, limited compute, or risk of overfitting, we show that stable coexistence and mutualistic symbiosis between groups becomes possible. We analyze these possibilities in terms of their stability and feasibility, present mechanisms that can sustain their existence, and empirically demonstrate our findings.
Authors: Mohit Pandey, Gopeshh Subbaraj, Artem Cherkasov, Martin Ester, Emmanuel Bengio
Abstract: Generative Flow Networks (GFlowNets) have recently emerged as a suitable framework for generating diverse and high-quality molecular structures by learning from rewards treated as unnormalized distributions. Previous works in this framework often restrict exploration by using predefined molecular fragments as building blocks, limiting the chemical space that can be accessed. In this work, we introduce Atomic GFlowNets (A-GFNs), a foundational generative model leveraging individual atoms as building blocks to explore drug-like chemical space more comprehensively. We propose an unsupervised pre-training approach using drug-like molecule datasets, which teaches A-GFNs about inexpensive yet informative molecular descriptors such as drug-likeliness, topological polar surface area, and synthetic accessibility scores. These properties serve as proxy rewards, guiding A-GFNs towards regions of chemical space that exhibit desirable pharmacological properties. We further implement a goal-conditioned finetuning process, which adapts A-GFNs to optimize for specific target properties. In this work, we pretrain A-GFN on a subset of ZINC dataset, and by employing robust evaluation metrics we show the effectiveness of our approach when compared to other relevant baseline methods for a wide range of drug design tasks. The code is accessible at https://github.com/diamondspark/AGFN.
Authors: Fateme Jamshidi, Mohammad Shahverdikondori, Negar Kiyavash
Abstract: We study multi-armed bandits under network interference, where each unit's reward depends on its own treatment and those of its neighbors in a given graph. This induces an exponentially large action space, making standard approaches computationally impractical. We propose a novel algorithm that uses the local graph structure to minimize regret. We derive a graph-dependent upper bound on cumulative regret that improves over prior work. Additionally, we provide the first lower bounds for bandits with arbitrary network interference, where each bound involves a distinct structural property of the graph. These bounds show that for both dense and sparse graphs, our algorithm is nearly optimal, with matching upper and lower bounds up to logarithmic factors. When the interference graph is unknown, a variant of our algorithm is Pareto optimal: no algorithm can uniformly outperform it across all instances. We complement our theoretical results with numerical experiments, showing that our approach outperforms the baseline methods.
Authors: Yimeng Zeng, Natalie Maus, Haydn Thomas Jones, Jeffrey Tao, Fangping Wan, Marcelo Der Torossian Torres, Cesar de la Fuente-Nunez, Ryan Marcus, Osbert Bastani, Jacob R. Gardner
Abstract: In multi-task Bayesian optimization, the goal is to leverage experience from optimizing existing tasks to improve the efficiency of optimizing new ones. While approaches using multi-task Gaussian processes or deep kernel transfer exist, the performance improvement is marginal when scaling beyond a moderate number of tasks. We introduce a novel approach leveraging large language models (LLMs) to learn from, and improve upon, previous optimization trajectories, scaling to approximately 1500 distinct tasks. Specifically, we propose a feedback loop in which an LLM is fine-tuned on the high quality solutions to specific tasks found by Bayesian optimization (BO). This LLM is then used to generate initialization points for future BO searches for new tasks. The trajectories of these new searches provide additional training data for fine-tuning the LLM, completing the loop. We evaluate our method on two distinct domains: database query optimization and antimicrobial peptide design. Results demonstrate that our approach creates a positive feedback loop, where the LLM's generated initializations gradually improve, leading to better optimization performance. As this feedback loop continues, we find that the LLM is eventually able to generate solutions to new tasks in just a few shots that are better than the solutions produced by "from scratch" by Bayesian optimization while simultaneously requiring significantly fewer oracle calls.
Authors: William L. Tong, Cengiz Pehlevan
Abstract: Equality reasoning is ubiquitous and purely abstract: sameness or difference may be evaluated no matter the nature of the underlying objects. As a result, same-different (SD) tasks have been extensively studied as a starting point for understanding abstract reasoning in humans and across animal species. With the rise of neural networks that exhibit striking apparent proficiency for abstractions, equality reasoning in these models has also gained interest. Yet despite extensive study, conclusions about equality reasoning vary widely and with little consensus. To clarify the underlying principles in learning SD tasks, we develop a theory of equality reasoning in multi-layer perceptrons (MLP). Following observations in comparative psychology, we propose a spectrum of behavior that ranges from conceptual to perceptual outcomes. Conceptual behavior is characterized by task-specific representations, efficient learning, and insensitivity to spurious perceptual details. Perceptual behavior is characterized by strong sensitivity to spurious perceptual details, accompanied by the need for exhaustive training to learn the task. We develop a mathematical theory to show that an MLP's behavior is driven by learning richness. Rich-regime MLPs exhibit conceptual behavior, whereas lazy-regime MLPs exhibit perceptual behavior. We validate our theoretical findings in vision SD experiments, showing that rich feature learning promotes success by encouraging hallmarks of conceptual behavior. Overall, our work identifies feature learning richness as a key parameter modulating equality reasoning, and suggests that equality reasoning in humans and animals may similarly depend on learning richness in neural circuits.
Authors: Lucas Gnecco-Heredia, Matteo Sammut, Muni Sreenivas Pydi, Rafael Pinot, Benjamin Negrevergne, Yann Chevaleyre
Abstract: Randomization as a mean to improve the adversarial robustness of machine learning models has recently attracted significant attention. Unfortunately, much of the theoretical analysis so far has focused on binary classification, providing only limited insights into the more complex multiclass setting. In this paper, we take a step toward closing this gap by drawing inspiration from the field of graph theory. Our analysis focuses on discrete data distributions, allowing us to cast the adversarial risk minimization problems within the well-established framework of set packing problems. By doing so, we are able to identify three structural conditions on the support of the data distribution that are necessary for randomization to improve robustness. Furthermore, we are able to construct several data distributions where (contrarily to binary classification) switching from a deterministic to a randomized solution significantly reduces the optimal adversarial risk. These findings highlight the crucial role randomization can play in enhancing robustness to adversarial attacks in multiclass classification.
Authors: Samuel Belkadi, Steve Hong, Marian Chen, Miruna Cretu, Charles Harris, Pietro Lio
Abstract: Autoregressive models excel in efficiency and plug directly into the transformer ecosystem, delivering robust generalization, predictable scalability, and seamless workflows such as fine-tuning and parallelized training. However, they require an explicit sequence order, which contradicts the unordered nature of graphs. In contrast, diffusion models maintain permutation invariance and enable one-shot generation but require up to thousands of denoising steps and additional features for expressivity, leading to high computational costs. Inspired by recent breakthroughs in image generation, especially the success of visual autoregressive methods, we propose MAG, a novel diffusion-free graph generation framework based on next-scale prediction. By leveraging a hierarchy of latent representations, the model progressively generates scales of the entire graph without the need for explicit node ordering. Experiments on both generic and molecular graph datasets demonstrated the potential of this method, achieving inference speedups of up to three orders of magnitude over state-of-the-art methods, while preserving high-quality generation.
Authors: Reza Sameni
Abstract: We study the geometry of Receiver Operating Characteristic (ROC) and Precision-Recall (PR) curves in binary classification problems. The key finding is that many of the most commonly used binary classification metrics are merely functions of the composition function $G := F_p \circ F_n^{-1}$, where $F_p(\cdot)$ and $F_n(\cdot)$ are the class-conditional cumulative distribution functions of the classifier scores in the positive and negative classes, respectively. This geometric perspective facilitates the selection of operating points, understanding the effect of decision thresholds, and comparison between classifiers. It also helps explain how the shapes and geometry of ROC/PR curves reflect classifier behavior, providing objective tools for building classifiers optimized for specific applications with context-specific constraints. We further explore the conditions for classifier dominance, present analytical and numerical examples demonstrating the effects of class separability and variance on ROC and PR geometries, and derive a link between the positive-to-negative class leakage function $G(\cdot)$ and the Kullback--Leibler divergence. The framework highlights practical considerations, such as model calibration, cost-sensitive optimization, and operating point selection under real-world capacity constraints, enabling more informed approaches to classifier deployment and decision-making.
Authors: Wei Xiong, Jiarui Yao, Yuhui Xu, Bo Pang, Lei Wang, Doyen Sahoo, Junnan Li, Nan Jiang, Tong Zhang, Caiming Xiong, Hanze Dong
Abstract: Reinforcement learning (RL) has become a prevailing approach for fine-tuning large language models (LLMs) on complex reasoning tasks. Among recent methods, GRPO stands out for its empirical success in training models such as DeepSeek-R1, yet the sources of its effectiveness remain poorly understood. In this work, we revisit GRPO from a reinforce-like algorithm perspective and analyze its core components. Surprisingly, we find that a simple rejection sampling baseline, RAFT, which trains only on positively rewarded samples, yields competitive performance than GRPO and PPO. Our ablation studies reveal that GRPO's main advantage arises from discarding prompts with entirely incorrect responses, rather than from its reward normalization. Motivated by this insight, we propose Reinforce-Rej, a minimal extension of policy gradient that filters both entirely incorrect and entirely correct samples. Reinforce-Rej improves KL efficiency and stability, serving as a lightweight yet effective alternative to more complex RL algorithms. We advocate RAFT as a robust and interpretable baseline, and suggest that future advances should focus on more principled designs for incorporating negative samples, rather than relying on them indiscriminately. Our findings provide guidance for future work in reward-based LLM post-training.
Authors: Cheng-Yen Hsieh, Xinyou Wang, Daiheng Zhang, Dongyu Xue, Fei Ye, Shujian Huang, Zaixiang Zheng, Quanquan Gu
Abstract: Multimodal protein language models (PLMs) integrate sequence and token-based structural information, serving as a powerful foundation for protein modeling, generation, and design. However, the reliance on tokenizing 3D structures into discrete tokens causes substantial loss of fidelity about fine-grained structural details and correlations. In this paper, we systematically elucidate the design space of multimodal PLMs to overcome their limitations. We identify tokenization loss and inaccurate structure token predictions by the PLMs as major bottlenecks. To address these, our proposed design space covers improved generative modeling, structure-aware architectures and representation learning, and data exploration. Our advancements approach finer-grained supervision, demonstrating that token-based multimodal PLMs can achieve robust structural modeling. The effective design methods dramatically improve the structure generation diversity, and notably, folding abilities of our 650M model by reducing the RMSD from 5.52 to 2.36 on PDB testset, even outperforming 3B baselines and on par with the specialized folding models. Project page and code: https://bytedance.github.io/dplm/dplm-2.1/.
Authors: Alan Lee, Harry Tong
Abstract: We propose reinforcement learning (RL) strategies tailored for reasoning in large language models (LLMs) under strict memory and compute limits, with a particular focus on compatibility with LoRA fine-tuning. Building on early policy gradient methods with baseline subtraction, we design critic-free methods that operate on a small, informative subset of output tokens to reduce memory usage and stabilize training. We introduce S-GRPO, a stochastic variant of Group Relative Policy Optimization, and T-SPMO, a token-level prefix matching approach for fine-grained credit assignment. Applied to Qwen2-1.5B, our methods raise accuracy on the SVAMP benchmark from 46% to over 70% and show strong performance on multi-digit multiplication. Surprisingly, full-token GRPO under LoRA fails to improve over the base model, suggesting that selective token-level optimization may act as an implicit regularizer in low-parameter training regimes.
Authors: Leming Zhou, Zuo Wang, Zhixuan Duan
Abstract: The comorbidities of hypertension impose a heavy burden on patients and society. Early identification is necessary to prompt intervention, but it remains a challenging task. This study aims to address this challenge by combining joint graph learning with network analysis. Motivated by this discovery, we develop a Conjoint Graph Representation Learning (CGRL) framework that: a) constructs two networks based on disease coding, including the patient network and the disease difference network. Three comorbidity network features were generated based on the basic difference network to capture the potential relationship between comorbidities and risk diseases; b) incorporates computational structure intervention and learning feature representation, CGRL was developed to predict the risks of diabetes and coronary heart disease in patients; and c) analysis the comorbidity patterns and exploring the pathways of disease progression, the pathological pathogenesis of diabetes and coronary heart disease may be revealed. The results show that the network features extracted based on the difference network are important, and the framework we proposed provides more accurate predictions than other strong models in terms of accuracy.
Authors: Saikat Barua, Mostafizur Rahman, Shehenaz Khaled, Md Jafor Sadek, Rafiul Islam, Shahnewaz Siddique
Abstract: The emergence of hybrid quantum-classical machine learning (HQML) models opens new horizons of computational intelligence but their fundamental complexity frequently leads to black box behavior that undermines transparency and reliability in their application. Although XAI for quantum systems still in its infancy, a major research gap is evident in robust global and local explainability approaches that are designed for HQML architectures that employ quantized feature encoding followed by classical learning. The gap is the focus of this work, which introduces QuXAI, an framework based upon Q-MEDLEY, an explainer for explaining feature importance in these hybrid systems. Our model entails the creation of HQML models incorporating quantum feature maps, the use of Q-MEDLEY, which combines feature based inferences, preserving the quantum transformation stage and visualizing the resulting attributions. Our result shows that Q-MEDLEY delineates influential classical aspects in HQML models, as well as separates their noise, and competes well against established XAI techniques in classical validation settings. Ablation studies more significantly expose the virtues of the composite structure used in Q-MEDLEY. The implications of this work are critically important, as it provides a route to improve the interpretability and reliability of HQML models, thus promoting greater confidence and being able to engage in safer and more responsible use of quantum-enhanced AI technology. Our code and experiments are open-sourced at: https://github.com/GitsSaikat/QuXAI
Authors: Shihao Zhang, Haoyu Zhang, Ian Colbert, Rayan Saab
Abstract: We introduce Qronos -- a new state-of-the-art post-training quantization algorithm that sequentially rounds and updates neural network weights. Qronos not only explicitly corrects errors due to both weight and activation quantization, but also errors resulting from quantizing previous layers. Our iterative algorithm is based on an interpretable and disciplined optimization framework that subsumes and surpasses existing data-driven approaches. At each step, Qronos alternates between error correction and diffusion via optimal update rules. Importantly, we prove that Qronos admits an efficient implementation that uses the Cholesky decomposition for solving least-squares problems. We also demonstrate that Qronos is compatible with existing transformation techniques such as Hadamard-based incoherence processing and weight-activation scaling equalization, among others. We evaluate Qronos using recent autoregressive language generation models in the Llama3 family; Qronos consistently outperforms previous state-of-the-art adaptive rounding methods when quantizing the weights, activations, and/or KV caches.
Authors: Xiaoqing Zhang, Huabin Zheng, Ang Lv, Yuhan Liu, Zirui Song, Xiuying Chen, Rui Yan, Flood Sung
Abstract: Large language models (LLMs) have been observed to suddenly exhibit advanced reasoning abilities during reinforcement learning (RL), resembling an ``aha moment'' triggered by simple outcome-based rewards. While RL has proven effective in eliciting such breakthroughs in tasks involving mathematics, coding, and vision, it faces significant challenges in multi-scenario games. The diversity of game rules, interaction modes, and environmental complexities often leads to policies that perform well in one scenario but fail to generalize to others. Simply combining multiple scenarios during training introduces additional challenges, such as training instability and poor performance. To overcome these challenges, we propose Divide-Fuse-Conquer, a framework designed to enhance generalization in multi-scenario RL. This approach starts by heuristically grouping games based on characteristics such as rules and difficulties. Specialized models are then trained for each group to excel at games in the group is what we refer to as the divide step. Next, we fuse model parameters from different groups as a new model, and continue training it for multiple groups, until the scenarios in all groups are conquered. Experiments across 18 TextArena games show that Qwen2.5-32B-Align trained with the Divide-Fuse-Conquer strategy reaches a performance level comparable to Claude3.5, achieving 7 wins and 4 draws. We hope our approach can inspire future research on using reinforcement learning to improve the generalization of LLMs.
Authors: Patrick Leask, Neel Nanda, Noura Al Moubayed
Abstract: Sparse autoencoders (SAEs) are a popular method for decomposing Large Langage Models (LLM) activations into interpretable latents. However, due to their substantial training cost, most academic research uses open-source SAEs which are only available for a restricted set of models of up to 27B parameters. SAE latents are also learned from a dataset of activations, which means they do not transfer between models. Motivated by relative representation similarity measures, we introduce Inference-Time Decomposition of Activations (ITDA) models, an alternative method for decomposing language model activations. To train an ITDA, we greedily construct a dictionary of language model activations on a dataset of prompts, selecting those activations which were worst approximated by matching pursuit on the existing dictionary. ITDAs can be trained in just 1% of the time required for SAEs, using 1% of the data. This allowed us to train ITDAs on Llama-3.1 70B and 405B on a single consumer GPU. ITDAs can achieve similar reconstruction performance to SAEs on some target LLMs, but generally incur a performance penalty. However, ITDA dictionaries enable cross-model comparisons, and a simple Jaccard similarity index on ITDA dictionaries outperforms existing methods like CKA, SVCCA, and relative representation similarity metrics. ITDAs provide a cheap alternative to SAEs where computational resources are limited, or when cross model comparisons are necessary. Code available at https://github.com/pleask/itda.
Authors: Tina Khezresmaeilzadeh, Parsa Razmara, Seyedarmin Azizi, Mohammad Erfan Sadeghi, Erfan Baghaei Potraghloo
Abstract: Stock price prediction remains a complex and high-stakes task in financial analysis, traditionally addressed using statistical models or, more recently, language models. In this work, we introduce VISTA (Vision-Language Inference for Stock Time-series Analysis), a novel, training-free framework that leverages Vision-Language Models (VLMs) for multi-modal stock forecasting. VISTA prompts a VLM with both textual representations of historical stock prices and their corresponding line charts to predict future price values. By combining numerical and visual modalities in a zero-shot setting and using carefully designed chain-of-thought prompts, VISTA captures complementary patterns that unimodal approaches often miss. We benchmark VISTA against standard baselines, including ARIMA and text-only LLM-based prompting methods. Experimental results show that VISTA outperforms these baselines by up to 89.83%, demonstrating the effectiveness of multi-modal inference for stock time-series analysis and highlighting the potential of VLMs in financial forecasting tasks without requiring task-specific training.
Authors: Mingyuan Wu, Jingcheng Yang, Jize Jiang, Meitang Li, Kaizhuo Yan, Hanchao Yu, Minjia Zhang, Chengxiang Zhai, Klara Nahrstedt
Abstract: Reinforcement Learning Finetuning (RFT) has significantly advanced the reasoning capabilities of large language models (LLMs) by enabling long chains of thought, self-correction, and effective tool use. While recent works attempt to extend RFT to vision-language models (VLMs), these efforts largely produce text-only reasoning conditioned on static image inputs, falling short of true multimodal reasoning in the response. In contrast, test-time methods like Visual Sketchpad incorporate visual steps but lack training mechanisms. We introduce VTool-R1, the first framework that trains VLMs to generate multimodal chains of thought by interleaving text and intermediate visual reasoning steps. VTool-R1 integrates Python-based visual editing tools into the RFT process, enabling VLMs to learn when and how to generate visual reasoning steps that benefit final reasoning. Trained with outcome-based rewards tied to task accuracy, our approach elicits strategic visual tool use for reasoning without relying on process-based supervision. Experiments on structured visual question answering over charts and tables show that VTool-R1 enhances reasoning performance by teaching VLMs to "think with images" and generate multimodal chain of thoughts with tools.
Authors: Joonkyu Kim, Yejin Kim, Jy-yong Sohn
Abstract: In continual learning scenarios, catastrophic forgetting of previously learned tasks is a critical issue, making it essential to effectively measure such forgetting. Recently, there has been growing interest in focusing on representation forgetting, the forgetting measured at the hidden layer. In this paper, we provide the first theoretical analysis of representation forgetting and use this analysis to better understand the behavior of continual learning. First, we introduce a new metric called representation discrepancy, which measures the difference between representation spaces constructed by two snapshots of a model trained through continual learning. We demonstrate that our proposed metric serves as an effective surrogate for the representation forgetting while remaining analytically tractable. Second, through mathematical analysis of our metric, we derive several key findings about the dynamics of representation forgetting: the forgetting occurs more rapidly to a higher degree as the layer index increases, while increasing the width of the network slows down the forgetting process. Third, we support our theoretical findings through experiments on real image datasets, including Split-CIFAR100 and ImageNet1K.
Authors: Shifeng Xie, Aref Einizade, Jhony H. Giraldo
Abstract: Graph Representation Learning (GRL) is a fundamental task in machine learning, aiming to encode high-dimensional graph-structured data into low-dimensional vectors. Self-Supervised Learning (SSL) methods are widely used in GRL because they can avoid expensive human annotation. In this work, we propose a novel Subgraph Gaussian Embedding Contrast (SubGEC) method. Our approach introduces a subgraph Gaussian embedding module, which adaptively maps subgraphs to a structured Gaussian space, ensuring the preservation of input subgraph characteristics while generating subgraphs with a controlled distribution. We then employ optimal transport distances, more precisely the Wasserstein and Gromov-Wasserstein distances, to effectively measure the similarity between subgraphs, enhancing the robustness of the contrastive learning process. Extensive experiments across multiple benchmarks demonstrate that \method~outperforms or presents competitive performance against state-of-the-art approaches. Our findings provide insights into the design of SSL methods for GRL, emphasizing the importance of the distribution of the generated contrastive pairs.
Authors: Yanzhao Hou, Jiaxiang Geng, Boyu Li, Xiaofeng Tao, Juncheng Wang, Xiaodong Xu, Bing Luo
Abstract: Federated LoRA has emerged as a promising technique for efficiently fine-tuning large language models (LLMs) on distributed devices by reducing the number of trainable parameters. However, existing approaches often inadequately overlook the theoretical and practical implications of system and data heterogeneity, thereby failing to optimize the overall training efficiency, particularly in terms of wall-clock time. In this paper, we propose an adaptive federated LoRA strategy with independent client sampling to minimize the convergence wall-clock time of federated fine-tuning under both computation and communication heterogeneity. We first derive a new convergence bound for federated LoRA with arbitrary and independent client sampling, notably without requiring the stringent bounded gradient assumption. Then, we introduce an adaptive bandwidth allocation scheme that accounts for heterogeneous client resources and system bandwidth constraints. Based on the derived theory, we formulate and solve a non-convex optimization problem to jointly determine the LoRA sketching ratios and sampling probabilities, aiming to minimize wall-clock convergence time. An efficient and low-complexity algorithm is developed to approximate the solution. Finally, extensive experiments demonstrate that our approach significantly reduces wall-clock training time compared to state-of-the-art methods across various models and datasets.
Authors: Abhisek Ray, Lukas Esterle
Abstract: Federated Learning (FL) has emerged as a powerful paradigm for training machine learning models across distributed data sources while preserving data locality. However, the privacy of local data is always a pivotal concern and has received a lot of attention in recent research on the FL regime. Moreover, the lack of domain heterogeneity and client-specific segregation in the benchmarks remains a critical bottleneck for rigorous evaluation. In this paper, we introduce ModelNet, a novel image classification dataset constructed from the embeddings extracted from a pre-trained ResNet50 model. First, we modify the CIFAR100 dataset into three client-specific variants, considering three domain heterogeneities (homogeneous, heterogeneous, and random). Subsequently, we train each client-specific subset of all three variants on the pre-trained ResNet50 model to save model parameters. In addition to multi-domain image data, we propose a new hypothesis to define the FL algorithm that can access the anonymized model parameters to preserve the local privacy in a more effective manner compared to existing ones. ModelNet is designed to simulate realistic FL settings by incorporating non-IID data distributions and client diversity design principles in the mainframe for both conventional and futuristic graph-driven FL algorithms. The three variants are ModelNet-S, ModelNet-D, and ModelNet-R, which are based on homogeneous, heterogeneous, and random data settings, respectively. To the best of our knowledge, we are the first to propose a cross-environment client-specific FL dataset along with the graph-based variant. Extensive experiments based on domain shifts and aggregation strategies show the effectiveness of the above variants, making it a practical benchmark for classical and graph-based FL research. The dataset and related code are available online.
Authors: Junquan Huang, Zong-Gan Chen, Yuncheng Jiang, Zhi-Hui Zhan
Abstract: Neural traveling salesman problem (TSP) solvers face two critical challenges: poor generalization for scalable TSPs and high training costs. To address these challenges, we propose a new Rescaling Graph Convolutional Network (RsGCN). Focusing on the scale-dependent features (i.e., features varied with problem scales) related to nodes and edges that influence the sensitivity of GCNs to the problem scales, a Rescaling Mechanism in RsGCN enhances the generalization capability by (1) rescaling adjacent nodes to construct a subgraph with a uniform number of adjacent nodes for each node across various scales of TSPs, which stabilizes the graph message aggregation; (2) rescaling subgraph edges to adjust the lengths of subgraph edges to the same magnitude, which maintains numerical consistency. In addition, an efficient training strategy with a mixed-scale dataset and bidirectional loss is used in RsGCN. To fully exploit the heatmaps generated by RsGCN, we design an efficient post-search algorithm termed Re2Opt, in which a reconstruction process based on adaptive weight is incorporated to help avoid local optima. Based on a combined architecture of RsGCN and Re2Opt, our solver achieves remarkable generalization and low training cost: with only 3 epochs of training on the mixed-scale dataset containing instances with up to 100 nodes, it can be generalized successfully to 10K-node instances without any fine-tuning. Extensive experiments demonstrate our state-of-the-art performance across uniform distribution instances of 9 different scales from 20 to 10K nodes and 78 real-world instances from TSPLIB, while requiring the fewest learnable parameters and training epochs among neural competitors.
Authors: Zeng Wang, Minghao Shao, Rupesh Karn, Likhitha Mankali, Jitendra Bhandari, Ramesh Karri, Ozgur Sinanoglu, Muhammad Shafique, Johann Knechtel
Abstract: Large Language Models (LLMs) offer transformative capabilities for hardware design automation, particularly in Verilog code generation. However, they also pose significant data security challenges, including Verilog evaluation data contamination, intellectual property (IP) design leakage, and the risk of malicious Verilog generation. We introduce SALAD, a comprehensive assessment that leverages machine unlearning to mitigate these threats. Our approach enables the selective removal of contaminated benchmarks, sensitive IP and design artifacts, or malicious code patterns from pre-trained LLMs, all without requiring full retraining. Through detailed case studies, we demonstrate how machine unlearning techniques effectively reduce data security risks in LLM-aided hardware design.
Authors: Baihe Huang, Shanda Li, Tianhao Wu, Yiming Yang, Ameet Talwalkar, Kannan Ramchandran, Michael I. Jordan, Jiantao Jiao
Abstract: Test-time scaling paradigms have significantly advanced the capabilities of large language models (LLMs) on complex tasks. Despite their empirical success, theoretical understanding of the sample efficiency of various test-time strategies -- such as self-consistency, best-of-$n$, and self-correction -- remains limited. In this work, we first establish a separation result between two repeated sampling strategies: self-consistency requires $\Theta(1/\Delta^2)$ samples to produce the correct answer, while best-of-$n$ only needs $\Theta(1/\Delta)$, where $\Delta < 1$ denotes the probability gap between the correct and second most likely answers. Next, we present an expressiveness result for the self-correction approach with verifier feedback: it enables Transformers to simulate online learning over a pool of experts at test time. Therefore, a single Transformer architecture can provably solve multiple tasks without prior knowledge of the specific task associated with a user query, extending the representation theory of Transformers from single-task to multi-task settings. Finally, we empirically validate our theoretical results, demonstrating the practical effectiveness of self-correction methods.
Authors: Sajjad Abdoli, Freeman Lewin, Gediminas Vasiliauskas, Fabian Schonholz
Abstract: The development of modern Artificial Intelligence (AI) models, particularly diffusion-based models employed in computer vision and image generation tasks, is undergoing a paradigmatic shift in development methodologies. Traditionally dominated by a "Model Centric" approach, in which performance gains were primarily pursued through increasingly complex model architectures and hyperparameter optimization, the field is now recognizing a more nuanced "Data-Centric" approach. This emergent framework foregrounds the quality, structure, and relevance of training data as the principal driver of model performance. To operationalize this paradigm shift, we introduce the DataSeeds.AI sample dataset (the "DSD"), initially comprised of approximately 10,610 high-quality human peer-ranked photography images accompanied by extensive multi-tier annotations. The DSD is a foundational computer vision dataset designed to usher in a new standard for commercial image datasets. Representing a small fraction of DataSeeds.AI's 100 million-plus image catalog, the DSD provides a scalable foundation necessary for robust commercial and multimodal AI development. Through this in-depth exploratory analysis, we document the quantitative improvements generated by the DSD on specific models against known benchmarks and make the code and the trained models used in our evaluation publicly available.
Authors: Fardis Nadimi, Payam Abdisarabshali, Kasra Borazjani, Jacob Chakareski, Seyyedali Hosseinalipour
Abstract: Extended reality (XR) systems, which consist of virtual reality (VR), augmented reality (AR), and mixed reality (XR), offer a transformative interface for immersive, multi-modal, and embodied human-computer interaction. In this paper, we envision that multi-modal multi-task (M3T) federated foundation models (FedFMs) can offer transformative capabilities for XR systems through integrating the representational strength of M3T foundation models (FMs) with the privacy-preserving model training principles of federated learning (FL). We present a modular architecture for FedFMs, which entails different coordination paradigms for model training and aggregations. Central to our vision is the codification of XR challenges that affect the implementation of FedFMs under the SHIFT dimensions: (1) Sensor and modality diversity, (2) Hardware heterogeneity and system-level constraints, (3) Interactivity and embodied personalization, (4) Functional/task variability, and (5) Temporality and environmental variability. We illustrate the manifestation of these dimensions across a set of emerging and anticipated applications of XR systems. Finally, we propose evaluation metrics, dataset requirements, and design tradeoffs necessary for the development of resource-aware FedFMs in XR. This perspective aims to chart the technical and conceptual foundations for context-aware privacy-preserving intelligence in the next generation of XR systems.
Authors: Xudong Zhang, Renato Cordeiro de Amorim
Abstract: Unsupervised feature selection is critical for improving clustering performance in high-dimensional data, where irrelevant features can obscure meaningful structure. In this work, we introduce the Minkowski weighted $k$-means++, a novel initialisation strategy for the Minkowski Weighted $k$-means. Our initialisation selects centroids probabilistically using feature relevance estimates derived from the data itself. Building on this, we propose two new feature selection algorithms, FS-MWK++, which aggregates feature weights across a range of Minkowski exponents to identify stable and informative features, and SFS-MWK++, a scalable variant based on subsampling. We support our approach with a theoretical guarantee under mild assumptions and extensive experiments showing that our methods consistently outperform existing alternatives. Our software can be found at https://github.com/xzhang4-ops1/FSMWK.
Authors: Benjamin von Berg, Bernhard K. Aichernig
Abstract: AALpy is a well-established open-source automata learning library written in Python with a focus on active learning of systems with IO behavior. It provides a wide range of state-of-the-art algorithms for different automaton types ranging from fully deterministic to probabilistic automata. In this work, we present the recent addition of a generalized implementation of an important method from the domain of passive automata learning: state-merging in the red-blue framework. Using a common internal representation for different automaton types allows for a general and highly configurable implementation of the red-blue framework. We describe how to define and execute state-merging algorithms using AALpy, which reduces the implementation effort for state-merging algorithms mainly to the definition of compatibility criteria and scoring. This aids the implementation of both existing and novel algorithms. In particular, defining some existing state-merging algorithms from the literature with AALpy only takes a few lines of code.
Authors: Haochen Song, Dominik Hofer, Rania Islambouli, Laura Hawkins, Ananya Bhattacharjee, Meredith Franklin, Joseph Jay Williams
Abstract: Machine learning approaches, such as contextual multi-armed bandit (cMAB) algorithms, offer a promising strategy to reduce sedentary behavior by delivering personalized interventions to encourage physical activity. However, cMAB algorithms typically require large participant samples to learn effectively and may overlook key psychological factors that are not explicitly encoded in the model. In this study, we propose a hybrid approach that combines cMAB for selecting intervention types with large language models (LLMs) to personalize message content. We evaluate four intervention types: behavioral self-monitoring, gain-framed, loss-framed, and social comparison, each delivered as a motivational message aimed at increasing motivation for physical activity and daily step count. Message content is further personalized using dynamic contextual factors including daily fluctuations in self-efficacy, social influence, and regulatory focus. Over a seven-day trial, participants receive daily messages assigned by one of four models: cMAB alone, LLM alone, combined cMAB with LLM personalization (cMABxLLM), or equal randomization (RCT). Outcomes include daily step count and message acceptance, assessed via ecological momentary assessments (EMAs). We apply a causal inference framework to evaluate the effects of each model. Our findings offer new insights into the complementary roles of LLM-based personalization and cMAB adaptation in promoting physical activity through personalized behavioral messaging.
Authors: Mengsong Wu, YaFei Wang, Yidong Ming, Yuqi An, Yuwei Wan, Wenliang Chen, Binbin Lin, Yuqiang Li, Tong Xie, Dongzhan Zhou
Abstract: Large language models (LLMs) have recently demonstrated promising capabilities in chemistry tasks while still facing challenges due to outdated pretraining knowledge and the difficulty of incorporating specialized chemical expertise. To address these issues, we propose an LLM-based agent that synergistically integrates 137 external chemical tools created ranging from basic information retrieval to complex reaction predictions, and a dataset curation pipeline to generate the dataset ChemToolBench that facilitates both effective tool selection and precise parameter filling during fine-tuning and evaluation. We introduce a Hierarchical Evolutionary Monte Carlo Tree Search (HE-MCTS) framework, enabling independent optimization of tool planning and execution. By leveraging self-generated data, our approach supports step-level fine-tuning (FT) of the policy model and training task-adaptive PRM and ORM that surpass GPT-4o. Experimental evaluations demonstrate that our approach significantly improves performance in Chemistry QA and discovery tasks, offering a robust solution to integrate specialized tools with LLMs for advanced chemical applications. All datasets and code are available at https://github.com/AI4Chem/ChemistryAgent .
Authors: Jo\~ao Manoel Herrera Pinheiro, Suzana Vilas Boas de Oliveira, Thiago Henrique Segreto Silva, Pedro Antonio Rabelo Saraiva, Enzo Ferreira de Souza, Leonardo Andr\'e Ambrosio, Marcelo Becker
Abstract: This research addresses the critical lack of comprehensive studies on feature scaling by systematically evaluating 12 scaling techniques - including several less common transformations - across 14 different Machine Learning algorithms and 16 datasets for classification and regression tasks. We meticulously analyzed impacts on predictive performance (using metrics such as accuracy, MAE, MSE, and $R^2$) and computational costs (training time, inference time, and memory usage). Key findings reveal that while ensemble methods (such as Random Forest and gradient boosting models like XGBoost, CatBoost and LightGBM) demonstrate robust performance largely independent of scaling, other widely used models such as Logistic Regression, SVMs, TabNet, and MLPs show significant performance variations highly dependent on the chosen scaler. This extensive empirical analysis, with all source code, experimental results, and model parameters made publicly available to ensure complete transparency and reproducibility, offers model-specific crucial guidance to practitioners on the need for an optimal selection of feature scaling techniques.
Authors: Diptarka Chakraborty, Kushagra Chatterjee, Debarati Das, Tien Long Nguyen, Romina Nobahari
Abstract: Consensus clustering, a fundamental task in machine learning and data analysis, aims to aggregate multiple input clusterings of a dataset, potentially based on different non-sensitive attributes, into a single clustering that best represents the collective structure of the data. In this work, we study this fundamental problem through the lens of fair clustering, as introduced by Chierichetti et al. [NeurIPS'17], which incorporates the disparate impact doctrine to ensure proportional representation of each protected group in the dataset within every cluster. Our objective is to find a consensus clustering that is not only representative but also fair with respect to specific protected attributes. To the best of our knowledge, we are the first to address this problem and provide a constant-factor approximation. As part of our investigation, we examine how to minimally modify an existing clustering to enforce fairness -- an essential postprocessing step in many clustering applications that require fair representation. We develop an optimal algorithm for datasets with equal group representation and near-linear time constant factor approximation algorithms for more general scenarios with different proportions of two group sizes. We complement our approximation result by showing that the problem is NP-hard for two unequal-sized groups. Given the fundamental nature of this problem, we believe our results on Closest Fair Clustering could have broader implications for other clustering problems, particularly those for which no prior approximation guarantees exist for their fair variants.
Authors: Val Andrei Fajardo, David B. Emerson, Amandeep Singh, Veronica Chatrath, Marcelo Lotif, Ravi Theja, Alex Cheung, Izuki Matsuba
Abstract: Retrieval-augmented generation (RAG) systems have been shown to be effective in addressing many of the drawbacks of relying solely on the parametric memory of large language models. Recent work has demonstrated that RAG systems can be improved via fine-tuning of their retriever and generator models. In this work, we introduce FedRAG, a framework for fine-tuning RAG systems across centralized and federated architectures. FedRAG supports state-of-the-art fine-tuning methods, offering a simple and intuitive interface and a seamless conversion from centralized to federated training tasks. FedRAG is also deeply integrated with the modern RAG ecosystem, filling a critical gap in available tools.
Authors: Hao Hu, Xinqi Wang, Simon Shaolei Du
Abstract: We introduce a novel task of clustering trajectories from offline reinforcement learning (RL) datasets, where each cluster center represents the policy that generated its trajectories. By leveraging the connection between the KL-divergence of offline trajectory distributions and a mixture of policy-induced distributions, we formulate a natural clustering objective. To solve this, we propose Policy-Guided K-means (PG-Kmeans) and Centroid-Attracted Autoencoder (CAAE). PG-Kmeans iteratively trains behavior cloning (BC) policies and assigns trajectories based on policy generation probabilities, while CAAE resembles the VQ-VAE framework by guiding the latent representations of trajectories toward the vicinity of specific codebook entries to achieve clustering. Theoretically, we prove the finite-step convergence of PG-Kmeans and identify a key challenge in offline trajectory clustering: the inherent ambiguity of optimal solutions due to policy-induced conflicts, which can result in multiple equally valid but structurally distinct clusterings. Experimentally, we validate our methods on the widely used D4RL dataset and custom GridWorld environments. Our results show that both PG-Kmeans and CAAE effectively partition trajectories into meaningful clusters. They offer a promising framework for policy-based trajectory clustering, with broad applications in offline RL and beyond.
Authors: William Anderson, Seung Whan Chung, Youngsoo Choi
Abstract: Determining accurate numerical solutions of partial differential equations (PDEs) is an important task in many scientific disciplines. However, solvers can be computationally expensive, leading to the development of reduced-order models (ROMs). Recently, Latent Space Dynamics Identification (LaSDI) was proposed as a data-driven, non-intrusive ROM framework. LaSDI compresses the training data using an autoencoder and learns a system of user-chosen ordinary differential equations (ODEs), which govern the latent space dynamics. This allows for rapid predictions by interpolating and evolving the low-dimensional ODEs in the latent space. While LaSDI has produced effective ROMs for numerous problems, the autoencoder can have difficulty accurately reconstructing training data while also satisfying the imposed dynamics in the latent space, particularly in complex or high-frequency regimes. To address this, we propose multi-stage Latent Space Dynamics Identification (mLaSDI). With mLaSDI, several autoencoders are trained sequentially in stages, where each autoencoder learns to correct the error of the previous stages. We find that applying mLaSDI with small autoencoders results in lower prediction and reconstruction errors, while also reducing training time compared to LaSDI.
Authors: Yeonju Ro, Zhenyu Zhang, Souvik Kundu, Zhangyang Wang, Aditya Akella
Abstract: Large language models (LLMs) excel at capturing global token dependencies via self-attention but face prohibitive compute and memory costs on lengthy inputs. While sub-quadratic methods (e.g., linear attention) can reduce these costs, they often degrade accuracy due to overemphasizing recent tokens. In this work, we first propose dual-state linear attention (DSLA), a novel design that maintains two specialized hidden states-one for preserving historical context and one for tracking recency-thereby mitigating the short-range bias typical of linear-attention architectures. To further balance efficiency and accuracy under dynamic workload conditions, we introduce DSLA-Serve, an online adaptive distillation framework that progressively replaces Transformer layers with DSLA layers at inference time, guided by a sensitivity-based layer ordering. DSLA-Serve uses a chained fine-tuning strategy to ensure that each newly converted DSLA layer remains consistent with previously replaced layers, preserving the overall quality. Extensive evaluations on commonsense reasoning, long-context QA, and text summarization demonstrate that DSLA-Serve yields 2.3x faster inference than Llama2-7B and 3.0x faster than the hybrid Zamba-7B, while retaining comparable performance across downstream tasks. Our ablation studies show that DSLA's dual states capture both global and local dependencies, addressing the historical-token underrepresentation seen in prior linear attentions. Codes are available at https://github.com/utnslab/DSLA-Serve.
Authors: Yang Yang, Du Yin, Hao Xue, Flora Salim
Abstract: Spatial-temporal causal time series (STC-TS) involve region-specific temporal observations driven by causally relevant covariates and interconnected across geographic or network-based spaces. Existing methods often model spatial and temporal dynamics independently and overlook causality-driven probabilistic forecasting, limiting their predictive power. To address this, we propose STOAT (Spatial-Temporal Probabilistic Causal Inference Network), a novel framework for probabilistic forecasting in STC-TS. The proposed method extends a causal inference approach by incorporating a spatial relation matrix that encodes interregional dependencies (e.g. proximity or connectivity), enabling spatially informed causal effect estimation. The resulting latent series are processed by deep probabilistic models to estimate the parameters of the distributions, enabling calibrated uncertainty modeling. We further explore multiple output distributions (e.g., Gaussian, Student's-$t$, Laplace) to capture region-specific variability. Experiments on COVID-19 data across six countries demonstrate that STOAT outperforms state-of-the-art probabilistic forecasting models (DeepAR, DeepVAR, Deep State Space Model, etc.) in key metrics, particularly in regions with strong spatial dependencies. By bridging causal inference and geospatial probabilistic forecasting, STOAT offers a generalizable framework for complex spatial-temporal tasks, such as epidemic management.
Authors: Cecilia Casolo, S\"oren Becker, Niki Kilbertus
Abstract: Dynamical systems modeling is a core pillar of scientific inquiry across natural and life sciences. Increasingly, dynamical system models are learned from data, rendering identifiability a paramount concept. For systems that are not identifiable from data, no guarantees can be given about their behavior under new conditions and inputs, or about possible control mechanisms to steer the system. It is known in the community that "linear ordinary differential equations (ODE) are almost surely identifiable from a single trajectory." However, this only holds for dense matrices. The sparse regime remains underexplored, despite its practical relevance with sparsity arising naturally in many biological, social, and physical systems. In this work, we address this gap by characterizing the identifiability of sparse linear ODEs. Contrary to the dense case, we show that sparse systems are unidentifiable with a positive probability in practically relevant sparsity regimes and provide lower bounds for this probability. We further study empirically how this theoretical unidentifiability manifests in state-of-the-art methods to estimate linear ODEs from data. Our results corroborate that sparse systems are also practically unidentifiable. Theoretical limitations are not resolved through inductive biases or optimization dynamics. Our findings call for rethinking what can be expected from data-driven dynamical system modeling and allows for quantitative assessments of how much to trust a learned linear ODE.
Authors: Johan Erbani, Sonia Ben Mokhtar, Pierre-Edouard Portier, Elod Egyed-Zsigmond, Diana Nurbakova
Abstract: Federated learning (FL) is a machine learning paradigm that enables multiple data holders to collaboratively train a machine learning model without sharing their training data with external parties. In this paradigm, workers locally update a model and share with a central server their updated gradients (or model parameters). While FL seems appealing from a privacy perspective, it opens a number of threats from a security perspective as (Byzantine) participants can contribute poisonous gradients (or model parameters) harming model convergence. Byzantine-resilient FL addresses this issue by ensuring that the training proceeds as if Byzantine participants were absent. Towards this purpose, common strategies ignore outlier gradients during model aggregation, assuming that Byzantine gradients deviate more from honest gradients than honest gradients do from each other. However, in heterogeneous settings, honest gradients may differ significantly, making it difficult to distinguish honest outliers from Byzantine ones. In this paper, we introduce the Worker Label Alignement Loss (WoLA), a weighted loss that aligns honest worker gradients despite data heterogeneity, which facilitates the identification of Byzantines' gradients. This approach significantly outperforms state-of-the-art methods in heterogeneous settings. In this paper, we provide both theoretical insights and empirical evidence of its effectiveness.
Authors: Alejandro Carderera, Sebastian Pokutta
Abstract: Constrained second-order convex optimization algorithms are the method of choice when a high accuracy solution to a problem is needed, due to their local quadratic convergence. These algorithms require the solution of a constrained quadratic subproblem at every iteration. We present the \emph{Second-Order Conditional Gradient Sliding} (SOCGS) algorithm, which uses a projection-free algorithm to solve the constrained quadratic subproblems inexactly. When the feasible region is a polytope the algorithm converges quadratically in primal gap after a finite number of linearly convergent iterations. Once in the quadratic regime the SOCGS algorithm requires $\mathcal{O}(\log(\log 1/\varepsilon))$ first-order and Hessian oracle calls and $\mathcal{O}(\log (1/\varepsilon) \log(\log1/\varepsilon))$ linear minimization oracle calls to achieve an $\varepsilon$-optimal solution. This algorithm is useful when the feasible region can only be accessed efficiently through a linear optimization oracle, and computing first-order information of the function, although possible, is costly.
Authors: Hyeon Jeon, Micha\"el Aupetit, Soohyun Lee, Kwon Ko, Youngtaek Kim, Ghulam Jilani Quadri, Jinwook Seo
Abstract: Brushing is a common interaction technique in 2D scatterplots, allowing users to select clustered points within a continuous, enclosed region for further analysis or filtering. However, applying conventional brushing to 2D representations of multidimensional (MD) data, i.e., Multidimensional Projections (MDPs), can lead to unreliable cluster analysis due to MDP-induced distortions that inaccurately represent the cluster structure of the original MD data. To alleviate this problem, we introduce a novel brushing technique for MDPs called Distortion-aware brushing. As users perform brushing, Distortion-aware brushing corrects distortions around the currently brushed points by dynamically relocating points in the projection, pulling data points close to the brushed points in MD space while pushing distant ones apart. This dynamic adjustment helps users brush MD clusters more accurately, leading to more reliable cluster analysis. Our user studies with 24 participants show that Distortion-aware brushing significantly outperforms previous brushing techniques for MDPs in accurately separating clusters in the MD space and remains robust against distortions. We further demonstrate the effectiveness of our technique through two use cases: (1) conducting cluster analysis of geospatial data and (2) interactively labeling MD clusters.
Authors: Jeff M. Phillips, Hasan Pourmahmood-Aghababa
Abstract: We introduce the notion of an $\varepsilon$-cover for a kernel range space. A kernel range space concerns a set of points $X \subset \mathbb{R}^d$ and the space of all queries by a fixed kernel (e.g., a Gaussian kernel $K(p,\cdot) = \exp(-\|p-\cdot\|^2)$, where $p \in \mathbb{R}^d$). For a point set $X$ of size $n$, a query returns a vector of values $R_p \in \mathbb{R}^n$, where the $i$th coordinate $(R_p)_i = K(p,x_i)$ for $x_i \in X$. An $\varepsilon$-cover is a subset of points $Q \subset \mathbb{R}^d$ so for any $p \in \mathbb{R}^d$ that $\frac{1}{n} \|R_p - R_q\|_1\leq \varepsilon$ for some $q \in Q$. This is a smooth analog of Haussler's notion of $\varepsilon$-covers for combinatorial range spaces (e.g., defined by subsets of points within a ball query) where the resulting vectors $R_p$ are in $\{0,1\}^n$ instead of $[0,1]^n$. The kernel versions of these range spaces show up in data analysis tasks where the coordinates may be uncertain or imprecise, and hence one wishes to add some flexibility in the notion of inside and outside of a query range. Our main result is that, unlike combinatorial range spaces, the size of kernel $\varepsilon$-covers is independent of the input size $n$ and dimension $d$. We obtain a bound of $2^{\tilde O(1/\varepsilon^2)}$, where $\tilde{O}(f(1/\varepsilon))$ hides log factors in $(1/\varepsilon)$ that can depend on the kernel. This implies that by relaxing the notion of boundaries in range queries, eventually the curse of dimensionality disappears, and may help explain the success of machine learning in very high-dimensions. We also complement this result with a lower bound of almost $(1/\varepsilon)^{\Omega(1/\varepsilon)}$, showing the exponential dependence on $1/\varepsilon$ is necessary.
Authors: AmirEhsan Khorashadizadeh, Valentin Debarnot, Tianlin Liu, Ivan Dokmani\'c
Abstract: Deep learning has become the state-of-the-art approach to medical tomographic imaging. A common approach is to feed the result of a simple inversion, for example the backprojection, to a multiscale convolutional neural network (CNN) which computes the final reconstruction. Despite good results on in-distribution test data, this often results in overfitting certain large-scale structures and poor generalization on out-of-distribution (OOD) samples. Moreover, the memory and computational complexity of multiscale CNNs scale unfavorably with image resolution, making them impractical for application at realistic clinical resolutions. In this paper, we introduce Glimpse, a local coordinate-based neural network for computed tomography which reconstructs a pixel value by processing only the measurements associated with the neighborhood of the pixel. Glimpse significantly outperforms successful CNNs on OOD samples, while achieving comparable or better performance on in-distribution test data and maintaining a memory footprint almost independent of image resolution; 5GB memory suffices to train on 1024x1024 images which is orders of magnitude less than CNNs. Glimpse is fully differentiable and can be used plug-and-play in arbitrary deep learning architectures, enabling feats such as correcting miscalibrated projection orientations. Our implementation and Google Colab demo can be accessed at https://github.com/swing-research/Glimpse.
Authors: Kahraman Kostas, Mike Just, Michael A. Lones
Abstract: Previous research on behavior-based attack detection for networks of IoT devices has resulted in machine learning models whose ability to adapt to unseen data is limited and often not demonstrated. This paper presents IoTGeM, an approach for modeling IoT network attacks that focuses on generalizability, yet also leads to better detection and performance. We first introduce an improved rolling window approach for feature extraction. To reduce overfitting, we then apply a multi-step feature selection process where a Genetic Algorithm (GA) is uniquely guided by exogenous feedback from a separate, independent dataset. To prevent common data leaks that have limited previous models, we build and test our models using strictly isolated train and test datasets. The resulting models are rigorously evaluated using a diverse portfolio of machine learning algorithms and datasets. Our window-based models demonstrate superior generalization compared to traditional flow-based models, particularly when tested on unseen datasets. On these stringent, cross-dataset tests, IoTGeM achieves F1 scores of 99\% for ACK, HTTP, SYN, MHD, and PS attacks, as well as a 94\% F1 score for UDP attacks. Finally, we build confidence in the models by using the SHAP (SHapley Additive exPlanations) explainable AI technique, allowing us to identify the specific features that underlie the accurate detection of attacks.
Authors: Giovanni S. Alberti, Luca Ratti, Matteo Santacesaria, Silvia Sciutto
Abstract: In inverse problems, it is widely recognized that the incorporation of a sparsity prior yields a regularization effect on the solution. This approach is grounded on the a priori assumption that the unknown can be appropriately represented in a basis with a limited number of significant components, while most coefficients are close to zero. This occurrence is frequently observed in real-world scenarios, such as with piecewise smooth signals. In this study, we propose a probabilistic sparsity prior formulated as a mixture of degenerate Gaussians, capable of modeling sparsity with respect to a generic basis. Under this premise, we design a neural network that can be interpreted as the Bayes estimator for linear inverse problems. Additionally, we put forth both a supervised and an unsupervised training strategy to estimate the parameters of this network. To evaluate the effectiveness of our approach, we conduct a numerical comparison with commonly employed sparsity-promoting regularization techniques, namely LASSO, group LASSO, iterative hard thresholding, and sparse coding/dictionary learning. Notably, our reconstructions consistently exhibit lower mean square error values across all $1$D datasets utilized for the comparisons, even in cases where the datasets significantly deviate from a Gaussian mixture model.
Authors: Tor Lattimore
Abstract: Bandit convex optimisation is a fundamental framework for studying zeroth-order convex optimisation. This book covers the many tools used for this problem, including cutting plane methods, interior point methods, continuous exponential weights, gradient descent and online Newton step. The nuances between the many assumptions and setups are explained. Although there is not much truly new here, some existing tools are applied in novel ways to obtain new algorithms. A few bounds are improved in minor ways.
Authors: Sara Abdali, Richard Anarfi, CJ Barberan, Jia He, Erfan Shayegani
Abstract: Large language models (LLMs) have significantly transformed the landscape of Natural Language Processing (NLP). Their impact extends across a diverse spectrum of tasks, revolutionizing how we approach language understanding and generations. Nevertheless, alongside their remarkable utility, LLMs introduce critical security and risk considerations. These challenges warrant careful examination to ensure responsible deployment and safeguard against potential vulnerabilities. This research paper thoroughly investigates security and privacy concerns related to LLMs from five thematic perspectives: security and privacy concerns, vulnerabilities against adversarial attacks, potential harms caused by misuses of LLMs, mitigation strategies to address these challenges while identifying limitations of current strategies. Lastly, the paper recommends promising avenues for future research to enhance the security and risk management of LLMs.
Authors: Chaoyi Deng, Jialong Wu, Ningya Feng, Jianmin Wang, Mingsheng Long
Abstract: Effective code optimization in compilers is crucial for computer and software engineering. The success of these optimizations primarily depends on the selection and ordering of the optimization passes applied to the code. While most compilers rely on a fixed sequence of optimization passes, current methods to find the optimal sequence either employ impractically slow search algorithms or learning methods that struggle to generalize to code unseen during training. We introduce CompilerDream, a model-based reinforcement learning approach to general code optimization. CompilerDream comprises a compiler world model that accurately simulates the intrinsic properties of optimization passes and an agent trained on this model to produce effective optimization strategies. By training on a large-scale program dataset, CompilerDream is equipped to serve as a general code optimizer across various application scenarios and source-code languages. Our extensive experiments first highlight CompilerDream's strong optimization capabilities for autotuning, where it leads the CompilerGym leaderboard. More importantly, the zero-shot generalization ability of large-scale trained compiler world model and agent, excels across diverse datasets, surpassing LLVM's built-in optimizations and other state-of-the-art methods in both settings of value prediction and end-to-end code optimization.
Authors: Eden Saig, Ohad Einav, Inbal Talgam-Cohen
Abstract: While the success of large language models (LLMs) increases demand for machine-generated text, current pay-per-token pricing schemes create a misalignment of incentives known in economics as moral hazard: Text-generating agents have strong incentive to cut costs by preferring a cheaper model over the cutting-edge one, and this can be done "behind the scenes" since the agent performs inference internally. In this work, we approach this issue from an economic perspective, by proposing a pay-for-performance, contract-based framework for incentivizing quality. We study a principal-agent game where the agent generates text using costly inference, and the contract determines the principal's payment for the text according to an automated quality evaluation. Since standard contract theory is inapplicable when internal inference costs are unknown, we introduce cost-robust contracts. As our main theoretical contribution, we characterize optimal cost-robust contracts through a direct correspondence to optimal composite hypothesis tests from statistics, generalizing a result of Saig et al. (NeurIPS'23). We evaluate our framework empirically by deriving contracts for a range of objectives and LLM evaluation benchmarks, and find that cost-robust contracts sacrifice only a marginal increase in objective value compared to their cost-aware counterparts.
Authors: Tennessee Hickling, Dennis Prangle
Abstract: Normalizing flows are a flexible class of probability distributions, expressed as transformations of a simple base distribution. A limitation of standard normalizing flows is representing distributions with heavy tails, which arise in applications to both density estimation and variational inference. A popular current solution to this problem is to use a heavy tailed base distribution. We argue this can lead to poor performance due to the difficulty of optimising neural networks, such as normalizing flows, under heavy tailed input. We propose an alternative, "tail transform flow" (TTF), which uses a Gaussian base distribution and a final transformation layer which can produce heavy tails. Experimental results show this approach outperforms current methods, especially when the target distribution has large dimension or tail weight.
Authors: Yushan Zhu, Wen Zhang, Zhiqiang Liu, Mingyang Chen, Lei Liang, Huajun Chen
Abstract: Knowledge Graph Embedding (KGE) is a common approach for Knowledge Graphs (KGs) in AI tasks. Embedding dimensions depend on application scenarios. Requiring a new dimension means training a new KGE model from scratch, increasing cost and limiting efficiency and flexibility. In this work, we propose a novel KGE training framework MED. It allows one training to obtain a croppable KGE model for multiple scenarios with different dimensional needs. Sub-models of required dimensions can be directly cropped and used without extra training. In MED, we propose a mutual learning mechanism to improve the low-dimensional sub-models and make high-dimensional sub-models retain the low-dimensional sub-models' capacity, an evolutionary improvement mechanism to promote the high-dimensional sub-models to master the triple that the low-dimensional sub-models can not, and a dynamic loss weight to adaptively balance the multiple losses. Experiments on 4 KGE models across 4 standard KG completion datasets, 3 real-world scenarios using a large-scale KG, and extending MED to the BERT language model demonstrate its effectiveness, high efficiency, and flexible extensibility.
Authors: Zexian Huang, Kourosh Khoshelham, Martin Tomko
Abstract: Geometric shape classification of vector polygons remains a challenging task in spatial analysis. Previous studies have primarily focused on deep learning approaches for rasterized vector polygons, while the study of discrete polygon representations and corresponding learning methods remains underexplored. In this study, we investigate a graph-based representation of vector polygons and propose a simple graph message-passing framework, PolyMP, along with its densely self-connected variant, PolyMP-DSC, to learn more expressive and robust latent representations of polygons. This framework hierarchically captures self-looped graph information and learns geometric-invariant features for polygon shape classification. Through extensive experiments, we demonstrate that combining a permutation-invariant graph message-passing neural network with a densely self-connected mechanism achieves robust performance on benchmark datasets, including synthetic glyphs and real-world building footprints, outperforming several baseline methods. Our findings indicate that PolyMP and PolyMP-DSC effectively capture expressive geometric features that remain invariant under common transformations, such as translation, rotation, scaling, and shearing, while also being robust to trivial vertex removals. Furthermore, we highlight the strong generalization ability of the proposed approach, enabling the transfer of learned geometric features from synthetic glyph polygons to real-world building footprints.
Authors: Terrance Liu, Zhiwei Steven Wu
Abstract: While past works have shown how uncertainty quantification can be applied to large language model (LLM) outputs, the question of whether resulting uncertainty guarantees still hold within sub-groupings of data remains open. In our work, given some long-form text generated by an LLM, we study uncertainty at both the level of individual claims contained within the output (via calibration) and across the entire output itself (via conformal prediction). Using biography generation as a testbed for this study, we derive a set of (demographic) attributes (e.g., whether some text describes a man or woman) for each generation to form such "subgroups" of data. We find that although canonical methods for both types of uncertainty quantification perform well when measuring across the entire dataset, such guarantees break down when examining particular subgroups. Having established this issue, we invoke group-conditional methods for uncertainty quantification -- multicalibration and multivalid conformal prediction -- and find that across a variety of approaches, additional subgroup information consistently improves calibration and conformal prediction within subgroups (while crucially retaining guarantees across the entire dataset). As the problems of calibration, conformal prediction, and their multi-group counterparts have not been extensively explored in the context of long-form text generation, we consider these results to form a benchmark for this setting.
Authors: Richard Liu, Nicholas T. Williams, Kara E. Rudolph, Iv\'an D\'iaz
Abstract: Causal mediation analyses investigate the mechanisms through which causes exert their effects, and are therefore central to scientific progress. The literature on the non-parametric definition and identification of mediational effects in rigourous causal models has grown significantly in recent years, and there has been important progress to address challenges in the interpretation and identification of such effects. Despite great progress in the causal inference front, statistical methodology for non-parametric estimation has lagged behind, with few or no methods available for tackling non-parametric estimation in the presence of multiple, continuous, or high-dimensional mediators. In this paper we show that the identification formulas for six popular non-parametric approaches to mediation analysis proposed in recent years can be recovered from just two statistical estimands. We leverage this finding to propose an all-purpose one-step estimation algorithm that can be coupled with machine learning in any mediation study that uses any of these six definitions of mediation. The estimators have desirable properties, such as $\sqrt{n}$-convergence and asymptotic normality. Estimating the first-order correction for the one-step estimator requires estimation of complex density ratios on the potentially high-dimensional mediators, a challenge that is solved using recent advancements in so-called Riesz learning. We illustrate the properties of our methods in a simulation study and illustrate its use on real data to estimate the extent to which pain management practices mediate the total effect of having a chronic pain disorder on opioid use disorder.
Authors: Yannis Bendi-Ouis, Dan Dutartre, Xavier Hinaut
Abstract: Since the release of ChatGPT in November 2022, large language models (LLMs) have seen considerable success, including in the open-source community, with many open-weight models available. However, the requirements to deploy such a service are often unknown and difficult to evaluate in advance. To facilitate this process, we conducted numerous tests at the Centre Inria de l'Universit\'e de Bordeaux. In this article, we propose a comparison of the performance of several models of different sizes (mainly Mistral and LLaMa) depending on the available GPUs, using vLLM, a Python library designed to optimize the inference of these models. Our results provide valuable information for private and public groups wishing to deploy LLMs, allowing them to evaluate the performance of different models based on their available hardware. This study thus contributes to facilitating the adoption and use of these large language models in various application domains.
Authors: Xingrui Gu, Guanren Qiao, Chuyi Jiang
Abstract: Traditional reinforcement learning (RL) methods mainly rely on trial-and-error exploration, often lacking mechanisms to guide agents toward more informative decision-making and struggling to leverage past experiences, resulting in low sample efficiency. To overcome this issue, we propose an innovative framework inspired by cognitive principles: Cognitive Belief-Driven Reinforcement Learning (CBD-RL). By incorporating cognitive heuristics, CBD-RL transforms conventional trial-and-error learning into a more structured and guided learning paradigm, simulating the human reasoning process. This framework's core is a belief system that optimizes action probabilities by integrating feedback with prior experience, thus enhancing decision making under uncertainty. It also organizes state-action pairs into meaningful categories, promoting generalization and improving sample efficiency. The concrete implementations of this framework, CBDQ, CBDPPO, and CBDSAC, demonstrate superior performance in discrete and continuous action spaces in diverse environments such as Atari and MuJoCo. By bridging cognitive science and reinforcement learning, this research opens a new avenue for developing RL systems that are more interpretable, efficient, and cognitively inspired.
Authors: Jason Yang, Aadyot Bhatnagar, Jeffrey A. Ruffolo, Ali Madani
Abstract: The conditional generation of proteins with desired functions is a key goal for generative models. Existing methods based on prompting of protein language models (PLMs) can generate proteins conditioned on a target functionality, such as a desired enzyme family. However, these methods are limited to simple, tokenized conditioning and have not been shown to generalize to unseen functions. In this study, we propose ProCALM (Protein Conditionally Adapted Language Model), an approach for the conditional generation of proteins using adapters to PLMs. While previous methods have used adapters for structure-conditioned generation from PLMs, our implementation of ProCALM involves finetuning ProGen2 to condition generation based on versatile representations of protein function-e.g. enzyme family, taxonomy, or natural language descriptions. ProCALM matches or exceeds the performance of existing methods at conditional sequence generation from target functions. Impressively, it can also generalize to rare and unseen functions. Overall, ProCALM is a flexible and computationally efficient approach, and we expect that it can be extended to a wide range of generative language models.
Authors: Mo\"ise Blanchard
Abstract: Classical results in statistical learning typically consider two extreme data-generating models: i.i.d. instances from an unknown distribution, or fully adversarial instances, often much more challenging statistically. To bridge the gap between these models, recent work introduced the smoothed framework, in which at each iteration an adversary generates instances from a distribution constrained to have density bounded by $\sigma^{-1}$ compared to some fixed base measure $\mu$. This framework interpolates between the i.i.d. and adversarial cases, depending on the value of $\sigma$. For the classical online prediction problem, most prior results in smoothed online learning rely on the arguably strong assumption that the base measure $\mu$ is known to the learner, contrasting with standard settings in the PAC learning or consistency literature. We consider the general agnostic problem in which the base measure is unknown and values are arbitrary. Along this direction, Block et al. showed that empirical risk minimization has sublinear regret under the well-specified assumption. We propose an algorithm R-Cover based on recursive coverings which is the first to guarantee sublinear regret for agnostic smoothed online learning without prior knowledge of $\mu$. For classification, we prove that R-Cover has adaptive regret $\tilde O(\sqrt{dT/\sigma})$ for function classes with VC dimension $d$, which is optimal up to logarithmic factors. For regression, we establish that R-Cover has sublinear oblivious regret for function classes with polynomial fat-shattering dimension growth.
Authors: Ali Kashefi
Abstract: Kolmogorov-Arnold Networks (KANs) have recently gained attention as an alternative to traditional Multilayer Perceptrons (MLPs) in deep learning frameworks. KANs have been integrated into various deep learning architectures such as convolutional neural networks, graph neural networks, and transformers, with their performance evaluated. However, their effectiveness within point-cloud-based neural networks remains unexplored. To address this gap, we incorporate KANs into PointNet for the first time to evaluate their performance on 3D point cloud classification and segmentation tasks. Specifically, we introduce PointNet-KAN, built upon two key components. First, it employs KANs instead of traditional MLPs. Second, it retains the core principle of PointNet by using shared KAN layers and applying symmetric functions for global feature extraction, ensuring permutation invariance with respect to the input features. In traditional MLPs, the goal is to train the weights and biases with fixed activation functions; however, in KANs, the goal is to train the activation functions themselves. We use Jacobi polynomials to construct the KAN layers. We extensively and systematically evaluate PointNet-KAN across various polynomial degrees and special types such as the Lagrange, Chebyshev, and Gegenbauer polynomials. Our results show that PointNet-KAN achieves competitive performance compared to PointNet with MLPs on benchmark datasets for 3D object classification and part and semantic segmentation, despite employing a shallower and simpler network architecture. We also study a hybrid PointNet model incorporating both KAN and MLP layers. We hope this work serves as a foundation and provides guidance for integrating KANs, as an alternative to MLPs, into more advanced point cloud processing architectures.
Authors: Yuri Gardinazzi, Karthik Viswanathan, Giada Panerai, Alessio Ansuini, Alberto Cazzaniga, Matteo Biagetti
Abstract: Understanding the decision-making processes of large language models is critical given their widespread applications. To achieve this, we aim to connect a formal mathematical framework -- zigzag persistence from topological data analysis -- with practical and easily applicable algorithms. Zigzag persistence is particularly effective for characterizing data as it dynamically transforms across model layers. Within this framework, we introduce topological descriptors that measure how topological features, $p$-dimensional holes, persist and evolve throughout the layers. Unlike methods that assess each layer individually and then aggregate the results, our approach directly tracks the full evolutionary path of these features. This offers a statistical perspective on how prompts are rearranged and their relative positions changed in the representation space, providing insights into the system's operation as an integrated whole. To demonstrate the expressivity and applicability of our framework, we highlight how sensitive these descriptors are to different models and a variety of datasets. As a showcase application to a downstream task, we use zigzag persistence to establish a criterion for layer pruning, achieving results comparable to state-of-the-art methods while preserving the system-level perspective.
Authors: Giannis Fikioris, Robert Kleinberg, Yoav Kolumbus, Raunak Kumar, Yishay Mansour, \'Eva Tardos
Abstract: In many repeated auction settings, participants care not only about how frequently they win but also how their winnings are distributed over time. This problem arises in various practical domains where avoiding congested demand is crucial, such as online retail sales and compute services, as well as in advertising campaigns that require sustained visibility over time. We introduce a simple model of this phenomenon, modeling it as a budgeted auction where the value of a win is a concave function of the time since the last win. This implies that for a given number of wins, even spacing over time is optimal. We also extend our model and results to the case when not all wins result in "conversions" (realization of actual gains), and the probability of conversion depends on a context. The goal is to maximize and evenly space conversions rather than just wins. We study the optimal policies for this setting in second-price auctions and offer learning algorithms for the bidders that achieve low regret against the optimal bidding policy in a Bayesian online setting. Our main result is a computationally efficient online learning algorithm that achieves $\tilde O(\sqrt T)$ regret. We achieve this by showing that an infinite-horizon Markov decision process (MDP) with the budget constraint in expectation is essentially equivalent to our problem, even when limiting that MDP to a very small number of states. The algorithm achieves low regret by learning a bidding policy that chooses bids as a function of the context and the system's state, which will be the time elapsed since the last win (or conversion). We show that state-independent strategies incur linear regret even without uncertainty of conversions. We complement this by showing that there are state-independent strategies that, while still having linear regret, achieve a $(1-\frac 1 e)$ approximation to the optimal reward.
Authors: Yangxinyu Xie, Xiang Li, Tanwi Mallick, Weijie J. Su, Ruixun Zhang
Abstract: Watermarking language models is essential for distinguishing between human and machine-generated text and thus maintaining the integrity and trustworthiness of digital communication. We present a novel green/red list watermarking approach that partitions the token set into ``green'' and ``red'' lists, subtly increasing the generation probability for green tokens. To correct token distribution bias, our method employs maximal coupling, using a uniform coin flip to decide whether to apply bias correction, with the result embedded as a pseudorandom watermark signal. Theoretical analysis confirms this approach's unbiased nature and robust detection capabilities. Experimental results show that it outperforms prior techniques by preserving text quality while maintaining high detectability, and it demonstrates resilience to targeted modifications aimed at improving text quality. This research provides a promising watermarking solution for language models, balancing effective detection with minimal impact on text quality.
Authors: Xuandong Zhao, Sam Gunn, Miranda Christ, Jaiden Fairoze, Andres Fabrega, Nicholas Carlini, Sanjam Garg, Sanghyun Hong, Milad Nasr, Florian Tramer, Somesh Jha, Lei Li, Yu-Xiang Wang, Dawn Song
Abstract: As the outputs of generative AI (GenAI) techniques improve in quality, it becomes increasingly challenging to distinguish them from human-created content. Watermarking schemes are a promising approach to address the problem of distinguishing between AI and human-generated content. These schemes embed hidden signals within AI-generated content to enable reliable detection. While watermarking is not a silver bullet for addressing all risks associated with GenAI, it can play a crucial role in enhancing AI safety and trustworthiness by combating misinformation and deception. This paper presents a comprehensive overview of watermarking techniques for GenAI, beginning with the need for watermarking from historical and regulatory perspectives. We formalize the definitions and desired properties of watermarking schemes and examine the key objectives and threat models for existing approaches. Practical evaluation strategies are also explored, providing insights into the development of robust watermarking techniques capable of resisting various attacks. Additionally, we review recent representative works, highlight open challenges, and discuss potential directions for this emerging field. By offering a thorough understanding of watermarking in GenAI, this work aims to guide researchers in advancing watermarking methods and applications, and support policymakers in addressing the broader implications of GenAI.
Authors: Justin Le
Abstract: We provide an overview of the diffusion model as a method to generate new samples. Generative models have been recently adopted for tasks such as art generation (Stable Diffusion, Dall-E) and text generation (ChatGPT). Diffusion models in particular apply noise to sample data and then "reverse" this noising process to generate new samples. We will formally define these noising and denoising processes, then present algorithms to train and generate with a diffusion model. Afterward, we will explore a potential application of diffusion models in improving classifier performance on imbalanced data.
Authors: Junyang Cai, Serdar Kadioglu, Bistra Dilkina
Abstract: Mixed-integer programming (MIP) is a powerful paradigm for modeling and solving various important combinatorial optimization problems. Recently, learning-based approaches have shown a potential to speed up MIP solving via offline training that then guides important design decisions during the search. However, a significant drawback of these methods is their heavy reliance on offline training, which requires collecting training datasets and computationally costly training epochs yet offering only limited generalization to unseen (larger) instances. In this paper, we propose Balans, an adaptive meta-solver for MIPs with online learning capability that does not require any supervision or apriori training. At its core, Balans is based on adaptive large-neighborhood search, operating on top of an MIP solver by successive applications of destroy and repair neighborhood operators. During the search, the selection among different neighborhood definitions is guided on the fly for the instance at hand via multi-armed bandit algorithms. Our extensive experiments on hard optimization instances show that Balans offers significant performance gains over the default MIP solver, is better than committing to any single best neighborhood, and improves over the state-of-the-art large-neighborhood search for MIPs. Finally, we release Balans as a highly configurable, MIP solver agnostic, open-source software.
Authors: Junyang Cai, Taoan Huang, Bistra Dilkina
Abstract: Mixed Integer Linear Programs (MILPs) are highly flexible and powerful tools for modeling and solving complex real-world combinatorial optimization problems. Recently, machine learning (ML)-guided approaches have demonstrated significant potential in improving MILP-solving efficiency. However, these methods typically rely on separate offline data collection and training processes, which limits their scalability and adaptability. This paper introduces the first multi-task learning framework for ML-guided MILP solving. The proposed framework provides MILP embeddings helpful in guiding MILP solving across solvers (e.g., Gurobi and SCIP) and across tasks (e.g., Branching and Solver configuration). Through extensive experiments on three widely used MILP benchmarks, we demonstrate that our multi-task learning model performs similarly to specialized models within the same distribution. Moreover, it significantly outperforms them in generalization across problem sizes and tasks.
Authors: Alexander Nikulin, Ilya Zisman, Denis Tarasov, Nikita Lyubaykin, Andrei Polubarov, Igor Kiselev, Vladislav Kurenkov
Abstract: Recently, latent action learning, pioneered by Latent Action Policies (LAPO), have shown remarkable pre-training efficiency on observation-only data, offering potential for leveraging vast amounts of video available on the web for embodied AI. However, prior work has focused on distractor-free data, where changes between observations are primarily explained by ground-truth actions. Unfortunately, real-world videos contain action-correlated distractors that may hinder latent action learning. Using Distracting Control Suite (DCS) we empirically investigate the effect of distractors on latent action learning and demonstrate that LAPO struggle in such scenario. We propose LAOM, a simple LAPO modification that improves the quality of latent actions by 8x, as measured by linear probing. Importantly, we show that providing supervision with ground-truth actions, as few as 2.5% of the full dataset, during latent action learning improves downstream performance by 4.2x on average. Our findings suggest that integrating supervision during Latent Action Models (LAM) training is critical in the presence of distractors, challenging the conventional pipeline of first learning LAM and only then decoding from latent to ground-truth actions.
Authors: Naveen Raman, Holly Wiberg
Abstract: Rising provider turnover results in frequently needing to rematch patients with available providers. However, the rematching process is cumbersome for both patients and health systems, resulting in labor-intensive and ad hoc reassignments. We propose a novel patient-provider matching approach to address this issue by offering patients limited provider menus. The goal is to maximize match quality across the system while preserving patient choice. We frame this as a novel variant of assortment optimization, where patient-specific provider menus are offered upfront, and patients respond in a random sequence to make their selections. This hybrid offline-online setting is understudied in previous literature and captures system dynamics across various domains. We first demonstrate that a greedy baseline policy--which offers all providers to all patients--can maximize the match rate but lead to low-quality matches. Based on this, we construct a set of policies and demonstrate that the best policy depends on problem specifics, such as a patient's willingness to match and the ratio of patients to providers. On real-world data, our proposed policy improves average match quality by 13% over a greedy solution by tailoring assortments based on patient characteristics. Our analysis reveals a tradeoff between menu size and system-wide match quality, highlighting the value of balancing patient choice with centralized planning.
Authors: Mark Russinovich, Ahmed Salem
Abstract: Recent copyright agreements between AI companies and content creators underscore the need for fine-grained control over language models' ability to reproduce copyrighted text. Existing defenses-ranging from aggressive unlearning to simplistic output filters-either sacrifice model utility or inadequately address verbatim leakage. We introduce Obliviate, a lightweight post-training method that surgically suppresses exact reproduction of specified sequences while preserving semantic understanding. Obliviate first identifies memorized passages and then, for each target token, minimally adjusts the model's output distribution via a Kullback-Leibler divergence penalty to drive down the probability of exact reproduction. Simultaneously, we enforce a consistency loss on non-target tokens to retain the model's fluency and task performance. We evaluate Obliviate on four popular 6-8B-parameter models (LLaMA-3.1, LLaMA-3.1-Instruct, Qwen-2.5, and Yi-1.5) using synthetic memorization benchmarks and organic copyrighted excerpts (e.g., Moby Dick, Frankenstein, Alice in Wonderland and Les Miserables). Across all settings, Obliviate reduces verbatim recall by two orders of magnitude (e.g., from hundreds of words to fewer than 12) while degrading downstream accuracy by at most 1% on HellaSwag, MMLU, TruthfulQA, and Winogrande. Furthermore, we benchmark Obliviate aganist different unlearning and copyright techniques using the MUSE and CoTaEval benchmarks. These results position Obliviate as a practical, high-fidelity solution for copyright compliance in deployed LLMs.
Authors: Shayleen Reynolds, Hengzhi He, Dung Daniel T. Ngo, Saheed Obitayo, Niccol\`o Dalmasso, Guang Cheng, Vamsi K. Potluru, Manuela Veloso
Abstract: In recent years, LLM watermarking has emerged as an attractive safeguard against AI-generated content, with promising applications in many real-world domains. However, there are growing concerns that the current LLM watermarking schemes are vulnerable to expert adversaries wishing to reverse-engineer the watermarking mechanisms. Prior work in breaking or stealing LLM watermarks mainly focuses on the distribution-modifying algorithm of Kirchenbauer et al. (2023), which perturbs the logit vector before sampling. In this work, we focus on reverse-engineering the other prominent LLM watermarking scheme, distortion-free watermarking (Kuditipudi et al. 2024), which preserves the underlying token distribution by using a hidden watermarking key sequence. We demonstrate that, even under a more sophisticated watermarking scheme, it is possible to compromise the LLM and carry out a spoofing attack, i.e. generate a large number of (potentially harmful) texts that can be attributed to the original watermarked LLM. Specifically, we propose using adaptive prompting and a sorting-based algorithm to accurately recover the underlying secret key for watermarking the LLM. Our empirical findings on LLAMA-3.1-8B-Instruct, Mistral-7B-Instruct, Gemma-7b, and OPT-125M challenge the current theoretical claims on the robustness and usability of the distortion-free watermarking techniques.
Authors: Zhaowei Zhang, Fengshuo Bai, Qizhi Chen, Chengdong Ma, Mingzhi Wang, Haoran Sun, Zilong Zheng, Yaodong Yang
Abstract: How to align large language models (LLMs) with user preferences from a static general dataset has been frequently studied. However, user preferences are usually personalized, changing, and diverse regarding culture, values, or time. This leads to the problem that the actual user preferences often do not coincide with those trained by the model developers in the practical use of LLMs. Since we cannot collect enough data and retrain for every demand, researching efficient real-time preference adaptation methods based on the backbone LLMs during test time is important. To this end, we introduce Amulet, a novel, training-free framework that formulates the decoding process of every token as a separate online learning problem with the guidance of simple user-provided prompts, thus enabling real-time optimization to satisfy users' personalized preferences. To reduce the computational cost brought by this optimization process for each token, we additionally provide a closed-form solution for each iteration step of the optimization process, thereby reducing the computational time cost to a negligible level. The detailed experimental results demonstrate that Amulet can achieve significant performance improvements in rich settings with combinations of different LLMs, datasets, and user preferences, while maintaining acceptable computational efficiency.
Authors: Xuandong Zhao, Will Cai, Tianneng Shi, David Huang, Licong Lin, Song Mei, Dawn Song
Abstract: Existing training-time safety alignment techniques for large language models (LLMs) remain vulnerable to jailbreak attacks. Direct preference optimization (DPO), a widely deployed alignment method, exhibits limitations in both experimental and theoretical contexts as its loss function proves suboptimal for refusal learning. Through gradient-based analysis, we identify these shortcomings and propose an improved safety alignment that disentangles DPO objectives into two components: (1) robust refusal training, which encourages refusal even when partial unsafe generations are produced, and (2) targeted unlearning of harmful knowledge. This approach significantly increases LLM robustness against a wide range of jailbreak attacks, including prefilling, suffix, and multi-turn attacks across both in-distribution and out-of-distribution scenarios. Furthermore, we introduce a method to emphasize critical refusal tokens by incorporating a reward-based token-level weighting mechanism for refusal learning, which further improves the robustness against adversarial exploits. Our research also suggests that robustness to jailbreak attacks is correlated with token distribution shifts in the training process and internal representations of refusal and harmful tokens, offering valuable directions for future research in LLM safety alignment. The code is available at https://github.com/wicai24/DOOR-Alignment
Authors: Antonio M. Mercado-Mart\'inez, Beatriz Soret, Antonio Jurado-Navas
Abstract: The Agile Earth Observation Satellite Scheduling Problem (AEOSSP) entails finding the subset of observation targets to be scheduled along the satellite's orbit while meeting operational constraints of time, energy and memory. The problem of deciding what and when to observe is inherently complex, and becomes even more challenging when considering several issues that compromise the quality of the captured images, such as cloud occlusion, atmospheric turbulence, and image resolution. This paper presents a Deep Reinforcement Learning (DRL) approach for addressing the AEOSSP with time-dependent profits, integrating these three factors to optimize the use of energy and memory resources. The proposed method involves a dual decision-making process: selecting the sequence of targets and determining the optimal observation time for each. Our results demonstrate that the proposed algorithm reduces the capture of images that fail to meet quality requirements by > 60% and consequently decreases energy waste from attitude maneuvers by up to 78%, all while maintaining strong observation performance.
Authors: Tobias Morocutti, Florian Schmid, Jonathan Greif, Francesco Foscarin, Gerhard Widmer
Abstract: We target the problem of developing new low-complexity networks for the sound event detection task. Our goal is to meticulously analyze the performance-complexity trade-off, aiming to be competitive with the large state-of-the-art models, at a fraction of the computational requirements. We find that low-complexity convolutional models previously proposed for audio tagging can be effectively adapted for event detection (which requires frame-wise prediction) by adjusting convolutional strides, removing the global pooling, and, importantly, adding a sequence model before the (now frame-wise) classification heads. Systematic experiments reveal that the best choice for the sequence model type depends on which complexity metric is most important for the given application. We also investigate the impact of enhanced training strategies such as knowledge distillation. In the end, we show that combined with an optimized training strategy, we can reach event detection performance comparable to state-of-the-art transformers while requiring only around 5% of the parameters. We release all our pre-trained models and the code for reproducing this work to support future research in low-complexity sound event detection at https://github.com/theMoro/EfficientSED.
Authors: Zeng Wang, Minghao Shao, Jitendra Bhandari, Likhitha Mankali, Ramesh Karri, Ozgur Sinanoglu, Muhammad Shafique, Johann Knechtel
Abstract: Large Language Models (LLMs) have revolutionized code generation, achieving exceptional results on various established benchmarking frameworks. However, concerns about data contamination - where benchmark data inadvertently leaks into pre-training or fine-tuning datasets - raise questions about the validity of these evaluations. While this issue is known, limiting the industrial adoption of LLM-driven software engineering, hardware coding has received little to no attention regarding these risks. For the first time, we analyze state-of-the-art (SOTA) evaluation frameworks for Verilog code generation (VerilogEval and RTLLM), using established methods for contamination detection (CCD and Min-K% Prob). We cover SOTA commercial and open-source LLMs (CodeGen2.5, Minitron 4b, Mistral 7b, phi-4 mini, LLaMA-{1,2,3.1}, GPT-{2,3.5,4o}, Deepseek-Coder, and CodeQwen 1.5), in baseline and fine-tuned models (RTLCoder and Verigen). Our study confirms that data contamination is a critical concern. We explore mitigations and the resulting trade-offs for code quality vs fairness (i.e., reducing contamination toward unbiased benchmarking).
Authors: Wei Fang, Yang Zhang, Kaizhi Qian, James Glass, Yada Zhu
Abstract: Large language models (LLMs) are increasingly integrated with specialized external tools, yet many tasks demand zero-shot tool usage with minimal or noisy documentation. Existing solutions rely on manual rewriting or labeled data for validation, making them inapplicable in true zero-shot settings. To address these challenges, we propose PLAY2PROMPT, an automated framework that systematically "plays" with each tool to explore its input-output behaviors. Through this iterative trial-and-error process, PLAY2PROMPT refines tool documentation and generates usage examples without any labeled data. These examples not only guide LLM inference but also serve as validation to further enhance tool utilization. Extensive experiments on real-world tasks demonstrate that PLAY2PROMPT significantly improves zero-shot tool performance across both open and closed models, offering a scalable and effective solution for domain-specific tool integration.
Authors: Yike Yuan, Ziyu Wang, Zihao Huang, Defa Zhu, Xun Zhou, Jingyi Yu, Qiyang Min
Abstract: Diffusion models have emerged as mainstream framework in visual generation. Building upon this success, the integration of Mixture of Experts (MoE) methods has shown promise in enhancing model scalability and performance. In this paper, we introduce Race-DiT, a novel MoE model for diffusion transformers with a flexible routing strategy, Expert Race. By allowing tokens and experts to compete together and select the top candidates, the model learns to dynamically assign experts to critical tokens. Additionally, we propose per-layer regularization to address challenges in shallow layer learning, and router similarity loss to prevent mode collapse, ensuring better expert utilization. Extensive experiments on ImageNet validate the effectiveness of our approach, showcasing significant performance gains while promising scaling properties.
Authors: Matthias Herp, Johannes Brachem, Michael Altenbuchinger, Thomas Kneib
Abstract: Graphical Transformation Models (GTMs) are introduced as a novel approach to effectively model multivariate data with intricate marginals and complex dependency structures non-parametrically, while maintaining interpretability through the identification of varying conditional independencies. GTMs extend multivariate transformation models by replacing the Gaussian copula with a custom-designed multivariate transformation, offering two major advantages. Firstly, GTMs can capture more complex interdependencies using penalized splines, which also provide an efficient regularization scheme. Secondly, we demonstrate how to approximately regularize GTMs using a lasso penalty towards pairwise conditional independencies, akin to Gaussian graphical models. The model's robustness and effectiveness are validated through simulations, showcasing its ability to accurately learn parametric vine copulas and identify conditional independencies. Additionally, the model is applied to a benchmark astrophysics dataset, where the GTM demonstrates favorable performance compared to non-parametric vine copulas in learning complex multivariate distributions.
Authors: Roie Kazoom, Raz Lapid, Moshe Sipper, Ofer Hadar
Abstract: Adversarial patch attacks pose a major threat to vision systems by embedding localized perturbations that mislead deep models. Traditional defense methods often require retraining or fine-tuning, making them impractical for real-world deployment. We propose a training-free Visual Retrieval-Augmented Generation (VRAG) framework that integrates Vision-Language Models (VLMs) for adversarial patch detection. By retrieving visually similar patches and images that resemble stored attacks in a continuously expanding database, VRAG performs generative reasoning to identify diverse attack types, all without additional training or fine-tuning. We extensively evaluate open-source large-scale VLMs, including Qwen-VL-Plus, Qwen2.5-VL-72B, and UI-TARS-72B-DPO, alongside Gemini-2.0, a closed-source model. Notably, the open-source UI-TARS-72B-DPO model achieves up to 95 percent classification accuracy, setting a new state-of-the-art for open-source adversarial patch detection. Gemini-2.0 attains the highest overall accuracy, 98 percent, but remains closed-source. Experimental results demonstrate VRAG's effectiveness in identifying a variety of adversarial patches with minimal human annotation, paving the way for robust, practical defenses against evolving adversarial patch attacks.
Authors: Ziwei Yang, Takeyuki Tamura
Abstract: In genome-scale constraint-based metabolic models, gene deletion strategies are crucial for achieving growth-coupled production, where cell growth and target metabolite production are simultaneously achieved. While computational methods for calculating gene deletions have been widely explored and contribute to developing gene deletion strategy databases, current approaches are limited in leveraging new data-driven paradigms, such as machine learning, for more efficient strain design. Therefore, it is necessary to propose a fundamental framework for this objective. In this study, we first formulate the problem of gene deletion strategy prediction and then propose a framework for predicting gene deletion strategies for growth-coupled production in genome-scale metabolic models. The proposed framework leverages deep learning algorithms to learn and integrate sequential gene and metabolite data representation, enabling the automatic gene deletion strategy prediction. Computational experiment results demonstrate the feasibility of the proposed framework, showing substantial improvements over baseline methods. Specifically, the proposed framework achieves a 14.69%, 22.52%, and 13.03% increase in overall accuracy across three metabolic models of different scales under study, while maintaining balanced precision and recall in predicting gene deletion statuses. The source code and examples for the framework are publicly available at https://github.com/MetNetComp/DeepGDel.
Authors: Cassidy Laidlaw, Eli Bronstein, Timothy Guo, Dylan Feng, Lukas Berglund, Justin Svegliato, Stuart Russell, Anca Dragan
Abstract: Assistance games are a promising alternative to reinforcement learning from human feedback (RLHF) for training AI assistants. Assistance games resolve key drawbacks of RLHF, such as incentives for deceptive behavior, by explicitly modeling the interaction between assistant and user as a two-player game where the assistant cannot observe their shared goal. Despite their potential, assistance games have only been explored in simple settings. Scaling them to more complex environments is difficult because it requires both solving intractable decision-making problems under uncertainty and accurately modeling human users' behavior. We present the first scalable approach to solving assistance games and apply it to a new, challenging Minecraft-based assistance game with over $10^{400}$ possible goals. Our approach, AssistanceZero, extends AlphaZero with a neural network that predicts human actions and rewards, enabling it to plan under uncertainty. We show that AssistanceZero outperforms model-free RL algorithms and imitation learning in the Minecraft-based assistance game. In a human study, our AssistanceZero-trained assistant significantly reduces the number of actions participants take to complete building tasks in Minecraft. Our results suggest that assistance games are a tractable framework for training effective AI assistants in complex environments. Our code and models are available at https://github.com/cassidylaidlaw/minecraft-building-assistance-game.
URLs: https://github.com/cassidylaidlaw/minecraft-building-assistance-game.
Authors: William Corrias, Fabio De Gaspari, Dorjan Hitaj, Luigi V. Mancini
Abstract: Recent advances in generative models have led to their application in password guessing, with the aim of replicating the complexity, structure, and patterns of human-created passwords. Despite their potential, inconsistencies and inadequate evaluation methodologies in prior research have hindered meaningful comparisons and a comprehensive, unbiased understanding of their capabilities. This paper introduces MAYA, a unified, customizable, plug-and-play benchmarking framework designed to facilitate the systematic characterization and benchmarking of generative password-guessing models in the context of trawling attacks. Using MAYA, we conduct a comprehensive assessment of six state-of-the-art approaches, which we re-implemented and adapted to ensure standardization. Our evaluation spans eight real-world password datasets and covers an exhaustive set of advanced testing scenarios, totaling over 15,000 compute hours. Our findings indicate that these models effectively capture different aspects of human password distribution and exhibit strong generalization capabilities. However, their effectiveness varies significantly with long and complex passwords. Through our evaluation, sequential models consistently outperform other generative architectures and traditional password-guessing tools, demonstrating unique capabilities in generating accurate and complex guesses. Moreover, the diverse password distributions learned by the models enable a multi-model attack that outperforms the best individual model. By releasing MAYA, we aim to foster further research, providing the community with a new tool to consistently and reliably benchmark generative password-guessing models. Our framework is publicly available at https://github.com/williamcorrias/MAYA-Password-Benchmarking.
URLs: https://github.com/williamcorrias/MAYA-Password-Benchmarking.
Authors: Anirudhan Badrinath, Alex Yang, Kousik Rajesh, Prabhat Agarwal, Jaewon Yang, Haoyu Chen, Jiajing Xu, Charles Rosenberg
Abstract: Representation learning, a task of learning latent vectors to represent entities, is a key task in improving search and recommender systems in web applications. Various representation learning methods have been developed, including graph-based approaches for relationships among entities, sequence-based methods for capturing the temporal evolution of user activities, and content-based models for leveraging text and visual content. However, the development of a unifying framework that integrates these diverse techniques to support multiple applications remains a significant challenge. This paper presents OmniSage, a large-scale representation framework that learns universal representations for a variety of applications at Pinterest. OmniSage integrates graph neural networks with content-based models and user sequence models by employing multiple contrastive learning tasks to effectively process graph data, user sequence data, and content signals. To support the training and inference of OmniSage, we developed an efficient infrastructure capable of supporting Pinterest graphs with billions of nodes. The universal representations generated by OmniSage have significantly enhanced user experiences on Pinterest, leading to an approximate 2.5% increase in sitewide repins (saves) across five applications. This paper highlights the impact of unifying representation learning methods, and we make the model code publicly available at https://github.com/pinterest/atg-research/tree/main/omnisage.
URLs: https://github.com/pinterest/atg-research/tree/main/omnisage.
Authors: Hanrui Wang, Shuo Wang, Chun-Shien Lu, Isao Echizen
Abstract: Face recognition technology presents serious privacy risks due to its reliance on sensitive and immutable biometric data. To address these concerns, such systems typically convert raw facial images into embeddings, which are traditionally viewed as privacy-preserving. However, model inversion attacks challenge this assumption by reconstructing private facial images from embeddings, highlighting a critical vulnerability in face recognition systems. Most existing inversion methods require training a separate generator for each target model, making them computationally intensive. In this work, we introduce DiffUMI, a diffusion-based universal model inversion attack that requires no additional training. DiffUMI is the first approach to successfully leverage unconditional face generation without relying on model-specific generators. It surpasses state-of-the-art attacks by 15.5% and 9.82% in success rate on standard and privacy-preserving face recognition systems, respectively. Furthermore, we propose a novel use of out-of-domain detection (OODD), demonstrating for the first time that model inversion can differentiate between facial and non-facial embeddings using only the embedding space.
Authors: Cornelius Hake, Christian Friedrich
Abstract: This study examines the digital value chain in automotive manufacturing, focusing on the identification, software flashing, customization, and commissioning of electronic control units in vehicle networks. A novel precedence graph design is proposed to optimize this process chain using an automated scheduling algorithm, which combines structured data extraction from heterogeneous sources via natural language processing and classification techniques with mixed integer linear programming for efficient graph generation. The results show significant improvements in key metrics. The algorithm reduces the number of production stations equipped with expensive hardware and software to execute digital value chain processes, while also increasing capacity utilization through efficient scheduling and reduced idle time. Task parallelization is optimized, resulting in streamlined workflows and increased throughput. Compared to the traditional scheduling method, the automated approach has reduced preparation time by 50% and reduced scheduling activities, as it now takes two minutes to create the precedence graph. The flexibility of the algorithm's constraints allows for vehicle-specific configurations while maintaining high responsiveness, eliminating backup stations and facilitating the integration of new topologies. Automated scheduling significantly outperforms manual methods in efficiency, functionality, and adaptability.
Authors: Trisanth Srinivasan, Santosh Patapati
Abstract: Robust navigation in diverse environments and domains requires both accurate state estimation and transparent decision making. We present PhysNav-DG, a novel framework that integrates classical sensor fusion with the semantic power of vision-language models. Our dual-branch architecture predicts navigation actions from multi-sensor inputs while simultaneously generating detailed chain-of-thought explanations. A modified Adaptive Kalman Filter dynamically adjusts its noise parameters based on environmental context. It leverages several streams of raw sensor data along with semantic insights from models such as LLaMA 3.2 11B and BLIP-2. To evaluate our approach, we introduce the MD-NEX Benchmark, a novel multi-domain dataset that unifies indoor navigation, autonomous driving, and social navigation tasks with ground-truth actions and human-validated explanations. Extensive experiments and ablations show that PhysNav-DG improves navigation success rates by over 20% and achieves high efficiency, with explanations that are both highly grounded and clear. This work connects high-level semantic reasoning and geometric planning for safer and more trustworthy autonomous systems.
Authors: Sajjad Rezvani Boroujeni, Hossein Abedi, Tom Bush
Abstract: Visual defect detection in industrial glass manufacturing remains a critical challenge due to the low frequency of defective products, leading to imbalanced datasets that limit the performance of deep learning models and computer vision systems. This paper presents a novel approach using Denoising Diffusion Probabilistic Models (DDPMs) to generate synthetic defective glass product images for data augmentation, effectively addressing class imbalance issues in manufacturing quality control and automated visual inspection. The methodology significantly enhances image classification performance of standard CNN architectures (ResNet50V2, EfficientNetB0, and MobileNetV2) in detecting anomalies by increasing the minority class representation. Experimental results demonstrate substantial improvements in key machine learning metrics, particularly in recall for defective samples across all tested deep neural network architectures while maintaining perfect precision. The most dramatic improvement was observed in ResNet50V2's overall classification accuracy, which increased from seventy-eight percent to ninety-three percent when trained with the augmented data. This work provides a scalable, cost effective approach to enhancing automated defect detection in glass manufacturing that can potentially be extended to other industrial quality assurance systems and industries with similar class imbalance challenges.
Authors: Derrick Gilchrist Edward Manoharan, Anubha Goel, Alexandros Iosifidis, Henri Hansen, Juho Kanniainen
Abstract: The spreading dynamics in social networks are often studied under the assumption that individuals' statuses, whether informed or infected, are fully observable. However, in many real-world situations, such statuses remain unobservable, which is crucial for determining an individual's potential to further spread the infection. While this final status is hidden, intermediate indicators such as symptoms of infection are observable and provide important insights into the spread process. We propose a partial observability-aware Machine Learning framework to learn the characteristics of the spreading model. We term the method Distribution Classification, which utilizes the power of classifiers to infer the underlying transmission dynamics. We evaluate our method on two types of synthetic networks and extend the study to a real-world insider trading network. Results show that the method performs well, especially on complex networks with high cyclic connectivity, supporting its utility in analyzing real-world spreading phenomena where direct observation of individual statuses is not possible.
Authors: Marcel Arpogaus, Thomas Kneib, Thomas Nagler, David R\"ugamer
Abstract: Density regression models allow a comprehensive understanding of data by modeling the complete conditional probability distribution. While flexible estimation approaches such as normalizing flows (NF) work particularly well in multiple dimensions, interpreting the input-output relationship of such models is often difficult, due to the black-box character of deep learning models. In contrast, existing statistical methods for multivariate outcomes such as multivariate conditional transformation models (MCTM) are restricted in flexibility and are often not expressive enough to represent complex multivariate probability distributions. In this paper, we combine MCTM with state-of-the-art and autoregressive NF to leverage the transparency of MCTM for modeling interpretable feature effects on the marginal distributions in the first step and the flexibility of neural-network-based NF techniques to account for complex and non-linear relationships in the joint data distribution. We demonstrate our method's versatility in various numerical experiments and compare it with MCTM and other NF models on both simulated and real-world data.
Authors: Thinh Pham, Nguyen Nguyen, Pratibha Zunjare, Weiyuan Chen, Yu-Min Tseng, Tu Vu
Abstract: We introduce SealQA, a new challenge benchmark for evaluating SEarch-Augmented Language models on fact-seeking questions where web search yields conflicting, noisy, or unhelpful results. SealQA comes in three flavors: (1) Seal-0 (main) and (2) Seal-Hard, which assess factual accuracy and reasoning capabilities, with Seal-0 focusing on the most challenging questions where chat models (e.g., GPT-4.1) typically achieve near-zero accuracy; and (3) LongSeal, which extends SealQA to test long-context, multi-document reasoning in "needle-in-a-haystack" settings. Our evaluation reveals critical limitations in current models: Even frontier LLMs perform poorly across all SealQA flavors. On Seal-0, frontier agentic models equipped with tools like o3 and o4-mini achieve only 17.1% and 6.3% accuracy, respectively, at their best reasoning efforts. We find that advanced reasoning models such as DeepSeek-R1-671B and o3-mini are highly vulnerable to noisy search results. Notably, increasing test-time compute does not yield reliable gains across o3-mini, o4-mini, and o3, with performance often plateauing or even declining early. Additionally, while recent models are less affected by the "lost-in-the-middle" issue, they still fail to reliably identify relevant documents in LongSeal when faced with numerous distractors. To facilitate future work, we release SealQA at huggingface.co/datasets/vtllms/sealqa.
Authors: Yulei Qin, Gang Li, Zongyi Li, Zihan Xu, Yuchen Shi, Zhekai Lin, Xiao Cui, Ke Li, Xing Sun
Abstract: Existing large language models (LLMs) face challenges of following complex instructions, especially when multiple constraints are present and organized in paralleling, chaining, and branching structures. One intuitive solution, namely chain-of-thought (CoT), is expected to universally improve capabilities of LLMs. However, we find that the vanilla CoT exerts a negative impact on performance due to its superficial reasoning pattern of simply paraphrasing the instructions. It fails to peel back the compositions of constraints for identifying their relationship across hierarchies of types and dimensions. To this end, we propose a systematic method to boost LLMs in dealing with complex instructions via incentivizing reasoning for test-time compute scaling. First, we stem from the decomposition of complex instructions under existing taxonomies and propose a reproducible data acquisition method. Second, we exploit reinforcement learning (RL) with verifiable rule-centric reward signals to cultivate reasoning specifically for instruction following. We address the shallow, non-essential nature of reasoning under complex instructions via sample-wise contrast for superior CoT enforcement. We also exploit behavior cloning of experts to facilitate steady distribution shift from fast-thinking LLMs to skillful reasoners. Extensive evaluations on seven comprehensive benchmarks confirm the validity of the proposed method, where a 1.5B LLM achieves 11.74% gains with performance comparable to a 8B LLM. Codes and data are available at https://github.com/yuleiqin/RAIF.
Authors: Li Liu, Heng Yong
Abstract: Recently, machine learning methods have gained significant traction in scientific computing, particularly for solving Partial Differential Equations (PDEs). However, methods based on deep neural networks (DNNs) often lack convergence guarantees and computational efficiency compared to traditional numerical schemes. This work introduces DeePoly, a novel framework that transforms the solution paradigm from pure non-convex parameter optimization to a two-stage approach: first employing a DNN to capture complex global features, followed by linear space optimization with combined DNN-extracted features (Scoper) and polynomial basis functions (Sniper). This strategic combination leverages the complementary strengths of both methods -- DNNs excel at approximating complex global features (i.e., high-gradient features) and stabilize the polynomial approximation while polynomial bases provide high-precision local corrections with convergence guarantees. Theoretical analysis and numerical experiments demonstrate that this approach significantly enhances both high-order accuracy and efficiency across diverse problem types while maintaining mesh-free and scheme-free properties. This paper also serves as a theoretical exposition for the open-source project DeePoly.
Authors: Alex Pan, Mary-Anne Williams
Abstract: The dominant way of judging Large Language Models (LLMs) has been to ask how well they can recall explicit facts from very long inputs. While today's best models achieve near perfect recall, this masks a harder skill: performing multi-step reasoning and tracking intermediate state that never appears verbatim. We introduce Verbose ListOps (VLO), a benchmark that embeds deterministic ListOps computations inside narrative camouflage and, crucially, allows step-level evaluation of every intermediate result. Experiments show that models which solve raw ListOps with approximately 100% accuracy collapse on VLO after only 10,000 tokens. By exposing where a model's reasoning chain first diverges, VLO moves assessment beyond sheer context length and toward genuine comprehension. VLO's generation pipeline is task-agnostic: it can weave any deterministically verifiable reasoning schema -- arithmetic, symbolic, abductive, inductive or defeasible -- into narrative form. This makes VLO a reusable test-bed for the next wave of reasoning-centric model designs, not merely those with step-explicit scaffolds.
Authors: Yijia Shao, Humishka Zope, Yucheng Jiang, Jiaxin Pei, David Nguyen, Erik Brynjolfsson, Diyi Yang
Abstract: The rapid rise of compound AI systems (a.k.a., AI agents) is reshaping the labor market, raising concerns about job displacement, diminished human agency, and overreliance on automation. Yet, we lack a systematic understanding of the evolving landscape. In this paper, we address this gap by introducing a novel auditing framework to assess which occupational tasks workers want AI agents to automate or augment, and how those desires align with the current technological capabilities. Our framework features an audio-enhanced mini-interview to capture nuanced worker desires and introduces the Human Agency Scale (HAS) as a shared language to quantify the preferred level of human involvement. Using this framework, we construct the WORKBank database, building on the U.S. Department of Labor's O*NET database, to capture preferences from 1,500 domain workers and capability assessments from AI experts across over 844 tasks spanning 104 occupations. Jointly considering the desire and technological capability divides tasks in WORKBank into four zones: Automation "Green Light" Zone, Automation "Red Light" Zone, R&D Opportunity Zone, Low Priority Zone. This highlights critical mismatches and opportunities for AI agent development. Moving beyond a simple automate-or-not dichotomy, our results reveal diverse HAS profiles across occupations, reflecting heterogeneous expectations for human involvement. Moreover, our study offers early signals of how AI agent integration may reshape the core human competencies, shifting from information-focused skills to interpersonal ones. These findings underscore the importance of aligning AI agent development with human desires and preparing workers for evolving workplace dynamics.
Authors: Marco Di Gennaro, Giovanni De Lucia, Stefano Longari, Stefano Zanero, Michele Carminati
Abstract: Federated Learning has emerged as a privacy-oriented alternative to centralized Machine Learning, enabling collaborative model training without direct data sharing. While extensively studied for neural networks, the security and privacy implications of tree-based models remain underexplored. This work introduces TimberStrike, an optimization-based dataset reconstruction attack targeting horizontally federated tree-based models. Our attack, carried out by a single client, exploits the discrete nature of decision trees by using split values and decision paths to infer sensitive training data from other clients. We evaluate TimberStrike on State-of-the-Art federated gradient boosting implementations across multiple frameworks, including Flower, NVFlare, and FedTree, demonstrating their vulnerability to privacy breaches. On a publicly available stroke prediction dataset, TimberStrike consistently reconstructs between 73.05% and 95.63% of the target dataset across all implementations. We further analyze Differential Privacy, showing that while it partially mitigates the attack, it also significantly degrades model performance. Our findings highlight the need for privacy-preserving mechanisms specifically designed for tree-based Federated Learning systems, and we provide preliminary insights into their design.
Authors: Kuo Yang, Xingjie Yang, Linhui Yu, Qing Xu, Yan Fang, Xu Wang, Zhengyang Zhou, Yang Wang
Abstract: Large Language Model (LLM)-driven Multi-agent systems (Mas) have recently emerged as a powerful paradigm for tackling complex real-world tasks. However, existing Mas construction methods typically rely on manually crafted interaction mechanisms or heuristic rules, introducing human biases and constraining the autonomous ability. Even with recent advances in adaptive Mas construction, existing systems largely remain within the paradigm of semi-autonomous patterns. In this work, we propose MasHost, a Reinforcement Learning (RL)-based framework for autonomous and query-adaptive Mas design. By formulating Mas construction as a graph search problem, our proposed MasHost jointly samples agent roles and their interactions through a unified probabilistic sampling mechanism. Beyond the accuracy and efficiency objectives pursued in prior works, we introduce component rationality as an additional and novel design principle in Mas. To achieve this multi-objective optimization, we propose Hierarchical Relative Policy Optimization (HRPO), a novel RL strategy that collaboratively integrates group-relative advantages and action-wise rewards. To our knowledge, our proposed MasHost is the first RL-driven framework for autonomous Mas graph construction. Extensive experiments on six benchmarks demonstrate that MasHost consistently outperforms most competitive baselines, validating its effectiveness, efficiency, and structure rationality.
Authors: Yu Guan, Zhiyu Yin, Haoyu Chen, Sheng Cheng, Chaojie Yang, Kun Qian, Tianyin Xu, Yang Zhang, Hanyu Zhao, Yong Li, Wei Lin, Dennis Cai, Ennan Zhai
Abstract: Troubleshooting performance problems of large model training (LMT) is immensely challenging, due to unprecedented scales of modern GPU clusters, the complexity of software-hardware interactions, and the data intensity of the training process. Existing troubleshooting approaches designed for traditional distributed systems or datacenter networks fall short and can hardly apply to real-world training systems. In this paper, we present PerfTracker, the first online troubleshooting system utilizing fine-grained profiling, to diagnose performance issues of large-scale model training in production. PerfTracker can diagnose performance issues rooted in both hardware (e.g., GPUs and their interconnects) and software (e.g., Python functions and GPU operations). It scales to LMT on modern GPU clusters. PerfTracker effectively summarizes runtime behavior patterns of fine-grained LMT functions via online profiling, and leverages differential observability to localize the root cause with minimal production impact. PerfTracker has been deployed as a production service for large-scale GPU clusters of O(10, 000) GPUs (product homepage https://help.aliyun.com/zh/pai/user-guide/perftracker-online-performance-analysis-diagnostic-tool). It has been used to diagnose a variety of difficult performance issues.
URLs: https://help.aliyun.com/zh/pai/user-guide/perftracker-online-performance-analysis-diagnostic-tool).
Authors: Hee Suk Yoon, Eunseop Yoon, Mark Hasegawa-Johnson, Sungwoong Kim, Chang D. Yoo
Abstract: We introduce ConfPO, a method for preference learning in Large Language Models (LLMs) that identifies and optimizes preference-critical tokens based solely on the training policy's confidence, without requiring any auxiliary models or compute. Unlike prior Direct Alignment Algorithms (DAAs) such as Direct Preference Optimization (DPO), which uniformly adjust all token probabilities regardless of their relevance to preference, ConfPO focuses optimization on the most impactful tokens. This targeted approach improves alignment quality while mitigating overoptimization (i.e., reward hacking) by using the KL divergence budget more efficiently. In contrast to recent token-level methods that rely on credit-assignment models or AI annotators, raising concerns about scalability and reliability, ConfPO is simple, lightweight, and model-free. Experimental results on challenging alignment benchmarks, including AlpacaEval 2 and Arena-Hard, demonstrate that ConfPO consistently outperforms uniform DAAs across various LLMs, delivering better alignment with zero additional computational overhead.
Authors: Songze Li, Mingxuan Zhang, Kang Wei, Shouling Ji
Abstract: Deep reinforcement learning (DRL) has achieved remarkable success in a wide range of sequential decision-making domains, including robotics, healthcare, smart grids, and finance. Recent research demonstrates that attackers can efficiently exploit system vulnerabilities during the training phase to execute backdoor attacks, producing malicious actions when specific trigger patterns are present in the state observations. However, most existing backdoor attacks rely primarily on simplistic and heuristic trigger configurations, overlooking the potential efficacy of trigger optimization. To address this gap, we introduce TooBadRL (Trigger Optimization to Boost Effectiveness of Backdoor Attacks on DRL), the first framework to systematically optimize DRL backdoor triggers along three critical axes, i.e., temporal, spatial, and magnitude. Specifically, we first introduce a performance-aware adaptive freezing mechanism for injection timing. Then, we formulate dimension selection as a cooperative game, utilizing Shapley value analysis to identify the most influential state variable for the injection dimension. Furthermore, we propose a gradient-based adversarial procedure to optimize the injection magnitude under environment constraints. Evaluations on three mainstream DRL algorithms and nine benchmark tasks show that TooBadRL significantly improves attack success rates, while ensuring minimal degradation of normal task performance. These results highlight the previously underappreciated importance of principled trigger optimization in DRL backdoor attacks. The source code of TooBadRL can be found at https://github.com/S3IC-Lab/TooBadRL.
Authors: Pierre Vernimmen, Fran\c{c}ois Glineur
Abstract: This work assesses both empirically and theoretically, using the performance estimation methodology, how robust different first-order optimization methods are when subject to relative inexactness in their gradient computations. Relative inexactness occurs, for example, when compressing the gradient using fewer bits of information, which happens when dealing with large-scale problems on GPUs. Three major families of methods are analyzed: constant step gradient descent, long-step methods, and accelerated methods. The latter two are first shown to be theoretically not robust to inexactness. Then, a semi-heuristic shortening factor is introduced to improve their theoretical guarantees. All methods are subsequently tested on a concrete inexact problem, with two different types of relative inexactness, and it is observed that both accelerated methods are much more robust than expected, and that the shortening factor significantly helps the long-step methods. In the end, all shortened methods appear to be promising, even in this inexact setting.
Authors: Chengpeng Li, Zhengyang Tang, Ziniu Li, Mingfeng Xue, Keqin Bao, Tian Ding, Ruoyu Sun, Benyou Wang, Xiang Wang, Junyang Lin, Dayiheng Liu
Abstract: Large Reasoning Models (LRMs) like o1 and DeepSeek-R1 have shown remarkable progress in natural language reasoning with long chain-of-thought (CoT), yet they remain inefficient or inaccurate when handling complex mathematical operations. Addressing these limitations through computational tools (e.g., computation libraries and symbolic solvers) is promising, but it introduces a technical challenge: Code Interpreter (CI) brings external knowledge beyond the model's internal text representations, thus the direct combination is not efficient. This paper introduces CoRT, a post-training framework for teaching LRMs to leverage CI effectively and efficiently. As a first step, we address the data scarcity issue by synthesizing code-integrated reasoning data through Hint-Engineering, which strategically inserts different hints at appropriate positions to optimize LRM-CI interaction. We manually create 30 high-quality samples, upon which we post-train models ranging from 1.5B to 32B parameters, with supervised fine-tuning, rejection fine-tuning and reinforcement learning. Our experimental results demonstrate that Hint-Engineering models achieve 4\% and 8\% absolute improvements on DeepSeek-R1-Distill-Qwen-32B and DeepSeek-R1-Distill-Qwen-1.5B respectively, across five challenging mathematical reasoning datasets. Furthermore, Hint-Engineering models use about 30\% fewer tokens for the 32B model and 50\% fewer tokens for the 1.5B model compared with the natural language models. The models and code are available at https://github.com/ChengpengLi1003/CoRT.