Authors: Tianlei Chen, Yuxiao Chen, Yang Li, Feifei Wang
Abstract: Automated International Classification of Diseases (ICD) coding aims to assign multiple disease codes to clinical documents, constituting a crucial multi-label text classification task in healthcare informatics. However, the task is challenging due to its large label space (10,000 to 20,000 codes) and long-tail distribution, where a few codes dominate while many rare codes lack sufficient training data. To address this, we propose a learning method that models fine-grained co-occurrence relationships among codes. Specifically, we construct a Directed Bipartite Graph Encoder with disjoint sets of common and rare code nodes. To facilitate a one-way information flow, edges are directed exclusively from common to rare codes. The nature of these connections is defined by a probability-based bias, which is derived from the conditional probability of a common code co-occurring given the presence of a rare code. This bias is then injected into the encoder's attention module, a process we term Co-occurrence Encoding. This structure empowers the graph encoder to enrich rare code representations by aggregating latent comorbidity information reflected in the statistical co-occurrence of their common counterparts. To ensure high-quality input to the graph, we utilize a large language model (LLM) to generate comprehensive descriptions for codes, enriching initial embeddings with clinical context and comorbidity information, serving as external knowledge for the statistical co-occurrence relationships in the code system. Experiments on three automated ICD coding benchmark datasets demonstrate that our method achieves state-of-the-art performance with particularly notable improvements in Macro-F1, which is the key metric for long-tail classification.
Authors: Todd Morrill, Aahlad Puli, Murad Megjhani, Soojin Park, Richard Zemel
Abstract: Deep mixture-of-experts models have attracted a lot of attention for survival analysis problems, particularly for their ability to cluster similar patients together. In practice, grouping often comes at the expense of key metrics such calibration error and predictive accuracy. This is due to the restrictive inductive bias that mixture-of-experts imposes, that predictions for individual patients must look like predictions for the group they're assigned to. Might we be able to discover patient group structure, where it exists, while improving calibration and predictive accuracy? In this work, we introduce several discrete-time deep mixture-of-experts (MoE) based architectures for survival analysis problems, one of which achieves all desiderata: clustering, calibration, and predictive accuracy. We show that a key differentiator between this array of MoEs is how expressive their experts are. We find that more expressive experts that tailor predictions per patient outperform experts that rely on fixed group prototypes.
Authors: George Stamatelis, George C. Alexandropoulos
Abstract: This paper presents the Jump Markov Filtering Network (JMFNet), a novel model-based deep learning framework for real-time state-state estimation in jump Markov systems with unknown noise statistics and mode transition dynamics. A hybrid architecture comprising two Recurrent Neural Networks (RNNs) is proposed: one for mode prediction and another for filtering that is based on a mode-augmented version of the recently presented KalmanNet architecture. The proposed RNNs are trained jointly using an alternating least squares strategy that enables mutual adaptation without supervision of the latent modes. Extensive numerical experiments on linear and nonlinear systems, including target tracking, pendulum angle tracking, Lorenz attractor dynamics, and a real-life dataset demonstrate that the proposed JMFNet framework outperforms classical model-based filters (e.g., interacting multiple models and particle filters) as well as model-free deep learning baselines, particularly in non-stationary and high-noise regimes. It is also showcased that JMFNet achieves a small yet meaningful improvement over the KalmanNet framework, which becomes much more pronounced in complicated systems or long trajectories. Finally, the method's performance is empirically validated to be consistent and reliable, exhibiting low sensitivity to initial conditions, hyperparameter selection, as well as to incorrect model knowledge
Authors: Valentino F. Foit, David W. Hogg, Soledad Villar
Abstract: Many machine learning tasks in the natural sciences are precisely equivariant to particular symmetries. Nonetheless, equivariant methods are often not employed, perhaps because training is perceived to be challenging, or the symmetry is expected to be learned, or equivariant implementations are seen as hard to build. Group averaging is an available technique for these situations. It happens at test time; it can make any trained model precisely equivariant at a (often small) cost proportional to the size of the group; it places no requirements on model structure or training. It is known that, under mild conditions, the group-averaged model will have a provably better prediction accuracy than the original model. Here we show that an inexpensive group averaging can improve accuracy in practice. We take well-established benchmark machine learning models of differential equations in which certain symmetries ought to be obeyed. At evaluation time, we average the models over a small group of symmetries. Our experiments show that this procedure always decreases the average evaluation loss, with improvements of up to 37\% in terms of the VRMSE. The averaging produces visually better predictions for continuous dynamics. This short paper shows that, under certain common circumstances, there are no disadvantages to imposing exact symmetries; the ML4PS community should consider group averaging as a cheap and simple way to improve model accuracy.
Authors: Hadi Keramati, Morteza Sadeghi, Rajeev K. Jaiman
Abstract: This study presents a generative optimization framework based on a guided denoising diffusion probabilistic model (DDPM) that leverages surrogate gradients to generate heat sink designs minimizing pressure drop while maintaining surface temperatures below a specified threshold. Geometries are represented using boundary representations of multiple fins, and a multi-fidelity approach is employed to generate training data. Using this dataset, along with vectors representing the boundary representation geometries, we train a denoising diffusion probabilistic model to generate heat sinks with characteristics consistent with those observed in the data. We train two different residual neural networks to predict the pressure drop and surface temperature for each geometry. We use the gradients of these surrogate models with respect to the design variables to guide the geometry generation process toward satisfying the low-pressure and surface temperature constraints. This inference-time guidance directs the generative process toward heat sink designs that not only prevent overheating but also achieve lower pressure drops compared to traditional optimization methods such as CMA-ES. In contrast to traditional black-box optimization approaches, our method is scalable, provided sufficient training data is available. Unlike traditional topology optimization methods, once the model is trained and the heat sink world model is saved, inference under new constraints (e.g., temperature) is computationally inexpensive and does not require retraining. Samples generated using the guided diffusion model achieve pressure drops up to 10 percent lower than the limits obtained by traditional black-box optimization methods. This work represents a step toward building a foundational generative model for electronics cooling.
Authors: Yuchen Huang, Sijia Li, Minghao Liu, Wei Liu, Shijue Huang, Zhiyuan Fan, Hou Pong Chan, Yi R. Fung
Abstract: LLM-based agents can autonomously accomplish complex tasks across various domains. However, to further cultivate capabilities such as adaptive behavior and long-term decision-making, training on static datasets built from human-level knowledge is insufficient. These datasets are costly to construct and lack both dynamism and realism. A growing consensus is that agents should instead interact directly with environments and learn from experience through reinforcement learning. We formalize this iterative process as the Generation-Execution-Feedback (GEF) loop, where environments generate tasks to challenge agents, return observations in response to agents' actions during task execution, and provide evaluative feedback on rollouts for subsequent learning. Under this paradigm, environments function as indispensable producers of experiential data, highlighting the need to scale them toward greater complexity, realism, and interactivity. In this survey, we systematically review representative methods for environment scaling from a pioneering environment-centric perspective and organize them along the stages of the GEF loop, namely task generation, task execution, and feedback. We further analyze benchmarks, implementation strategies, and applications, consolidating fragmented advances and outlining future research directions for agent intelligence.
Authors: Ruiyang Ma, Yunhao Zhou, Yipeng Wang, Yi Liu, Zhengyuan Shi, Ziyang Zheng, Kexin Chen, Zhiqiang He, Lingwei Yan, Gang Chen, Qiang Xu, Guojie Luo
Abstract: There is a growing body of work on using Graph Neural Networks (GNNs) to learn representations of circuits, focusing primarily on their static characteristics. However, these models fail to capture circuit runtime behavior, which is crucial for tasks like circuit verification and optimization. To address this limitation, we introduce DR-GNN (DynamicRTL-GNN), a novel approach that learns RTL circuit representations by incorporating both static structures and multi-cycle execution behaviors. DR-GNN leverages an operator-level Control Data Flow Graph (CDFG) to represent Register Transfer Level (RTL) circuits, enabling the model to capture dynamic dependencies and runtime execution. To train and evaluate DR-GNN, we build the first comprehensive dynamic circuit dataset, comprising over 6,300 Verilog designs and 63,000 simulation traces. Our results demonstrate that DR-GNN outperforms existing models in branch hit prediction and toggle rate prediction. Furthermore, its learned representations transfer effectively to related dynamic circuit tasks, achieving strong performance in power estimation and assertion prediction.
Authors: Mingkuan Zhao, Wentao Hu, Jiayin Wang, Xin Lai, Tianchen Huang, Yuheng Min, Rui Yan, Xiaoyan Zhu
Abstract: The design of Large Language Models (LLMs) has long been hampered by a fundamental conflict within their core attention mechanism: its remarkable expressivity is built upon a computational complexity of $O(H \cdot N^2)$ that grows quadratically with the context size ($N$) and linearly with the number of heads ($H$). This standard implementation harbors significant computational redundancy, as all heads independently compute attention over the same sequence space. Existing sparse methods, meanwhile, often trade information integrity for computational efficiency. To resolve this efficiency-performance trade-off, we propose SPAttention, whose core contribution is the introduction of a new paradigm we term Principled Structural Sparsity. SPAttention does not merely drop connections but instead reorganizes the computational task by partitioning the total attention workload into balanced, non-overlapping distance bands, assigning each head a unique segment. This approach transforms the multi-head attention mechanism from $H$ independent $O(N^2)$ computations into a single, collaborative $O(N^2)$ computation, fundamentally reducing complexity by a factor of $H$. The structured inductive bias compels functional specialization among heads, enabling a more efficient allocation of computational resources from redundant modeling to distinct dependencies across the entire sequence span. Extensive empirical validation on the OLMoE-1B-7B and 0.25B-1.75B model series demonstrates that while delivering an approximately two-fold increase in training throughput, its performance is on par with standard dense attention, even surpassing it on select key metrics, while consistently outperforming representative sparse attention methods including Longformer, Reformer, and BigBird across all evaluation metrics.
Authors: Tingyang Wei, Jiao Liu, Abhishek Gupta, Chin Chun Ooi, Puay Siew Tan, Yew-Soon Ong
Abstract: Many real-world applications require solving families of expensive multi-objective optimization problems~(EMOPs) under varying operational conditions. This gives rise to parametric expensive multi-objective optimization problems (P-EMOPs) where each task parameter defines a distinct optimization instance. Current multi-objective Bayesian optimization methods have been widely used for finding finite sets of Pareto optimal solutions for individual tasks. However, P-EMOPs present a fundamental challenge: the continuous task parameter space can contain infinite distinct problems, each requiring separate expensive evaluations. This demands learning an inverse model that can directly predict optimized solutions for any task-preference query without expensive re-evaluation. This paper introduces the first parametric multi-objective Bayesian optimizer that learns this inverse model by alternating between (1) acquisition-driven search leveraging inter-task synergies and (2) generative solution sampling via conditional generative models. This approach enables efficient optimization across related tasks and finally achieves direct solution prediction for unseen parameterized EMOPs without additional expensive evaluations. We theoretically justify the faster convergence by leveraging inter-task synergies through task-aware Gaussian processes. Meanwhile, empirical studies in synthetic and real-world benchmarks further verify the effectiveness of our alternating framework.
Authors: Mohammad Alipour-Vaezi, Huaiyang Zhong, Kwok-Leung Tsui, Sajad Khodadadian
Abstract: Reinforcement Learning (RL) has achieved tremendous success in recent years. However, the classical foundations of RL do not account for the risk sensitivity of the objective function, which is critical in various fields, including healthcare and finance. A popular approach to incorporate risk sensitivity is to optimize a specific quantile of the cumulative reward distribution. In this paper, we develop UCB-QRL, an optimistic learning algorithm for the $\tau$-quantile objective in finite-horizon Markov decision processes (MDPs). UCB-QRL is an iterative algorithm in which, at each iteration, we first estimate the underlying transition probability and then optimize the quantile value function over a confidence ball around this estimate. We show that UCB-QRL yields a high-probability regret bound $\mathcal O\left((2/\kappa)^{H+1}H\sqrt{SATH\log(2SATH/\delta)}\right)$ in the episodic setting with $S$ states, $A$ actions, $T$ episodes, and $H$ horizons. Here, $\kappa>0$ is a problem-dependent constant that captures the sensitivity of the underlying MDP's quantile value.
Authors: Junwei Ma, Nour Shaheen, Alex Labach, Amine Mhedhbi, Frank Hutter, Anthony L. Caterini, Valentin Thomas
Abstract: Deep tabular modelling increasingly relies on in-context learning where, during inference, a model receives a set of $(x,y)$ pairs as context and predicts labels for new inputs without weight updates. We challenge the prevailing view that broad generalization here requires pre-training on large synthetic corpora (e.g., TabPFN priors) or a large collection of real data (e.g., TabDPT training datasets), discovering that a relatively small amount of data suffices for generalization. We find that simple self-supervised pre-training on just a \emph{single} real table can produce surprisingly strong transfer across heterogeneous benchmarks. By systematically pre-training and evaluating on many diverse datasets, we analyze what aspects of the data are most important for building a Tabular Foundation Model (TFM) generalizing across domains. We then connect this to the pre-training procedure shared by most TFMs and show that the number and quality of \emph{tasks} one can construct from a dataset is key to downstream performance.
Authors: Samuel Maddock, Shripad Gade, Graham Cormode, Will Bullock
Abstract: State-of-the-art differentially private synthetic tabular data has been defined by adaptive 'select-measure-generate' frameworks, exemplified by methods like AIM. These approaches iteratively measure low-order noisy marginals and fit graphical models to produce synthetic data, enabling systematic optimisation of data quality under privacy constraints. Graphical models, however, are inefficient for high-dimensional data because they require substantial memory and must be retrained from scratch whenever the graph structure changes, leading to significant computational overhead. Recent methods, like GEM, overcome these limitations by using generator neural networks for improved scalability. However, empirical comparisons have mostly focused on small datasets, limiting real-world applicability. In this work, we introduce GEM+, which integrates AIM's adaptive measurement framework with GEM's scalable generator network. Our experiments show that GEM+ outperforms AIM in both utility and scalability, delivering state-of-the-art results while efficiently handling datasets with over a hundred columns, where AIM fails due to memory and computational overheads.
Authors: Pedro Dall'Antonia, Tiago da Silva, Daniel Augusto de Souza, C\'esar Lincoln C. Mattos, Diego Mesquita
Abstract: Generative Flow Networks (GFlowNets) are powerful samplers for compositional objects that, by design, sample proportionally to a given non-negative reward. Nonetheless, in practice, they often struggle to explore the reward landscape evenly: trajectories toward easy-to-reach regions dominate training, while hard-to-reach modes receive vanishing or uninformative gradients, leading to poor coverage of high-reward areas. We address this imbalance with Boosted GFlowNets, a method that sequentially trains an ensemble of GFlowNets, each optimizing a residual reward that compensates for the mass already captured by previous models. This residual principle reactivates learning signals in underexplored regions and, under mild assumptions, ensures a monotone non-degradation property: adding boosters cannot worsen the learned distribution and typically improves it. Empirically, Boosted GFlowNets achieve substantially better exploration and sample diversity on multimodal synthetic benchmarks and peptide design tasks, while preserving the stability and simplicity of standard trajectory-balance training.
Authors: Tairan Huang, Yulin Jin, Junxu Liu, Qingqing Ye, Haibo Hu
Abstract: Visual reinforcement learning has achieved remarkable progress in visual control and robotics, but its vulnerability to adversarial perturbations remains underexplored. Most existing black-box attacks focus on vector-based or discrete-action RL, and their effectiveness on image-based continuous control is limited by the large action space and excessive environment queries. We propose SEBA, a sample-efficient framework for black-box adversarial attacks on visual RL agents. SEBA integrates a shadow Q model that estimates cumulative rewards under adversarial conditions, a generative adversarial network that produces visually imperceptible perturbations, and a world model that simulates environment dynamics to reduce real-world queries. Through a two-stage iterative training procedure that alternates between learning the shadow model and refining the generator, SEBA achieves strong attack performance while maintaining efficiency. Experiments on MuJoCo and Atari benchmarks show that SEBA significantly reduces cumulative rewards, preserves visual fidelity, and greatly decreases environment interactions compared to prior black-box and white-box methods.
Authors: Weiqin Chen, Nhan Huu Pham, Michael Robert Glass, Long Hai Vu, Gaetano Rossiello, Dharmashankar Subramanian, Santiago Paternain
Abstract: Reinforcement learning (RL) has demonstrated significant promise in enhancing the reasoning capabilities of Text2SQL LLMs, especially with advanced algorithms such as GRPO and DAPO. However, the performance of these methods is highly sensitive to the design of reward functions. Inappropriate rewards can lead to reward hacking, where models exploit loopholes in the reward structure to achieve high scores without genuinely solving the task. This work considers a constrained RL framework for Text2SQL that incorporates natural and interpretable reward and constraint signals, while dynamically balancing trade-offs among them during the training. We establish the theoretical guarantees of our constrained RL framework and our numerical experiments on the well-known Text2SQL datasets substantiate the improvement of our approach over the state-of-the-art RL-trained LLMs.
Authors: Marco Angioli, Christopher J. Kymn, Antonello Rosato, Amy Loutfi, Mauro Olivieri, Denis Kleyko
Abstract: The modular composite representation (MCR) is a computing model that represents information with high-dimensional integer vectors using modular arithmetic. Originally proposed as a generalization of the binary spatter code model, it aims to provide higher representational power while remaining a lighter alternative to models requiring high-precision components. Despite this potential, MCR has received limited attention. Systematic analyses of its trade-offs and comparisons with other models are lacking, sustaining the perception that its added complexity outweighs the improved expressivity. In this work, we revisit MCR by presenting its first extensive evaluation, demonstrating that it achieves a unique balance of capacity, accuracy, and hardware efficiency. Experiments measuring capacity demonstrate that MCR outperforms binary and integer vectors while approaching complex-valued representations at a fraction of their memory footprint. Evaluation on 123 datasets confirms consistent accuracy gains and shows that MCR can match the performance of binary spatter codes using up to 4x less memory. We investigate the hardware realization of MCR by showing that it maps naturally to digital logic and by designing the first dedicated accelerator. Evaluations on basic operations and 7 selected datasets demonstrate a speedup of up to 3 orders of magnitude and significant energy reductions compared to software implementation. When matched for accuracy against binary spatter codes, MCR achieves on average 3.08x faster execution and 2.68x lower energy consumption. These findings demonstrate that, although MCR requires more sophisticated operations than binary spatter codes, its modular arithmetic and higher per-component precision enable lower dimensionality. When realized with dedicated hardware, this results in a faster, more energy-efficient, and high-precision alternative to existing models.
Authors: Qian-Ze Zhu, Paul Raccuglia, Michael P. Brenner
Abstract: Solving partial differential equations (PDEs) can be prohibitively expensive using traditional numerical methods. Deep learning-based surrogate models typically specialize in a single PDE with fixed parameters. We present a framework for equation-aware emulation that generalizes to unseen PDEs, conditioning a neural model on a vector encoding representing the terms in a PDE and their coefficients. We present a baseline of four distinct modeling technqiues, trained on a family of 1D PDEs from the APEBench suite. Our approach achieves strong performance on parameter sets held out from the training distribution, with strong stability for rollout beyond the training window, and generalization to an entirely unseen PDE. This work was developed as part of a broader effort exploring AI systems that automate the creation of expert-level empirical software for scorable scientific tasks. The data and codebase are available at https://github.com/google-research/generalized-pde-emulator.
URLs: https://github.com/google-research/generalized-pde-emulator.
Authors: Bernardo Perrone Ribeiro, Jana Faganeli Pucer
Abstract: Radar-based precipitation nowcasting, the task of forecasting short-term precipitation fields from previous radar images, is a critical problem for flood risk management and decision-making. While deep learning has substantially advanced this field, two challenges remain fundamental: the uncertainty of atmospheric dynamics and the efficient modeling of high-dimensional data. Diffusion models have shown strong promise by producing sharp, reliable forecasts, but their iterative sampling process is computationally prohibitive for time-critical applications. We introduce FlowCast, the first model to apply Conditional Flow Matching (CFM) to precipitation nowcasting. Unlike diffusion, CFM learns a direct noise-to-data mapping, enabling rapid, high-fidelity sample generation with drastically fewer function evaluations. Our experiments demonstrate that FlowCast establishes a new state-of-the-art in predictive accuracy. A direct comparison further reveals the CFM objective is both more accurate and significantly more efficient than a diffusion objective on the same architecture, maintaining high performance with significantly fewer sampling steps. This work positions CFM as a powerful and practical alternative for high-dimensional spatiotemporal forecasting.
Authors: Joana Tirana, Dimitra Tsigkari, David Solans Noguero, Nicolas Kourtellis
Abstract: In Split Federated Learning (SFL), the clients collaboratively train a model with the help of a server by splitting the model into two parts. Part-1 is trained locally at each client and aggregated by the aggregator at the end of each round. Part-2 is trained at a server that sequentially processes the intermediate activations received from each client. We study the phenomenon of catastrophic forgetting (CF) in SFL in the presence of data heterogeneity. In detail, due to the nature of SFL, local updates of part-1 may drift away from global optima, while part-2 is sensitive to the processing sequence, similar to forgetting in continual learning (CL). Specifically, we observe that the trained model performs better in classes (labels) seen at the end of the sequence. We investigate this phenomenon with emphasis on key aspects of SFL, such as the processing order at the server and the cut layer. Based on our findings, we propose Hydra, a novel mitigation method inspired by multi-head neural networks and adapted for the SFL's setting. Extensive numerical evaluations show that Hydra outperforms baselines and methods from the literature.
Authors: Bram Grooten, Patrick MacAlpine, Kaushik Subramanian, Peter Stone, Peter R. Wurman
Abstract: Generalization to unseen environments is a significant challenge in the field of robotics and control. In this work, we focus on contextual reinforcement learning, where agents act within environments with varying contexts, such as self-driving cars or quadrupedal robots that need to operate in different terrains or weather conditions than they were trained for. We tackle the critical task of generalizing to out-of-distribution (OOD) settings, without access to explicit context information at test time. Recent work has addressed this problem by training a context encoder and a history adaptation module in separate stages. While promising, this two-phase approach is cumbersome to implement and train. We simplify the methodology and introduce SPARC: single-phase adaptation for robust control. We test SPARC on varying contexts within the high-fidelity racing simulator Gran Turismo 7 and wind-perturbed MuJoCo environments, and find that it achieves reliable and robust OOD generalization.
Authors: Houming Wu, Ling Chen
Abstract: Training large language models (LLMs) is fundamentally constrained by limited device memory and costly inter-device communication. Although pipeline parallelism alleviates memory pressure by partitioning models across devices, it incurs activation communication overhead that scales linearly with sequence length, limiting efficiency in long-context training. Recent weight-passing approaches (e.g., WeiPipe) mitigate this by transmitting model weights instead of activations, but suffer from redundant peer-to-peer (P2P) transfers and underutilized intra-node bandwidth. We propose TawPipe--topology-aware weight pipeline parallelism, which exploits hierarchical bandwidth in distributed clusters for improved communication efficiency. TawPipe: (i) groups devices based on topology to optimize intra-node collective and inter-node P2P communication; (ii) assigns each device a fixed shard of model weights and gradients, avoiding redundant transfers; and (iii) overlaps communication with computation to hide latency. Unlike global collective operations used in fully sharded data parallelism (FSDP), TawPipe confines most communication within node boundaries, significantly reducing cross-node traffic. Extensive experiments on up to 24 GPUs with LLaMA-style models show that TawPipe achieves superior throughput and scalability compared to state-of-the-art baselines.
Authors: Sarthak Khanna, Armin Berger, Muskaan Chopra, Rafet Sifa
Abstract: Financial markets are inherently non-stationary: structural breaks and macroeconomic regime shifts often cause forecasting models to fail when deployed out of distribution (OOD). Conventional multimodal approaches that simply fuse numerical indicators and textual sentiment rarely adapt to such shifts. We introduce macro-contextual retrieval, a retrieval-augmented forecasting framework that grounds each prediction in historically analogous macroeconomic regimes. The method jointly embeds macro indicators (e.g., CPI, unemployment, yield spread, GDP growth) and financial news sentiment in a shared similarity space, enabling causal retrieval of precedent periods during inference without retraining. Trained on seventeen years of S&P 500 data (2007-2023) and evaluated OOD on AAPL (2024) and XOM (2024), the framework consistently narrows the CV to OOD performance gap. Macro-conditioned retrieval achieves the only positive out-of-sample trading outcomes (AAPL: PF=1.18, Sharpe=0.95; XOM: PF=1.16, Sharpe=0.61), while static numeric, text-only, and naive multimodal baselines collapse under regime shifts. Beyond metric gains, retrieved neighbors form interpretable evidence chains that correspond to recognizable macro contexts, such as inflationary or yield-curve inversion phases, supporting causal interpretability and transparency. By operationalizing the principle that "financial history may not repeat, but it often rhymes," this work demonstrates that macro-aware retrieval yields robust, explainable forecasts under distributional change. All datasets, models, and source code are publicly available.
Authors: Guy Blanc, Yizhi Huang, Tal Malkin, Rocco A. Servedio
Abstract: We consider the relative abilities and limitations of computationally efficient algorithms for learning in the presence of noise, under two well-studied and challenging adversarial noise models for learning Boolean functions: malicious noise, in which an adversary can arbitrarily corrupt a random subset of examples given to the learner; and nasty noise, in which an adversary can arbitrarily corrupt an adversarially chosen subset of examples given to the learner. We consider both the distribution-independent and fixed-distribution settings. Our main results highlight a dramatic difference between these two settings: For distribution-independent learning, we prove a strong equivalence between the two noise models: If a class ${\cal C}$ of functions is efficiently learnable in the presence of $\eta$-rate malicious noise, then it is also efficiently learnable in the presence of $\eta$-rate nasty noise. In sharp contrast, for the fixed-distribution setting we show an arbitrarily large separation: Under a standard cryptographic assumption, for any arbitrarily large value $r$ there exists a concept class for which there is a ratio of $r$ between the rate $\eta_{malicious}$ of malicious noise that polynomial-time learning algorithms can tolerate, versus the rate $\eta_{nasty}$ of nasty noise that such learning algorithms can tolerate. To offset the negative result for the fixed-distribution setting, we define a broad and natural class of algorithms, namely those that ignore contradictory examples (ICE). We show that for these algorithms, malicious noise and nasty noise are equivalent up to a factor of two in the noise rate: Any efficient ICE learner that succeeds with $\eta$-rate malicious noise can be converted to an efficient learner that succeeds with $\eta/2$-rate nasty noise. We further show that the above factor of two is necessary, again under a standard cryptographic assumption.
Authors: Mahdi Samaee, Mehran Yazdi, Daniel Massicotte
Abstract: Automated sleep stage classification from polysomnography remains limited by the lack of expressive temporal hierarchies, challenges in multimodal EEG and EOG fusion, and the limited interpretability of deep learning models. We propose NeuroLingua, a language-inspired framework that conceptualizes sleep as a structured physiological language. Each 30-second epoch is decomposed into overlapping 3-second subwindows ("tokens") using a CNN-based tokenizer, enabling hierarchical temporal modeling through dual-level Transformers: intra-segment encoding of local dependencies and inter-segment integration across seven consecutive epochs (3.5 minutes) for extended context. Modality-specific embeddings from EEG and EOG channels are fused via a Graph Convolutional Network, facilitating robust multimodal integration. NeuroLingua is evaluated on the Sleep-EDF Expanded and ISRUC-Sleep datasets, achieving state-of-the-art results on Sleep-EDF (85.3% accuracy, 0.800 macro F1, and 0.796 Cohen's kappa) and competitive performance on ISRUC (81.9% accuracy, 0.802 macro F1, and 0.755 kappa), matching or exceeding published baselines in overall and per-class metrics. The architecture's attention mechanisms enhance the detection of clinically relevant sleep microevents, providing a principled foundation for future interpretability, explainability, and causal inference in sleep research. By framing sleep as a compositional language, NeuroLingua unifies hierarchical sequence modeling and multimodal fusion, advancing automated sleep staging toward more transparent and clinically meaningful applications.
Authors: Nikolay Blagoev, O\u{g}uzhan Ersoy, Lydia Yiyu Chen
Abstract: Group Relative Policy Optimization (GRPO) has demonstrated great utilization in post-training of Large Language Models (LLMs). In GRPO, prompts are answered by the model and, through reinforcement learning, preferred completions are learnt. Owing to the small communication volume, GRPO is inherently suitable for decentralised training as the prompts can be concurrently answered by multiple nodes and then exchanged in the forms of strings. In this work, we present the first adversarial attack in decentralised GRPO. We demonstrate that malicious parties can poison such systems by injecting arbitrary malicious tokens in benign models in both out-of-context and in-context attacks. Using empirical examples of math and coding tasks, we show that adversarial attacks can easily poison the benign nodes, polluting their local LLM post-training, achieving attack success rates up to 100% in as few as 50 iterations. We propose two ways to defend against these attacks, depending on whether all users train the same model or different models. We show that these defenses can achieve stop rates of up to 100%, making the attack impossible.
Authors: Pablo Ruiz-Morales, Dries Vanoost, Davy Pissoort, Mathias Verbeke
Abstract: Joint-Embedding Predictive Architectures (JEPAs), a powerful class of self-supervised models, exhibit an unexplained ability to cluster time-series data by their underlying dynamical regimes. We propose a novel theoretical explanation for this phenomenon, hypothesizing that JEPA's predictive objective implicitly drives it to learn the invariant subspace of the system's Koopman operator. We prove that an idealized JEPA loss is minimized when the encoder represents the system's regime indicator functions, which are Koopman eigenfunctions. This theory was validated on synthetic data with known dynamics, demonstrating that constraining the JEPA's linear predictor to be a near-identity operator is the key inductive bias that forces the encoder to learn these invariants. We further discuss that this constraint is critical for selecting this interpretable solution from a class of mathematically equivalent but entangled optima, revealing the predictor's role in representation disentanglement. This work demystifies a key behavior of JEPAs, provides a principled connection between modern self-supervised learning and dynamical systems theory, and informs the design of more robust and interpretable time-series models.
Authors: Fulong Yao, Wanqing Zhao, Chao Zheng, Xiaofei Han
Abstract: Recent advances in deep forecasting models have achieved remarkable performance, yet most approaches still struggle to provide both accurate predictions and interpretable insights into temporal dynamics. This paper proposes CaReTS, a novel multi-task learning framework that combines classification and regression tasks for multi-step time series forecasting problems. The framework adopts a dual-stream architecture, where a classification branch learns the stepwise trend into the future, while a regression branch estimates the corresponding deviations from the latest observation of the target variable. The dual-stream design provides more interpretable predictions by disentangling macro-level trends from micro-level deviations in the target variable. To enable effective learning in output prediction, deviation estimation, and trend classification, we design a multi-task loss with uncertainty-aware weighting to adaptively balance the contribution of each task. Furthermore, four variants (CaReTS1--4) are instantiated under this framework to incorporate mainstream temporal modelling encoders, including convolutional neural networks (CNNs), long short-term memory networks (LSTMs), and Transformers. Experiments on real-world datasets demonstrate that CaReTS outperforms state-of-the-art (SOTA) algorithms in forecasting accuracy, while achieving higher trend classification performance.
Authors: Tianmeng Hu, Yongzheng Cui, Rui Tang, Biao Luo, Ke Li
Abstract: Value decomposition is a central approach in multi-agent reinforcement learning (MARL), enabling centralized training with decentralized execution by factorizing the global value function into local values. To ensure individual-global-max (IGM) consistency, existing methods either enforce monotonicity constraints, which limit expressive power, or adopt softer surrogates at the cost of algorithmic complexity. In this work, we present a dynamical systems analysis of non-monotonic value decomposition, modeling learning dynamics as continuous-time gradient flow. We prove that, under approximately greedy exploration, all zero-loss equilibria violating IGM consistency are unstable saddle points, while only IGM-consistent solutions are stable attractors of the learning dynamics. Extensive experiments on both synthetic matrix games and challenging MARL benchmarks demonstrate that unconstrained, non-monotonic factorization reliably recovers IGM-optimal solutions and consistently outperforms monotonic baselines. Additionally, we investigate the influence of temporal-difference targets and exploration strategies, providing actionable insights for the design of future value-based MARL algorithms.
Authors: Ting Cai, Kirthevasan Kandasamy
Abstract: Best arm identification (BAI) aims to identify the highest-performance arm among a set of $K$ arms by collecting stochastic samples from each arm. In real-world problems, the best arm needs to satisfy additional feasibility constraints. While there is limited prior work on BAI with feasibility constraints, they typically assume the performance and constraints are observed simultaneously on each pull of an arm. However, this assumption does not reflect most practical use cases, e.g., in drug discovery, we wish to find the most potent drug whose toxicity and solubility are below certain safety thresholds. These safety experiments can be conducted separately from the potency measurement. Thus, this requires designing BAI algorithms that not only decide which arm to pull but also decide whether to test for the arm's performance or feasibility. In this work, we study feasible BAI which allows a decision-maker to choose a tuple $(i,\ell)$, where $i\in [K]$ denotes an arm and $\ell$ denotes whether she wishes to test for its performance ($\ell=0$) or any of its $N$ feasibility constraints ($\ell\in[N]$). We focus on the fixed confidence setting, which is to identify the \textit{feasible} arm with the \textit{highest performance}, with a probability of at least $1-\delta$. We propose an efficient algorithm and upper-bound its sample complexity, showing our algorithm can naturally adapt to the problem's difficulty and eliminate arms by worse performance or infeasibility, whichever is easier. We complement this upper bound with a lower bound showing that our algorithm is \textit{asymptotically ($\delta\rightarrow 0$) optimal}. Finally, we empirically show that our algorithm outperforms other state-of-the-art BAI algorithms in both synthetic and real-world datasets.
Authors: Arthur Castello Branco de Oliveira, Dhruv Jatkar, Eduardo Sontag
Abstract: This paper investigates how the compositional structure of neural networks shapes their optimization landscape and training dynamics. We analyze the gradient flow associated with overparameterized optimization problems, which can be interpreted as training a neural network with linear activations. Remarkably, we show that the global convergence properties can be derived for any cost function that is proper and real analytic. We then specialize the analysis to scalar-valued cost functions, where the geometry of the landscape can be fully characterized. In this setting, we demonstrate that key structural features -- such as the location and stability of saddle points -- are universal across all admissible costs, depending solely on the overparameterized representation rather than on problem-specific details. Moreover, we show that convergence can be arbitrarily accelerated depending on the initialization, as measured by an imbalance metric introduced in this work. Finally, we discuss how these insights may generalize to neural networks with sigmoidal activations, showing through a simple example which geometric and dynamical properties persist beyond the linear case.
Authors: Mingkun Yang, Ran Zhu, Qing Wang, Jie Yang
Abstract: Split Federated Learning is a system-efficient federated learning paradigm that leverages the rich computing resources at a central server to train model partitions. Data heterogeneity across silos, however, presents a major challenge undermining the convergence speed and accuracy of the global model. This paper introduces Step-wise Momentum Fusion (SMoFi), an effective and lightweight framework that counteracts gradient divergence arising from data heterogeneity by synchronizing the momentum buffers across server-side optimizers. To control gradient divergence over the training process, we design a staleness-aware alignment mechanism that imposes constraints on gradient updates of the server-side submodel at each optimization step. Extensive validations on multiple real-world datasets show that SMoFi consistently improves global model accuracy (up to 7.1%) and convergence speed (up to 10.25$\times$). Furthermore, SMoFi has a greater impact with more clients involved and deeper learning models, making it particularly suitable for model training in resource-constrained contexts.
Authors: Ilias Diakonikolas, Mingchen Ma, Lisheng Ren, Christos Tzamos
Abstract: Learning intersections of halfspaces is a central problem in Computational Learning Theory. Even for just two halfspaces, it remains a major open question whether learning is possible in polynomial time with respect to the margin $\gamma$ of the data points and their dimensionality $d$. The best-known algorithms run in quasi-polynomial time $d^{O(\log(1/\gamma))}$, and it has been shown that this complexity is unavoidable for any algorithm relying solely on correlational statistical queries (CSQ). In this work, we introduce a novel algorithm that provably circumvents the CSQ hardness barrier. Our approach applies to a broad class of distributions satisfying a natural, previously studied, factorizability assumption. Factorizable distributions lie between distribution-specific and distribution-free settings, and significantly extend previously known tractable cases. Under these distributions, we show that CSQ-based methods still require quasipolynomial time even for weakly learning, whereas our algorithm achieves $poly(d,1/\gamma)$ time by leveraging more general statistical queries (SQ), establishing a strong separation between CSQ and SQ for this simple realizable PAC learning problem. Our result is grounded in a rigorous analysis utilizing a novel duality framework that characterizes the moment tensor structure induced by the marginal distributions. Building on these structural insights, we propose new, efficient learning algorithms. These algorithms combine a refined variant of Jennrich's Algorithm with PCA over random projections of the moment tensor, along with a gradient-descent-based non-convex optimization framework.
Authors: Lequan Lin, Dai Shi, Andi Han, Feng Chen, Qiuzheng Chen, Jiawen Li, Zhaoyang Li, Jiyuan Li, Zhenbang Sun, Junbin Gao
Abstract: Supervised learning relies on high-quality labeled data, but obtaining such data through human annotation is both expensive and time-consuming. Recent work explores using large language models (LLMs) for annotation, but LLM-generated labels still fall short of human-level quality. To address this problem, we propose the Annotation with Critical Thinking (ACT) data pipeline, where LLMs serve not only as annotators but also as judges to critically identify potential errors. Human effort is then directed towards reviewing only the most "suspicious" cases, significantly improving the human annotation efficiency. Our major contributions are as follows: (1) ACT is applicable to a wide range of domains, including natural language processing (NLP), computer vision (CV), and multimodal understanding, by leveraging multimodal-LLMs (MLLMs). (2) Through empirical studies, we derive 7 insights on how to enhance annotation quality while efficiently reducing the human cost, and then translate these findings into user-friendly guidelines. (3) We theoretically analyze how to modify the loss function so that models trained on ACT data achieve similar performance to those trained on fully human-annotated data. Our experiments show that the performance gap can be reduced to less than 2% on most benchmark datasets while saving up to 90% of human costs.
Authors: Fr\'ed\'eric Berdoz, Peer Rheinboldt, Roger Wattenhofer
Abstract: Speculative decoding accelerates language model inference by separating generation into fast drafting and parallel verification. Its main limitation is drafter-verifier misalignment, which limits token acceptance and reduces overall effectiveness. While small drafting heads trained from scratch compensate with speed, they struggle when verification dominates latency or when inputs are out of distribution. In contrast, pretrained drafters, though slower, achieve higher acceptance rates thanks to stronger standalone generation capabilities, making them competitive when drafting latency is negligible relative to verification or communication overhead. In this work, we aim to improve the acceptance rates of pretrained drafters by introducing a lightweight dynamic alignment mechanism: a steering vector computed from the verifier's hidden states and injected into the pretrained drafter. Compared to existing offline alignment methods such as distillation, our approach boosts the number of accepted tokens by up to 35\% under standard sampling and 22\% under greedy sampling, all while incurring negligible computational overhead. Importantly, our approach can be retrofitted to existing architectures and pretrained models, enabling rapid adoption.
Authors: Dianzhi Yu, Conghao Xiong, Yankai Chen, Wenqian Cui, Xinni Zhang, Yifei Zhang, Hao Chen, Joseph J. Y. Sung, Irwin King
Abstract: Survival prediction of cancers is crucial for clinical practice, as it informs mortality risks and influences treatment plans. However, a static model trained on a single dataset fails to adapt to the dynamically evolving clinical environment and continuous data streams, limiting its practical utility. While continual learning (CL) offers a solution to learn dynamically from new datasets, existing CL methods primarily focus on unimodal inputs and suffer from severe catastrophic forgetting in survival prediction. In real-world scenarios, multimodal inputs often provide comprehensive and complementary information, such as whole slide images and genomics; and neglecting inter-modal correlations negatively impacts the performance. To address the two challenges of catastrophic forgetting and complex inter-modal interactions between gigapixel whole slide images and genomics, we propose ConSurv, the first multimodal continual learning (MMCL) method for survival analysis. ConSurv incorporates two key components: Multi-staged Mixture of Experts (MS-MoE) and Feature Constrained Replay (FCR). MS-MoE captures both task-shared and task-specific knowledge at different learning stages of the network, including two modality encoders and the modality fusion component, learning inter-modal relationships. FCR further enhances learned knowledge and mitigates forgetting by restricting feature deviation of previous data at different levels, including encoder-level features of two modalities and the fusion-level representations. Additionally, we introduce a new benchmark integrating four datasets, Multimodal Survival Analysis Incremental Learning (MSAIL), for comprehensive evaluation in the CL setting. Extensive experiments demonstrate that ConSurv outperforms competing methods across multiple metrics.
Authors: James Jin Kang, Dang Bui, Thanh Pham, Huo-Chong Ling
Abstract: The growing use of large language models in sensitive domains has exposed a critical weakness: the inability to ensure that private information can be permanently forgotten. Yet these systems still lack reliable mechanisms to guarantee that sensitive information can be permanently removed once it has been used. Retraining from the beginning is prohibitively costly, and existing unlearning methods remain fragmented, difficult to verify, and often vulnerable to recovery. This paper surveys recent research on machine unlearning for LLMs and considers how far current approaches can address these challenges. We review methods for evaluating whether forgetting has occurred, the resilience of unlearned models against adversarial attacks, and mechanisms that can support user trust when model complexity or proprietary limits restrict transparency. Technical solutions such as differential privacy, homomorphic encryption, federated learning, and ephemeral memory are examined alongside institutional safeguards including auditing practices and regulatory frameworks. The review finds steady progress, but robust and verifiable unlearning is still unresolved. Efficient techniques that avoid costly retraining, stronger defenses against adversarial recovery, and governance structures that reinforce accountability are needed if LLMs are to be deployed safely in sensitive applications. By integrating technical and organizational perspectives, this study outlines a pathway toward AI systems that can be required to forget, while maintaining both privacy and public trust.
Authors: Manh Nguyen, Dung Nguyen, Dai Do, Svetha Venkatesh, Hung Le
Abstract: Reinforcement learning (RL) finetuning is crucial to aligning large language models (LLMs), but the process is notoriously unstable and exhibits high variance across model checkpoints. In practice, selecting the best checkpoint is challenging: evaluating checkpoints on the validation set during training is computationally expensive and requires a good validation set, while relying on the final checkpoint provides no guarantee of good performance. We introduce an uncertainty-guided approach for checkpoint selection (UGCS) that avoids these pitfalls. Our method identifies hard question-answer pairs using per-sample uncertainty and ranks checkpoints by how well they handle these challenging cases. By averaging the rewards of the top-uncertain samples over a short training window, our method produces a stable and discriminative signal without additional forward passes or significant computation overhead. Experiments across three datasets and three LLMs demonstrate that it consistently identifies checkpoints with stronger generalization, outperforming traditional strategies such as relying on training or validation performance. These results highlight that models solving their hardest tasks with low uncertainty are the most reliable overall.
Authors: Hyung-Jun Moon, Sung-Bae Cho
Abstract: Continual learning methods used to force neural networks to process sequential tasks in isolation, preventing them from leveraging useful inter-task relationships and causing them to repeatedly relearn similar features or overly differentiate them. To address this problem, we propose a fully differentiable, exemplar-free expandable method composed of two complementary memories: One learns common features that can be used across all tasks, and the other combines the shared features to learn discriminative characteristics unique to each sample. Both memories are differentiable so that the network can autonomously learn latent representations for each sample. For each task, the memory adjustment module adaptively prunes critical slots and minimally expands capacity to accommodate new concepts, and orthogonal regularization enforces geometric separation between preserved and newly learned memory components to prevent interference. Experiments on CIFAR-10, CIFAR-100, and Tiny-ImageNet show that the proposed method outperforms 14 state-of-the-art methods for class-incremental learning, achieving final accuracies of 55.13\%, 37.24\%, and 30.11\%, respectively. Additional analysis confirms that, through effective integration and utilization of knowledge, the proposed method can increase average performance across sequential tasks, and it produces feature extraction results closest to the upper bound, thus establishing a new milestone in continual learning.
Authors: Shengfei Wei, Suyuan Liu, Jun Wang, Ke Liang, Miaomiao Li, Lei Luo
Abstract: Fair clustering is crucial for mitigating bias in unsupervised learning, yet existing algorithms often suffer from quadratic or super-quadratic computational complexity, rendering them impractical for large-scale datasets. To bridge this gap, we introduce the Anchor-based Fair Clustering Framework (AFCF), a novel, general, and plug-and-play framework that empowers arbitrary fair clustering algorithms with linear-time scalability. Our approach first selects a small but representative set of anchors using a novel fair sampling strategy. Then, any off-the-shelf fair clustering algorithm can be applied to this small anchor set. The core of our framework lies in a novel anchor graph construction module, where we formulate an optimization problem to propagate labels while preserving fairness. This is achieved through a carefully designed group-label joint constraint, which we prove theoretically ensures that the fairness of the final clustering on the entire dataset matches that of the anchor clustering. We solve this optimization efficiently using an ADMM-based algorithm. Extensive experiments on multiple large-scale benchmarks demonstrate that AFCF drastically accelerates state-of-the-art methods, which reduces computational time by orders of magnitude while maintaining strong clustering performance and fairness guarantees.
Authors: Xiaoda Wang, Kaiqiao Han, Yuhao Xu, Xiao Luo, Yizhou Sun, Wei Wang, Carl Yang
Abstract: Cardiovascular disease (CVD) is a leading cause of mortality worldwide. Electrocardiograms (ECGs) are the most widely used non-invasive tool for cardiac assessment, yet large, well-annotated ECG corpora are scarce due to cost, privacy, and workflow constraints. Generating ECGs can be beneficial for the mechanistic understanding of cardiac electrical activity, enable the construction of large, heterogeneous, and unbiased datasets, and facilitate privacy-preserving data sharing. Generating realistic ECG signals from clinical context is important yet underexplored. Recent work has leveraged diffusion models for text-to-ECG generation, but two challenges remain: (i) existing methods often overlook the physiological simulator knowledge of cardiac activity; and (ii) they ignore broader, experience-based clinical knowledge grounded in real-world practice. To address these gaps, we propose SE-Diff, a novel physiological simulator and experience enhanced diffusion model for comprehensive ECG generation. SE-Diff integrates a lightweight ordinary differential equation (ODE)-based ECG simulator into the diffusion process via a beat decoder and simulator-consistent constraints, injecting mechanistic priors that promote physiologically plausible waveforms. In parallel, we design an LLM-powered experience retrieval-augmented strategy to inject clinical knowledge, providing more guidance for ECG generation. Extensive experiments on real-world ECG datasets demonstrate that SE-Diff improves both signal fidelity and text-ECG semantic alignment over baselines, proving its superiority for text-to-ECG generation. We further show that the simulator-based and experience-based knowledge also benefit downstream ECG classification.
Authors: Weilin Wan, Fan Yi, Weizhong Zhang, Quan Zhou, Cheng Jin
Abstract: Modern deep neural networks rely heavily on massive model weights and training samples, incurring substantial computational costs. Weight pruning and coreset selection are two emerging paradigms proposed to improve computational efficiency. In this paper, we first explore the interplay between redundant weights and training samples through a transparent analysis: redundant samples, particularly noisy ones, cause model weights to become unnecessarily overtuned to fit them, complicating the identification of irrelevant weights during pruning; conversely, irrelevant weights tend to overfit noisy data, undermining coreset selection effectiveness. To further investigate and harness this interplay in deep learning, we develop a Simultaneous Weight and Sample Tailoring mechanism (SWaST) that alternately performs weight pruning and coreset selection to establish a synergistic effect in training. During this investigation, we observe that when simultaneously removing a large number of weights and samples, a phenomenon we term critical double-loss can occur, where important weights and their supportive samples are mistakenly eliminated at the same time, leading to model instability and nearly irreversible degradation that cannot be recovered in subsequent training. Unlike classic machine learning models, this issue can arise in deep learning due to the lack of theoretical guarantees on the correctness of weight pruning and coreset selection, which explains why these paradigms are often developed independently. We mitigate this by integrating a state preservation mechanism into SWaST, enabling stable joint optimization. Extensive experiments reveal a strong synergy between pruning and coreset selection across varying prune rates and coreset sizes, delivering accuracy boosts of up to 17.83% alongside 10% to 90% FLOPs reductions.
Authors: Hossein Rouhvarzi, Anastasis Kratsios
Abstract: Incremental flow-based denoising models have reshaped generative modelling, but their empirical advantage still lacks a rigorous approximation-theoretic foundation. We show that incremental generation is necessary and sufficient for universal flow-based generation on the largest natural class of self-maps of $[0,1]^d$ compatible with denoising pipelines, namely the orientation-preserving homeomorphisms of $[0,1]^d$. All our guarantees are uniform on the underlying maps and hence imply approximation both samplewise and in distribution. Using a new topological-dynamical argument, we first prove an impossibility theorem: the class of all single-step autonomous flows, independently of the architecture, width, depth, or Lipschitz activation of the underlying neural network, is meagre and therefore not universal in the space of orientation-preserving homeomorphisms of $[0,1]^d$. By exploiting algebraic properties of autonomous flows, we conversely show that every orientation-preserving Lipschitz homeomorphism on $[0,1]^d$ can be approximated at rate $\mathcal{O}(n^{-1/d})$ by a composition of at most $K_d$ such flows, where $K_d$ depends only on the dimension. Under additional smoothness assumptions, the approximation rate can be made dimension-free, and $K_d$ can be chosen uniformly over the class being approximated. Finally, by linearly lifting the domain into one higher dimension, we obtain structured universal approximation results for continuous functions and for probability measures on $[0,1]^d$, the latter realized as pushforwards of empirical measures with vanishing $1$-Wasserstein error.
Authors: Brian B. Moser, Shalini Strode, Federico Raue, Stanislav Frolov, Krzysztof Adamkiewicz, Arundhati Shanbhag, Joachim Folk, Tobias C. Nauen, Andreas Dengel
Abstract: Dataset distillation (DD) promises compact yet faithful synthetic data, but existing approaches often inherit the inductive bias of a single teacher model. As dataset size increases, this bias drives generation toward overly smooth, homogeneous samples, reducing intra-class diversity and limiting generalization. We present PRISM (PRIors from diverse Source Models), a framework that disentangles architectural priors during synthesis. PRISM decouples the logit-matching and regularization objectives, supervising them with different teacher architectures: a primary model for logits and a stochastic subset for batch-normalization (BN) alignment. On ImageNet-1K, PRISM consistently and reproducibly outperforms single-teacher methods (e.g., SRe2L) and recent multi-teacher variants (e.g., G-VBSM) at low- and mid-IPC regimes. The generated data also show significantly richer intra-class diversity, as reflected by a notable drop in cosine similarity between features. We further analyze teacher selection strategies (pre- vs. intra-distillation) and introduce a scalable cross-class batch formation scheme for fast parallel synthesis. Code will be released after the review period.
Authors: Jiazhen Chen, Xiuqin Liang, Sichao Fu, Zheng Ma, Weihua Ou
Abstract: Unsupervised graph anomaly detection (GAD) has received increasing attention in recent years, which aims to identify data anomalous patterns utilizing only unlabeled node information from graph-structured data. However, prevailing unsupervised GAD methods typically presuppose complete node attributes and structure information, a condition hardly satisfied in real-world scenarios owing to privacy, collection errors or dynamic node arrivals. Existing standard imputation schemes risk "repairing" rare anomalous nodes so that they appear normal, thereby introducing imputation bias into the detection process. In addition, when both node attributes and edges are missing simultaneously, estimation errors in one view can contaminate the other, causing cross-view interference that further undermines the detection performance. To overcome these challenges, we propose M$^2$V-UGAD, a multiple missing values-resistant unsupervised GAD framework on incomplete graphs. Specifically, a dual-pathway encoder is first proposed to independently reconstruct missing node attributes and graph structure, thereby preventing errors in one view from propagating to the other. The two pathways are then fused and regularized in a joint latent space so that normals occupy a compact inner manifold while anomalies reside on an outer shell. Lastly, to mitigate imputation bias, we sample latent codes just outside the normal region and decode them into realistic node features and subgraphs, providing hard negative examples that sharpen the decision boundary. Experiments on seven public benchmarks demonstrate that M$^2$V-UGAD consistently outperforms existing unsupervised GAD methods across varying missing rates.
Authors: Ethan Hirschowitz, Fabio Ramos
Abstract: Improving competent robot policies with on-policy RL is often hampered by noisy, low-signal gradients. We revisit Evolution Strategies (ES) as a policy-gradient proxy and localize exploration with bounded, antithetic triangular perturbations, suitable for policy refinement. We propose Triangular-Distribution ES (TD-ES) which pairs bounded triangular noise with a centered-rank finite-difference estimator to deliver stable, parallelizable, gradient-free updates. In a two-stage pipeline -- PPO pretraining followed by TD-ES refinement -- this preserves early sample efficiency while enabling robust late-stage gains. Across a suite of robotic manipulation tasks, TD-ES raises success rates by 26.5% relative to PPO and greatly reduces variance, offering a simple, compute-light path to reliable refinement.
Authors: Hu Zhang, Zhien Dai, Zhaohui Tang, Yongfang Xie
Abstract: Time series forecasting is essential across diverse domains. While MLP-based methods have gained attention for achieving Transformer-comparable performance with fewer parameters and better robustness, they face critical limitations including loss of weak seasonal signals, capacity constraints in weight-sharing MLPs, and insufficient channel fusion in channel-independent strategies. To address these challenges, we propose MDMLP-EIA (Multi-domain Dynamic MLPs with Energy Invariant Attention) with three key innovations. First, we develop an adaptive fused dual-domain seasonal MLP that categorizes seasonal signals into strong and weak components. It employs an adaptive zero-initialized channel fusion strategy to minimize noise interference while effectively integrating predictions. Second, we introduce an energy invariant attention mechanism that adaptively focuses on different feature channels within trend and seasonal predictions across time steps. This mechanism maintains constant total signal energy to align with the decomposition-prediction-reconstruction framework and enhance robustness against disturbances. Third, we propose a dynamic capacity adjustment mechanism for channel-independent MLPs. This mechanism scales neuron count with the square root of channel count, ensuring sufficient capacity as channels increase. Extensive experiments across nine benchmark datasets demonstrate that MDMLP-EIA achieves state-of-the-art performance in both prediction accuracy and computational efficiency.
Authors: Sha Zhao, Mingyi Peng, Haiteng Jiang, Tao Li, Shijian Li, Gang Pan
Abstract: Scalable and generalizable analysis of brain activity is essential for advancing both clinical diagnostics and cognitive research. Electroencephalography (EEG), a non-invasive modality with high temporal resolution, has been widely used for brain states analysis. However, most existing EEG models are usually tailored for individual specific tasks, limiting their utility in realistic scenarios where EEG analysis often involves multi-task and continuous reasoning. In this work, we introduce EEGAgent, a general-purpose framework that leverages large language models (LLMs) to schedule and plan multiple tools to automatically complete EEG-related tasks. EEGAgent is capable of performing the key functions: EEG basic information perception, spatiotemporal EEG exploration, EEG event detection, interaction with users, and EEG report generation. To realize these capabilities, we design a toolbox composed of different tools for EEG preprocessing, feature extraction, event detection, etc. These capabilities were evaluated on public datasets, and our EEGAgent can support flexible and interpretable EEG analysis, highlighting its potential for real-world clinical applications.
Authors: Pengqian Lu, Jie Lu, Anjin Liu, En Yu, Guangquan Zhang
Abstract: Existing drift detection methods focus on designing sensitive test statistics. They treat the detection threshold as a fixed hyperparameter, set once to balance false alarms and late detections, and applied uniformly across all datasets and over time. However, maintaining model performance is the key objective from the perspective of machine learning, and we observe that model performance is highly sensitive to this threshold. This observation inspires us to investigate whether a dynamic threshold could be provably better. In this paper, we prove that a threshold that adapts over time can outperform any single fixed threshold. The main idea of the proof is that a dynamic strategy, constructed by combining the best threshold from each individual data segment, is guaranteed to outperform any single threshold that apply to all segments. Based on the theorem, we propose a Dynamic Threshold Determination algorithm. It enhances existing drift detection frameworks with a novel comparison phase to inform how the threshold should be adjusted. Extensive experiments on a wide range of synthetic and real-world datasets, including both image and tabular data, validate that our approach substantially enhances the performance of state-of-the-art drift detectors.
Authors: Ziqing Yin, Xuanjing Chen, Xi Zhang
Abstract: The rapid proliferation of AI-generated content (AIGC) has reshaped the dynamics of digital marketing and online consumer behavior. However, predicting the diffusion trajectory and market impact of such content remains challenging due to data heterogeneity, non linear propagation mechanisms, and evolving consumer interactions. This study proposes an AI driven Decision Support System (DSS) that integrates multi source data including social media streams, marketing expenditure records, consumer engagement logs, and sentiment dynamics using a hybrid Graph Neural Network (GNN) and Temporal Transformer framework. The model jointly learns the content diffusion structure and temporal influence evolution through a dual channel architecture, while causal inference modules disentangle the effects of marketing stimuli on return on investment (ROI) and market visibility. Experiments on large scale real-world datasets collected from multiple online platforms such as Twitter, TikTok, and YouTube advertising show that our system outperforms existing baselines in all six metrics. The proposed DSS enhances marketing decisions by providing interpretable real-time insights into AIGC driven content dissemination and market growth patterns.
Authors: Dimitrios Sinodinos, Jack Yi Wei, Narges Armanfard
Abstract: Tabular data is the most abundant data type in the world, powering systems in finance, healthcare, e-commerce, and beyond. As tabular datasets grow and span multiple related targets, there is an increasing need to exploit shared task information for improved multitask generalization. Multitask learning (MTL) has emerged as a powerful way to improve generalization and efficiency, yet most existing work focuses narrowly on large-scale recommendation systems, leaving its potential in broader tabular domains largely underexplored. Also, existing MTL approaches for tabular data predominantly rely on multi-layer perceptron-based backbones, which struggle to capture complex feature interactions and often fail to scale when data is abundant, a limitation that transformer architectures have overcome in other domains. Motivated by this, we introduce MultiTab-Net, the first multitask transformer architecture specifically designed for large tabular data. MultiTab-Net employs a novel multitask masked-attention mechanism that dynamically models feature-feature dependencies while mitigating task competition. Through extensive experiments, we show that MultiTab-Net consistently achieves higher multitask gain than existing MTL architectures and single-task transformers across diverse domains including large-scale recommendation data, census-like socioeconomic data, and physics datasets, spanning a wide range of task counts, task types, and feature modalities. In addition, we contribute MultiTab-Bench, a generalized multitask synthetic dataset generator that enables systematic evaluation of multitask dynamics by tuning task count, task correlations, and relative task complexity. Our code is publicly available at https://github.com/Armanfard-Lab/MultiTab.
Authors: Saumya Shah, Zi-Yu Khoo, Abel Yang, St\'ephane Bressan
Abstract: This work explores using the physics-inspired AI Feynman symbolic regression algorithm to automatically rediscover a fundamental equation in astronomy -- the Equation of the Centre. Through the introduction of observational and inductive biases corresponding to the physical nature of the system through data preprocessing and search space restriction, AI Feynman was successful in recovering the first-order analytical form of this equation from lunar ephemerides data. However, this manual approach highlights a key limitation in its reliance on expert-driven coordinate system selection. We therefore propose an automated preprocessing extension to find the canonical coordinate system. Results demonstrate that targeted domain knowledge embedding enables symbolic regression to rediscover physical laws, but also highlight further challenges in constraining symbolic regression to derive physics equations when leveraging domain knowledge through tailored biases.
Authors: Fushuo Huo
Abstract: The rapid evolution of machine learning has propelled neural networks to unprecedented success across diverse domains. In particular, multimodal learning has emerged as a transformative paradigm, leveraging complementary information from heterogeneous data streams (e.g., text, vision, audio) to advance contextual reasoning and intelligent decision-making. Despite these advancements, current neural network-based models often fall short in open-world environments characterized by inherent unpredictability, where unpredictable environmental composition dynamics, incomplete modality inputs, and spurious distributions relations critically undermine system reliability. While humans naturally adapt to such dynamic, ambiguous scenarios, artificial intelligence systems exhibit stark limitations in robustness, particularly when processing multimodal signals under real-world complexity. This study investigates the fundamental challenge of multimodal learning robustness in open-world settings, aiming to bridge the gap between controlled experimental performance and practical deployment requirements.
Authors: Alireza F. Pour, Shai Ben-David
Abstract: We address the general task of learning with a set of candidate models that is too large to have a uniform convergence of empirical estimates to true losses. While the common approach to such challenges is SRM (or regularization) based learning algorithms, we propose a novel learning paradigm that relies on stronger incorporation of empirical data and requires less algorithmic decisions to be based on prior assumptions. We analyze the generalization capabilities of our approach and demonstrate its merits in several common learning assumptions, including similarity of close points, clustering of the domain into highly label-homogeneous regions, Lipschitzness assumptions of the labeling rule, and contrastive learning assumptions. Our approach allows utilizing such assumptions without the need to know their true parameters a priori.
Authors: Hui Dou, Lei Jin, Yuxuan Zhou, Jiang He, Yiwen Zhang
Abstract: The performance of modern DBMSs such as MySQL and PostgreSQL heavily depends on the configuration of performance-critical knobs. Manual tuning these knobs is laborious and inefficient due to the complex and high-dimensional nature of the configuration space. Among the automated tuning methods, reinforcement learning (RL)-based methods have recently sought to improve the DBMS knobs tuning process from several different perspectives. However, they still encounter challenges with slow convergence speed during offline training. In this paper, we mainly focus on how to leverage the valuable tuning hints contained in various textual documents such as DBMS manuals and web forums to improve the offline training of RL-based methods. To this end, we propose an efficient DBMS knobs tuning framework named DemoTuner via a novel LLM-assisted demonstration reinforcement learning method. Specifically, to comprehensively and accurately mine tuning hints from documents, we design a structured chain of thought prompt to employ LLMs to conduct a condition-aware tuning hints extraction task. To effectively integrate the mined tuning hints into RL agent training, we propose a hint-aware demonstration reinforcement learning algorithm HA-DDPGfD in DemoTuner. As far as we know, DemoTuner is the first work to introduce the demonstration reinforcement learning algorithm for DBMS knobs tuning. Experimental evaluations conducted on MySQL and PostgreSQL across various workloads demonstrate the significant advantages of DemoTuner in both performance improvement and online tuning cost reduction over three representative baselines including DB-BERT, GPTuner and CDBTune. Additionally, DemoTuner also exhibits superior adaptability to application scenarios with unknown workloads.
Authors: Pilsung Kang
Abstract: Classical approaches often treat interaction as engineered product terms or as emergent patterns in flexible models, offering little control over how synergy or antagonism arises. We take a quantum-inspired view: following the Born rule (probability as squared amplitude), \emph{coherent} aggregation sums complex amplitudes before squaring, creating an interference cross-term, whereas an \emph{incoherent} proxy sums squared magnitudes and removes it. In a minimal linear-amplitude model, this cross-term equals the standard potential-outcome interaction contrast \(\Delta_{\mathrm{INT}}\) in a \(2\times 2\) factorial design, giving relative phase a direct, mechanism-level control over synergy versus antagonism. We instantiate this idea in a lightweight \emph{Interference Kernel Classifier} (IKC) and introduce two diagnostics: \emph{Coherent Gain} (log-likelihood gain of coherent over the incoherent proxy) and \emph{Interference Information} (the induced Kullback-Leibler gap). A controlled phase sweep recovers the identity. On a high-interaction synthetic task (XOR), IKC outperforms strong baselines under paired, budget-matched comparisons; on real tabular data (\emph{Adult} and \emph{Bank Marketing}) it is competitive overall but typically trails the most capacity-rich baseline in paired differences. Holding learned parameters fixed, toggling aggregation from incoherent to coherent consistently improves negative log-likelihood, Brier score, and expected calibration error, with positive Coherent Gain on both datasets.
Authors: Chaofan Zhu, Xiaobing Rui, Zhixiao Wang
Abstract: Imbalanced node classification is a critical challenge in graph learning, where most existing methods typically utilize Graph Neural Networks (GNNs) to learn node representations. These methods can be broadly categorized into the data-level and the algorithm-level. The former aims to synthesize minority-class nodes to mitigate quantity imbalance, while the latter tries to optimize the learning process to highlight minority classes. However, neither category addresses the inherently imbalanced graph structure, which is a fundamental factor that incurs majority-class dominance and minority-class assimilation in GNNs. Our theoretical analysis further supports this critical insight. Therefore, we propose GraphSB (Graph Structural Balance), a novel framework that incorporates Structural Balance as a key strategy to address the underlying imbalanced graph structure before node synthesis. Structural Balance performs a two-stage structure optimization: Structure Enhancement that adaptively builds similarity-based edges to strengthen connectivity of minority-class nodes, and Relation Diffusion that captures higher-order dependencies while amplifying signals from minority classes. Thus, GraphSB balances structural distribution before node synthesis, enabling more effective learning in GNNs. Extensive experiments demonstrate that GraphSB significantly outperforms the state-of-the-art methods. More importantly, the proposed Structural Balance can be seamlessly integrated into state-of-the-art methods as a simple plug-and-play module, increasing their accuracy by an average of 3.67\%.
Authors: Noam Koren, Ralf J. J. Mackenbach, Ruud J. G. van Sloun, Kira Radinsky, Daniel Freedman
Abstract: Neural operators have emerged as a promising paradigm for learning solution operators of partial differential equa- tions (PDEs) directly from data. Existing methods, such as those based on Fourier or graph techniques, make strong as- sumptions about the structure of the kernel integral opera- tor, assumptions which may limit expressivity. We present SVD-NO, a neural operator that explicitly parameterizes the kernel by its singular-value decomposition (SVD) and then carries out the integral directly in the low-rank basis. Two lightweight networks learn the left and right singular func- tions, a diagonal parameter matrix learns the singular values, and a Gram-matrix regularizer enforces orthonormality. As SVD-NO approximates the full kernel, it obtains a high de- gree of expressivity. Furthermore, due to its low-rank struc- ture the computational complexity of applying the operator remains reasonable, leading to a practical system. In exten- sive evaluations on five diverse benchmark equations, SVD- NO achieves a new state of the art. In particular, SVD-NO provides greater performance gains on PDEs whose solutions are highly spatially variable. The code of this work is publicly available at https://github.com/2noamk/SVDNO.git.
Authors: Ruichu Cai, Xiaokai Huang, Wei Chen, Zijian Li, Zhifeng Hao
Abstract: Inferring causal relationships from observed data is an important task, yet it becomes challenging when the data is subject to various external interferences. Most of these interferences are the additional effects of external factors on observed variables. Since these external factors are often unknown, we introduce latent variables to represent these unobserved factors that affect the observed data. Specifically, to capture the causal strength and adjacency information, we propose a new temporal latent variable structural causal model, incorporating causal strength and adjacency coefficients that represent the causal relationships between variables. Considering that expert knowledge can provide information about unknown interferences in certain scenarios, we develop a method that facilitates the incorporation of prior knowledge into parameter learning based on Variational Inference, to guide the model estimation. Experimental results demonstrate the stability and accuracy of our proposed method.
Authors: Yun Wang, Lingyun Yang, Senhao Yu, Yixiao Wang, Ruixing Li, Zhixiang Wei, James Yen, Zhengwei Qi
Abstract: Mixture-of-Experts (MoE) architectures scale language models by activating only a subset of specialized expert networks for each input token, thereby reducing the number of floating-point operations. However, the growing size of modern MoE models causes their full parameter sets to exceed GPU memory capacity; for example, Mixtral-8x7B has 45 billion parameters and requires 87 GB of memory even though only 14 billion parameters are used per token. Existing systems alleviate this limitation by offloading inactive experts to CPU memory, but transferring experts across the PCIe interconnect incurs significant latency (about 10 ms). Prefetching heuristics aim to hide this latency by predicting which experts are needed, but prefetch failures introduce significant stalls and amplify inference latency. In the event of a prefetch failure, prior work offers two primary solutions: either fetch the expert on demand, which incurs a long stall due to the PCIe bottleneck, or drop the expert from the computation, which significantly degrades model accuracy. The critical challenge, therefore, is to maintain both high inference speed and model accuracy when prefetching fails.
Authors: Zijing Liu, Bin Feng, He Cao, Yu Li
Abstract: Protein structure tokenization converts 3D structures into discrete or vectorized representations, enabling the integration of structural and sequence data. Despite many recent works on structure tokenization, the properties of the underlying discrete representations are not well understood. In this work, we first demonstrate that the successful utilization of structural tokens in a language model for structure prediction depends on using rich, pre-trained sequence embeddings to bridge the semantic gap between the sequence and structural "language". The analysis of the structural vocabulary itself then reveals significant semantic redundancy, where multiple distinct tokens correspond to nearly identical local geometries, acting as "structural synonyms". This redundancy, rather than being a flaw, can be exploited with a simple "synonym swap" strategy to generate diverse conformational ensembles by perturbing a predicted structure with its structural synonyms. This computationally lightweight method accurately recapitulates protein flexibility, performing competitively with state-of-the-art models. Our study provides fundamental insights into the nature of discrete protein structure representations and introduces a powerful, near-instantaneous method for modeling protein dynamics. Source code is available in https://github.com/IDEA-XL/TokenMD.
Authors: Muhammad Kashif, Shaf Khalid, Alberto Marchisio, Nouhaila Innan, Muhammad Shafique
Abstract: Hybrid Quantum Neural Networks (HQNNs), which combine parameterized quantum circuits with classical neural layers, are emerging as promising models in the noisy intermediate-scale quantum (NISQ) era. While quantum circuits are not naturally measured in floating point operations (FLOPs), most HQNNs (in NISQ era) are still trained on classical simulators where FLOPs directly dictate runtime and scalability. Hence, FLOPs represent a practical and viable metric to measure the computational complexity of HQNNs. In this work, we introduce FAQNAS, a FLOPs-aware neural architecture search (NAS) framework that formulates HQNN design as a multi-objective optimization problem balancing accuracy and FLOPs. Unlike traditional approaches, FAQNAS explicitly incorporates FLOPs into the optimization objective, enabling the discovery of architectures that achieve strong performance while minimizing computational cost. Experiments on five benchmark datasets (MNIST, Digits, Wine, Breast Cancer, and Iris) show that quantum FLOPs dominate accuracy improvements, while classical FLOPs remain largely fixed. Pareto-optimal solutions reveal that competitive accuracy can often be achieved with significantly reduced computational cost compared to FLOPs-agnostic baselines. Our results establish FLOPs-awareness as a practical criterion for HQNN design in the NISQ era and as a scalable principle for future HQNN systems.
Authors: Shuxin Zhuang, Linjian Meng, Shuxin Li, Minming Li, Youzhi Zhang
Abstract: Urban Network Security Games (UNSGs), which model the strategic allocation of limited security resources on city road networks, are critical for urban safety. However, finding a Nash Equilibrium (NE) in large-scale UNSGs is challenging due to their massive and combinatorial action spaces. One common approach to addressing these games is the Policy-Space Response Oracle (PSRO) framework, which requires computing best responses (BR) at each iteration. However, precisely computing exact BRs is impractical in large-scale games, and employing reinforcement learning to approximate BRs inevitably introduces errors, which limits the overall effectiveness of the PSRO methods. Recent advancements in leveraging non-convex stochastic optimization to approximate an NE offer a promising alternative to the burdensome BR computation. However, utilizing existing stochastic optimization techniques with an unbiased loss function for UNSGs remains challenging because the action spaces are too vast to be effectively represented by neural networks. To address these issues, we introduce Tree-based Stochastic Optimization (TSO), a framework that bridges the gap between the stochastic optimization paradigm for NE-finding and the demands of UNSGs. Specifically, we employ the tree-based action representation that maps the whole action space onto a tree structure, addressing the challenge faced by neural networks in representing actions when the action space cannot be enumerated. We then incorporate this representation into the loss function and theoretically demonstrate its equivalence to the unbiased loss function. To further enhance the quality of the converged solution, we introduce a sample-and-prune mechanism that reduces the risk of being trapped in suboptimal local optima. Extensive experimental results indicate the superiority of TSO over other baseline algorithms in addressing the UNSGs.
Authors: Leonardo Pesce, Jiawen Wei, Gianmarco Mengaldo
Abstract: Post-hoc explainability methods are a subset of Machine Learning (ML) that aim to provide a reason for why a model behaves in a certain way. In this paper, we show a new black-box model-agnostic adversarial attack for post-hoc explainable Artificial Intelligence (XAI), particularly in the image domain. The goal of the attack is to modify the original explanations while being undetected by the human eye and maintain the same predicted class. In contrast to previous methods, we do not require any access to the model or its weights, but only to the model's computed predictions and explanations. Additionally, the attack is accomplished in a single step while significantly changing the provided explanations, as demonstrated by empirical evaluation. The low requirements of our method expose a critical vulnerability in current explainability methods, raising concerns about their reliability in safety-critical applications. We systematically generate attacks based on the explanations generated by post-hoc explainability methods (saliency maps, integrated gradients, and DeepLIFT SHAP) for pretrained ResNet-18 and ViT-B16 on ImageNet. The results show that our attacks could lead to dramatically different explanations without changing the predictive probabilities. We validate the effectiveness of our attack, compute the induced change based on the explanation with mean absolute difference, and verify the closeness of the original image and the corrupted one with the Structural Similarity Index Measure (SSIM).
Authors: Abu Sufian, Cosimo Distante, Marco Leo, Hanan Salam
Abstract: Text-to-image (T2I) generative models are largely used in AI-powered real-world applications and value creation. However, their strategic deployment raises critical concerns for responsible AI management, particularly regarding the reproduction and amplification of race- and gender-related stereotypes that can undermine organizational ethics. In this work, we investigate whether such societal biases are systematically encoded within the pretrained latent spaces of state-of-the-art T2I models. We conduct an empirical study across the five most popular open-source models, using ten neutral, profession-related prompts to generate 100 images per profession, resulting in a dataset of 5,000 images evaluated by diverse human assessors representing different races and genders. We demonstrate that all five models encode and amplify pronounced societal skew: caregiving and nursing roles are consistently feminized, while high-status professions such as corporate CEO, politician, doctor, and lawyer are overwhelmingly represented by males and mostly White individuals. We further identify model-specific patterns, such as QWEN-Image's near-exclusive focus on East Asian outputs, Kandinsky's dominance of White individuals, and SDXL's comparatively broader but still biased distributions. These results provide critical insights for AI project managers and practitioners, enabling them to select equitable AI models and customized prompts that generate images in alignment with the principles of responsible AI. We conclude by discussing the risks of these biases and proposing actionable strategies for bias mitigation in building responsible GenAI systems.
Authors: Yiming Tang, Abhijeet Sinha, Dianbo Liu
Abstract: Although recent generative models are remarkably capable of producing instruction-following and realistic outputs, they remain prone to notable physical plausibility failures. Though critical in applications, these physical plausibility errors often escape detection by existing evaluation methods. Furthermore, no framework exists for automatically identifying and interpreting specific physical error patterns in natural language, preventing targeted model improvements. We introduce Matryoshka Transcoders, a novel framework for the automatic discovery and interpretation of physical plausibility features in generative models. Our approach extends the Matryoshka representation learning paradigm to transcoder architectures, enabling hierarchical sparse feature learning at multiple granularity levels. By training on intermediate representations from a physical plausibility classifier and leveraging large multimodal models for interpretation, our method identifies diverse physics-related failure modes without manual feature engineering, achieving superior feature relevance and feature accuracy compared to existing approaches. We utilize the discovered visual patterns to establish a benchmark for evaluating physical plausibility in generative models. Our analysis of eight state-of-the-art generative models provides valuable insights into how these models fail to follow physical constraints, paving the way for further model improvements.
Authors: Jieting Wang, Xiaolei Shang, Feijiang Li, Furong Peng
Abstract: Time series forecasting relies on predicting future values from historical data, yet most state-of-the-art approaches-including transformer and multilayer perceptron-based models-optimize using Mean Squared Error (MSE), which has two fundamental weaknesses: its point-wise error computation fails to capture temporal relationships, and it does not account for inherent noise in the data. To overcome these limitations, we introduce the Residual-Informed Loss (RI-Loss), a novel objective function based on the Hilbert-Schmidt Independence Criterion (HSIC). RI-Loss explicitly models noise structure by enforcing dependence between the residual sequence and a random time series, enabling more robust, noise-aware representations. Theoretically, we derive the first non-asymptotic HSIC bound with explicit double-sample complexity terms, achieving optimal convergence rates through Bernstein-type concentration inequalities and Rademacher complexity analysis. This provides rigorous guarantees for RI-Loss optimization while precisely quantifying kernel space interactions. Empirically, experiments across eight real-world benchmarks and five leading forecasting models demonstrate improvements in predictive performance, validating the effectiveness of our approach. Code will be made publicly available to ensure reproducibility.
Authors: Yuancheng Sun, Yuxuan Ren, Zhaoming Chen, Xu Han, Kang Liu, Qiwei Ye
Abstract: Accurate exploration of protein conformational ensembles is essential for uncovering function but remains hard because molecular-dynamics (MD) simulations suffer from high computational costs and energy-barrier trapping. This paper presents Energy Preference Optimization (EPO), an online refinement algorithm that turns a pretrained protein ensemble generator into an energy-aware sampler without extra MD trajectories. Specifically, EPO leverages stochastic differential equation sampling to explore the conformational landscape and incorporates a novel energy-ranking mechanism based on list-wise preference optimization. Crucially, EPO introduces a practical upper bound to efficiently approximate the intractable probability of long sampling trajectories in continuous-time generative models, making it easily adaptable to existing pretrained generators. On Tetrapeptides, ATLAS, and Fast-Folding benchmarks, EPO successfully generates diverse and physically realistic ensembles, establishing a new state-of-the-art in nine evaluation metrics. These results demonstrate that energy-only preference signals can efficiently steer generative models toward thermodynamically consistent conformational ensembles, providing an alternative to long MD simulations and widening the applicability of learned potentials in structural biology and drug discovery.
Authors: Outongyi Lv, Yewei Yuan, Nana Liu
Abstract: Reinforcement learning (RL) with limited samples is common in real-world applications. However, offline RL performance under this constraint is often suboptimal. We consider an alternative approach to dealing with limited samples by introducing the Quantum Metric Encoder (QME). In this methodology, instead of applying the RL framework directly on the original states and rewards, we embed the states into a more compact and meaningful representation, where the structure of the encoding is inspired by quantum circuits. For classical data, QME is a classically simulable, trainable unitary embedding and thus serves as a quantum-inspired module, on a classical device. For quantum data in the form of quantum states, QME can be implemented directly on quantum hardware, allowing for training without measurement or re-encoding. We evaluated QME on three datasets, each limited to 100 samples. We use Soft-Actor-Critic (SAC) and Implicit-Q-Learning (IQL), two well-known RL algorithms, to demonstrate the effectiveness of our approach. From the experimental results, we find that training offline RL agents on QME-embedded states with decoded rewards yields significantly better performance than training on the original states and rewards. On average across the three datasets, for maximum reward performance, we achieve a 116.2% improvement for SAC and 117.6% for IQL. We further investigate the $\Delta$-hyperbolicity of our framework, a geometric property of the state space known to be important for the RL training efficacy. The QME-embedded states exhibit low $\Delta$-hyperbolicity, suggesting that the improvement after embedding arises from the modified geometry of the state space induced by QME. Thus, the low $\Delta$-hyperbolicity and the corresponding effectiveness of QME could provide valuable information for developing efficient offline RL methods under limited-sample conditions.
Authors: Eklavya Sarkar, Mathew Magimai. -Doss
Abstract: Animal vocalizations contain sequential structures that carry important communicative information, yet most computational bioacoustics studies average the extracted frame-level features across the temporal axis, discarding the order of the sub-units within a vocalization. This paper investigates whether discrete acoustic token sequences, derived through vector quantization and gumbel-softmax vector quantization of extracted self-supervised speech model representations can effectively capture and leverage temporal information. To that end, pairwise distance analysis of token sequences generated from HuBERT embeddings shows that they can discriminate call-types and callers across four bioacoustics datasets. Sequence classification experiments using $k$-Nearest Neighbour with Levenshtein distance show that the vector-quantized token sequences yield reasonable call-type and caller classification performances, and hold promise as alternative feature representations towards leveraging sequential information in animal vocalizations.
Authors: Jieting Wang, Huimei Shi, Feijiang Li, Xiaolei Shang
Abstract: Time series forecasting is an important task that involves analyzing temporal dependencies and underlying patterns (such as trends, cyclicality, and seasonality) in historical data to predict future values or trends. Current deep learning-based forecasting models primarily employ Mean Squared Error (MSE) loss functions for regression modeling. Despite enabling direct value prediction, this method offers no uncertainty estimation and exhibits poor outlier robustness. To address these limitations, we propose OCE-TS, a novel ordinal classification approach for time series forecasting that replaces MSE with Ordinal Cross-Entropy (OCE) loss, preserving prediction order while quantifying uncertainty through probability output. Specifically, OCE-TS begins by discretizing observed values into ordered intervals and deriving their probabilities via a parametric distribution as supervision signals. Using a simple linear model, we then predict probability distributions for each timestep. The OCE loss is computed between the cumulative distributions of predicted and ground-truth probabilities, explicitly preserving ordinal relationships among forecasted values. Through theoretical analysis using influence functions, we establish that cross-entropy (CE) loss exhibits superior stability and outlier robustness compared to MSE loss. Empirically, we compared OCE-TS with five baseline models-Autoformer, DLinear, iTransformer, TimeXer, and TimeBridge-on seven public time series datasets. Using MSE and Mean Absolute Error (MAE) as evaluation metrics, the results demonstrate that OCE-TS consistently outperforms benchmark models. The code will be published.
Authors: Cheng Kevin Qu, Andrew Ly, Pulin Gong
Abstract: Attention mechanisms underpin the computational power of Transformer models, which have achieved remarkable success across diverse domains. Yet understanding and extending the principles underlying self-attention remains a key challenge for advancing artificial intelligence. Drawing inspiration from the multiscale dynamics of biological attention and from dynamical systems theory, we introduce Fractional Neural Attention (FNA), a principled, neuroscience-inspired framework for multiscale information processing. FNA models token interactions through L\'evy diffusion governed by the fractional Laplacian, intrinsically realizing simultaneous short- and long-range dependencies across multiple scales. This mechanism yields greater expressivity and faster information mixing, advancing the foundational capacity of Transformers. Theoretically, we show that FNA's dynamics are governed by the fractional diffusion equation, and that the resulting attention networks exhibit larger spectral gaps and shorter path lengths -- mechanistic signatures of enhanced computational efficiency. Empirically, FNA achieves competitive text-classification performance even with a single layer and a single head; it also improves performance in image processing and neural machine translation. Finally, the diffusion map algorithm from geometric harmonics enables dimensionality reduction of FNA weights while preserving the intrinsic structure of embeddings and hidden states. Together, these results establish FNA as a principled mechanism connecting self-attention, stochastic dynamics, and geometry, providing an interpretable, biologically grounded foundation for powerful, neuroscience-inspired AI.
Authors: Xi Yang, Han Zhang, Zhijian Lin, Yibiao Hu, Hong Han
Abstract: Out-of-context misinformation (OOC) is a low-cost form of misinformation in news reports, which refers to place authentic images into out-of-context or fabricated image-text pairings. This problem has attracted significant attention from researchers in recent years. Current methods focus on assessing image-text consistency or generating explanations. However, these approaches assume that the training and test data are drawn from the same distribution. When encountering novel news domains, models tend to perform poorly due to the lack of prior knowledge. To address this challenge, we propose \textbf{VDT} to enhance the domain adaptation capability for OOC misinformation detection by learning domain-invariant features and test-time training mechanisms. Domain-Invariant Variational Align module is employed to jointly encodes source and target domain data to learn a separable distributional space domain-invariant features. For preserving semantic integrity, we utilize domain consistency constraint module to reconstruct the source and target domain latent distribution. During testing phase, we adopt the test-time training strategy and confidence-variance filtering module to dynamically updating the VAE encoder and classifier, facilitating the model's adaptation to the target domain distribution. Extensive experiments conducted on the benchmark dataset NewsCLIPpings demonstrate that our method outperforms state-of-the-art baselines under most domain adaptation settings.
Authors: Yue Chen, Jianfeng Lu, Shuqing Cao, Wei Wang, Gang Li, Guanghui Wen
Abstract: While semi-asynchronous federated learning (SAFL) combines the efficiency of synchronous training with the flexibility of asynchronous updates, it inherently suffers from participation bias, which is further exacerbated by non-IID data distributions. More importantly, hierarchical architecture shifts participation from individual clients to client groups, thereby further intensifying this issue. Despite notable advancements in SAFL research, most existing works still focus on conventional cloud-end architectures while largely overlooking the critical impact of non-IID data on scheduling across the cloud-edge-client hierarchy. To tackle these challenges, we propose FedCure, an innovative semi-asynchronous Federated learning framework that leverages coalition construction and participation-aware scheduling to mitigate participation bias with non-IID data. Specifically, FedCure operates through three key rules: (1) a preference rule that optimizes coalition formation by maximizing collective benefits and establishing theoretically stable partitions to reduce non-IID-induced performance degradation; (2) a scheduling rule that integrates the virtual queue technique with Bayesian-estimated coalition dynamics, mitigating efficiency loss while ensuring mean rate stability; and (3) a resource allocation rule that enhances computational efficiency by optimizing client CPU frequencies based on estimated coalition dynamics while satisfying delay requirements. Comprehensive experiments on four real-world datasets demonstrate that FedCure improves accuracy by up to 5.1x compared with four state-of-the-art baselines, while significantly enhancing efficiency with the lowest coefficient of variation 0.0223 for per-round latency and maintaining long-term balance across diverse scenarios.
Authors: Daniel Herbst, Lea Karbeska, Divyanshu Kumar, Akanksha Ahuja, Fatemeh Gholamzadeh Nasrabadi, Fabrizio Frasca
Abstract: While promising, graph reasoners based on Large Language Models (LLMs) lack built-in invariance to symmetries in graph representations. Operating on sequential graph serializations, LLMs can produce different outputs under node reindexing, edge reordering, or formatting changes, raising robustness concerns. We systematically analyze these effects, studying how fine-tuning impacts encoding sensitivity as well generalization on unseen tasks. We propose a principled decomposition of graph serializations into node labeling, edge encoding, and syntax, and evaluate LLM robustness to variations of each of these factors on a comprehensive benchmarking suite. We also contribute a novel set of spectral tasks to further assess generalization abilities of fine-tuned reasoners. Results show that larger (non-fine-tuned) models are more robust. Fine-tuning reduces sensitivity to node relabeling but may increase it to variations in structure and format, while it does not consistently improve performance on unseen tasks.
Authors: Oliver Dippel, Alexei Lisitsa, Bei Peng
Abstract: Transformers have demonstrated exceptional in-context learning (ICL) capabilities, enabling applications across natural language processing, computer vision, and sequential decision-making. In reinforcement learning, ICL reframes learning as a supervised problem, facilitating task adaptation without parameter updates. Building on prior work leveraging transformers for sequential decision-making, we propose Heuristic Transformer (HT), an in-context reinforcement learning (ICRL) approach that augments the in-context dataset with a belief distribution over rewards to achieve better decision-making. Using a variational auto-encoder (VAE), a low-dimensional stochastic variable is learned to represent the posterior distribution over rewards, which is incorporated alongside an in-context dataset and query states as prompt to the transformer policy. We assess the performance of HT across the Darkroom, Miniworld, and MuJoCo environments, showing that it consistently surpasses comparable baselines in terms of both effectiveness and generalization. Our method presents a promising direction to bridge the gap between belief-based augmentations and transformer-based decision-making.
Authors: Qian Jin, Yumeng Liu, Yuqi Jiang, Qi Sun, Cheng Zhuo
Abstract: Reliable, generalizable data foundations are critical for enabling large-scale models in computational lithography. However, essential tasks-mask generation, rule violation detection, and layout optimization-are often handled in isolation, hindered by scarce datasets and limited modeling approaches. To address these challenges, we introduce Unitho, a unified multi-task large vision model built upon the Transformer architecture. Trained on a large-scale industrial lithography simulation dataset with hundreds of thousands of cases, Unitho supports end-to-end mask generation, lithography simulation, and rule violation detection. By enabling agile and high-fidelity lithography simulation, Unitho further facilitates the construction of robust data foundations for intelligent EDA. Experimental results validate its effectiveness and generalizability, with performance substantially surpassing academic baselines.
Authors: Adrien Lafage, Olivier Laurent, Firas Gabetni, Gianni Franchi
Abstract: Deep Neural Networks (DNNs) have demonstrated remarkable performance across various domains, including computer vision and natural language processing. However, they often struggle to accurately quantify the uncertainty of their predictions, limiting their broader adoption in critical real-world applications. Uncertainty Quantification (UQ) for Deep Learning seeks to address this challenge by providing methods to improve the reliability of uncertainty estimates. Although numerous techniques have been proposed, a unified tool offering a seamless workflow to evaluate and integrate these methods remains lacking. To bridge this gap, we introduce Torch-Uncertainty, a PyTorch and Lightning-based framework designed to streamline DNN training and evaluation with UQ techniques and metrics. In this paper, we outline the foundational principles of our library and present comprehensive experimental results that benchmark a diverse set of UQ methods across classification, segmentation, and regression tasks. Our library is available at https://github.com/ENSTA-U2IS-AI/Torch-Uncertainty
Authors: Yuping Yan, Yuhan Xie, Yuanshuai Li, Yingchao Yu, Lingjuan Lyu, Yaochu Jin
Abstract: Since Multimodal Large Language Models (MLLMs) are increasingly being integrated into everyday tools and intelligent agents, growing concerns have arisen regarding their possible output of unsafe contents, ranging from toxic language and biased imagery to privacy violations and harmful misinformation. Current safety benchmarks remain highly limited in both modality coverage and performance evaluations, often neglecting the extensive landscape of content safety. In this work, we introduce OutSafe-Bench, the first most comprehensive content safety evaluation test suite designed for the multimodal era. OutSafe-Bench includes a large-scale dataset that spans four modalities, featuring over 18,000 bilingual (Chinese and English) text prompts, 4,500 images, 450 audio clips and 450 videos, all systematically annotated across nine critical content risk categories. In addition to the dataset, we introduce a Multidimensional Cross Risk Score (MCRS), a novel metric designed to model and assess overlapping and correlated content risks across different categories. To ensure fair and robust evaluation, we propose FairScore, an explainable automated multi-reviewer weighted aggregation framework. FairScore selects top-performing models as adaptive juries, thereby mitigating biases from single-model judgments and enhancing overall evaluation reliability. Our evaluation of nine state-of-the-art MLLMs reveals persistent and substantial safety vulnerabilities, underscoring the pressing need for robust safeguards in MLLMs.
Authors: Fuyuan Cao, Jiaxuan Zhang, Xiaoli Li
Abstract: Estimating Individual Treatment Effects (ITE) from observational data is challenging due to confounding bias. Most studies tackle this bias by balancing distributions globally, but ignore individual heterogeneity and fail to capture the local structure that represents the natural clustering among individuals, which ultimately compromises ITE estimation. While instance-level alignment methods consider heterogeneity, they similarly overlook the local structure information. To address these issues, we propose an end-to-end Multi-\textbf{P}rototype alignment method for \textbf{ITE} estimation (\textbf{PITE}). PITE effectively captures local structure within groups and enforces cross-group alignment, thereby achieving robust ITE estimation. Specifically, we first define prototypes as cluster centroids based on similar individuals under the same treatment. To identify local similarity and the distribution consistency, we perform instance-to-prototype matching to assign individuals to the nearest prototype within groups, and design a multi-prototype alignment strategy to encourage the matched prototypes to be close across treatment arms in the latent space. PITE not only reduces distribution shift through fine-grained, prototype-level alignment, but also preserves the local structures of treated and control groups, which provides meaningful constraints for ITE estimation. Extensive evaluations on benchmark datasets demonstrate that PITE outperforms 13 state-of-the-art methods, achieving more accurate and robust ITE estimation.
Authors: Qingao Yi, Jiaang Duan, Hanwen Hu, Qin Hua, Haiyan Zhao, Shiyou Qian, Dingyu Yang, Jian Cao, Jinghua Tang, Yinghao Yu, Chenzhi Liao, Kangjin Wang, Liping Zhang
Abstract: Training large language models (LLMs) poses significant challenges regarding computational resources and memory capacity. Although distributed training techniques help mitigate these issues, they still suffer from considerable communication overhead. Existing approaches primarily rely on static gradient compression to enhance communication efficiency; however, these methods neglect the dynamic nature of evolving gradients during training, leading to performance degradation. Accelerating LLM training via compression without sacrificing performance remains a challenge. In this paper, we propose an entropy-driven dynamic gradient compression framework called EDGC. The core concept is to adjust the compression rate during LLM training based on the evolving trends of gradient entropy, taking into account both compression efficiency and error. EDGC consists of three key components.First, it employs a down-sampling method to efficiently estimate gradient entropy, reducing computation overhead. Second, it establishes a theoretical model linking compression rate with gradient entropy, enabling more informed compression decisions. Lastly, a window-based adjustment mechanism dynamically adapts the compression rate across pipeline stages, improving communication efficiency and maintaining model performance. We implemented EDGC on a 32-NVIDIA-V100 cluster and a 64-NVIDIA-H100 cluster to train GPT2-2.5B and GPT2-12.1B, respectively. The results show that EDGC significantly reduces communication latency and training time by up to 46.45% and 16.13% while preserving LLM accuracy.
Authors: Zicheng Hu, Yuchen Wang, Cheng Chen
Abstract: Decentralized cooperative multi-agent multi-armed bandits (DeCMA2B) considers how multiple agents collaborate in a decentralized multi-armed bandit setting. Though this problem has been extensively studied in previous work, most existing methods remain susceptible to various adversarial attacks. In this paper, we first study DeCMA2B with adversarial corruption, where an adversary can corrupt reward observations of all agents with a limited corruption budget. We propose a robust algorithm, called DeMABAR, which ensures that each agent's individual regret suffers only an additive term proportional to the corruption budget. Then we consider a more realistic scenario where the adversary can only attack a small number of agents. Our theoretical analysis shows that the DeMABAR algorithm can also almost completely eliminate the influence of adversarial attacks and is inherently robust in the Byzantine setting, where an unknown fraction of the agents can be Byzantine, i.e., may arbitrarily select arms and communicate wrong information. We also conduct numerical experiments to illustrate the robustness and effectiveness of the proposed method.
Authors: Joel Wendin, Claudio Altafini
Abstract: The paper surveys recent progresses in understanding the dynamics and loss landscape of the gradient flow equations associated to deep linear neural networks, i.e., the gradient descent training dynamics (in the limit when the step size goes to 0) of deep neural networks missing the activation functions and subject to quadratic loss functions. When formulated in terms of the adjacency matrix of the neural network, as we do in the paper, these gradient flow equations form a class of converging matrix ODEs which is nilpotent, polynomial, isospectral, and with conservation laws. The loss landscape is described in detail. It is characterized by infinitely many global minima and saddle points, both strict and nonstrict, but lacks local minima and maxima. The loss function itself is a positive semidefinite Lyapunov function for the gradient flow, and its level sets are unbounded invariant sets of critical points, with critical values that correspond to the amount of singular values of the input-output data learnt by the gradient along a certain trajectory. The adjacency matrix representation we use in the paper allows to highlight the existence of a quotient space structure in which each critical value of the loss function is represented only once, while all other critical points with the same critical value belong to the fiber associated to the quotient space. It also allows to easily determine stable and unstable submanifolds at the saddle points, even when the Hessian fails to obtain them.
Authors: Arnab Bhattacharyya, Davin Choo, Philips George John, Themis Gouleakis
Abstract: Given i.i.d.~samples from an unknown distribution $P$, the goal of distribution learning is to recover the parameters of a distribution that is close to $P$. When $P$ belongs to the class of product distributions on the Boolean hypercube $\{0,1\}^d$, it is known that $\Omega(d/\varepsilon^2)$ samples are necessary to learn $P$ within total variation (TV) distance $\varepsilon$. We revisit this problem when the learner is also given as advice the parameters of a product distribution $Q$. We show that there is an efficient algorithm to learn $P$ within TV distance $\varepsilon$ that has sample complexity $\tilde{O}(d^{1-\eta}/\varepsilon^2)$, if $\|\mathbf{p} - \mathbf{q}\|_1 < \varepsilon d^{0.5 - \Omega(\eta)}$. Here, $\mathbf{p}$ and $\mathbf{q}$ are the mean vectors of $P$ and $Q$ respectively, and no bound on $\|\mathbf{p} - \mathbf{q}\|_1$ is known to the algorithm a priori.
Authors: Yixi Chen, Weixuan Liang, Tianrui Liu, Jun-Jie Huang, Ao Li, Xueling Zhu, Xinwang Liu
Abstract: Kernel power $k$-means (KPKM) leverages a family of means to mitigate local minima issues in kernel $k$-means. However, KPKM faces two key limitations: (1) the computational burden of the full kernel matrix restricts its use on extensive data, and (2) the lack of authentic centroid-sample assignment learning reduces its noise robustness. To overcome these challenges, we propose RFF-KPKM, introducing the first approximation theory for applying random Fourier features (RFF) to KPKM. RFF-KPKM employs RFF to generate efficient, low-dimensional feature maps, bypassing the need for the whole kernel matrix. Crucially, we are the first to establish strong theoretical guarantees for this combination: (1) an excess risk bound of $\mathcal{O}(\sqrt{k^3/n})$, (2) strong consistency with membership values, and (3) a $(1+\varepsilon)$ relative error bound achievable using the RFF of dimension $\mathrm{poly}(\varepsilon^{-1}\log k)$. Furthermore, to improve robustness and the ability to learn multiple kernels, we propose IP-RFF-MKPKM, an improved possibilistic RFF-based multiple kernel power $k$-means. IP-RFF-MKPKM ensures the scalability of MKPKM via RFF and refines cluster assignments by combining the merits of the possibilistic membership and fuzzy membership. Experiments on large-scale datasets demonstrate the superior efficiency and clustering accuracy of the proposed methods compared to the state-of-the-art alternatives.
Authors: Yunpeng Zhai, Shuchang Tao, Cheng Chen, Anni Zou, Ziqian Chen, Qingxu Fu, Shinji Mai, Li Yu, Jiaji Deng, Zouying Cao, Zhaoyang Liu, Bolin Ding, Jingren Zhou
Abstract: Autonomous agents powered by large language models (LLMs) have the potential to significantly enhance human productivity by reasoning, using tools, and executing complex tasks in diverse environments. However, current approaches to developing such agents remain costly and inefficient, as they typically require manually constructed task datasets and reinforcement learning (RL) pipelines with extensive random exploration. These limitations lead to prohibitively high data-construction costs, low exploration efficiency, and poor sample utilization. To address these challenges, we present AgentEvolver, a self-evolving agent system that leverages the semantic understanding and reasoning capabilities of LLMs to drive autonomous agent learning. AgentEvolver introduces three synergistic mechanisms: (i) self-questioning, which enables curiosity-driven task generation in novel environments, reducing dependence on handcrafted datasets; (ii) self-navigating, which improves exploration efficiency through experience reuse and hybrid policy guidance; and (iii) self-attributing, which enhances sample efficiency by assigning differentiated rewards to trajectory states and actions based on their contribution. By integrating these mechanisms into a unified framework, AgentEvolver enables scalable, cost-effective, and continual improvement of agent capabilities. Preliminary experiments indicate that AgentEvolver achieves more efficient exploration, better sample utilization, and faster adaptation compared to traditional RL-based baselines.
Authors: Feng Wang, Tianxiang Chen, Shuyue Wei, Qian Chu, Yi Zhang, Yifan Sun, Zhiming Zheng
Abstract: Spatio-temporal graphs are powerful tools for modeling complex dependencies in traffic time series. However, the distributed nature of real-world traffic data across multiple stakeholders poses significant challenges in modeling and reconstructing inter-client spatial dependencies while adhering to data locality constraints. Existing methods primarily address static dependencies, overlooking their dynamic nature and resulting in suboptimal performance. In response, we propose Federated Spatio-Temporal Graph with Dynamic Inter-Client Dependencies (FedSTGD), a framework designed to model and reconstruct dynamic inter-client spatial dependencies in federated learning. FedSTGD incorporates a federated nonlinear computation decomposition module to approximate complex graph operations. This is complemented by a graph node embedding augmentation module, which alleviates performance degradation arising from the decomposition. These modules are coordinated through a client-server collective learning protocol, which decomposes dynamic inter-client spatial dependency learning tasks into lightweight, parallelizable subtasks. Extensive experiments on four real-world datasets demonstrate that FedSTGD achieves superior performance over state-of-the-art baselines in terms of RMSE, MAE, and MAPE, approaching that of centralized baselines. Ablation studies confirm the contribution of each module in addressing dynamic inter-client spatial dependencies, while sensitivity analysis highlights the robustness of FedSTGD to variations in hyperparameters.
Authors: Darsh Pareek, Umesh Kumar, Ruthu Rao, Ravi Janjam
Abstract: Deep Neural Networks (DNNs) rely on inherent fluctuations in their internal parameters (weights and biases) to effectively navigate the complex optimization landscape and achieve robust performance. While these fluctuations are recognized as crucial for escaping local minima and improving generalization, their precise relationship with fundamental hyperparameters remains underexplored. A significant knowledge gap exists concerning how the learning rate, a critical parameter governing the training process, directly influences the dynamics of these neural fluctuations. This study systematically investigates the impact of varying learning rates on the magnitude and character of weight and bias fluctuations within a neural network. We trained a model using distinct learning rates and analyzed the corresponding parameter fluctuations in conjunction with the network's final accuracy. Our findings aim to establish a clear link between the learning rate's value, the resulting fluctuation patterns, and overall model performance. By doing so, we provide deeper insights into the optimization process, shedding light on how the learning rate mediates the crucial exploration-exploitation trade-off during training. This work contributes to a more nuanced understanding of hyperparameter tuning and the underlying mechanics of deep learning.
Authors: Thomas Decker, Volker Tresp, Florian Buettner
Abstract: Perturbation-based explanations are widely utilized to enhance the transparency of machine-learning models in practice. However, their reliability is often compromised by the unknown model behavior under the specific perturbations used. This paper investigates the relationship between uncertainty calibration - the alignment of model confidence with actual accuracy - and perturbation-based explanations. We show that models systematically produce unreliable probability estimates when subjected to explainability-specific perturbations and theoretically prove that this directly undermines global and local explanation quality. To address this, we introduce ReCalX, a novel approach to recalibrate models for improved explanations while preserving their original predictions. Empirical evaluations across diverse models and datasets demonstrate that ReCalX consistently reduces perturbation-specific miscalibration most effectively while enhancing explanation robustness and the identification of globally important input features.
Authors: \c{C}a\u{g}r{\i} Eser, Zeynep Sonat Baltac{\i}, Emre Akba\c{s}, Sinan Kalkan
Abstract: Imbalance in classification tasks is commonly quantified by the cardinalities of examples across classes. This, however, disregards the presence of redundant examples and inherent differences in the learning difficulties of classes. Alternatively, one can use complex measures such as training loss and uncertainty, which, however, depend on training a machine learning model. Our paper proposes using data Intrinsic Dimensionality (ID) as an easy-to-compute, model-free measure of imbalance that can be seamlessly incorporated into various imbalance mitigation methods. Our results across five different datasets with a diverse range of imbalance ratios show that ID consistently outperforms cardinality-based re-weighting and re-sampling techniques used in the literature. Moreover, we show that combining ID with cardinality can further improve performance. Code: https://github.com/cagries/IDIM.
Authors: Ruxi Deng, Wenxuan Bao, Tianxin Wei, Jingrui He
Abstract: Pretrained VLMs exhibit strong zero-shot classification capabilities, but their predictions degrade significantly under common image corruptions. To improve robustness, many test-time adaptation (TTA) methods adopt positive data augmentation (PDA), which generates multiple views of each test sample to reduce prediction variance. However, these methods suffer from two key limitations. First, it introduces considerable computational overhead due to the large number of augmentations required per image. Second, it fails to mitigate prediction bias, where the model tends to predict certain classes disproportionately under corruption, as PDA operates on corrupted inputs and typically does not remove the corruption itself. To address these challenges, we propose Panda, a novel TTA method based on negative data augmentation (NDA). Unlike positive augmentations that preserve object semantics, Panda generates negative augmentations by disrupting semantic content. It divides images into patches and randomly assembles them from a shared patch pool. These negatively augmented images retain corruption-specific features while discarding object-relevant signals. We then subtract the mean feature of these negative samples from the original image feature, effectively suppressing corruption-related components while preserving class-relevant information. This mitigates prediction bias under distribution shifts. Panda allows augmentation to be shared across samples within a batch, resulting in minimal computational overhead. Panda can be seamlessly integrated into existing test-time adaptation frameworks and substantially improve their robustness. Our experiments indicate that Panda delivers superior performance compared to PDA methods, and a wide range of TTA methods exhibit significantly enhanced performance when integrated with Panda. Our code is available at https://github.com/ruxideng/Panda .
Authors: Stanislav Selitskiy
Abstract: We propose loss function week enforcement of the velocity relations between time-series points in the Kinematic-Informed artificial Neural Networks (KINN) for long-term stock prediction. Problems of the series volatility, Out-of-Distribution (OOD) test data, and outliers in training data are addressed by (Artificial Neural Networks) ANN's learning not only future points prediction but also by learning velocity relations between the points, such a way as avoiding unrealistic spurious predictions. The presented loss function penalizes not only errors between predictions and supervised label data, but also errors between the next point prediction and the previous point plus velocity prediction. The loss function is tested on the multiple popular and exotic AR ANN architectures, and around fifteen years of Dow Jones function demonstrated statistically meaningful improvement across the normalization-sensitive activation functions prone to spurious behaviour in the OOD data conditions. Results show that such architecture addresses the issue of the normalization in the auto-regressive models that break the data topology by weakly enforcing the data neighbourhood proximity (relation) preservation during the ANN transformation.
Authors: Daryl Noupa Yongueng, Hamidou Tembine
Abstract: Normalization is a key point in transformer training . In Dynamic Tanh (DyT), the author demonstrated that Tanh can be used as an alternative layer normalization (LN) and confirmed the effectiveness of the idea. But Tanh itself faces orthogonality, linearity and distortion problems. Due to that, his proposition cannot be reliable. So we propose a Holonorm (hn) which has residual connections and nonlinearity. Holonorm is suitable for replacing Tanh in the context of normalization. Although the HoloNorm expression could be similar to the softsign function in dimension one, softsign is a componentwise function which is not good for tensors and vectors of great dimension. Holonorm preserves the orthogonality, the direction, the invertibility of the signal. Holonorm is also a suitable metric, maps all vectors into the open unit ball. This prevents exploding activations and improves stability in deep Transformer models. In this work, we have meticulously examined the normalization in transformers and say that Holonorm, a generalized form of softsign function suited as a normalization function first.Second, defined between 0 and 1 hn serves as a percentage, and $1 - \text{Holonorm}$ is its complement, making it better understandable in evaluating a model.
Authors: Benjamin Yu, Vincenzo Lordi, Daniel Schwalbe-Koda
Abstract: Machine learning interatomic potentials (MLIPs) balance high accuracy and lower costs compared to density functional theory calculations, but their performance often depends on the size and diversity of training datasets. Large datasets improve model accuracy and generalization but are computationally expensive to produce and train on, while smaller datasets risk discarding rare but important atomic environments and compromising MLIP accuracy/reliability. Here, we develop an information-theoretical framework to quantify the efficiency of dataset compression methods and propose an algorithm that maximizes this efficiency. By framing atomistic dataset compression as an instance of the minimum set cover (MSC) problem over atom-centered environments, our method identifies the smallest subset of structures that contains as much information as possible from the original dataset while pruning redundant information. The approach is extensively demonstrated on the GAP-20 and TM23 datasets, and validated on 64 varied datasets from the ColabFit repository. Across all cases, MSC consistently retains outliers, preserves dataset diversity, and reproduces the long-tail distributions of forces even at high compression rates, outperforming other subsampling methods. Furthermore, MLIPs trained on MSC-compressed datasets exhibit reduced error for out-of-distribution data even in low-data regimes. We explain these results using an outlier analysis and show that such quantitative conclusions could not be achieved with conventional dimensionality reduction methods. The algorithm is implemented in the open-source QUESTS package and can be used for several tasks in atomistic modeling, from data subsampling, outlier detection, and training improved MLIPs at a lower cost.
Authors: Emmanuel Asiedu Brempong, Mohammed Alewi Hassen, MohamedElfatih MohamedKhair, Vusumuzi Dube, Santiago Hincapie Potes, Olivia Graham, Amanie Brik, Amy McGovern, George Huffman, Jason Hickey
Abstract: Accurate precipitation estimation is critical for hydrological applications, especially in the Global South where ground-based observation networks are sparse and forecasting skill is limited. Existing satellite-based precipitation products often rely on the longwave infrared channel alone or are calibrated with data that can introduce significant errors, particularly at sub-daily timescales. This study introduces Oya, a novel real-time precipitation retrieval algorithm utilizing the full spectrum of visible and infrared (VIS-IR) observations from geostationary (GEO) satellites. Oya employs a two-stage deep learning approach, combining two U-Net models: one for precipitation detection and another for quantitative precipitation estimation (QPE), to address the inherent data imbalance between rain and no-rain events. The models are trained using high-resolution GPM Combined Radar-Radiometer Algorithm (CORRA) v07 data as ground truth and pre-trained on IMERG-Final retrievals to enhance robustness and mitigate overfitting due to the limited temporal sampling of CORRA. By leveraging multiple GEO satellites, Oya achieves quasi-global coverage and demonstrates superior performance compared to existing competitive regional and global precipitation baselines, offering a promising pathway to improved precipitation monitoring and forecasting.
Authors: Rishi Singhal, Jung-Eun Kim
Abstract: Layer Normalization (LayerNorm) is one of the fundamental components in transformers that stabilizes training and improves optimization. In recent times, Pre-LayerNorm transformers have become the preferred choice over Post-LayerNorm transformers due to their stable gradient flow. However, the impact of LayerNorm on learning and memorization across these architectures remains unclear. In this work, we investigate how LayerNorm influences memorization and learning for Pre- and Post-LayerNorm transformers. We identify that LayerNorm serves as a key factor for stable learning in Pre-LayerNorm transformers, while in Post-LayerNorm transformers, it impacts memorization. Our analysis reveals that eliminating LayerNorm parameters in Pre-LayerNorm models exacerbates memorization and destabilizes learning, while in Post-LayerNorm models, it effectively mitigates memorization by restoring genuine labels. We further precisely identify that early layers LayerNorm are the most critical over middle/later layers and their influence varies across Pre and Post LayerNorm models. We have validated it through 13 models across 6 Vision and Language datasets. These insights shed new light on the role of LayerNorm in shaping memorization and learning in transformers.
Authors: Reginald Zhiyan Chen, Heng-Sheng Chang, Prashant G. Mehta
Abstract: Hidden Markov Models (HMMs) are fundamental for modeling sequential data, yet learning their parameters from observations remains challenging. Classical methods like the Baum-Welch (EM) algorithm are computationally intensive and prone to local optima, while modern spectral algorithms offer provable guarantees but may produce probability outputs outside valid ranges. This work introduces Belief Net, a novel framework that learns HMM parameters through gradient-based optimization by formulating the HMM's forward filter as a structured neural network. Unlike black-box Transformer models, Belief Net's learnable weights are explicitly the logits of the initial distribution, transition matrix, and emission matrix, ensuring full interpretability. The model processes observation sequences using a decoder-only architecture and is trained end-to-end with standard autoregressive next-observation prediction loss. On synthetic HMM data, Belief Net achieves superior convergence speed compared to Baum-Welch, successfully recovering parameters in both undercomplete and overcomplete settings where spectral methods fail. Comparisons with Transformer-based models are also presented on real-world language data.
Authors: Garapati Keerthana, Manik Gupta
Abstract: Personalized decision systems in healthcare and behavioral support often rely on static rule-based or engagement-maximizing heuristics that overlook users' emotional context and ethical constraints. Such approaches risk recommending insensitive or unsafe interventions, especially in domains involving serious mental illness, substance use disorders, or depression. To address this limitation, we propose a Responsible Reinforcement Learning (RRL) framework that integrates emotional and contextual understanding with ethical considerations into the sequential decision-making process. RRL formulates personalization as a Constrained Markov Decision Process (CMDP), where the agent optimizes engagement and adherence while ensuring emotional alignment and ethical safety. We introduce a multi-objective reward function that explicitly balances short-term behavioral engagement with long-term user well-being, and define an emotion-informed state representation that captures fluctuations in emotional readiness, affect, and risk. The proposed architecture can be instantiated with any RL algorithm (e.g., DQN, PPO) augmented with safety constraints or Lagrangian regularization. Conceptually, this framework operationalizes empathy and responsibility within machine learning policy optimization, bridging safe RL, affective computing and responsible AI. We discuss the implications of this approach for human-centric domains such as behavioral health, education, and digital therapeutics, and outline simulation-based validation paths for future empirical work. This paper aims to initiate a methodological conversation about ethically aligned reinforcement learning for emotionally aware and trustworthy personalization systems.
Authors: Fengsheng Lin, Shengyi Yan, Trac Duy Tran
Abstract: We present a semi-unified sparse dictionary learning framework that bridges the gap between classical sparse models and modern deep architectures. Specifically, the method integrates strict Top-$K$ LISTA and its convex FISTA-based variant (LISTAConv) into the discriminative LC-KSVD2 model, enabling co-evolution between the sparse encoder and the dictionary under supervised or unsupervised regimes. This unified design retains the interpretability of traditional sparse coding while benefiting from efficient, differentiable training. We further establish a PALM-style convergence analysis for the convex variant, ensuring theoretical stability under block alternation. Experimentally, our method achieves 95.6\% on CIFAR-10, 86.3\% on CIFAR-100, and 88.5\% on TinyImageNet with faster convergence and lower memory cost ($<$4GB GPU). The results confirm that the proposed LC-KSVD2 + LISTA/LISTAConv pipeline offers an interpretable and computationally efficient alternative for modern deep architectures.
Authors: Yuval Shapira, Dana Drachsler-Cohen
Abstract: Few-pixel attacks mislead a classifier by modifying a few pixels of an image. Their perturbation space is an $\ell_0$-ball, which is not convex, unlike $\ell_p$-balls for $p\geq1$. However, existing local robustness verifiers typically scale by relying on linear bound propagation, which captures convex perturbation spaces. We show that the convex hull of an $\ell_0$-ball is the intersection of its bounding box and an asymmetrically scaled $\ell_1$-like polytope. The volumes of the convex hull and this polytope are nearly equal as the input dimension increases. We then show a linear bound propagation that precisely computes bounds over the convex hull and is significantly tighter than bound propagations over the bounding box or our $\ell_1$-like polytope. This bound propagation scales the state-of-the-art $\ell_0$ verifier on its most challenging robustness benchmarks by 1.24x-7.07x, with a geometric mean of 3.16.
Authors: Miles Wang-Henderson, Ben Kaufman, Edward Williams, Ryan Pederson, Matteo Rossi, Owen Howell, Carl Underkoffler, Narbe Mardirossian, John Parkhill
Abstract: Batched synthesis and testing of molecular designs is the key bottleneck of drug development. There has been great interest in leveraging biomolecular foundation models as surrogates to accelerate this process. In this work, we show how to obtain scalable probabilistic surrogates of binding affinity for use in Batch Bayesian Optimization (Batch BO). This demands parallel acquisition functions that hedge between designs and the ability to rapidly sample from a joint predictive density to approximate them. Through the framework of Epistemic Neural Networks (ENNs), we obtain scalable joint predictive distributions of binding affinity on top of representations taken from large structure-informed models. Key to this work is an investigation into the importance of prior networks in ENNs and how to pretrain them on synthetic data to improve downstream performance in Batch BO. Their utility is demonstrated by rediscovering known potent EGFR inhibitors on a semi-synthetic benchmark in up to 5x fewer iterations, as well as potent inhibitors from a real-world small-molecule library in up to 10x fewer iterations, offering a promising solution for large-scale drug discovery applications.
Authors: Avrim Blum, Marten Garicano, Kavya Ravichandran, Dravyansh Sharma
Abstract: The improving multi-armed bandits problem is a formal model for allocating effort under uncertainty, motivated by scenarios such as investing research effort into new technologies, performing clinical trials, and hyperparameter selection from learning curves. Each pull of an arm provides reward that increases monotonically with diminishing returns. A growing line of work has designed algorithms for improving bandits, albeit with somewhat pessimistic worst-case guarantees. Indeed, strong lower bounds of $\Omega(k)$ and $\Omega(\sqrt{k})$ multiplicative approximation factors are known for both deterministic and randomized algorithms (respectively) relative to the optimal arm, where $k$ is the number of bandit arms. In this work, we propose two new parameterized families of bandit algorithms and bound the sample complexity of learning the near-optimal algorithm from each family using offline data. The first family we define includes the optimal randomized algorithm from prior work. We show that an appropriately chosen algorithm from this family can achieve stronger guarantees, with optimal dependence on $k$, when the arm reward curves satisfy additional properties related to the strength of concavity. Our second family contains algorithms that both guarantee best-arm identification on well-behaved instances and revert to worst case guarantees on poorly-behaved instances. Taking a statistical learning perspective on the bandit rewards optimization problem, we achieve stronger data-dependent guarantees without the need for actually verifying whether the assumptions are satisfied.
Authors: Tommaso Castellani, Naimeng Ye, Daksh Mittal, Thomson Yen, Hongseok Namkoong
Abstract: AI agents increasingly rely on external tools to solve complex, long-horizon tasks. Advancing such agents requires reproducible evaluation and large-scale training in controllable, diverse, and realistic tool-use environments. However, real-world APIs are limited in availability, domain coverage, and stability, often requiring access keys and imposing rate limits, which render them impractical for stable evaluation or scalable training. To address these challenges, we introduce SynthTools, a flexible and scalable framework for generating synthetic tool ecosystems. Our framework consists of three core components: Tool Generation for automatic and scalable creation of diverse tools, Tool Simulation to emulate realistic tool behaviors, and Tool Audit to ensure correctness and consistency of tool simulation. To illustrate its scalability, we show that SynthTools can readily produce toolsets that span twice as many domains and twice as many tools per domain as prior work. Furthermore, the tool simulation and tool audit components demonstrate strong reliability, achieving $94\%$ and $99\%$ accuracy respectively. Finally, we construct downstream tasks from the generated tools that even state-of-the-art models struggle to complete. By enabling scalable, diverse, and reliable tool ecosystems, SynthTools provides a practical path toward large-scale training and stable evaluation of tool-use agents. Our code is available at https://github.com/namkoong-lab/SynthTools.
Authors: Abraham Francisco Arellano Tavara, Umesh Kumar, Jathurshan Pradeepkumar, Jimeng Sun
Abstract: Variants of Uncertain Significance (VUS) limit the clinical utility of prostate cancer genomics by delaying diagnosis and therapy when evidence for pathogenicity or benignity is incomplete. Progress is further limited by inconsistent annotations across sources and the absence of a prostate-specific benchmark for fair comparison. We introduce Prostate-VarBench, a curated pipeline for creating prostate-specific benchmarks that integrates COSMIC (somatic cancer mutations), ClinVar (expert-curated clinical variants), and TCGA-PRAD (prostate tumor genomics from The Cancer Genome Atlas) into a harmonized dataset of 193,278 variants supporting patient- or gene-aware splits to prevent data leakage. To ensure data integrity, we corrected a Variant Effect Predictor (VEP) issue that merged multiple transcript records, introducing ambiguity in clinical significance fields. We then standardized 56 interpretable features across eight clinically relevant tiers, including population frequency, variant type, and clinical context. AlphaMissense pathogenicity scores were incorporated to enhance missense variant classification and reduce VUS uncertainty. Building on this resource, we trained an interpretable TabNet model to classify variant pathogenicity, whose step-wise sparse masks provide per-case rationales consistent with molecular tumor board review practices. On the held-out test set, the model achieved 89.9% accuracy with balanced class metrics, and the VEP correction yields an 6.5% absolute reduction in VUS.
Authors: Xuan Son Nguyen, Aymeric Histace, Nistor Grozavu
Abstract: Riemannian symmetric spaces (RSS) such as hyperbolic spaces and symmetric positive definite (SPD) manifolds have become popular spaces for representation learning. In this paper, we propose a novel approach for building discriminative neural networks on Siegel spaces, a family of RSS that is largely unexplored in machine learning tasks. For classification applications, one focus of recent works is the construction of multiclass logistic regression (MLR) and fully-connected (FC) layers for hyperbolic and SPD neural networks. Here we show how to build such layers for Siegel neural networks. Our approach relies on the quotient structure of those spaces and the notation of vector-valued distance on RSS. We demonstrate the relevance of our approach on two applications, i.e., radar clutter classification and node classification. Our results successfully demonstrate state-of-the-art performance across all datasets.
Authors: Johannes Kiechle, Stefan M. Fischer, Daniel M. Lang, Cosmin I. Bercea, Matthew J. Nyflot, Lina Felsner, Julia A. Schnabel, Jan C. Peeken
Abstract: The growing number of medical tomography examinations has necessitated the development of automated methods capable of extracting comprehensive imaging features to facilitate downstream tasks such as tumor characterization, while assisting physicians in managing their growing workload. However, 3D medical image classification remains a challenging task due to the complex spatial relationships and long-range dependencies inherent in volumetric data. Training models from scratch suffers from low data regimes, and the absence of 3D large-scale multimodal datasets has limited the development of 3D medical imaging foundation models. Recent studies, however, have highlighted the potential of 2D vision foundation models, originally trained on natural images, as powerful feature extractors for medical image analysis. Despite these advances, existing approaches that apply 2D models to 3D volumes via slice-based decomposition remain suboptimal. Conventional volume slicing strategies, which rely on canonical planes such as axial, sagittal, or coronal, may inadequately capture the spatial extent of target structures when these are misaligned with standardized viewing planes. Furthermore, existing slice-wise aggregation strategies rarely account for preserving the volumetric structure, resulting in a loss of spatial coherence across slices. To overcome these limitations, we propose TomoGraphView, a novel framework that integrates omnidirectional volume slicing with spherical graph-based feature aggregation. We publicly share our accessible code base at http://github.com/compai-lab/2025-MedIA-kiechle and provide a user-friendly library for omnidirectional volume slicing at https://pypi.org/project/OmniSlicer.
URLs: http://github.com/compai-lab/2025-MedIA-kiechle, https://pypi.org/project/OmniSlicer.
Authors: Yu. Yu. Dubenskaya, S. P. Polyakov, A. P. Kryukov, A. P. Demichev, E. O. Gres, E. B. Postnikov, A. Yu. Razumov, P. A. Volchugov, D. P. Zhurov
Abstract: The aim of extensive air shower (EAS) analysis is to reconstruct the physical parameters of the primary particle that initiated the shower. The TAIGA experiment is a hybrid detector system that combines several imaging atmospheric Cherenkov telescopes (IACTs) and an array of non-imaging Cherenkov detectors (TAIGA-HiSCORE) for EAS detection. Because the signals recorded by different detector types differ in physical nature, the direct merging of data is unfeasible, which complicates multimodal analysis. Currently, to analyze data from the IACTs and TAIGA-HiSCORE, a set of auxiliary parameters specific to each detector type is calculated from the recorded signals. These parameters are chosen empirically, so there is no certainty that they retain all important information and are the best suited for the respective problems. We propose to use autoencoders (AE) for the analysis of TAIGA experimental data and replace the conventionally used auxiliary parameters with the parameters of the AE latent space. The advantage of the AE latent space parameters is that they preserve essential physics from experimental data without prior assumptions. This approach also holds potential for enabling seamless integration of heterogeneous IACT and HiSCORE data through a joint latent space. To reconstruct the parameters of the primary particle of the EAS from the latent space of the AE, a separate artificial neural network is used. In this paper, the proposed approach is used to reconstruct the energy of the EAS primary particles based on Monte Carlo simulation data for TAIGA-HiSCORE. The dependence of the energy determination accuracy on the dimensionality of the latent space is analyzed, and these results are also compared with the results obtained by the conventional technique. It is shown that when using the AE latent space, the energy of the primary particle is reconstructed with satisfactory accuracy.
Authors: Hannah Lu, Llu{\i}s Salo-Salgado, Ruben Juanes
Abstract: Fault zones exhibit complex and heterogeneous permeability structures influenced by stratigraphic, compositional, and structural factors, making them critical yet uncertain components in subsurface flow modeling. In this study, we investigate how lithological controls influence fault permeability using the PREDICT framework: a probabilistic workflow that couples stochastic fault geometry generation, physically constrained material placement, and flow-based upscaling. The flow-based upscaling step, however, is a very computationally expensive component of the workflow and presents a major bottleneck that makes global sensitivity analysis (GSA) intractable, as it requires millions of model evaluations. To overcome this challenge, we develop a neural network surrogate to emulate the flow-based upscaling step. This surrogate model dramatically reduces the computational cost while maintaining high accuracy, thereby making GSA feasible. The surrogate-model-enabled GSA reveals new insights into the effects of lithological controls on fault permeability. In addition to identifying dominant parameters and negligible ones, the analysis uncovers significant nonlinear interactions between parameters that cannot be captured by traditional local sensitivity methods.
Authors: Felix B. Mueller, Jan F. Meier, Timo Lueddecke, Richard Vogg, Roger L. Freixanet, Valentin Hassler, Tiffany Bosshard, Elif Karakoc, William J. O'Hearn, Sofia M. Pereira, Sandro Sehner, Kaja Wierucka, Judith Burkart, Claudia Fichtel, Julia Fischer, Alexander Gail, Catherine Hobaiter, Julia Ostner, Liran Samuni, Oliver Sch\"ulke, Neda Shahidi, Erin G. Wessling, Alexander S. Ecker
Abstract: Non-human primates are our closest living relatives, and analyzing their behavior is central to research in cognition, evolution, and conservation. Computer vision could greatly aid this research, but existing methods often rely on human-centric pretrained models and focus on single datasets, which limits generalization. We address this limitation by shifting from a model-centric to a data-centric approach and introduce PriVi, a large-scale primate-centric video pretraining dataset. PriVi contains 424 hours of curated video, combining 174 hours from behavioral research across 11 settings with 250 hours of diverse web-sourced footage, assembled through a scalable data curation pipeline. We pretrain V-JEPA on PriVi to learn primate-specific representations and evaluate it using a lightweight frozen classifier. Across four benchmark datasets, ChimpACT, BaboonLand, PanAf500, and ChimpBehave, our approach consistently outperforms prior work, including fully finetuned baselines, and scales favorably with fewer labels. These results demonstrate that primate-centric pretraining substantially improves data efficiency and generalization, making it a promising approach for low-label applications. Code, models, and the majority of the dataset will be made available.
Authors: Katie Matton, Purvaja Balaji, Hamzeh Ghasemzadeh, Jameson C. Cooper, Daryush D. Mehta, Jarrad H. Van Stan, Robert E. Hillman, Rosalind Picard, John Guttag, S. Mazdak Abulnaga
Abstract: Phonotrauma refers to vocal fold tissue damage resulting from exposure to forces during voicing. It occurs on a continuum from mild to severe, and treatment options can vary based on severity. Assessment of severity involves a clinician's expert judgment, which is costly and can vary widely in reliability. In this work, we present the first method for automatically classifying phonotrauma severity from vocal fold images. To account for the ordinal nature of the labels, we adopt a widely used ordinal regression framework. To account for label uncertainty, we propose a novel modification to ordinal regression loss functions that enables them to operate on soft labels reflecting annotator rating distributions. Our proposed soft ordinal regression method achieves predictive performance approaching that of clinical experts, while producing well-calibrated uncertainty estimates. By providing an automated tool for phonotrauma severity assessment, our work can enable large-scale studies of phonotrauma, ultimately leading to improved clinical understanding and patient care.
Authors: Sujay Nair, Evan Coleman, Sherrie Wang, Elsa Olivetti
Abstract: Minerals play a critical role in the advanced energy technologies necessary for decarbonization, but characterizing mineral deposits hidden underground remains costly and challenging. Inspired by recent progress in generative modeling, we develop a learning method which infers the locations of minerals by masking and infilling geospatial maps of resource availability. We demonstrate this technique using mineral data for the conterminous United States, and train performant models, with the best achieving Dice coefficients of $0.31 \pm 0.01$ and recalls of $0.22 \pm 0.02$ on test data at 1$\times$1 mi$^2$ spatial resolution. One major advantage of our approach is that it can easily incorporate auxiliary data sources for prediction which may be more abundant than mineral data. We highlight the capabilities of our model by adding input layers derived from geophysical sources, along with a nation-wide ground survey of soils originally intended for agronomic purposes. We find that employing such auxiliary features can improve inference performance, while also enabling model evaluation in regions with no recorded minerals.
Authors: Craig Michoski, Matthew Waller, Brian Sammuli, Zeyu Li, Tapan Ganatma Nakkina, Raffi Nazikian, Sterling Smith, David Orozco, Dongyang Kuang, Martin Foltin, Erik Olofsson, Mike Fredrickson, Jerry Louis-Jeune, David R. Hatch, Todd A. Oliver, Mitchell Clark, Steph-Yves Louis
Abstract: Fusion energy research increasingly depends on the ability to integrate heterogeneous, multimodal datasets from high-resolution diagnostics, control systems, and multiscale simulations. The sheer volume and complexity of these datasets demand the development of new tools capable of systematically harmonizing and extracting knowledge across diverse modalities. The Data Fusion Labeler (dFL) is introduced as a unified workflow instrument that performs uncertainty-aware data harmonization, schema-compliant data fusion, and provenance-rich manual and automated labeling at scale. By embedding alignment, normalization, and labeling within a reproducible, operator-order-aware framework, dFL reduces time-to-analysis by greater than 50X (e.g., enabling >200 shots/hour to be consistently labeled rather than a handful per day), enhances label (and subsequently training) quality, and enables cross-device comparability. Case studies from DIII-D demonstrate its application to automated ELM detection and confinement regime classification, illustrating its potential as a core component of data-driven discovery, model validation, and real-time control in future burning plasma devices.
Authors: Stelios Zarifis, Ioannis Chalkiadakis, Artemis Chardouveli, Vasiliki Moutzouri, Aggelos Sotirchos, Katerina Papadimitriou, Panagiotis Filntisis, Niki Efthymiou, Petros Maragos, Katerina Pastra
Abstract: Inspired by infant development, we propose a Reinforcement Learning (RL) framework for autonomous self-exploration in a robotic agent, Baby Sophia, using the BabyBench simulation environment. The agent learns self-touch and hand regard behaviors through intrinsic rewards that mimic an infant's curiosity-driven exploration of its own body. For self-touch, high-dimensional tactile inputs are transformed into compact, meaningful representations, enabling efficient learning. The agent then discovers new tactile contacts through intrinsic rewards and curriculum learning that encourage broad body coverage, balance, and generalization. For hand regard, visual features of the hands, such as skin-color and shape, are learned through motor babbling. Then, intrinsic rewards encourage the agent to perform novel hand motions, and follow its hands with its gaze. A curriculum learning setup from single-hand to dual-hand training allows the agent to reach complex visual-motor coordination. The results of this work demonstrate that purely curiosity-based signals, with no external supervision, can drive coordinated multimodal learning, imitating an infant's progression from random motor babbling to purposeful behaviors.
Authors: Huanhuan Zhao, Connor Vernachio, Laxmi Bhurtel, Wooin Yang, Ruben Millan-Solsona, Spenser R. Brown, Marti Checa, Komal Sharma Agrawal, Adam M. Guss, Liam Collins, Wonhee Ko, Arpan Biswas
Abstract: Microscopy such as Scanning Tunneling Microscopy (STM), Atomic Force Microscopy (AFM) and Scanning Electron Microscopy (SEM) are essential tools in material imaging at micro- and nanoscale resolutions to extract physical knowledge and materials structure-property relationships. However, tuning microscopy controls (e.g. scanning speed, current setpoint, tip bias etc.) to obtain a high-quality of images is a non-trivial and time-consuming effort. On the other hand, with sub-standard images, the key features are not accurately discovered due to noise and artifacts, leading to erroneous analysis. Existing denoising models mostly build on generalizing the weak signals as noises while the strong signals are enhanced as key features, which is not always the case in microscopy images, thus can completely erase a significant amount of hidden physical information. To address these limitations, we propose a global denoising model (GDM) to smartly remove artifacts of microscopy images while preserving weaker but physically important features. The proposed model is developed based on 1) first designing a two-imaging input channel of non-pair and goal specific pre-processed images with user-defined trade-off information between two channels and 2) then integrating a loss function of pixel- and fast Fourier-transformed (FFT) based on training the U-net model. We compared the proposed GDM with the non-FFT denoising model over STM-generated images of Copper(Cu) and Silicon(Si) materials, AFM-generated Pantoea sp.YR343 bio-film images and SEM-generated plastic degradation images. We believe this proposed workflow can be extended to improve other microscopy image quality and will benefit the experimentalists with the proposed design flexibility to smartly tune via domain-experts preferences.
Authors: C. LeMay, A. Lane, J. Seales, M. Winstead, S. Baty
Abstract: Our research investigates how Natural Language Processing (NLP) can be used to extract main topics from a larger corpus of written data, as applied to the case of identifying signaling themes in Presidential Directives (PDs) from the Reagan through Clinton administrations. Analysts and NLP both identified relevant documents, demonstrating the potential utility of NLPs in research involving large written corpuses. However, we also identified discrepancies between NLP and human-labeled results that indicate a need for more research to assess the validity of NLP in this use case. The research was conducted in 2023, and the rapidly evolving landscape of AIML means existing tools have improved and new tools have been developed; this research displays the inherent capabilities of a potentially dated AI tool in emerging social science applications.
Authors: Mahsa Mitcheff, Siamul Karim Khan, Adam Czajka
Abstract: Developing reliable iris recognition and presentation attack detection methods requires diverse datasets that capture realistic variations in iris features and a wide spectrum of anomalies. Because of the rich texture of iris images, which spans a wide range of spatial frequencies, synthesizing same-identity iris images while controlling specific attributes remains challenging. In this work, we introduce a new iris image augmentation strategy by traversing a generative model's latent space toward latent codes that represent same-identity samples but with some desired iris image properties manipulated. The latent space traversal is guided by a gradient of specific geometrical, textural, or quality-related iris image features (e.g., sharpness, pupil size, iris size, or pupil-to-iris ratio) and preserves the identity represented by the image being manipulated. The proposed approach can be easily extended to manipulate any attribute for which a differentiable loss term can be formulated. Additionally, our approach can use either randomly generated images using either a pre-train GAN model or real-world iris images. We can utilize GAN inversion to project any given iris image into the latent space and obtain its corresponding latent code.
Authors: Zag ElSayed, Grace Westerkamp, Jack Yanchen Liu, Ernest Pedapati
Abstract: Neurodevelopmental disorders such as Fragile X Syndrome (FXS) and Autism Spectrum Disorder (ASD) are characterized by disrupted cortical oscillatory activity, particularly in the alpha and gamma frequency bands. These abnormalities are linked to deficits in attention, sensory processing, and cognitive function. In this work, we present an adaptive machine learning-based brain-computer interface (BCI) system designed to modulate neural oscillations through frequency-specific auditory stimulation to enhance cognitive readiness in individuals with FXS. EEG data were recorded from 38 participants using a 128-channel system under a stimulation paradigm consisting of a 30-second baseline (no stimulus) followed by 60-second auditory entrainment episodes at 7Hz, 9Hz, 11Hz, and 13Hz. A comprehensive analysis of power spectral features (Alpha, Gamma, Delta, Theta, Beta) and cross-frequency coupling metrics (Alpha-Gamma, Alpha-Beta, etc.) was conducted. The results identified Peak Alpha Power, Peak Gamma Power, and Alpha Power per second per channel as the most discriminative biomarkers. The 13Hz stimulation condition consistently elicited a significant increase in Alpha activity and suppression of Gamma activity, aligning with our optimization objective. A supervised machine learning framework was developed to predict EEG responses and dynamically adjust stimulation parameters, enabling real-time, subject-specific adaptation. This work establishes a novel EEG-driven optimization framework for cognitive neuromodulation, providing a foundational model for next-generation AI-integrated BCI systems aimed at personalized neurorehabilitation in FXS and related disorders.
Authors: Alessandro V. M. Oliveira
Abstract: This paper presents and discusses forms of estimation by regularized regression and model selection using the LASSO method - Least Absolute Shrinkage and Selection Operator. LASSO is recognized as one of the main supervised learning methods applied to high-dimensional econometrics, allowing work with large volumes of data and multiple correlated controls. Conceptual issues related to the consequences of high dimensionality in modern econometrics and the principle of sparsity, which underpins regularization procedures, are addressed. The study examines the main post-double selection and post-regularization models, including variations applied to instrumental variable models. A brief description of the lassopack routine package, its syntaxes, and examples of HD, HDS (High-Dimension Sparse), and IV-HDS models, with combinations involving fixed effects estimators, is also presented. Finally, the potential application of the approach in research focused on air transport is discussed, with emphasis on an empirical study on the operational efficiency of airlines and aircraft fuel consumption.
Authors: Rik Adriaensen, Lucas Van Praet, Jessa Bekker, Robin Manhaeve, Pieter Delobelle, Maarten Buyl
Abstract: Operationalizing definitions of fairness is difficult in practice, as multiple definitions can be incompatible while each being arguably desirable. Instead, it may be easier to directly describe algorithmic bias through ad-hoc assumptions specific to a particular real-world task, e.g., based on background information on systemic biases in its context. Such assumptions can, in turn, be used to mitigate this bias during training. Yet, a framework for incorporating such assumptions that is simultaneously principled, flexible, and interpretable is currently lacking. Our approach is to formalize bias assumptions as programs in ProbLog, a probabilistic logic programming language that allows for the description of probabilistic causal relationships through logic. Neurosymbolic extensions of ProbLog then allow for easy integration of these assumptions in a neural network's training process. We propose a set of templates to express different types of bias and show the versatility of our approach on synthetic tabular datasets with known biases. Using estimates of the bias distortions present, we also succeed in mitigating algorithmic bias in real-world tabular and image data. We conclude that ProbLog4Fairness outperforms baselines due to its ability to flexibly model the relevant bias assumptions, where other methods typically uphold a fixed bias type or notion of fairness.
Authors: Aaron Miller, Sahil Kommalapati, Robert Moser, Petros Koumoutsakos
Abstract: Accurate and generalizable Reynolds-averaged Navier-Stokes (RANS) models for turbulent flows rely on effective closures. We introduce tensor-based, symmetry aware closures using equivariant neural networks (ENNs) and present an algorithm for enforcing algebraic contraction relations among tensor components. The modeling approach builds on the structure tensor framework introduced by Kassinos and Reynolds to learn closures in the rapid distortion theory setting. Experiments show that ENNs can effectively learn relationships involving high-order tensors, meeting or exceeding the performance of existing models in tasks such as predicting the rapid pressure-strain correlation. Our results show that ENNs provide a physically consistent alternative to classical tensor basis models, enabling end-to-end learning of unclosed terms in RANS and fast exploration of model dependencies.
Authors: Dilli Prasad Sharma, Xiaowei Sun, Liang Xue, Xiaodong Lin, Pulei Xiong
Abstract: The widespread integration of Artificial Intelligence of Things (AIoT) in smart home environments has amplified the demand for transparent and interpretable machine learning models. To foster user trust and comply with emerging regulatory frameworks, the Explainable AI (XAI) methods, particularly post-hoc techniques such as SHapley Additive exPlanations (SHAP), and Local Interpretable Model-Agnostic Explanations (LIME), are widely employed to elucidate model behavior. However, recent studies have shown that these explanation methods can inadvertently expose sensitive user attributes and behavioral patterns, thereby introducing new privacy risks. To address these concerns, we propose a novel privacy-preserving approach based on SHAP entropy regularization to mitigate privacy leakage in explainable AIoT applications. Our method incorporates an entropy-based regularization objective that penalizes low-entropy SHAP attribution distributions during training, promoting a more uniform spread of feature contributions. To evaluate the effectiveness of our approach, we developed a suite of SHAP-based privacy attacks that strategically leverage model explanation outputs to infer sensitive information. We validate our method through comparative evaluations using these attacks alongside utility metrics on benchmark smart home energy consumption datasets. Experimental results demonstrate that SHAP entropy regularization substantially reduces privacy leakage compared to baseline models, while maintaining high predictive accuracy and faithful explanation fidelity. This work contributes to the development of privacy-preserving explainable AI techniques for secure and trustworthy AIoT applications.
Authors: Eshika Pathak, Ahmed Aboudonia, Sandeep Banik, Naira Hovakimyan
Abstract: Dynamical system (DS)-based learning from demonstration (LfD) is a powerful tool for generating motion plans in the operation (`task') space of robotic systems. However, the realization of the generated motion plans is often compromised by a ''task-execution mismatch'', where unmodeled dynamics, persistent disturbances, and system latency cause the robot's actual task-space state to diverge from the desired motion trajectory. We propose a novel task-level robust control architecture, L1-augmented Dynamical Systems (L1-DS), that explicitly handles the task-execution mismatch in tracking a nominal motion plan generated by any DS-based LfD scheme. Our framework augments any DS-based LfD model with a nominal stabilizing controller and an L1 adaptive controller. Furthermore, we introduce a windowed Dynamic Time Warping (DTW)-based target selector, which enables the nominal stabilizing controller to handle temporal misalignment for improved phase-consistent tracking. We demonstrate the efficacy of our architecture on the LASA and IROS handwriting datasets.
Authors: Salvish Goomanee, Andi Han, Pratik Jawanpuria, Bamdev Mishra
Abstract: This work extends the recently introduced Alpha-Procrustes family of Riemannian metrics for symmetric positive definite (SPD) matrices by incorporating generalized versions of the Bures-Wasserstein (GBW), Log-Euclidean, and Wasserstein distances. While the Alpha-Procrustes framework has unified many classical metrics in both finite- and infinite- dimensional settings, it previously lacked the structural components necessary to realize these generalized forms. We introduce a formalism based on unitized Hilbert-Schmidt operators and an extended Mahalanobis norm that allows the construction of robust, infinite-dimensional generalizations of GBW and Log-Hilbert-Schmidt distances. Our approach also incorporates a learnable regularization parameter that enhances geometric stability in high-dimensional comparisons. Preliminary experiments reproducing benchmarks from the literature demonstrate the improved performance of our generalized metrics, particularly in scenarios involving comparisons between datasets of varying dimension and scale. This work lays a theoretical and computational foundation for advancing robust geometric methods in machine learning, statistical inference, and functional data analysis.
Authors: Konstantinos M. Dafnis, Dimitris N. Metaxas
Abstract: Vision-Language Models (VLMs) excel at zero-shot inference but often degrade under test-time domain shifts. For this reason, episodic test-time adaptation strategies have recently emerged as powerful techniques for adapting VLMs to a single unlabeled image. However, existing adaptation strategies, such as test-time prompt tuning, typically require backpropagating through large encoder weights or altering core model components. In this work, we introduce Spectrum-Aware Test-Time Steering (STS), a lightweight adaptation framework that extracts a spectral subspace from the textual embeddings to define principal semantic directions and learns to steer latent representations in a spectrum-aware manner by adapting a small number of per-sample shift parameters to minimize entropy across augmented views. STS operates entirely at inference in the latent space, without backpropagation through or modification of the frozen encoders. Building on standard evaluation protocols, our comprehensive experiments demonstrate that STS largely surpasses or compares favorably against state-of-the-art test-time adaptation methods, while introducing only a handful of additional parameters and achieving inference speeds up to 8x faster with a 12x smaller memory footprint than conventional test-time prompt tuning. The code is available at https://github.com/kdafnis/STS.
Authors: Nikunj Gupta, Bill Guo, Rajgopal Kannan, Viktor K. Prasanna
Abstract: Large Language Models (LLMs) deliver state-of-the-art performance across many tasks but impose high computational and memory costs, limiting their deployment in resource-constrained or real-time settings. To address this, we propose HierRouter, a hierarchical routing approach that dynamically assembles inference pipelines from a pool of specialized, lightweight language models. Formulated as a finite-horizon Markov Decision Process (MDP), our approach trains a Proximal Policy Optimization (PPO)-based reinforcement learning agent to iteratively select which models to invoke at each stage of multi-hop inference. The agent conditions on the evolving context and accumulated cost to make context-aware routing decisions. Experiments with three open-source candidate LLMs across six benchmarks, including QA, code generation, and mathematical reasoning, show that HierRouter improves response quality by up to 2.4x compared to using individual models independently, while incurring only a minimal additional inference cost on average. These results highlight the promise of hierarchical routing for cost-efficient, high-performance LLM inference. All codes can be found here https://github.com/ Nikunj-Gupta/hierouter.
URLs: https://github.com/
Authors: Keshara Weerasinghe, Xueren Ge, Tessa Heick, Lahiru Nuwan Wijayasingha, Anthony Cortez, Abhishek Satpathy, John Stankovic, Homa Alemzadeh
Abstract: Emergency Medical Services (EMS) are critical to patient survival in emergencies, but first responders often face intense cognitive demands in high-stakes situations. AI cognitive assistants, acting as virtual partners, have the potential to ease this burden by supporting real-time data collection and decision making. In pursuit of this vision, we introduce EgoEMS, the first end-to-end, high-fidelity, multimodal, multiperson dataset capturing over 20 hours of realistic, procedural EMS activities from an egocentric view in 233 simulated emergency scenarios performed by 62 participants, including 46 EMS professionals. Developed in collaboration with EMS experts and aligned with national standards, EgoEMS is captured using an open-source, low-cost, and replicable data collection system and is annotated with keysteps, timestamped audio transcripts with speaker diarization, action quality metrics, and bounding boxes with segmentation masks. Emphasizing realism, the dataset includes responder-patient interactions reflecting real-world emergency dynamics. We also present a suite of benchmarks for real-time multimodal keystep recognition and action quality estimation, essential for developing AI support tools for EMS. We hope EgoEMS inspires the research community to push the boundaries of intelligent EMS systems and ultimately contribute to improved patient outcomes.
Authors: Shunan Sheng, Bohan Wu, Bennett Zhu, Sinho Chewi, Aram-Alexandre Pooladian
Abstract: Structured variational inference constitutes a core methodology in modern statistical applications. Unlike mean-field variational inference, the approximate posterior is assumed to have interdependent structure. We consider the natural setting of star-structured variational inference, where a root variable impacts all the other ones. We prove the first results for existence, uniqueness, and self-consistency of the variational approximation. In turn, we derive quantitative approximation error bounds for the variational approximation to the posterior, extending prior work from the mean-field setting to the star-structured setting. We also develop a gradient-based algorithm with provable guarantees for computing the variational approximation using ideas from optimal transport theory. We explore the implications of our results for Gaussian measures and hierarchical Bayesian models, including generalized linear models with location family priors and spike-and-slab priors with one-dimensional debiasing. As a by-product of our analysis, we develop new stability results for star-separable transport maps which might be of independent interest.
Authors: Hao Xu, Yuntian Chen, Dongxiao Zhang
Abstract: Constitutive models are fundamental to solid mechanics and materials science, underpinning the quantitative description and prediction of material responses under diverse loading conditions. Traditional phenomenological models, which are derived through empirical fitting, often lack generalizability and rely heavily on expert intuition and predefined functional forms. In this work, we propose a graph-based equation discovery framework for the automated discovery of constitutive laws directly from multisource experimental data. This framework expresses equations as directed graphs, where nodes represent operators and variables, edges denote computational relations, and edge features encode parametric dependencies. This enables the generation and optimization of free-form symbolic expressions with undetermined material-specific parameters. Through the proposed framework, we have discovered new constitutive models for strain-rate effects in alloy steel materials and the deformation behavior of lithium metal. Compared with conventional empirical models, these new models exhibit compact analytical structures and achieve higher accuracy. The proposed graph-based equation discovery framework provides a generalizable and interpretable approach for data-driven scientific modelling, particularly in contexts where traditional empirical formulations are inadequate for representing complex physical phenomena.
Authors: Minrui Luo, Weihang Xu, Xiang Gao, Maryam Fazel, Simon Shaolei Du
Abstract: Gradient descent dynamics on the deep matrix factorization problem is extensively studied as a simplified theoretical model for deep neural networks. Although the convergence theory for two-layer matrix factorization is well-established, no global convergence guarantee for general deep matrix factorization under random initialization has been established to date. To address this gap, we provide a polynomial-time global convergence guarantee for randomly initialized gradient descent on four-layer matrix factorization, given certain conditions on the target matrix and a standard balanced regularization term. Our analysis employs new techniques to show saddle-avoidance properties of gradient decent dynamics, and extends previous theories to characterize the change in eigenvalues of layer weights.
Authors: Mustafa Munir, Md Mostafijur Rahman, Radu Marculescu
Abstract: Vision Graph Neural Networks (ViGs) offer a new direction for advancements in vision architectures. While powerful, ViGs often face substantial computational challenges stemming from their graph construction phase, which can hinder their efficiency. To address this issue we propose AdaptViG, an efficient and powerful hybrid Vision GNN that introduces a novel graph construction mechanism called Adaptive Graph Convolution. This mechanism builds upon a highly efficient static axial scaffold and a dynamic, content-aware gating strategy called Exponential Decay Gating. This gating mechanism selectively weighs long-range connections based on feature similarity. Furthermore, AdaptViG employs a hybrid strategy, utilizing our efficient gating mechanism in the early stages and a full Global Attention block in the final stage for maximum feature aggregation. Our method achieves a new state-of-the-art trade-off between accuracy and efficiency among Vision GNNs. For instance, our AdaptViG-M achieves 82.6% top-1 accuracy, outperforming ViG-B by 0.3% while using 80% fewer parameters and 84% fewer GMACs. On downstream tasks, AdaptViG-M obtains 45.8 mIoU, 44.8 APbox, and 41.1 APmask, surpassing the much larger EfficientFormer-L7 by 0.7 mIoU, 2.2 APbox, and 2.1 APmask, respectively, with 78% fewer parameters.
Authors: Shahid Amin, Syed Pervez Hussnain Shah
Abstract: The remarkable progress in Artificial Intelligence (AI) is foundation-ally linked to a concurrent revolution in computer architecture. As AI models, particularly Deep Neural Networks (DNNs), have grown in complexity, their massive computational demands have pushed traditional architectures to their limits. This paper provides a structured review of this co-evolution, analyzing the architectural landscape designed to accelerate modern AI workloads. We explore the dominant architectural paradigms Graphics Processing Units (GPUs), Appli-cation-Specific Integrated Circuits (ASICs), and Field-Programmable Gate Ar-rays (FPGAs) by breaking down their design philosophies, key features, and per-formance trade-offs. The core principles essential for performance and energy efficiency, including dataflow optimization, advanced memory hierarchies, spar-sity, and quantization, are analyzed. Furthermore, this paper looks ahead to emerging technologies such as Processing-in-Memory (PIM) and neuromorphic computing, which may redefine future computation. By synthesizing architec-tural principles with quantitative performance data from industry-standard benchmarks, this survey presents a comprehensive picture of the AI accelerator landscape. We conclude that AI and computer architecture are in a symbiotic relationship, where hardware-software co-design is no longer an optimization but a necessity for future progress in computing.
Authors: Tao Jiang, Zichuan Lin, Lihe Li, Yi-Chen Li, Cong Guan, Lei Yuan, Zongzhang Zhang, Yang Yu, Deheng Ye
Abstract: Large transformer models, trained on diverse datasets, have demonstrated impressive few-shot performance on previously unseen tasks without requiring parameter updates. This capability has also been explored in Reinforcement Learning (RL), where agents interact with the environment to retrieve context and maximize cumulative rewards, showcasing strong adaptability in complex settings. However, in cooperative Multi-Agent Reinforcement Learning (MARL), where agents must coordinate toward a shared goal, decentralized policy deployment can lead to mismatches in task alignment and reward assignment, limiting the efficiency of policy adaptation. To address this challenge, we introduce Multi-agent In-context Coordination via Decentralized Memory Retrieval (MAICC), a novel approach designed to enhance coordination by fast adaptation. Our method involves training a centralized embedding model to capture fine-grained trajectory representations, followed by decentralized models that approximate the centralized one to obtain team-level task information. Based on the learned embeddings, relevant trajectories are retrieved as context, which, combined with the agents' current sub-trajectories, inform decision-making. During decentralized execution, we introduce a novel memory mechanism that effectively balances test-time online data with offline memory. Based on the constructed memory, we propose a hybrid utility score that incorporates both individual- and team-level returns, ensuring credit assignment across agents. Extensive experiments on cooperative MARL benchmarks, including Level-Based Foraging (LBF) and SMAC (v1/v2), show that MAICC enables faster adaptation to unseen tasks compared to existing methods. Code is available at https://github.com/LAMDA-RL/MAICC.
Authors: Yizheng Wang, Timon Rabczuk, Yinghua Liu
Abstract: Friction modeling plays a crucial role in achieving high-precision motion control in robotic operating systems. Traditional static friction models (such as the Stribeck model) are widely used due to their simple forms; however, they typically require predefined functional assumptions, which poses significant challenges when dealing with unknown functional structures. To address this issue, this paper proposes a physics-inspired machine learning approach based on the Kolmogorov Arnold Network (KAN) for static friction modeling of robotic joints. The method integrates spline activation functions with a symbolic regression mechanism, enabling model simplification and physical expression extraction through pruning and attribute scoring, while maintaining both high prediction accuracy and interpretability. We first validate the method's capability to accurately identify key parameters under known functional models, and further demonstrate its robustness and generalization ability under conditions with unknown functional structures and noisy data. Experiments conducted on both synthetic data and real friction data collected from a six-degree-of-freedom industrial manipulator show that the proposed method achieves a coefficient of determination greater than 0.95 across various tasks and successfully extracts concise and physically meaningful friction expressions. This study provides a new perspective for interpretable and data-driven robotic friction modeling with promising engineering applicability.
Authors: Haidong Huang, Haiyue Zhu. Jiayu Song, Xixin Zhao, Yaohua Zhou, Jiayi Zhang, Yuze Zhai, Xiaocong Li
Abstract: Offline-to-online reinforcement learning (O2O-RL) has emerged as a promising paradigm for safe and efficient robotic policy deployment but suffers from two fundamental challenges: limited coverage of multimodal behaviors and distributional shifts during online adaptation. We propose UEPO, a unified generative framework inspired by large language model pretraining and fine-tuning strategies. Our contributions are threefold: (1) a multi-seed dynamics-aware diffusion policy that efficiently captures diverse modalities without training multiple models; (2) a dynamic divergence regularization mechanism that enforces physically meaningful policy diversity; and (3) a diffusion-based data augmentation module that enhances dynamics model generalization. On the D4RL benchmark, UEPO achieves +5.9\% absolute improvement over Uni-O4 on locomotion tasks and +12.4\% on dexterous manipulation, demonstrating strong generalization and scalability.
Authors: Philipp Seeberger, Steffen Freisinger, Tobias Bocklet, Korbinian Riedhammer
Abstract: Due to the rapid growth of social media platforms, these tools have become essential for monitoring information during ongoing disaster events. However, extracting valuable insights requires real-time processing of vast amounts of data. A major challenge in existing systems is their exposure to event-related biases, which negatively affects their ability to generalize to emerging events. While recent advancements in debiasing and causal learning offer promising solutions, they remain underexplored in the disaster event domain. In this work, we approach bias mitigation through a causal lens and propose a method to reduce event- and domain-related biases, enhancing generalization to future events. Our approach outperforms multiple baselines by up to +1.9% F1 and significantly improves a PLM-based classifier across three disaster classification tasks.
Authors: J. Javier Alonso-Ramos, Ignacio Aguilera-Martos, Andr\'es Herrera-Poyatos, Francisco Herrera
Abstract: The performance of Machine Learning (ML) models, particularly those operating within the Interpretable Artificial Intelligence (Interpretable AI) framework, is significantly affected by the presence of noise in both training and production data. Denoising has therefore become a critical preprocessing step, typically categorized into instance removal and instance correction techniques. However, existing correction approaches often degrade performance or oversimplify the problem by altering the original data distribution. This leads to unrealistic scenarios and biased models, which is particularly problematic in contexts where interpretable AI models are employed, as their interpretability depends on the fidelity of the underlying data patterns. In this paper, we argue that defining noise independently of the solution may be ineffective, as its nature can vary significantly across tasks and datasets. Using a task-specific high quality solution as a reference can provide a more precise and adaptable noise definition. To this end, we propose DenoGrad, a novel Gradient-based instance Denoiser framework that leverages gradients from an accurate Deep Learning (DL) model trained on the target data -- regardless of the specific task -- to detect and adjust noisy samples. Unlike conventional approaches, DenoGrad dynamically corrects noisy instances, preserving problem's data distribution, and improving AI models robustness. DenoGrad is validated on both tabular and time series datasets under various noise settings against the state-of-the-art. DenoGrad outperforms existing denoising strategies, enhancing the performance of interpretable IA models while standing out as the only high quality approach that preserves the original data distribution.
Authors: Jianghan Zhu, Yaoxin Wu, Zhuoyi Lin, Zhengyuan Zhang, Haiyan Yin, Zhiguang Cao, Senthilnath Jayavelu, Xiaoli Li
Abstract: Recent advances in Neural Combinatorial Optimization (NCO) methods have significantly improved the capability of neural solvers to handle synthetic routing instances. Nonetheless, existing neural solvers typically struggle to generalize effectively from synthetic, uniformly-distributed training data to real-world VRP scenarios, including widely recognized benchmark instances from TSPLib and CVRPLib. To bridge this generalization gap, we present Evolutionary Realistic Instance Synthesis (EvoReal), which leverages an evolutionary module guided by large language models (LLMs) to generate synthetic instances characterized by diverse and realistic structural patterns. Specifically, the evolutionary module produces synthetic instances whose structural attributes statistically mimics those observed in authentic real-world instances. Subsequently, pre-trained NCO models are progressively refined, firstly aligning them with these structurally enriched synthetic distributions and then further adapting them through direct fine-tuning on actual benchmark instances. Extensive experimental evaluations demonstrate that EvoReal markedly improves the generalization capabilities of state-of-the-art neural solvers, yielding a notable reduced performance gap compared to the optimal solutions on the TSPLib (1.05%) and CVRPLib (2.71%) benchmarks across a broad spectrum of problem scales.
Authors: Aswin Arun, Christo Kurisummoottil Thomas, Rimalpudi Sarvendranath, Walid Saad
Abstract: Despite the advantages of multi-agent reinforcement learning (MARL) for wireless use case such as medium access control (MAC), their real-world deployment in Internet of Things (IoT) is hindered by their sample inefficiency. To alleviate this challenge, one can leverage model-based reinforcement learning (MBRL) solutions, however, conventional MBRL approaches rely on black-box models that are not interpretable and cannot reason. In contrast, in this paper, a novel causal model-based MARL framework is developed by leveraging tools from causal learn- ing. In particular, the proposed model can explicitly represent causal dependencies between network variables using structural causal models (SCMs) and attention-based inference networks. Interpretable causal models are then developed to capture how MAC control messages influence observations, how transmission actions determine outcomes, and how channel observations affect rewards. Data augmentation techniques are then used to generate synthetic rollouts using the learned causal model for policy optimization via proximal policy optimization (PPO). Analytical results demonstrate exponential sample complexity gains of causal MBRL over black-box approaches. Extensive simulations demonstrate that, on average, the proposed approach can reduce environment interactions by 58%, and yield faster convergence compared to model-free baselines. The proposed approach inherently is also shown to provide interpretable scheduling decisions via attention-based causal attribution, revealing which network conditions drive the policy. The resulting combination of sample efficiency and interpretability establishes causal MBRL as a practical approach for resource-constrained wireless systems.
Authors: Florian Ebmeier, Nicole Ludwig, Jannik Thuemmel, Georg Martius, Volker H. Franz
Abstract: Solar thermal systems (STS) present a promising avenue for low-carbon heat generation, with a well-running system providing heat at minimal cost and carbon emissions. However, STS can exhibit faults due to improper installation, maintenance, or operation, often resulting in a substantial reduction in efficiency or even damage to the system. As monitoring at the individual level is economically prohibitive for small-scale systems, automated monitoring and fault detection should be used to address such issues. Recent advances in data-driven anomaly detection, particularly in time series analysis, offer a cost-effective solution by leveraging existing sensors to identify abnormal system states. Here, we propose a probabilistic reconstruction-based framework for anomaly detection. We evaluate our method on the publicly available PaSTS dataset of operational domestic STS, which features real-world complexities and diverse fault types. Our experiments show that reconstruction-based methods can detect faults in domestic STS both qualitatively and quantitatively, while generalizing to previously unseen systems. We also demonstrate that our model outperforms both simple and more complex deep learning baselines. Additionally, we show that heteroscedastic uncertainty estimation is essential to fault detection performance. Finally, we discuss the engineering overhead required to unlock these improvements and make a case for simple deep learning models.
Authors: Patrick Feifel, Benedikt Franke, Frank Bonarens, Frank K\"oster, Arne Raulf, Friedhelm Schwenker
Abstract: Reliable pedestrian detection represents a crucial step towards automated driving systems. However, the current performance benchmarks exhibit weaknesses. The currently applied metrics for various subsets of a validation dataset prohibit a realistic performance evaluation of a DNN for pedestrian detection. As image segmentation supplies fine-grained information about a street scene, it can serve as a starting point to automatically distinguish between different types of errors during the evaluation of a pedestrian detector. In this work, eight different error categories for pedestrian detection are proposed and new metrics are proposed for performance comparison along these error categories. We use the new metrics to compare various backbones for a simplified version of the APD, and show a more fine-grained and robust way to compare models with each other especially in terms of safety-critical performance. We achieve SOTA on CityPersons-reasonable (without extra training data) by using a rather simple architecture.
Authors: Kai-Hendrik Cohrs, Zuzanna Osika, Maria Gonzalez-Calabuig, Vishal Nedungadi, Ruben Cartuyvels, Steffen Knoblauch, Joppe Massant, Shruti Nath, Patrick Ebel, Vasileios Sitokonstantinou
Abstract: Geospatial foundation models for Earth observation often fail to perform reliably in environments underrepresented during pretraining. We introduce SHRUG-FM, a framework for reliability-aware prediction that integrates three complementary signals: out-of-distribution (OOD) detection in the input space, OOD detection in the embedding space and task-specific predictive uncertainty. Applied to burn scar segmentation, SHRUG-FM shows that OOD scores correlate with lower performance in specific environmental conditions, while uncertainty-based flags help discard many poorly performing predictions. Linking these flags to land cover attributes from HydroATLAS shows that failures are not random but concentrated in certain geographies, such as low-elevation zones and large river areas, likely due to underrepresentation in pretraining data. SHRUG-FM provides a pathway toward safer and more interpretable deployment of GFMs in climate-sensitive applications, helping bridge the gap between benchmark performance and real-world reliability.
Authors: Nicolas Hoischen, Petar Bevanda, Max Beier, Stefan Sosnowski, Boris Houska, Sandra Hirche
Abstract: Continuous-time stochastic processes underlie many natural and engineered systems. In healthcare, autonomous driving, and industrial control, direct interaction with the environment is often unsafe or impractical, motivating offline reinforcement learning from historical data. However, there is limited statistical understanding of the approximation errors inherent in learning policies from offline datasets. We address this by linking reinforcement learning to the Hamilton-Jacobi-Bellman equation and proposing an operator-theoretic algorithm based on a simple dynamic programming recursion. Specifically, we represent our world model in terms of the infinitesimal generator of controlled diffusion processes learned in a reproducing kernel Hilbert space. By integrating statistical learning methods and operator theory, we establish global convergence of the value function and derive finite-sample guarantees with bounds tied to system properties such as smoothness and stability. Our theoretical and numerical results indicate that operator-based approaches may hold promise in solving offline reinforcement learning using continuous-time optimal control.
Authors: Prince Mensah, Pelumi Victor Aderinto, Ibrahim Salihu Yusuf, Arnu Pretorius
Abstract: Accurate retrieval of vegetation biophysical variables from satellite imagery is crucial for ecosystem monitoring and agricultural management. In this work, we propose a physics-informed Transformer-VAE architecture to invert the PROSAIL radiative transfer model for simultaneous estimation of key canopy parameters from Sentinel-2 data. Unlike previous hybrid approaches that require real satellite images for self-supevised training. Our model is trained exclusively on simulated data, yet achieves performance on par with state-of-the-art methods that utilize real imagery. The Transformer-VAE incorporates the PROSAIL model as a differentiable physical decoder, ensuring that inferred latent variables correspond to physically plausible leaf and canopy properties. We demonstrate retrieval of leaf area index (LAI) and canopy chlorophyll content (CCC) on real-world field datasets (FRM4Veg and BelSAR) with accuracy comparable to models trained with real Sentinel-2 data. Our method requires no in-situ labels or calibration on real images, offering a cost-effective and self-supervised solution for global vegetation monitoring. The proposed approach illustrates how integrating physical models with advanced deep networks can improve the inversion of RTMs, opening new prospects for large-scale, physically-constrained remote sensing of vegetation traits.
Authors: Francisco Cunha, Yves Lepage, Zied Bouraoui, Miguel Couceiro
Abstract: Analogical reasoning is a powerful inductive mechanism, widely used in human cognition and increasingly applied in artificial intelligence. Formal frameworks for analogical inference have been developed for Boolean domains, where inference is provably sound for affine functions and approximately correct for functions close to affine. These results have informed the design of analogy-based classifiers. However, they do not extend to regression tasks or continuous domains. In this paper, we revisit analogical inference from a foundational perspective. We first present a counterexample showing that existing generalization bounds fail even in the Boolean setting. We then introduce a unified framework for analogical reasoning in real-valued domains based on parameterized analogies defined via generalized means. This model subsumes both Boolean classification and regression, and supports analogical inference over continuous functions. We characterize the class of analogy-preserving functions in this setting and derive both worst-case and average-case error bounds under smoothness assumptions. Our results offer a general theory of analogical inference across discrete and continuous domains.
Authors: Tom Pan, Evan Dramko, Mitchell D. Miller, Anastasios Kyrillidis, George N. Phillips Jr
Abstract: Protein structure determination has long been one of the primary challenges of structural biology, to which deep machine learning (ML)-based approaches have increasingly been applied. However, these ML models generally do not incorporate the experimental measurements directly, such as X-ray crystallographic diffraction data. To this end, we explore an approach that more tightly couples these traditional crystallographic and recent ML-based methods, by training a hybrid 3-d vision transformer and convolutional network on inputs from both domains. We make use of two distinct input constructs / Patterson maps, which are directly obtainable from crystallographic data, and ``partial structure'' template maps derived from predicted structures deposited in the AlphaFold Protein Structure Database with subsequently omitted residues. With these, we predict electron density maps that are then post-processed into atomic models through standard crystallographic refinement processes. Introducing an initial dataset of small protein fragments taken from Protein Data Bank entries and placing them in hypothetical crystal settings, we demonstrate that our method is effective at both improving the phases of the crystallographic structure factors and completing the regions missing from partial structure templates, as well as improving the agreement of the electron density maps with the ground truth atomic structures.
Authors: Jonghun Lee, YongKyung Oh, Sungil Kim, Dong-Young Lim
Abstract: Neural Differential Equations (NDEs) excel at modeling continuous-time dynamics, effectively handling challenges such as irregular observations, missing values, and noise. Despite their advantages, NDEs face a fundamental challenge in adopting dropout, a cornerstone of deep learning regularization, making them susceptible to overfitting. To address this research gap, we introduce Continuum Dropout, a universally applicable regularization technique for NDEs built upon the theory of alternating renewal processes. Continuum Dropout formulates the on-off mechanism of dropout as a stochastic process that alternates between active (evolution) and inactive (paused) states in continuous time. This provides a principled approach to prevent overfitting and enhance the generalization capabilities of NDEs. Moreover, Continuum Dropout offers a structured framework to quantify predictive uncertainty via Monte Carlo sampling at test time. Through extensive experiments, we demonstrate that Continuum Dropout outperforms existing regularization methods for NDEs, achieving superior performance on various time series and image classification tasks. It also yields better-calibrated and more trustworthy probability estimates, highlighting its effectiveness for uncertainty-aware modeling.
Authors: Simon Donike, Cesar Aybar, Julio Contreras, Luis G\'omez-Chova
Abstract: We present OpenSR-SRGAN, an open and modular framework for single-image super-resolution in Earth Observation. The software provides a unified implementation of SRGAN-style models that is easy to configure, extend, and apply to multispectral satellite data such as Sentinel-2. Instead of requiring users to modify model code, OpenSR-SRGAN exposes generators, discriminators, loss functions, and training schedules through concise configuration files, making it straightforward to switch between architectures, scale factors, and band setups. The framework is designed as a practical tool and benchmark implementation rather than a state-of-the-art model. It ships with ready-to-use configurations for common remote sensing scenarios, sensible default settings for adversarial training, and built-in hooks for logging, validation, and large-scene inference. By turning GAN-based super-resolution into a configuration-driven workflow, OpenSR-SRGAN lowers the entry barrier for researchers and practitioners who wish to experiment with SRGANs, compare models in a reproducible way, and deploy super-resolution pipelines across diverse Earth-observation datasets.
Authors: Yunkai Zhang (Ryan), Qiang Zhang (Ryan), Feng (Ryan), Lin, Ruizhong Qiu, Hanchao Yu, Jason Liu, Yinglong Xia, Zhuoran Yu, Zeyu Zheng, Diji Yang
Abstract: Optimizing recommender systems for objectives beyond accuracy, such as diversity, novelty, and personalization, is crucial for long-term user satisfaction. To this end, industrial practitioners have accumulated vast amounts of structured domain knowledge, which we term human priors (e.g., item taxonomies, temporal patterns). This knowledge is typically applied through post-hoc adjustments during ranking or post-ranking. However, this approach remains decoupled from the core model learning, which is particularly undesirable as the industry shifts to end-to-end generative recommendation foundation models. On the other hand, many methods targeting these beyond-accuracy objectives often require architecture-specific modifications and discard these valuable human priors by learning user intent in a fully unsupervised manner. Instead of discarding the human priors accumulated over years of practice, we introduce a backbone-agnostic framework that seamlessly integrates these human priors directly into the end-to-end training of generative recommenders. With lightweight, prior-conditioned adapter heads inspired by efficient LLM decoding strategies, our approach guides the model to disentangle user intent along human-understandable axes (e.g., interaction types, long- vs. short-term interests). We also introduce a hierarchical composition strategy for modeling complex interactions across different prior types. Extensive experiments on three large-scale datasets demonstrate that our method significantly enhances both accuracy and beyond-accuracy objectives. We also show that human priors allow the backbone model to more effectively leverage longer context lengths and larger model sizes.
Authors: Deniz Yilmaz, Liangyu Wu, Julia Gonski
Abstract: Drift chambers have long been central to collider tracking, but future machines like a Higgs factory motivate higher granularity and cluster counting for particle ID, posing new data processing challenges. Machine learning (ML) at the "edge", or in cell-level readout, can dramatically reduce the off-detector data rate for high-granularity drift chambers by performing cluster counting at-source. We present machine learning algorithms for cluster counting in real-time readout of future drift chambers. These algorithms outperform traditional derivative-based techniques based on achievable pion-kaon separation. When synthesized to FPGA resources, they can achieve latencies consistent with real-time operation in a future Higgs factory scenario, thus advancing both R&D for future collider detectors as well as hardware-based ML for edge applications in high energy physics.
Authors: Stefano Maria Iacus, Giuseppe Porro
Abstract: Using 2.6 billion geolocated social-media posts (2014-2022) and a fine-tuned generative language model, we construct county-level indicators of life satisfaction and happiness for the United States. We document an apparent rural-urban paradox: rural counties express higher life satisfaction while urban counties exhibit greater happiness. We reconcile this by treating the two as distinct layers of subjective well-being, evaluative vs. hedonic, showing that each maps differently onto place, politics, and time. Republican-leaning areas appear more satisfied in evaluative terms, but partisan gaps in happiness largely flatten outside major metros, indicating context-dependent political effects. Temporal shocks dominate the hedonic layer: happiness falls sharply during 2020-2022, whereas life satisfaction moves more modestly. These patterns are robust across logistic and OLS specifications and align with well-being theory. Interpreted as associations for the population of social-media posts, the results show that large-scale, language-based indicators can resolve conflicting findings about the rural-urban divide by distinguishing the type of well-being expressed, offering a transparent, reproducible complement to traditional surveys.
Authors: Isabela Albuquerque, Ira Ktena, Olivia Wiles, Ivana Kaji\'c, Amal Rannen-Triki, Cristina Vasconcelos, Aida Nematzadeh
Abstract: Despite advances in generation quality, current text-to-image (T2I) models often lack diversity, generating homogeneous outputs. This work introduces a framework to address the need for robust diversity evaluation in T2I models. Our framework systematically assesses diversity by evaluating individual concepts and their relevant factors of variation. Key contributions include: (1) a novel human evaluation template for nuanced diversity assessment; (2) a curated prompt set covering diverse concepts with their identified factors of variation (e.g. prompt: An image of an apple, factor of variation: color); and (3) a methodology for comparing models in terms of human annotations via binomial tests. Furthermore, we rigorously compare various image embeddings for diversity measurement. Notably, our principled approach enables ranking of T2I models by diversity, identifying categories where they particularly struggle. This research offers a robust methodology and insights, paving the way for improvements in T2I model diversity and metric development.
Authors: Mohammadsina Almasi, Hadis Anahideh
Abstract: Equitably allocating limited resources in high-stakes domains-such as education, employment, and healthcare-requires balancing short-term utility with long-term impact, while accounting for delayed outcomes, hidden heterogeneity, and ethical constraints. However, most learning-based allocation frameworks either assume immediate feedback or ignore the complex interplay between individual characteristics and intervention dynamics. We propose a novel bi-level contextual bandit framework for individualized resource allocation under delayed feedback, designed to operate in real-world settings with dynamic populations, capacity constraints, and time-sensitive impact. At the meta level, the model optimizes subgroup-level budget allocations to satisfy fairness and operational constraints. At the base level, it identifies the most responsive individuals within each group using a neural network trained on observational data, while respecting cooldown windows and delayed treatment effects modeled via resource-specific delay kernels. By explicitly modeling temporal dynamics and feedback delays, the algorithm continually refines its policy as new data arrive, enabling more responsive and adaptive decision-making. We validate our approach on two real-world datasets from education and workforce development, showing that it achieves higher cumulative outcomes, better adapts to delay structures, and ensures equitable distribution across subgroups. Our results highlight the potential of delay-aware, data-driven decision-making systems to improve institutional policy and social welfare.
Authors: Zack Dewis, Yimin Zhu, Zhengsen Xu, Mabel Heffring, Saeid Taleghanidoozdoozan, Kaylee Xiao, Motasem Alkayid, Lincoln Linlin Xu
Abstract: Although Sentinel-2 based land use and land cover (LULC) classification is critical for various environmental monitoring applications, it is a very difficult task due to some key data challenges (e.g., spatial heterogeneity, context information, signature ambiguity). This paper presents a novel Multitask Glocal OBIA-Mamba (MSOM) for enhanced Sentinel-2 classification with the following contributions. First, an object-based image analysis (OBIA) Mamba model (OBIA-Mamba) is designed to reduce redundant computation without compromising fine-grained details by using superpixels as Mamba tokens. Second, a global-local (GLocal) dual-branch convolutional neural network (CNN)-mamba architecture is designed to jointly model local spatial detail and global contextual information. Third, a multitask optimization framework is designed to employ dual loss functions to balance local precision with global consistency. The proposed approach is tested on Sentinel-2 imagery in Alberta, Canada, in comparison with several advanced classification approaches, and the results demonstrate that the proposed approach achieves higher classification accuracy and finer details that the other state-of-the-art methods.
Authors: Benjamin L. Badger, Matthew Neligeorge
Abstract: Language prediction is constrained by informational entropy intrinsic to language, such that there exists a limit to how accurate any language model can become and equivalently a lower bound to language compression. The most efficient language compression algorithms today are causal (next token prediction) large language models, but the use of these models to form accurate estimates of language entropy is currently computationally infeasible. We introduce encoder-augmented causal decoder model architectures that exhibit superior training efficiency characteristics and achieve higher compression than causal transformers even when trained on modest hardware. We demonstrate how entropy estimates can be obtained on a per-token basis, and show that the generalization of models trained to approach the entropy of their training data necessarily exceeds the generalization of models trained to minimize loss beyond this value. We show empirically that causal models trained to approach but not exceed estimated per-token entropies exhibit greater generalization than models trained without taking entropy into account.
Authors: Haizhou Shi, Ye Liu, Bo Pang, Zeyu Leo Liu, Hao Wang, Silvio Savarese, Caiming Xiong, Yingbo Zhou, Semih Yavuz
Abstract: Large Language Models (LLMs) have demonstrated remarkable reasoning abilities, yet existing test-time frameworks often rely on coarse self-verification and self-correction, limiting their effectiveness on complex tasks. In this paper, we propose Socratic Self-Refine (SSR), a novel framework for fine-grained evaluation and precise refinement of LLM reasoning. Our proposed SSR decomposes model responses into verifiable (sub-question, sub-answer) pairs, enabling step-level confidence estimation through controlled re-solving and self-consistency checks. By pinpointing unreliable steps and iteratively refining them, SSR produces more accurate and interpretable reasoning chains. Empirical results across five reasoning benchmarks and three LLMs show that SSR consistently outperforms state-of-the-art iterative self-refinement baselines. Beyond performance gains, SSR provides a principled black-box approach for evaluating and understanding the internal reasoning processes of LLMs. Code is available at https://github.com/SalesforceAIResearch/socratic-self-refine-reasoning.
URLs: https://github.com/SalesforceAIResearch/socratic-self-refine-reasoning.
Authors: Ilyas Fatkhullin, Niao He, Guanghui Lan, Florian Wolf
Abstract: Constrained non-convex optimization is fundamentally challenging, as global solutions are generally intractable and constraint qualifications may not hold. However, in many applications, including safe policy optimization in control and reinforcement learning, such problems possess hidden convexity, meaning they can be reformulated as convex programs via a nonlinear invertible transformation. Typically such transformations are implicit or unknown, making the direct link with the convex program impossible. On the other hand, (sub-)gradients with respect to the original variables are often accessible or can be easily estimated, which motivates algorithms that operate directly in the original (non-convex) problem space using standard (sub-)gradient oracles. In this work, we develop the first algorithms to provably solve such non-convex problems to global minima. First, using a modified inexact proximal point method, we establish global last-iterate convergence guarantees with $\widetilde{\mathcal{O}}(\varepsilon^{-3})$ oracle complexity in non-smooth setting. For smooth problems, we propose a new bundle-level type method based on linearly constrained quadratic subproblems, improving the oracle complexity to $\widetilde{\mathcal{O}}(\varepsilon^{-1})$. Surprisingly, despite non-convexity, our methodology does not require any constraint qualifications, can handle hidden convex equality constraints, and achieves complexities matching those for solving unconstrained hidden convex optimization.
Authors: Edward Kim, Devan Shanker, Varun Bharadwaj, Hongbeen Park, Jinkyu Kim, Hazem Torfah, Daniel J Fremont, Sanjit A Seshia
Abstract: Simulation-based testing has become a crucial complement to road testing for ensuring the safety of cyber physical systems (CPS). As a result, significant research efforts have been directed toward identifying failure scenarios within simulation environments. However, a critical question remains. Are the AV failure scenarios discovered in simulation reproducible on actual systems in the real world? The sim-to-real gap caused by differences between simulated and real sensor data means that failure scenarios identified in simulation might either be artifacts of synthetic sensor data or actual issues that also occur with real sensor data. To address this, an effective approach to validating simulated failure scenarios is to locate occurrences of these scenarios within real-world datasets and verify whether the failure persists on the datasets. To this end, we introduce a formal definition of how labeled time series sensor data can match an abstract scenario, represented as a scenario program using the Scenic probabilistic programming language. We present a querying algorithm that, given a scenario program and a labeled dataset, identifies the subset of data that matches the specified scenario. Our experiment shows that our algorithm is more accurate and orders of magnitude faster in querying scenarios than the state-of-the-art commercial vision large language models, and can scale with the duration of queried time series data.
Authors: Jiang Liu, Jialian Wu, Xiaodong Yu, Yusheng Su, Prakamya Mishra, Gowtham Ramesh, Sudhanshu Ranjan, Chaitanya Manem, Ximeng Sun, Ze Wang, Pratik Prabhanjan Brahma, Zicheng Liu, Emad Barsoum
Abstract: Large language models (LLMs) have demonstrated remarkable performance across a wide range of tasks, yet the majority of high-performing models remain closed-source or partially open, limiting transparency and reproducibility. In this work, we introduce Instella, a family of fully open three billion parameter language models trained entirely on openly available data and codebase. Powered by AMD Instinct MI300X GPUs, Instella is developed through large-scale pre-training, general-purpose instruction tuning, and alignment with human preferences. Despite using substantially fewer pre-training tokens than many contemporaries, Instella achieves state-of-the-art results among fully open models and is competitive with leading open-weight models of comparable size. We further release two specialized variants: Instella-Long, capable of handling context lengths up to 128K tokens, and Instella-Math, a reasoning-focused model enhanced through supervised fine-tuning and reinforcement learning on mathematical tasks. Together, these contributions establish Instella as a transparent, performant, and versatile alternative for the community, advancing the goal of open and reproducible language modeling research.
Authors: Pascal Strauch, David M\"uller, Sammy Christen, Agon Serifi, Ruben Grandia, Espen Knoop, Moritz B\"acher
Abstract: Despite recent advances in robust locomotion, bipedal robots operating in the real world remain at risk of falling. While most research focuses on preventing such events, we instead concentrate on the phenomenon of falling itself. Specifically, we aim to reduce physical damage to the robot while providing users with control over a robot's end pose. To this end, we propose a robot agnostic reward function that balances the achievement of a desired end pose with impact minimization and the protection of critical robot parts during reinforcement learning. To make the policy robust to a broad range of initial falling conditions and to enable the specification of an arbitrary and unseen end pose at inference time, we introduce a simulation-based sampling strategy of initial and end poses. Through simulated and real-world experiments, our work demonstrates that even bipedal robots can perform controlled, soft falls.
Authors: Adrienne Tuynman, Ronald Ortner
Abstract: We present an approach for the quantification of the usefulness of transfer in reinforcement learning via regret bounds for a multi-agent setting. Considering a number of $\aleph$ agents operating in the same Markov decision process, however possibly with different reward functions, we consider the regret each agent suffers with respect to an optimal policy maximizing her average reward. We show that when the agents share their observations the total regret of all agents is smaller by a factor of $\sqrt{\aleph}$ compared to the case when each agent has to rely on the information collected by herself. This result demonstrates how considering the regret in multi-agent settings can provide theoretical bounds on the benefit of sharing observations in transfer learning.
Authors: Ali Muzaffar, Hani Ragab Hassen, Hind Zantout, Michael A Lones
Abstract: We report the findings of a reimplementation of 18 foundational studies in feature-based machine learning for Android malware detection, published during the period 2013-2023. These studies are reevaluated on a level playing field using a contemporary Android environment and a balanced dataset of 124,000 applications. Our findings show that feature-based approaches can still achieve detection accuracies beyond 98%, despite a considerable increase in the size of the underlying Android feature sets. We observe that features derived through dynamic analysis yield only a small benefit over those derived from static analysis, and that simpler models often out-perform more complex models. We also find that API calls and opcodes are the most productive static features within our evaluation context, network traffic is the most predictive dynamic feature, and that ensemble models provide an efficient means of combining models trained on static and dynamic features. Together, these findings suggest that simple, fast machine learning approaches can still be an effective basis for malware detection, despite the increasing focus on slower, more expensive machine learning models in the literature.
Authors: Vasilis Gkolemis, Christos Diou, Dimitris Kyriakopoulos, Konstantinos Tsopelas, Julia Herbinger, Hubert Baniecki, Dimitrios Rontogiannis, Loukas Kavouras, Maximilian Muschalik, Theodore Dalamagas, Eirini Ntoutsi, Bernd Bischl, Giuseppe Casalicchio
Abstract: Effector is a Python package for interpreting machine learning (ML) models that are trained on tabular data through global and regional feature effects. Global effects, like Partial Dependence Plot (PDP) and Accumulated Local Effects (ALE), are widely used for explaining tabular ML models due to their simplicity -- each feature's average influence on the prediction is summarized by a single 1D plot. However, when features are interacting, global effects can be misleading. Regional effects address this by partitioning the input space into disjoint subregions with minimal interactions within each and computing a separate regional effect per subspace. Regional effects are then visualized by a set of 1D plots per feature. Effector provides efficient implementations of state-of-the-art global and regional feature effects methods under a unified API. The package integrates seamlessly with major ML libraries like scikit-learn and PyTorch. It is designed to be modular and extensible, and comes with comprehensive documentation and tutorials. Effector is an open-source project publicly available on Github at https://github.com/givasile/effector.
Authors: Xulin Chen, Ruipeng Liu, Zhenyu Gan, Garrett E. Katz
Abstract: Uncertainties in transition dynamics pose a critical challenge in reinforcement learning (RL), often resulting in performance degradation of trained policies when deployed on hardware. Many robust RL approaches follow two strategies: enforcing smoothness in actor or actor-critic modules with Lipschitz regularization, or learning robust Bellman operators. However, the first strategy does not investigate the impact of critic-only Lipschitz regularization on policy robustness, while the second lacks comprehensive validation in real-world scenarios. Building on this gap and prior work, we propose PPO-PGDLC, an algorithm based on Proximal Policy Optimization (PPO) that integrates Projected Gradient Descent (PGD) with a Lipschitz-regularized critic (LC). The PGD component calculates the adversarial state within an uncertainty set to approximate the robust Bellman operator, and the Lipschitz-regularized critic further improves the smoothness of learned policies. Experimental results on two classic control tasks and one real-world robotic locomotion task demonstrates that, compared to several baseline algorithms, PPO-PGDLC achieves better performance and predicts smoother actions under environmental perturbations.
Authors: Arnab Bhattacharyya, Sutanu Gayen, Philips George John, Sayantan Sen, N. V. Vinodchandran
Abstract: This work establishes a novel link between the problem of PAC-learning high-dimensional graphical models and the task of (efficient) counting and sampling of graph structures, using an online learning framework. We observe that if we apply the exponentially weighted average (EWA) or randomized weighted majority (RWM) forecasters on a sequence of samples from a distribution P using the log loss function, the average regret incurred by the forecaster's predictions can be used to bound the expected KL divergence between P and the predictions. Known regret bounds for EWA and RWM then yield new sample complexity bounds for learning Bayes nets. Moreover, these algorithms can be made computationally efficient for several interesting classes of Bayes nets. Specifically, we give a new sample-optimal and polynomial time learning algorithm with respect to trees of unknown structure and the first polynomial sample and time algorithm for learning with respect to Bayes nets over a given chordal skeleton.
Authors: Indu Solomon, Aye Phyu Phyu Aung, Uttam Kumar, Senthilnath Jayavelu
Abstract: Continual learning (CL) enables models to adapt to evolving data streams without catastrophic forgetting, a fundamental requirement for real-world AI systems. However, the current methods often depend on large replay buffers or heavily annotated datasets which are impractical due to storage, privacy, and cost constraints. We propose CLTS (Continual Learning via Text-Image Synergy), a novel class-incremental framework that mitigates forgetting without storing real task data. CLTS leverages pre-trained vision-language models, BLIP (Bootstrapping Language-Image Pre-training) for caption generation and stable diffusion for sample generation. Each task is handled by a dedicated Task Head, while a Task Router learns to assign inputs to the correct Task Head using the generated data. On three benchmark datasets, CLTS improves average task accuracy by up to 54% and achieves 63 times better memory efficiency compared to four recent continual learning baselines, demonstrating improved retention and adaptability. CLTS introduces a novel perspective by integrating generative text-image augmentation for scalable continual learning.
Authors: Changhao Li, Yuchen Zhuang, Rushi Qiang, Haotian Sun, Hanjun Dai, Chao Zhang, Bo Dai
Abstract: Despite the impressive generative abilities of black-box large language models (LLMs), their inherent opacity hinders further advancements in capabilities such as reasoning, planning, and personalization. Existing works aim to enhance LLM capabilities via domain-specific adaptation, which require additional training on accessible model parameters, an infeasible option for black-box LLMs. To address this challenge, we introduce Matryoshka Pilot (M-Pilot), a lightweight white-box LLM controller that guides a large-scale black-box LLM generator by decomposing complex tasks into a series of intermediate outputs. Specifically, we consider the black-box LLM as an environment, with M-Pilot serving as a policy to provide intermediate guidance through prompts for driving the black-box LLM. M-Pilot is trained to pivot the outputs of the black-box LLM aligning with preferences during iterative interaction, which enables controllable multi-turn generation and self-improvement in optimizing intermediate guidance. Empirical evaluations on diverse tasks demonstrate that our method effectively enhances the capabilities of black-box LLMs in complex, long-horizon tasks. Our code is publicly available at: https://github.com/lichangh20/Matryoshka.
Authors: Ousmane Touat, Jezekael Brunon, Yacine Belal, Julien Nicolas, C\'esar Sabater, Mohamed Maouche, Sonia Ben Mokhtar
Abstract: The primary promise of decentralized learning is to allow users to engage in the training of machine learning models in a collaborative manner while keeping their data on their premises and without relying on any central entity. However, this paradigm necessitates the exchange of model parameters or gradients between peers. Such exchanges can be exploited to infer sensitive information about training data, which is achieved through privacy attacks (e.g., Membership Inference Attacks -- MIA). In order to devise effective defense mechanisms, it is important to understand the factors that increase/reduce the vulnerability of a given decentralized learning architecture to MIA. In this study, we extensively explore the vulnerability to MIA of various decentralized learning architectures by varying the graph structure (e.g., number of neighbors), the graph dynamics, and the aggregation strategy, across diverse datasets and data distributions. Our key finding, which to the best of our knowledge we are the first to report, is that the vulnerability to MIA is heavily correlated to (i) the local model mixing strategy performed by each node upon reception of models from neighboring nodes and (ii) the global mixing properties of the communication graph. We illustrate these results experimentally using four datasets and by theoretically analyzing the mixing properties of various decentralized architectures. We also empirically show that enhancing mixing properties is highly beneficial when combined with other privacy-preserving techniques such as Differential Privacy. Our paper draws a set of lessons learned for devising decentralized learning systems that reduce by design the vulnerability to MIA.
Authors: Zaikang Lin, Sei Chang, Aaron Zweig, Minseo Kang, Elham Azizi, David A. Knowles
Abstract: Modern high-throughput biological datasets with thousands of perturbations provide the opportunity for large-scale discovery of causal graphs that represent the regulatory interactions between genes. Differentiable causal graphical models have been proposed to infer a gene regulatory network (GRN) from large scale interventional datasets, capturing the causal gene regulatory relationships from genetic perturbations. However, existing models are limited in their expressivity and scalability while failing to address the dynamic nature of biological processes such as cellular differentiation. We propose PerturbODE, a novel framework that incorporates biologically informative neural ordinary differential equations (neural ODEs) to model cell state trajectories under perturbations and derive the causal GRN from the neural ODE's parameters. We demonstrate PerturbODE's efficacy in trajectory prediction and GRN inference across simulated and real over-expression datasets.
Authors: Zhuoxuan Liang, Wei Li, Dalin Zhang, Ziyu Jia, Yidan Chen, Zhihong Wang, Xiangping Zheng, Moustafa Youssef
Abstract: With the rapid growth of the Internet of Things and Cyber-Physical Systems, widespread sensor deployment has become essential. However, the high costs of building sensor networks limit their scale and coverage, making fine-grained deployment challenging. Inductive Spatio-Temporal Kriging (ISK) addresses this issue by introducing virtual sensors. Based on graph neural networks (GNNs) extracting the relationships between physical and virtual sensors, ISK can infer the measurements of virtual sensors from physical sensors. However, current ISK methods rely on conventional message-passing mechanisms and network architectures, without effectively extracting spatio-temporal features of physical sensors and focusing on representing virtual sensors. Additionally, existing graph construction methods face issues of sparse and noisy connections, destroying ISK performance. To address these issues, we propose DarkFarseer, a novel ISK framework with three key components. First, we propose the Neighbor Hidden Style Enhancement module with a style transfer strategy to enhance the representation of virtual nodes in a temporal-then-spatial manner to better extract the spatial relationships between physical and virtual nodes. Second, we propose Virtual-Component Contrastive Learning, which aims to enrich the node representation by establishing the association between the patterns of virtual nodes and the regional patterns within graph components. Lastly, we design a Similarity-Based Graph Denoising Strategy, which reduces the connectivity strength of noisy connections around virtual nodes and their neighbors based on their temporal information and regional spatial patterns. Extensive experiments demonstrate that DarkFarseer significantly outperforms existing ISK methods.
Authors: Shenglong Zhou, Ouya Wang, Ziyan Luo, Yongxu Zhu, Geoffrey Ye Li
Abstract: The recent advancement of foundation models (FMs) has brought about a paradigm shift, revolutionizing various sectors worldwide. The popular optimizers used to train these models are stochastic gradient descent-based algorithms, which face inherent limitations, such as slow convergence and stringent assumptions for convergence. In particular, data heterogeneity arising from distributed settings poses significant challenges to their theoretical and numerical performance. This paper develops an algorithm, PISA (Preconditioned Inexact Stochastic Alternating Direction Method of Multipliers). Grounded in rigorous theoretical guarantees, the algorithm converges under the sole assumption of Lipschitz continuity of the gradient on a bounded region, thereby removing the need for other conditions commonly imposed by stochastic methods. This capability enables the proposed algorithm to tackle the challenge of data heterogeneity effectively. Moreover, the algorithmic architecture enables scalable parallel computing and supports various preconditions, such as second-order information, second moment, and orthogonalized momentum by Newton-Schulz iterations. Incorporating the latter two preconditions in PISA yields two computationally efficient variants: SISA and NSISA. Comprehensive experimental evaluations for training or fine-tuning diverse deep models, including vision models, large language models, reinforcement learning models, generative adversarial networks, and recurrent neural networks, demonstrate superior numerical performance of SISA and NSISA compared to various state-of-the-art optimizers.
Authors: Kanishk Bhatia, Felix Koehler, Nils Thuerey
Abstract: The physics solvers employed for neural network training are primarily iterative, and hence, differentiating through them introduces a severe computational burden as iterations grow large. Inspired by works in bilevel optimization, we show that full accuracy of the network is achievable through physics significantly coarser than fully converged solvers. We propose Progressively Refined Differentiable Physics (PRDP), an approach that identifies the level of physics refinement sufficient for full training accuracy. By beginning with coarse physics, adaptively refining it during training, and stopping refinement at the level adequate for training, it enables significant compute savings without sacrificing network accuracy. Our focus is on differentiating iterative linear solvers for sparsely discretized differential operators, which are fundamental to scientific computing. PRDP is applicable to both unrolled and implicit differentiation. We validate its performance on a variety of learning scenarios involving differentiable physics solvers such as inverse problems, autoregressive neural emulators, and correction-based neural-hybrid solvers. In the challenging example of emulating the Navier-Stokes equations, we reduce training time by 62%.
Authors: Murong Yang, Shihui Ying, Yue Gao, Xin-Jian Xu
Abstract: Although hypergraph neural networks (HGNNs) have emerged as a powerful framework for analyzing complex datasets, their practical performance often remains limited. On one hand, existing networks typically employ a single type of attention mechanism, focusing on either structural or feature similarities during message passing. On the other hand, assuming that all nodes in current hypergraph models have the same level of overlap may lead to suboptimal generalization. To overcome these limitations, we propose a novel framework, overlap-aware meta-learning attention for hypergraph neural networks (OMA-HGNN). First, we introduce a hypergraph attention mechanism that integrates both structural and feature similarities. Specifically, we linearly combine their respective losses with weighted factors for the HGNN model. Second, we partition nodes into different tasks based on their diverse overlap levels and develop a multi-task Meta-Weight-Net (MWN) to determine the corresponding weighted factors. Third, we jointly train the internal MWN model with the losses from the external HGNN model and train the external model with the weighted factors from the internal model. To evaluate the effectiveness of OMA-HGNN, we conducted experiments on six real-world datasets and benchmarked its perfor-mance against nine state-of-the-art methods for node classification. The results demonstrate that OMA-HGNN excels in learning superior node representations and outperforms these baselines.
Authors: Jonas Elsborg, Luca Thiede, Al\'an Aspuru-Guzik, Tejs Vegge, Arghya Bhowmik
Abstract: We present the Electronic Tensor Reconstruction Algorithm (ELECTRA) - an equivariant model for predicting electronic charge densities using floating orbitals. Floating orbitals are a long-standing concept in the quantum chemistry community that promises more compact and accurate representations by placing orbitals freely in space, as opposed to centering all orbitals at the position of atoms. Finding the ideal placement of these orbitals requires extensive domain knowledge, though, which thus far has prevented widespread adoption. We solve this in a data-driven manner by training a Cartesian tensor network to predict the orbital positions along with orbital coefficients. This is made possible through a symmetry-breaking mechanism that is used to learn position displacements with lower symmetry than the input molecule while preserving the rotation equivariance of the charge density itself. Inspired by recent successes of Gaussian Splatting in representing densities in space, we are using Gaussian orbitals and predicting their weights and covariance matrices. Our method achieves a state-of-the-art balance between computational efficiency and predictive accuracy on established benchmarks. Furthermore, ELECTRA is able to lower the compute time required to arrive at converged DFT solutions - initializing calculations using our predicted densities yields an average 50.72 \% reduction in self-consistent field (SCF) iterations on unseen molecules.
Authors: Linjian Meng, Tianpei Yang, Youzhi Zhang, Zhenxing Ge, Yang Gao
Abstract: Counterfactual Regret Minimization (CFR) algorithms are widely used to compute a Nash equilibrium (NE) in two-player zero-sum imperfect-information extensive-form games (IIGs). Among them, Predictive CFR$^+$ (PCFR$^+$) is particularly powerful, achieving an exceptionally fast empirical convergence rate via the prediction in many games.However, the empirical convergence rate of PCFR$^+$ would significantly degrade if the prediction is inaccurate, leading to unstable performance on certain IIGs. To enhance the robustness of PCFR$^+$, we propose Asymmetric PCFR$^+$ (APCFR$^+$), which employs an adaptive asymmetry of step sizes between the updates of implicit and explicit accumulated counterfactual regrets to mitigate the impact of the prediction inaccuracy on convergence. We present a theoretical analysis demonstrating why APCFR$^+$ can enhance the robustness. To the best of our knowledge, we are the first to propose the asymmetry of step sizes, a simple yet novel technique that effectively improves the robustness of PCFR$^+$. Then, to reduce the difficulty of implementing APCFR$^+$ caused by the adaptive asymmetry, we propose a simplified version of APCFR$^+$ called Simple APCFR$^+$ (SAPCFR$^+$), which uses a fixed asymmetry of step sizes to enable only a single-line modification compared to original PCFR$^+$.Experimental results on five standard IIG benchmarks and two heads-up no-limit Texas Hold' em (HUNL) Subagems show that (i) both APCFR$^+$ and SAPCFR$^+$ outperform PCFR$^+$ in most of the tested games, (ii) SAPCFR$^+$ achieves a comparable empirical convergence rate with APCFR$^+$,and (iii) our approach can be generalized to improve other CFR algorithms, e.g., Discount CFR (DCFR).
Authors: Selim Jerad, Anej Svete, Jiaoda Li, Ryan Cotterell
Abstract: Understanding the expressive power of transformers has recently attracted attention, as it offers insights into their abilities and limitations. Many studies analyze unique hard attention transformers, where attention selects a single position that maximizes the attention scores. When multiple positions achieve the maximum score, either the rightmost or the leftmost of those is chosen. In this paper, we highlight the importance of this seeming triviality. Recently, finite-precision transformers with both leftmost- and rightmost-hard attention were shown to be equivalent to Linear Temporal Logic (LTL). We show that this no longer holds with only leftmost-hard attention -- in that case, they correspond to a \emph{strictly weaker} fragment of LTL. Furthermore, we show that models with leftmost-hard attention are equivalent to \emph{soft} attention, suggesting they may better approximate real-world transformers than right-attention models. These findings refine the landscape of transformer expressivity and underscore the role of attention directionality.
Authors: Alessio Micheli, Alejandro Moreo, Marco Podda, Fabrizio Sebastiani, William Simoni, Domenico Tortorella
Abstract: Network quantification (NQ) is the problem of estimating the proportions of nodes belonging to each class in subsets of unlabelled graph nodes. When prior probability shift is at play, this task cannot be effectively addressed by first classifying the nodes and then counting the class predictions. In addition, unlike non-relational quantification, NQ demands enhanced flexibility in order to capture a broad range of connectivity patterns, resilience to the challenge of heterophily, and scalability to large networks. In order to meet these stringent requirements, we introduce XNQ, a novel method that synergizes the flexibility and efficiency of the unsupervised node embeddings computed by randomized recursive Graph Neural Networks, with an Expectation-Maximization algorithm that provides a robust quantification-aware adjustment to the output probabilities of a calibrated node classifier. In an extensive evaluation, in which we also validate the design choices underpinning XNQ through comprehensive ablation experiments, we find that XNQ consistently and significantly improves on the best network quantification methods to date, thereby setting the new state of the art for this challenging task. XNQ also provides a training speed-up of up to 10x-100x over other methods based on graph learning.
Authors: Bruno Jacob, Ashish S. Nair, Amanda A. Howard, Jan Drgona, Panos Stinis
Abstract: Physics-informed neural networks (PINNs) have demonstrated promise as a framework for solving forward and inverse problems involving partial differential equations. Despite recent progress in the field, it remains challenging to quantify uncertainty in these networks. While techniques such as Bayesian PINNs (B-PINNs) provide a principled approach to capturing epistemic uncertainty through Bayesian inference, they can be computationally expensive for large-scale applications. In this work, we propose Epistemic Physics-Informed Neural Networks (E-PINNs), a framework that uses a small network, the epinet, to efficiently quantify epistemic uncertainty in PINNs. The proposed approach works as an add-on to existing, pre-trained PINNs with a small computational overhead. We demonstrate the applicability of the proposed framework in various test cases and compare the results with B-PINNs using Hamiltonian Monte Carlo (HMC) posterior estimation and dropout-equipped PINNs (Dropout-PINNs). In our experiments, E-PINNs achieve calibrated coverage with competitive sharpness at substantially lower cost. We demonstrate that when B-PINNs produce narrower bands, they under-cover in our tests. E-PINNs also show better calibration than Dropout-PINNs in these examples, indicating a favorable accuracy-efficiency trade-off.
Authors: Ivan Svetunkov, Anna Sroginis
Abstract: Effective demand forecasting is critical for inventory management, production planning, and decision making across industries. Selecting the appropriate model and suitable features to efficiently capture patterns in the data is one of the main challenges in demand forecasting. In reality, this becomes even more complicated when the recorded sales have zeroes, which can happen naturally or due to some anomalies, such as stockouts and recording errors. Mistreating the zeroes can lead to the application of inappropriate forecasting methods, and thus leading to poor decision making. Furthermore, the demand itself can have different fundamental characteristics, and being able to distinguish one type from another might bring substantial benefits in terms of accuracy and thus decision making. We propose a two-stage model-based classification framework that in the first step, identifies artificially occurring zeroes, and in the second, classifies demand to one of the possible types: regular/intermittent, intermittent smooth/lumpy, fractional/count. The framework relies on statistical modelling and information criteria. We argue that different types of demand need different features, and show empirically that they tend to increase the accuracy of the forecasting methods and reduce inventory costs compared to those applied directly to the dataset without the generated features and the two-stage framework.
Authors: Lucas N. Alegre, Ana L. C. Bazzan, Andr\'e Barreto, Bruno C. da Silva
Abstract: Multi-task reinforcement learning aims to quickly identify solutions for new tasks with minimal or no additional interaction with the environment. Generalized Policy Improvement (GPI) addresses this by combining a set of base policies to produce a new one that is at least as good -- though not necessarily optimal -- as any individual base policy. Optimality can be ensured, particularly in the linear-reward case, via techniques that compute a Convex Coverage Set (CCS). However, these are computationally expensive and do not scale to complex domains. The Option Keyboard (OK) improves upon GPI by producing policies that are at least as good -- and often better. It achieves this through a learned meta-policy that dynamically combines base policies. However, its performance critically depends on the choice of base policies. This raises a key question: is there an optimal set of base policies -- an optimal behavior basis -- that enables zero-shot identification of optimal solutions for any linear tasks? We solve this open problem by introducing a novel method that efficiently constructs such an optimal behavior basis. We show that it significantly reduces the number of base policies needed to ensure optimality in new tasks. We also prove that it is strictly more expressive than a CCS, enabling particular classes of non-linear tasks to be solved optimally. We empirically evaluate our technique in challenging domains and show that it outperforms state-of-the-art approaches, increasingly so as task complexity increases.
Authors: Nikolaos Louloudakis, Ajitha Rajan
Abstract: With over 760 stars on GitHub and being part of the official ONNX repository, the ONNX Optimizer is the default tool for applying graph-based optimizations to ONNX models. Despite its widespread use, its ability to maintain model accuracy during optimization has not been thoroughly investigated. In this work, we present OODTE, a utility designed to automatically and comprehensively evaluate the correctness of the ONNX Optimizer. OODTE adopts a straightforward yet powerful differential testing and evaluation methodology, which can be readily adapted for use with other compiler optimizers. Specifically, OODTE takes a collection of ONNX models, applies optimizations, and executes both the original and optimized versions across a user-defined input set, automatically capturing any issues encountered during optimization. When discrepancies in accuracy arise, OODTE iteratively isolates the responsible optimization pass by repeating the process at a finer granularity. We applied OODTE to 130 well-known models from the official ONNX Model Hub, spanning diverse tasks including classification, object detection, semantic segmentation, text summarization, question answering, and sentiment analysis. Our evaluation revealed that 9.2% of the model instances either caused the optimizer to crash or led to the generation of invalid models using default optimization strategies. Additionally, 30% of classification models and 16.6% of object detection and segmentation models exhibited differing outputs across original and optimized versions, whereas models focused on text-related tasks were generally robust to optimization. OODTE uncovered 15 issues-14 previously unknown-affecting 9 of 47 optimization passes and the optimizer overall. All issues were reported to the ONNX Optimizer team. OODTE offers a simple but effective framework for validating AI model optimizers, applicable beyond the ONNX ecosystem.
Authors: Matthew Raffel, Lizhong Chen
Abstract: The Kolmogorov-Arnold Network (KAN) has been gaining popularity as an alternative to the multi-layer perceptron (MLP) with its increased expressiveness and interpretability. Even so, the KAN suffers from being orders of magnitude slower due to its increased computational cost and training instability, limiting its applicability to larger-scale tasks. Recently, the Kolmogorov-Arnold Transformer (KAT) has been proposed, which can achieve FLOPs similar to the traditional Transformer with MLPs by leveraging Group-Rational KAN (GR-KAN). Unfortunately, despite the comparable FLOPs, our testing reveals that the KAT is still 123x slower in training speeds, indicating that there are other performance bottlenecks beyond FLOPs. In this paper, we conduct a series of experiments to understand the root cause of the slowdown in KAT. We uncover that the slowdown can be isolated to memory stalls, linked more specifically to inefficient gradient accumulations in the backward pass of GR-KAN. To address this memory bottleneck, we propose FlashKAT, which minimizes accesses to slow memory and the usage of atomic adds through a restructured kernel. Evaluations demonstrate that FlashKAT can achieve a training speedup of 86.5x compared with the state-of-the-art KAT, while reducing rounding errors in the computation of the gradients.
Authors: Kabir V. Dabholkar, Omri Barak
Abstract: Many natural systems, including neural circuits involved in decision making, are modeled as high-dimensional dynamical systems with multiple stable states. While existing analytical tools primarily describe behavior near stable equilibria, characterizing separatrices--the manifolds that delineate boundaries between different basins of attraction--remains challenging, particularly in high-dimensional settings. Here, we introduce a numerical framework leveraging Koopman Theory combined with Deep Neural Networks to effectively characterize separatrices. Specifically, we approximate Koopman Eigenfunctions (KEFs) associated with real positive eigenvalues, which vanish precisely at the separatrices. Utilizing these scalar KEFs, optimization methods efficiently locate separatrices even in complex systems. We demonstrate our approach on synthetic benchmarks, ecological network models, and high-dimensional recurrent neural networks trained on either neuroscience-inspired tasks or fit to real neural data. Moreover, we illustrate the practical utility of our method by designing optimal perturbations that can shift systems across separatrices, enabling predictions relevant to optogenetic stimulation experiments in neuroscience.
Authors: Shuze Chen, Tianyi Peng
Abstract: Value decomposition has long been a fundamental technique in multi-agent dynamic programming and reinforcement learning (RL). Specifically, the value function of a global state $(s_1,s_2,\ldots,s_N)$ is often approximated as the sum of local functions: $V(s_1,s_2,\ldots,s_N)\approx\sum_{i=1}^N V_i(s_i)$. This approach traces back to the index policy in restless multi-armed bandit problems and has found various applications in modern RL systems. However, the theoretical justification for why this decomposition works so effectively remains underexplored. In this paper, we uncover the underlying mathematical structure that enables value decomposition. We demonstrate that a multi-agent Markov decision process (MDP) permits value decomposition if and only if its transition matrix is not "entangled" -- a concept analogous to quantum entanglement in quantum physics. Drawing inspiration from how physicists measure quantum entanglement, we introduce how to measure the "Markov entanglement" for multi-agent MDPs and show that this measure can be used to bound the decomposition error in general multi-agent MDPs. Using the concept of Markov entanglement, we proved that a widely-used class of index policies is weakly entangled and enjoys a sublinear $\mathcal O(\sqrt{N})$ scale of decomposition error for $N$-agent systems. Finally, we show how Markov entanglement can be efficiently estimated in practice, providing practitioners with an empirical proxy for the quality of value decomposition.
Authors: Ruokai Yin, Yuhang Li, Donghyun Lee, Priyadarshini Panda
Abstract: Large language models (LLMs) deliver strong performance but are difficult to deploy due to high memory and compute costs. While pruning reduces these demands, most methods ignore activation sparsity observed at runtime. We reinterpret activation sparsity as dynamic structured weight sparsity and propose DuoGPT, a unified framework that constructs dual-sparse (spMspV) workloads by combining unstructured weight pruning with activation sparsity. To preserve accuracy, we extend the Optimal Brain Compression (OBC) framework with activation-aware calibration and introduce output residuals from the dense model as correction terms. We further optimize the solution for efficient GPU execution, enabling scalability to billion-parameter LLMs. Evaluations on LLaMA-2 and LLaMA-3 show that DuoGPT outperforms state-of-the-art structured pruning methods by up to 9.17% accuracy at an iso-speedup of 1.39$\times$ compared to the baseline dense model. Code is available at Github.
Authors: Alexander Theus, Alessandro Cabodi, Sotiris Anagnostidis, Antonio Orvieto, Sidak Pal Singh, Valentina Boeva
Abstract: Understanding the geometry of neural network loss landscapes is a central question in deep learning, with implications for generalization and optimization. A striking phenomenon is linear mode connectivity (LMC), where independently trained models can be connected by low- or zero-loss paths despite appearing to lie in separate loss basins. However, this is often obscured by symmetries in parameter space -- such as neuron permutations -- which make functionally equivalent models appear dissimilar. Prior work has predominantly focused on neuron reordering through permutations, but such approaches are limited in scope and fail to capture the richer symmetries exhibited by modern architectures such as Transformers. In this work, we introduce a unified framework that captures four symmetry classes -- permutations, semi-permutations, orthogonal transformations, and general invertible maps -- broadening the set of valid reparameterizations and subsuming many previous approaches as special cases. Crucially, this generalization enables, for the first time, the discovery of low- and zero-barrier linear interpolation paths between independently trained Vision Transformers and GPT-2 models. Furthermore, our framework extends beyond pairwise alignment to multi-model and width-heterogeneous settings, enabling alignment across architectures of different sizes. These results reveal deeper structure in the loss landscape and underscore the importance of symmetry-aware analysis for understanding model space geometry.
Authors: Kamil Faber, Marcin Pietro\'n, Dominik \.Zurek, Roberto Corizzo
Abstract: The recently proposed xLSTM is a powerful model that leverages expressive multiplicative gating and residual connections, providing the temporal capacity needed for long-horizon forecasting and representation learning. This architecture has demonstrated success in time series forecasting, lossless compression, and even large-scale language modeling tasks, where its linear memory footprint and fast inference make it a viable alternative to Transformers. Despite its growing popularity, no prior work has explored xLSTM for anomaly detection. In this work, we fill this gap by proposing xLSTMAD, the first anomaly detection method that integrates a full encoder-decoder xLSTM architecture, purpose-built for multivariate time series data. Our encoder processes input sequences to capture historical context, while the decoder is devised in two separate variants of the method. In the forecasting approach, the decoder iteratively generates forecasted future values xLSTMAD-F, while the reconstruction approach reconstructs the input time series from its encoded counterpart xLSTMAD-R. We investigate the performance of two loss functions: Mean Squared Error (MSE), and Soft Dynamic Time Warping (SoftDTW) to consider local reconstruction fidelity and global sequence alignment, respectively. We evaluate our method on the comprehensive TSB-AD-M benchmark, which spans 17 real-world datasets, using state-of-the-art challenging metrics such as VUS-PR. In our results, xLSTM showcases state-of-the-art accuracy, outperforming 23 popular anomaly detection baselines. Our paper is the first work revealing the powerful modeling capabilities of xLSTM for anomaly detection, paving the way for exciting new developments on this subject. Our code is available at: https://github.com/Nyderx/xlstmad
Authors: Aoming Liu, Reuben Tan, Boqing Gong, Bryan A. Plummer
Abstract: Token reduction accelerates Multimodal Large Language Models (MLLMs) by reducing excessive tokens, but overlooks structural redundancy differences, where critical and redundant modules process identical token loads. For fine-grained computation control, we define an ``operation" as the computation for a module to process a group of tokens and introduce the operation pruning framework to enable modules to selectively process tokens. Built on this framework, we propose Depth-wise Operation Pruning (DOP), a data-driven method that searches for strategies to prune redundant operations and save computational budget for critical modules to process more tokens than uniform allocation by minimizing divergence from the original model's output probability distribution on a small validation set while satisfying computational constraints. For efficient optimization, DOP applies depth-wise pruning to reduce policy space and uses an additive approximation to minimize required validation runs. Depth-wise pruning partitions operations by module type and token group, and prunes operations in deeper layers before those in shallower layers within each module-group pair. The additive approximation obtains individual divergences by independently varying each policy parameter, and then sums them to approximate the joint divergence of simultaneously changing all policy parameters, reducing required validation runs from exponential to linear with respect to the number of policy parameters. Comprehensive evaluations show that DOP establishes new state-of-the-art performance across 6 MLLMs and 13 benchmarks against 12 baselines. On LLaVA-Next-7B, DOP achieves 86\% TFLOPS reduction and 83\% latency reduction on real GPU with only 1\% performance loss. Our extensive ablation studies further demonstrate DOP's data and time efficiency as well as strong generalization capabilities.
Authors: Ando Shah, Rajveer Singh, Akram Zaytar, Girmaw Abebe Tadesse, Caleb Robinson, Negar Tafti, Stephen A. Wood, Rahul Dodhia, Juan M. Lavista Ferres
Abstract: Rice cultivation supplies half the world's population with staple food, while also being a major driver of freshwater depletion--consuming roughly a quarter of global freshwater--and accounting for approx. 48% of greenhouse gas emissions from croplands. In regions like Punjab, India, where groundwater levels are plummeting at 41.6 cm/year, adopting water-saving rice farming practices is critical. Direct-Seeded Rice (DSR) and Alternate Wetting and Drying (AWD) can cut irrigation water use by 20-40% without hurting yields, yet lack of spatial data on adoption impedes effective adaptation policy and climate action. We present a machine learning framework to bridge this data gap by monitoring sustainable rice farming at scale. In collaboration with agronomy experts and a large-scale farmer training program, we obtained ground-truth data from 1,400 fields across Punjab. Leveraging this partnership, we developed a novel dimensional classification approach that decouples sowing and irrigation practices, achieving F1 scores of 0.8 and 0.74 respectively, solely employing Sentinel-1 satellite imagery. Explainability analysis reveals that DSR classification is robust while AWD classification depends primarily on planting schedule differences, as Sentinel-1's 12-day revisit frequency cannot capture the higher frequency irrigation cycles characteristic of AWD practices. Applying this model across 3 million fields reveals spatial heterogeneity in adoption at the state level, highlighting gaps and opportunities for policy targeting. Our district-level adoption rates correlate well with government estimates (Spearman's $\rho$=0.69 and Rank Biased Overlap=0.77). This study provides policymakers and sustainability programs a powerful tool to track practice adoption, inform targeted interventions, and drive data-driven policies for water conservation and climate mitigation at regional scale.
Authors: Jian Gao, Jianshe Wu, JingYi Ding
Abstract: Learning representations for continuous-time dynamic graphs is critical for dynamic link prediction. While recent methods have become increasingly complex, the field lacks a strong and informative baseline to reliably gauge progress. This paper proposes HyperEvent, a simple approach that captures relative structural patterns in event sequences through an intuitive encoding mechanism. As a straightforward baseline, HyperEvent leverages relative structural encoding to identify meaningful event sequences without complex parameterization. By combining these interpretable features with a lightweight transformer classifier, HyperEvent reframes link prediction as event structure recognition. Despite its simplicity, HyperEvent achieves competitive results across multiple benchmarks, often matching the performance of more complex models. This work demonstrates that effective modeling can be achieved through simple structural encoding, providing a clear reference point for evaluating future advancements.
Authors: Ganesh Sundaram, Jonas Ulmen, Amjad Haider, Daniel G\"orges
Abstract: Deep neural networks (DNNs) offer significant flexibility and robust performance. This makes them ideal for building not only system models but also advanced neural network controllers (NNCs). However, their high complexity and computational needs often limit their use. Various model compression strategies have been developed over the past few decades to address these issues. These strategies are effective for general DNNs but do not directly apply to NNCs. NNCs need both size reduction and the retention of key application-specific performance features. In structured pruning, which removes groups of related elements, standard importance metrics often fail to protect these critical characteristics. In this paper, we introduce a novel framework for calculating importance metrics in pruning groups. This framework not only shrinks the model size but also considers various application-specific constraints. To find the best pruning coefficient for each group, we evaluate two approaches. The first approach involves simple exploration through grid search. The second utilizes gradient descent optimization, aiming to balance compression and task performance. We test our method in two use cases: one on an MNIST autoencoder and the other on a Temporal Difference Model Predictive Control (TDMPC) agent. Results show that the method effectively maintains application-relevant performance while achieving a significant reduction in model size.
Authors: Ivona Krchova, Mariana Vargas Vieyra, Mario Scriminaci, Andrey Sidorenko
Abstract: Machine learning development critically depends on access to high-quality data. However, increasing restrictions due to privacy, proprietary interests, and ethical concerns have created significant barriers to data accessibility. Synthetic data offers a viable solution by enabling safe, broad data usage without compromising sensitive information. This paper presents the MOSTLY AI Synthetic Data Software Development Kit (SDK), an open-source toolkit designed specifically for synthesizing high-quality tabular data. The SDK integrates robust features such as differential privacy guarantees, fairness-aware data generation, and automated quality assurance into a flexible and accessible Python interface. Leveraging the TabularARGN autoregressive framework, the SDK supports diverse data types and complex multi-table and sequential datasets, delivering competitive performance with notable improvements in speed and usability. Currently deployed both as a cloud service and locally installable software, the SDK has seen rapid adoption, highlighting its practicality in addressing real-world data bottlenecks and promoting widespread data democratization.
Authors: Moritz Piening, Robert Beinert
Abstract: The Gromov--Wasserstein (GW) distance and its fused extension (FGW) are powerful tools for comparing heterogeneous data. Their computation is, however, challenging since both distances are based on non-convex, quadratic optimal transport (OT) problems. Leveraging 1D OT, a sliced version of GW has been proposed to lower the computational burden. Unfortunately, this sliced version is restricted to Euclidean geometry and loses invariance to isometries, strongly limiting its application in practice. To overcome these issues, we propose a novel slicing technique for GW as well as for FGW that is based on an appropriate lower bound, hierarchical OT, and suitable quadrature rules for the underlying 1D OT problems. Our novel sliced FGW significantly reduces the numerical effort while remaining invariant to isometric transformations and allowing the comparison of arbitrary geometries. We show that our new distance actually defines a pseudo-metric for structured spaces that bounds FGW from below and study its interpolation properties between sliced Wasserstein and GW. Since we avoid the underlying quadratic program, our sliced distance is numerically more robust and reliable than the original GW and FGW distance; especially in the context of shape retrieval and graph isomorphism testing.
Authors: Eliot Beyler (SIERRA), Francis Bach (SIERRA)
Abstract: We provide new convergence guarantees in Wasserstein distance for diffusion-based generative models, covering both stochastic (DDPM-like) and deterministic (DDIM-like) sampling methods. We introduce a simple framework to analyze discretization, initialization, and score estimation errors. Notably, we derive the first Wasserstein convergence bound for the Heun sampler and improve existing results for the Euler sampler of the probability flow ODE. Our analysis emphasizes the importance of spatial regularity of the learned score function and argues for controlling the score error with respect to the true reverse process, in line with denoising score matching. We also incorporate recent results on smoothed Wasserstein distances to sharpen initialization error bounds.
Authors: Georgios Papoudakis, Thomas Coste, Jianye Hao, Jun Wang, Kun Shao
Abstract: Reinforcement learning (RL) using foundation models for policy approximations in multi-turn tasks remains challenging. We identify two main limitations related to sparse reward settings and policy gradient updates, based on which we formulate a key insight: updates from positive samples with high returns typically do not require policy regularisation, whereas updates from negative samples, reflecting undesirable behaviour, can harm model performance. This paper introduces Succeed or Learn Slowly (SoLS), a novel off-policy RL algorithm evaluated on mobile app control tasks. SoLS improves sample efficiency when fine-tuning foundation models for user interface navigation via a modified off-policy actor-critic approach, applying direct policy updates for positive samples and conservative, regularised updates for negative ones to prevent model degradation. We augment SoLS with Successful Transition Replay (STR), which prioritises learning from successful interactions, further improving sample efficiency. We evaluate SoLS on the AndroidWorld benchmark, where it significantly outperforms existing methods (at least 17% relative increase), including prompt-engineering and RL approaches, while requiring substantially fewer computational resources than GPT-4o-based methods with 5-60x faster inference.
Authors: Adrian Catalin Lutu, Ioana Pintilie, Elena Burceanu, Andrei Manolache
Abstract: We present ChronoGraph, a graph-structured multivariate time series forecasting dataset built from real-world production microservices. Each node is a service that emits a multivariate stream of system-level performance metrics, capturing CPU, memory, and network usage patterns, while directed edges encode dependencies between services. The primary task is forecasting future values of these signals at the service level. In addition, ChronoGraph provides expert-annotated incident windows as anomaly labels, enabling evaluation of anomaly detection methods and assessment of forecast robustness during operational disruptions. Compared to existing benchmarks from industrial control systems or traffic and air-quality domains, ChronoGraph uniquely combines (i) multivariate time series, (ii) an explicit, machine-readable dependency graph, and (iii) anomaly labels aligned with real incidents. We report baseline results spanning forecasting models, pretrained time-series foundation models, and standard anomaly detectors. ChronoGraph offers a realistic benchmark for studying structure-aware forecasting and incident-aware evaluation in microservice systems.
Authors: Filip Sondej, Yushi Yang
Abstract: Current unlearning and safety training methods consistently fail to remove dangerous knowledge from language models. We identify the root cause - unlearning targets representations which are too general - and develop a highly selective technique that unlearns robustly while preserving general performance. Our method performs PCA on activations and module-output gradients to identify subspaces containing common representations, then collapses these subspaces before computing unlearning updates, a technique we term Collapse of Irrelevant Representations (CIR). This avoids unlearning general knowledge and targets only representations specific to the facts being unlearned. When unlearning bio- and cyber-hazardous facts from Llama-3.1-8B, we achieve over 30x greater reduction in post-attack accuracy than the best baseline (Circuit Breakers), while disrupting general performance 30x less, and using less than 3 GPU-seconds per fact. Thus, by disentangling harmful and benign capabilities at the level of representations, CIR enables robust and non-disruptive unlearning.
Authors: Vishnu Narayanan Moothedath, Umang Agarwal, Umeshraja N, James Richard Gross, Jaya Prakash Champati, Sharayu Moharir
Abstract: We focus on a binary classification problem in an edge intelligence system where false negatives are more costly than false positives. The system has a compact, locally deployed model, which is supplemented by a larger, remote model, which is accessible via the network by incurring an offloading cost. For each sample, our system first uses the locally deployed model for inference. Based on the output of the local model, the sample may be offloaded to the remote model. This work aims to understand the fundamental trade-off between classification accuracy and the offloading costs within such a hierarchical inference (HI) system. To optimise this system, we propose an online learning framework that continuously adapts a pair of thresholds on the local model's confidence scores. These thresholds determine the prediction of the local model and whether a sample is classified locally or offloaded to the remote model. We present a closed-form solution for the setting where the local model is calibrated. For the more general case of uncalibrated models, we introduce H2T2, an online two-threshold hierarchical inference policy, and prove it achieves sublinear regret. H2T2 is model-agnostic, requires no training, and learns during the inference phase using limited feedback. Simulations on real-world datasets show that H2T2 consistently outperforms naive and single-threshold HI policies, sometimes even surpassing offline optima. The policy also demonstrates robustness to distribution shifts and adapts effectively to mismatched classifiers.
Authors: Alejandro Almod\'ovar, Patricia A. Apell\'aniz, Alba Garrido, Fernando Fern\'andez-Salvador, Santiago Zazo, Juan Parras
Abstract: Why should a clinician trust an Artificial Intelligence (AI) prediction? Despite the increasing accuracy of machine learning methods in medicine, the lack of transparency continues to hinder their adoption in clinical practice. In this work, we explore Kolmogorov-Arnold Networks (KANs) for clinical classification tasks on tabular data. In contrast to traditional neural networks, KANs are function-based architectures that offer intrinsic interpretability through transparent, symbolic representations. We introduce \emph{Logistic-KAN}, a flexible generalization of logistic regression, and \emph{Kolmogorov-Arnold Additive Model (KAAM)}, a simplified additive variant that delivers transparent, symbolic formulas. Unlike ``black-box'' models that require post-hoc explainability tools, our models support built-in patient-level insights, intuitive visualizations, and nearest-patient retrieval. Across multiple health datasets, our models match or outperform standard baselines, while remaining fully interpretable. These results position KANs as a promising step toward trustworthy AI that clinicians can understand, audit, and act upon. We release the code for reproducibility in \codeurl.
Authors: Hugo Math, Rainer Lienhart
Abstract: Understanding causality in event sequences where outcome labels such as diseases or system failures arise from preceding events like symptoms or error codes is critical. Yet remains an unsolved challenge across domains like healthcare or vehicle diagnostics. We introduce CARGO, a scalable multi-label causal discovery method for sparse, high-dimensional event sequences comprising of thousands of unique event types. Using two pretrained causal Transformers as domain-specific foundation models for event sequences. CARGO infers in parallel, per sequence one-shot causal graphs and aggregates them using an adaptive frequency fusion to reconstruct the global Markov boundaries of labels. This two-stage approach enables efficient probabilistic reasoning at scale while bypassing the intractable cost of full-dataset conditional independence testing. Our results on a challenging real-world automotive fault prediction dataset with over 29,100 unique event types and 474 imbalanced labels demonstrate CARGO's ability to perform structured reasoning.
Authors: Hugo Math, Robin Sch\"on, Rainer Lienhart
Abstract: Understanding causality in event sequences with thousands of sparse event types is critical in domains such as healthcare, cybersecurity, or vehicle diagnostics, yet current methods fail to scale. We present OSCAR, a one-shot causal autoregressive method that infers per-sequence Markov Boundaries using two pretrained Transformers as density estimators. This enables efficient, parallel causal discovery without costly global CI testing. On a real-world automotive dataset with 29,100 events and 474 labels, OSCAR recovers interpretable causal structures in minutes, while classical methods fail to scale, enabling practical scientific diagnostics at production scale.
Authors: Francesco Pappone, Donato Crisostomi, Emanuele Rodol\`a
Abstract: Recurrent-depth transformers scale test-time compute by iterating latent computations before emitting tokens. We study the geometry of these iterates and argue for a simple, two-scale operational picture: (i) within a looped block, updates act as small-scale refinements; (ii) across consecutive blocks, states undergo a larger-scale drift. Across training, our measurements show that loop steps become smaller and increasingly orthogonal to one another, indicating better local modeling of fine structure rather than merely pushing in a single direction. These dynamics motivate an early-exit mechanism based on the model's second-order difference in step-size, which we show is superior in terms of performance, stability and time-efficiency, when compared to the KL-divergence exit strategy of Geiping et al. and its naive first-order counterpart.
Authors: Andr\'e Longon, David Klindt, Meenakshi Khosla
Abstract: The superposition hypothesis states that single neurons may participate in representing multiple features in order for the neural network to represent more features than it has neurons. In neuroscience and AI, representational alignment metrics measure the extent to which different deep neural networks (DNNs) or brains represent similar information. In this work, we explore a critical question: does superposition interact with alignment metrics in any undesirable way? We hypothesize that models which represent the same features in different superposition arrangements, i.e., their neurons have different linear combinations of the features, will interfere with predictive mapping metrics (semi-matching, soft-matching, linear regression), producing lower alignment than expected. We develop a theory for how permutation metrics are dependent on superposition arrangements. This is tested by training sparse autoencoders (SAEs) to disentangle superposition in toy models, where alignment scores are shown to typically increase when a model's base neurons are replaced with its sparse overcomplete latent codes. We find similar increases for DNN-DNN and DNN-brain linear regression alignment in the visual domain. Our results suggest that superposition disentanglement is necessary for mapping metrics to uncover the true representational alignment between neural networks.
Authors: Milad Aghajohari, Kamran Chitsaz, Amirhossein Kazemnejad, Sarath Chandar, Alessandro Sordoni, Aaron Courville, Siva Reddy
Abstract: Reinforcement learning (RL) has recently become a strong recipe for training reasoning LLMs that produce long chains of thought (LongCoT). Yet the standard RL "thinking environment", where the state is the prompt plus all prior reasoning tokens, makes the state unbounded and forces attention-based policies to pay quadratic compute as thoughts lengthen. We revisit the environment itself. We propose Markovian Thinking, a paradigm in which the policy advances reasoning while conditioning on a constant-size state, decoupling thinking length from context size. As an immediate consequence this yields linear compute with constant memory. We instantiate this idea with Delethink, an RL environment that structures reasoning into fixed-size chunks. Within each chunk, the model thinks as usual; at the boundary, the environment resets the context and reinitializes the prompt with a short carryover. Through RL, the policy learns to write a textual state near the end of each chunk sufficient for seamless continuation of reasoning after reset. Trained in this environment, an R1-Distill 1.5B model reasons in 8K-token chunks yet thinks up to 24K tokens, matching or surpassing LongCoT-RL trained with a 24K budget. With test-time scaling, Delethink continues to improve where LongCoT plateaus. The effect of linear compute is substantial: we empirically estimate at 96K average thinking length LongCoT-RL costs 27 H100-months vs. 7 for Delethink. Analysis at RL initialization shows off-the-shelf reasoning models (1.5B-120B) often sample Markovian traces zero-shot across diverse benchmarks, providing positive samples that make RL effective at scale. Our results show that redesigning the thinking environment is a powerful lever: it enables very long reasoning without quadratic overhead and opens a path toward efficient, scalable reasoning LLMs.
Authors: Xingjian Wu, Xiangfei Qiu, Hanyin Cheng, Zhengyu Li, Jilin Hu, Chenjuan Guo, Bin Yang
Abstract: Time Series Forecasting has made significant progress with the help of Patching technique, which partitions time series into multiple patches to effectively retain contextual semantic information into a representation space beneficial for modeling long-term dependencies. However, conventional patching partitions a time series into adjacent patches, which causes a fixed representation space, thus resulting in insufficiently expressful representations. In this paper, we pioneer the exploration of constructing a selective representation space to flexibly include the most informative patches for forecasting. Specifically, we propose the Selective Representation Space (SRS) module, which utilizes the learnable Selective Patching and Dynamic Reassembly techniques to adaptively select and shuffle the patches from the contextual time series, aiming at fully exploiting the information of contextual time series to enhance the forecasting performance of patch-based models. To demonstrate the effectiveness of SRS module, we propose a simple yet effective SRSNet consisting of SRS and an MLP head, which achieves state-of-the-art performance on real-world datasets from multiple domains. Furthermore, as a novel plugin-and-play module, SRS can also enhance the performance of existing patch-based models. The resources are available at https://github.com/decisionintelligence/SRSNet.
Authors: Catherine Villeneuve, Benjamin Akera, M\'elisande Teng, David Rolnick
Abstract: Species distribution models (SDMs), which aim to predict species occurrence based on environmental variables, are widely used to monitor and respond to biodiversity change. Recent deep learning advances for SDMs have been shown to perform well on complex and heterogeneous datasets, but their effectiveness remains limited by spatial biases in the data. In this paper, we revisit deep SDMs from a Bayesian perspective and introduce BATIS, a novel and practical framework wherein prior predictions are updated iteratively using limited observational data. Models must appropriately capture both aleatoric and epistemic uncertainty to effectively combine fine-grained local insights with broader ecological patterns. We benchmark an extensive set of uncertainty quantification approaches on a novel dataset including citizen science observations from the eBird platform. Our empirical study shows how Bayesian deep learning approaches can greatly improve the reliability of SDMs in data-scarce locations, which can contribute to ecological understanding and conservation efforts.
Authors: Abhishek Panigrahi, Bingbin Liu, Sadhika Malladi, Sham Kakade, Surbhi Goel
Abstract: Knowledge distillation is an efficient strategy to use data generated by large "teacher" language models to train smaller capable "student" models, but selecting the optimal teacher for a specific student-task combination requires expensive trial-and-error. We propose a lightweight score called GRACE to quantify how effective a teacher will be for post-training a student model. GRACE measures distributional properties of the student's gradients without access to a verifier, teacher logits, teacher internals, or test data. From an information-theoretic perspective, GRACE connects to leave-one-out stability of gradient-based algorithms, which controls the generalization performance of the distilled students. On GSM8K and MATH, GRACE correlates strongly (up to 86% Spearman correlation) with the performance of the distilled LLaMA and OLMo students. In particular, training a student using the GRACE-selected teacher can improve the performance by up to 7.4% over naively using the best-performing teacher. Further, GRACE can provide guidance on crucial design choices in distillation, including (1) the best temperature to use when generating from the teacher, (2) the best teacher to use given a size constraint, and (3) the best teacher to use within a specific model family. Altogether, our findings demonstrate that GRACE can efficiently and effectively identify a strongly compatible teacher for a given student and provide fine-grained guidance on how to perform distillation.
Authors: Xiaokai Wang, Shaoyuan Huang, Yuting Li, Xiaofei Wang
Abstract: Deep neural networks (DNNs) form the cornerstone of modern AI services, supporting a wide range of applications, including autonomous driving, chatbots, and recommendation systems. As models increase in size and complexity, DNN workloads such as training and inference tasks impose unprecedented demands on distributed computing resources, making accurate runtime prediction essential for optimizing development and resource allocation. Traditional methods rely on additive computational unit models, limiting their accuracy and generalizability. In contrast, graph-enhanced modeling improves performance but significantly increases data collection costs. Therefore, there is a critical need for a method that strikes a balance between accuracy, generalizability, and data collection costs. To address these challenges, we propose ScaleDL, a novel runtime prediction framework that combines nonlinear layer-wise modeling with graph neural network (GNN)-based cross-layer interaction mechanism, enabling accurate DNN runtime prediction and hierarchical generalizability across different network architectures. Additionally, we employ the D-optimal method to reduce data collection costs. Experiments on the workloads of five popular DNN models demonstrate that ScaleDL enhances runtime prediction accuracy and generalizability, achieving 6 times lower MRE and 5 times lower RMSE compared to baseline models.
Authors: Linna Wang, Zhixuan You, Qihui Zhang, Jiunan Wen, Ji Shi, Yimin Chen, Yusen Wang, Fanqi Ding, Ziliang Feng, Li Lu
Abstract: Large Language Models (LLMs) and causal learning each hold strong potential for clinical decision making (CDM). However, their synergy remains poorly understood, largely due to the lack of systematic benchmarks evaluating their integration in clinical risk prediction. In real-world healthcare, identifying features with causal influence on outcomes is crucial for actionable and trustworthy predictions. While recent work highlights LLMs' emerging causal reasoning abilities, there lacks comprehensive benchmarks to assess their causal learning and performance informed by causal features in clinical risk prediction. To address this, we introduce REACT-LLM, a benchmark designed to evaluate whether combining LLMs with causal features can enhance clinical prognostic performance and potentially outperform traditional machine learning (ML) methods. Unlike existing LLM-clinical benchmarks that often focus on a limited set of outcomes, REACT-LLM evaluates 7 clinical outcomes across 2 real-world datasets, comparing 15 prominent LLMs, 6 traditional ML models, and 3 causal discovery (CD) algorithms. Our findings indicate that while LLMs perform reasonably in clinical prognostics, they have not yet outperformed traditional ML models. Integrating causal features derived from CD algorithms into LLMs offers limited performance gains, primarily due to the strict assumptions of many CD methods, which are often violated in complex clinical data. While the direct integration yields limited improvement, our benchmark reveals a more promising synergy.
Authors: Marcin Podhajski, Jan Dubi\'nski, Franziska Boenisch, Adam Dziedzic, Agnieszka Pr\k{e}gowska, Tomasz P. Michalak
Abstract: Current graph neural network (GNN) model-stealing methods rely heavily on queries to the victim model, assuming no hard query limits. However, in reality, the number of allowed queries can be severely limited. In this paper, we demonstrate how an adversary can extract a GNN with very limited interactions with the model. Our approach first enables the adversary to obtain the model backbone without making direct queries to the victim model and then to strategically utilize a fixed query limit to extract the most informative data. The experiments on eight real-world datasets demonstrate the effectiveness of the attack, even under a very restricted query limit and under defense against model extraction in place. Our findings underscore the need for robust defenses against GNN model extraction threats.
Authors: Chuanqing Pu, Feilong Fan, Nengling Tai, Yan Xu, Wentao Huang, Honglin Wen
Abstract: Power-logistics scheduling in modern seaports typically follow a predict-then-optimize pipeline. To enhance the decision quality of forecasts, decision-focused learning has been proposed, which aligns the training of forecasting models with downstream decision outcomes. However, this end-to-end design inherently restricts the value of forecasting models to only a specific task structure, and thus generalize poorly to evolving tasks induced by varying seaport vessel arrivals. We address this gap with a decision-focused continual learning framework that adapts online to a stream of scheduling tasks. Specifically, we introduce Fisher information based regularization to enhance cross-task generalization by preserving parameters critical to prior tasks. A differentiable convex surrogate is also developed to stabilize gradient backpropagation. The proposed approach enables learning a decision-aligned forecasting model across a varying tasks stream with a sustainable long-term computational burden. Experiments calibrated to the Jurong Port demonstrate superior decision performance and generalization over existing methods with reduced computational cost.
Authors: Junkai Lu, Peng Chen, Chenjuan Guo, Yang Shu, Meng Wang, Bin Yang
Abstract: Time series forecasting is critical for decision-making across dynamic domains such as energy, finance, transportation, and cloud computing. However, real-world time series often exhibit non-stationarity, including temporal distribution shifts and spectral variability, which pose significant challenges for long-term time series forecasting. In this paper, we propose DTAF, a dual-branch framework that addresses non-stationarity in both the temporal and frequency domains. For the temporal domain, the Temporal Stabilizing Fusion (TFS) module employs a non-stationary mix of experts (MOE) filter to disentangle and suppress temporal non-stationary patterns while preserving long-term dependencies. For the frequency domain, the Frequency Wave Modeling (FWM) module applies frequency differencing to dynamically highlight components with significant spectral shifts. By fusing the complementary outputs of TFS and FWM, DTAF generates robust forecasts that adapt to both temporal and frequency domain non-stationarity. Extensive experiments on real-world benchmarks demonstrate that DTAF outperforms state-of-the-art baselines, yielding significant improvements in forecasting accuracy under non-stationary conditions. All codes are available at https://github.com/PandaJunk/DTAF.
Authors: Jamison Moody, James Usevitch
Abstract: Kolmogorov-Arnold Networks (KANs) are a class of neural networks that have received increased attention in recent literature. In contrast to MLPs, KANs leverage parameterized, trainable activation functions and offer several benefits including improved interpretability and higher accuracy on learning symbolic equations. However, the original KAN architecture requires adjustments to the domain discretization of the network (called the "domain grid") during training, creating extra overhead for the user in the training process. Typical KAN layers are not designed with the ability to autonomously update their domains in a data-driven manner informed by the changing output ranges of previous layers. As an added benefit, this histogram algorithm may also be applied towards detecting out-of-distribution (OOD) inputs in a variety of settings. We demonstrate that AdaptKAN exceeds or matches the performance of prior KAN architectures and MLPs on four different tasks: learning scientific equations from the Feynman dataset, image classification from frozen features, learning a control Lyapunov function, and detecting OOD inputs on the OpenOOD v1.5 benchmark.
Authors: Rui Zhu, Yuexing Peng, George C. Alexandropoulos, Wenbo Wang, Wei Xiang
Abstract: The Method of Moments (MoM) is constrained by the usage of static, geometry-defined basis functions, such as the Rao-Wilton-Glisson (RWG) basis. This letter reframes electromagnetic modeling around a learnable basis representation rather than solving for the coefficients over a fixed basis. We first show that the RWG basis is essentially a static and piecewise-linear realization of the Kolmogorov-Arnold representation theorem. Inspired by this insight, we propose PhyKAN, a physics-informed Kolmogorov-Arnold Network (KAN) that generalizes RWG into a learnable and adaptive basis family. Derived from the EFIE, PhyKAN integrates a local KAN branch with a global branch embedded with Green's function priors to preserve physical consistency. It is demonstrated that, across canonical geometries, PhyKAN achieves sub-0.01 reconstruction errors as well as accurate, unsupervised radar cross section predictions, offering an interpretable, physics-consistent bridge between classical solvers and modern neural network models for electromagnetic modeling.
Authors: Tuhin Subhra De
Abstract: Variational Autoencoders and Generative Adversarial Networks remained the state-of-the-art (SOTA) generative models until 2022. Now they are superseded by diffusion based models. Efforts to improve traditional models have stagnated as a result. In old-school fashion, we explore image generation with conditional Variational Autoencoders (CVAE) to incorporate desired attributes within the images. VAEs are known to produce blurry images with less diversity, we refer a method that solve this issue by leveraging the variance of the gaussian decoder as a learnable parameter during training. Previous works on CVAEs assumed that the conditional distribution of the latent space given the labels is equal to the prior distribution, which is not the case in reality. We show that estimating it using normalizing flows results in better image generation than existing methods by reducing the FID by 5% and increasing log likelihood by 7.7% than the previous case.
Authors: Lijun Zhang, Suyuan Liu, Siwei Wang, Shengju Yu, Xueling Zhu, Miaomiao Li, Xinwang Liu
Abstract: Clustering is a fundamental task in unsupervised learning, but most existing methods heavily rely on hyperparameters such as the number of clusters or other sensitive settings, limiting their applicability in real-world scenarios. To address this long-standing challenge, we propose a novel and fully parameter-free clustering framework via Self-supervised Consensus Maximization, named SCMax. Our framework performs hierarchical agglomerative clustering and cluster evaluation in a single, integrated process. At each step of agglomeration, it creates a new, structure-aware data representation through a self-supervised learning task guided by the current clustering structure. We then introduce a nearest neighbor consensus score, which measures the agreement between the nearest neighbor-based merge decisions suggested by the original representation and the self-supervised one. The moment at which consensus maximization occurs can serve as a criterion for determining the optimal number of clusters. Extensive experiments on multiple datasets demonstrate that the proposed framework outperforms existing clustering approaches designed for scenarios with an unknown number of clusters.
Authors: Jiajie Su, Zihan Nan, Yunshan Ma, Xiaobo Xia, Xiaohua Feng, Weiming Liu, Xiaolin Zheng, Chaochao Chen
Abstract: Sequential Recommenders, which exploit dynamic user intents through interaction sequences, is vulnerable to adversarial attacks. While existing attacks primarily rely on data poisoning, they require large-scale user access or fake profiles thus lacking practicality. In this paper, we focus on the Profile Pollution Attack that subtly contaminates partial user interactions to induce targeted mispredictions. Previous PPA methods suffer from two limitations, i.e., i) over-reliance on sequence horizon impact restricts fine-grained perturbations on item transitions, and ii) holistic modifications cause detectable distribution shifts. To address these challenges, we propose a constrained reinforcement driven attack CREAT that synergizes a bi-level optimization framework with multi-reward reinforcement learning to balance adversarial efficacy and stealthiness. We first develop a Pattern Balanced Rewarding Policy, which integrates pattern inversion rewards to invert critical patterns and distribution consistency rewards to minimize detectable shifts via unbalanced co-optimal transport. Then we employ a Constrained Group Relative Reinforcement Learning paradigm, enabling step-wise perturbations through dynamic barrier constraints and group-shared experience replay, achieving targeted pollution with minimal detectability. Extensive experiments demonstrate the effectiveness of CREAT.
Authors: Emanuele Zappala
Abstract: Neural integral equations are deep learning models based on the theory of integral equations, where the model consists of an integral operator and the corresponding equation (of the second kind) which is learned through an optimization procedure. This approach allows to leverage the nonlocal properties of integral operators in machine learning, but it is computationally expensive. In this article, we introduce a framework for neural integral equations based on spectral methods that allows us to learn an operator in the spectral domain, resulting in a cheaper computational cost, as well as in high interpolation accuracy. We study the properties of our methods and show various theoretical guarantees regarding the approximation capabilities of the model, and convergence to solutions of the numerical methods. We provide numerical experiments to demonstrate the practical effectiveness of the resulting model.
Authors: Joohwan Ko, Kyurae Kim, Woo Chang Kim, Jacob R. Gardner
Abstract: Variational families with full-rank covariance approximations are known not to work well in black-box variational inference (BBVI), both empirically and theoretically. In fact, recent computational complexity results for BBVI have established that full-rank variational families scale poorly with the dimensionality of the problem compared to e.g. mean-field families. This is particularly critical to hierarchical Bayesian models with local variables; their dimensionality increases with the size of the datasets. Consequently, one gets an iteration complexity with an explicit $\mathcal{O}(N^2)$ dependence on the dataset size $N$. In this paper, we explore a theoretical middle ground between mean-field variational families and full-rank families: structured variational families. We rigorously prove that certain scale matrix structures can achieve a better iteration complexity of $\mathcal{O}\left(N\right)$, implying better scaling with respect to $N$. We empirically verify our theoretical results on large-scale hierarchical models.
Authors: Nathan Wycoff
Abstract: Regression trees have emerged as a preeminent tool for solving real-world regression problems due to their ability to deal with nonlinearities, interaction effects and sharp discontinuities. In this article, we rather study regression trees applied to well-behaved, differentiable functions, and determine the relationship between node parameters and the local gradient of the function being approximated. We find a simple estimate of the gradient which can be efficiently computed using quantities exposed by popular tree learning libraries. This allows the tools developed in the context of differentiable algorithms, like neural nets and Gaussian processes, to be deployed to tree-based models. To demonstrate this, we study measures of model sensitivity defined in terms of integrals of gradients and demonstrate how to compute them for regression trees using the proposed gradient estimates. Quantitative and qualitative numerical experiments reveal the capability of gradients estimated by regression trees to improve predictive analysis, solve tasks in uncertainty quantification, and provide interpretation of model behavior.
Authors: Susu Sun, Stefano Woerner, Andreas Maier, Lisa M. Koch, Christian F. Baumgartner
Abstract: Interpretability is crucial for machine learning algorithms in high-stakes medical applications. However, high-performing neural networks typically cannot explain their predictions. Post-hoc explanation methods provide a way to understand neural networks but have been shown to suffer from conceptual problems. Moreover, current research largely focuses on providing local explanations for individual samples rather than global explanations for the model itself. In this paper, we propose Attri-Net, an inherently interpretable model for multi-label classification that provides local and global explanations. Attri-Net first counterfactually generates class-specific attribution maps to highlight the disease evidence, then performs classification with logistic regression classifiers based solely on the attribution maps. Local explanations for each prediction can be obtained by interpreting the attribution maps weighted by the classifiers' weights. Global explanation of whole model can be obtained by jointly considering learned average representations of the attribution maps for each class (called the class centers) and the weights of the linear classifiers. To ensure the model is ``right for the right reason", we further introduce a mechanism to guide the model's explanations to align with human knowledge. Our comprehensive evaluations show that Attri-Net can generate high-quality explanations consistent with clinical knowledge while not sacrificing classification performance.
Authors: Xiaoxiao He, Quan Dao, Ligong Han, Song Wen, Minhao Bai, Di Liu, Han Zhang, Martin Renqiang Min, Felix Juefei-Xu, Chaowei Tan, Bo Liu, Kang Li, Hongdong Li, Junzhou Huang, Faez Ahmed, Akash Srivastava, Dimitris Metaxas
Abstract: Discrete diffusion models have achieved success in tasks like image generation and masked language modeling but face limitations in controlled content editing. We introduce DICE (Discrete Inversion for Controllable Editing), the first approach to enable precise inversion for discrete diffusion models, including multinomial diffusion and masked generative models. By recording noise sequences and masking patterns during the reverse diffusion process, DICE enables accurate reconstruction and flexible editing of discrete data without the need for predefined masks or attention manipulation. We demonstrate the effectiveness of DICE across both image and text domains, evaluating it on models such as VQ-Diffusion, Paella, and RoBERTa. Our results show that DICE preserves high data fidelity while enhancing editing capabilities, offering new opportunities for fine-grained content manipulation in discrete spaces.
Authors: David Yunis, Siyu Huo, Chulaka Gunasekara, Danish Contractor
Abstract: Large language models (LLMs) are deployed in a wide variety of user-facing applications. Typically, these deployments have some specific purpose, like answering questions grounded on documentation or acting as coding assistants, but they require general language understanding. In such deployments, LLMs should respond only to queries that align with the intended purpose and reject all other requests, such as generating poetry or answering questions about physics, a task we refer to as `scoping'. We conduct a comprehensive empirical evaluation of various methods, ranging from prompting, fine-tuning to preference learning and the recently proposed general alignment technique known as Circuit Breakers (CB). Across three families of language models and a broad variety of tasks, we show that it is possible to scope language models. We examine scoping for multiple topics, and fine-grained topics. We ablate diversity of irrelevant queries, layer different techniques, conduct adversarial evaluations and more. Among other results, we find that when diverse examples of irrelevant queries are available, simple supervised fine-tuning produces the best results, but when such diversity is low, Circuit Breakers perform quite well. One can often get the benefits of both methods by layering them in succession. We intend our study to serve as a practitioner's guide to scoping LLMs.
Authors: Claudius Krause, Michele Faucci Giannelli, Gregor Kasieczka, Benjamin Nachman, Dalila Salamani, David Shih, Anna Zaborowska, Oz Amram, Kerstin Borras, Matthew R. Buckley, Erik Buhmann, Thorsten Buss, Renato Paulo Da Costa Cardoso, Anthony L. Caterini, Nadezda Chernyavskaya, Federico A. G. Corchia, Jesse C. Cresswell, Sascha Diefenbacher, Etienne Dreyer, Vijay Ekambaram, Engin Eren, Florian Ernst, Luigi Favaro, Matteo Franchini, Frank Gaede, Eilam Gross, Shih-Chieh Hsu, Kristina Jaruskova, Benno K\"ach, Jayant Kalagnanam, Raghav Kansal, Taewoo Kim, Dmitrii Kobylianskii, Anatolii Korol, William Korcari, Dirk Kr\"ucker, Katja Kr\"uger, Marco Letizia, Shu Li, Qibin Liu, Xiulong Liu, Gabriel Loaiza-Ganem, Thandikire Madula, Peter McKeown, Isabell-A. Melzer-Pellmann, Vinicius Mikuni, Nam Nguyen, Ayodele Ore, Sofia Palacios Schweitzer, Ian Pang, Kevin Pedro, Tilman Plehn, Witold Pokorski, Huilin Qu, Piyush Raikwar, John A. Raine, Humberto Reyes-Gonzalez, Lorenzo Rinaldi, Brendan Leigh Ross, Moritz A. W. Scham, Simon Schnake, Chase Shimmin, Eli Shlizerman, Nathalie Soybelman, Mudhakar Srivatsa, Kalliopi Tsolaki, Sofia Vallecorsa, Kyongmin Yeo, Rui Zhang
Abstract: We present the results of the "Fast Calorimeter Simulation Challenge 2022" - the CaloChallenge. We study state-of-the-art generative models on four calorimeter shower datasets of increasing dimensionality, ranging from a few hundred voxels to a few tens of thousand voxels. The 31 individual submissions span a wide range of current popular generative architectures, including Variational AutoEncoders (VAEs), Generative Adversarial Networks (GANs), Normalizing Flows, Diffusion models, and models based on Conditional Flow Matching. We compare all submissions in terms of quality of generated calorimeter showers, as well as shower generation time and model size. To assess the quality we use a broad range of different metrics including differences in 1-dimensional histograms of observables, KPD/FPD scores, AUCs of binary classifiers, and the log-posterior of a multiclass classifier. The results of the CaloChallenge provide the most complete and comprehensive survey of cutting-edge approaches to calorimeter fast simulation to date. In addition, our work provides a uniquely detailed perspective on the important problem of how to evaluate generative models. As such, the results presented here should be applicable for other domains that use generative AI and require fast and faithful generation of samples in a large phase space.
Authors: Sudarshan Regmi
Abstract: Out-of-distribution (OOD) detection is critical to ensure the safe deployment of deep learning models in critical applications. Deep learning models can often misidentify OOD samples as in-distribution (ID) samples. This vulnerability worsens in the presence of spurious correlation in the training set. Likewise, in fine-grained classification settings, detection of fine-grained OOD samples becomes inherently challenging due to their high similarity to ID samples. However, current research on OOD detection has focused instead largely on relatively easier (conventional) cases. Even the few recent works addressing these challenging cases rely on carefully curated or synthesized outliers, ultimately requiring external data. This motivates our central research question: ``Can we innovate OOD detection training framework for fine-grained and spurious settings \textbf{without requiring any external data at all?}" In this work, we present a unified \textbf{A}pproach to \textbf{S}purious, fine-grained, and \textbf{C}onventional \textbf{OOD D}etection (\textbf{\ASCOOD}) that eliminates the reliance on external data. First, we synthesize virtual outliers from ID data by approximating the destruction of invariant features. Specifically, we propose to add gradient attribution values to ID inputs to disrupt invariant features while amplifying true-class logit, thereby synthesizing challenging near-manifold virtual outliers. Then, we simultaneously incentivize ID classification and predictive uncertainty towards virtual outliers. For this, we further propose to leverage standardized features with z-score normalization. ASCOOD effectively mitigates impact of spurious correlations and encourages capturing fine-grained attributes. Extensive experiments across \textbf{7} datasets and and comparisons with \textbf{30+} methods demonstrate merit of ASCOOD in spurious, fine-grained and conventional settings.
Authors: R{\i}za \"Oz\c{c}elik, Francesca Grisoni
Abstract: "How to evaluate the de novo designs proposed by a generative model?" Despite the transformative potential of generative deep learning in drug discovery, this seemingly simple question has no clear answer. The absence of standardized guidelines challenges both the benchmarking of generative approaches and the selection of molecules for prospective studies. In this work, we take a fresh - critical and constructive - perspective on de novo design evaluation. By training chemical language models, we analyze approximately 1 billion molecule designs and discover principles consistent across different neural networks and datasets. We uncover a key confounder: the size of the generated molecular library significantly impacts evaluation outcomes, often leading to misleading model comparisons. We find increasing the number of designs as a remedy and propose new and compute-efficient metrics to compute at large-scale. We also identify critical pitfalls in commonly used metrics - such as uniqueness and distributional similarity - that can distort assessments of generative performance. To address these issues, we propose new and refined strategies for reliable model comparison and design evaluation. Furthermore, when examining molecule selection and sampling strategies, our findings reveal the constraints to diversify the generated libraries and draw new parallels and distinctions between deep learning and drug discovery. We anticipate our findings to help reshape evaluation pipelines in generative drug discovery, paving the way for more reliable and reproducible generative modeling approaches.
Authors: Aur\'elien Decelle, Alfonso de Jes\'us Navas G\'omez, Beatriz Seoane
Abstract: Maximum entropy methods, rooted in the inverse Ising/Potts problem from statistical physics, are widely used to model pairwise interactions in complex systems across disciplines such as bioinformatics and neuroscience. While successful, these approaches often fail to capture higher-order interactions that are critical for understanding collective behavior. In contrast, modern machine learning methods can model such interactions, but their interpretability often comes at a prohibitive computational cost. Restricted Boltzmann Machines (RBMs) provide a computationally efficient alternative by encoding statistical correlations through hidden units in a bipartite architecture. In this work, we introduce a method that maps RBMs onto generalized Potts models, enabling the systematic extraction of interactions up to arbitrary order. Leveraging large-$N$ approximations, made tractable by the RBM's structure, we extract effective many-body couplings with minimal computational effort. We further propose a robust framework for recovering higher-order interactions in more complex generative models, and introduce a simple gauge-fixing scheme for the effective Potts representation. Validation on synthetic data demonstrates accurate recovery of two- and three-body interactions. Applied to protein sequence data, our method reconstructs contact maps with high fidelity and outperforms state-of-the-art inverse Potts models. These results establish RBMs as a powerful and efficient tool for modeling higher-order structure in high-dimensional categorical data.
Authors: Lars van der Laan, David Hubbard, Allen Tran, Nathan Kallus, Aur\'elien Bibaut
Abstract: Double Reinforcement Learning (DRL) enables efficient inference for policy values in nonparametric Markov decision processes (MDPs), but existing methods face two major obstacles: (1) they require stringent intertemporal overlap conditions on state trajectories, and (2) they rely on estimating high-dimensional occupancy density ratios. Motivated by problems in long-term causal inference, we extend DRL to a semiparametric setting and develop doubly robust, automatic estimators for general linear functionals of the Q-function in infinite-horizon, time-homogeneous MDPs. By imposing structure on the Q-function, we relax the overlap conditions required by nonparametric methods and obtain efficiency gains. The second obstacle--density-ratio estimation--typically requires computationally expensive and unstable min-max optimization. To address both challenges, we introduce superefficient nonparametric estimators whose limiting variance falls below the generalized Cramer-Rao bound. These estimators treat the Q-function as a one-dimensional summary of the state-action process, reducing high-dimensional overlap requirements to a single-dimensional condition. The procedure is simple to implement: estimate and calibrate the Q-function using fitted Q-iteration, then plug the result into the target functional, thereby avoiding density-ratio estimation altogether.
Authors: Yuwei Sun, Lu Mi, Ippei Fujisawa, Ruiqiao Mei, Jimin Chen, Siyu Zhu, Ryota Kanai
Abstract: Masking strategies commonly employed in natural language processing are still underexplored in vision tasks such as concept learning, where conventional methods typically rely on full images. However, using masked images diversifies perceptual inputs, potentially offering significant advantages in concept learning with large-scale Transformer models. To this end, we propose Multi-layer Concept Map (MCM), the first work to devise an efficient concept learning method based on masked images. In particular, we introduce an asymmetric concept learning architecture by establishing correlations between different encoder and decoder layers, updating concept tokens using backward gradients from reconstruction tasks. The learned concept tokens at various levels of granularity help either reconstruct the masked image patches by filling in gaps or guide the reconstruction results in a direction that reflects specific concepts. Moreover, we present both quantitative and qualitative results across a wide range of metrics, demonstrating that MCM significantly reduces computational costs by training on fewer than 75% of the total image patches while enhancing concept prediction performance. Additionally, editing specific concept tokens in the latent space enables targeted image generation from masked images, aligning both the visible contextual patches and the provided concepts. By further adjusting the testing time mask ratio, we could produce a range of reconstructions that blend the visible patches with the provided concepts, proportional to the chosen ratios.
Authors: Jiale Han, Xiaowu Dai, Yuhua Zhu
Abstract: Experience replay is a foundational technique in reinforcement learning that enhances learning stability by storing past experiences in a replay buffer and reusing them during training. Despite its practical success, its theoretical properties remain underexplored. In this paper, we present a theoretical framework that models experience replay using resampled $U$- and $V$-statistics, providing rigorous variance reduction guarantees. We apply this framework to policy evaluation tasks using the Least-Squares Temporal Difference (LSTD) algorithm and a Partial Differential Equation (PDE)-based model-free algorithm, demonstrating significant improvements in stability and efficiency, particularly in data-scarce scenarios. Beyond policy evaluation, we extend the framework to kernel ridge regression, showing that the experience replay-based method reduces the computational cost from the traditional $O(n^3)$ in time to as low as $O(n^2)$ in time while simultaneously reducing variance. Extensive numerical experiments validate our theoretical findings, demonstrating the broad applicability and effectiveness of experience replay in diverse machine learning tasks.
Authors: Hoang-Son Tran, Vladimir Petrovic, Remi Bardenet, Subhroshekhar Ghosh
Abstract: Negative dependence is becoming a key driver in advancing learning capabilities beyond the limits of traditional independence. Recent developments have evidenced support towards negatively dependent systems as a learning paradigm in a broad range of fundamental machine learning challenges including optimization, sampling, dimensionality reduction and sparse signal recovery, often surpassing the performance of current methods based on statistical independence. The most popular negatively dependent model has been that of determinantal point processes (DPPs), which have their origins in quantum theory. However, other models, such as perturbed lattice models, strongly Rayleigh measures, zeros of random functions have gained salience in various learning applications. In this article, we review this burgeoning field of research, as it has developed over the past two decades or so. We also present new results on applications of DPPs to the parsimonious representation of neural networks. In the limited scope of the article, we mostly focus on aspects of this area to which the authors contributed over the recent years, including applications to Monte Carlo methods, coresets and stochastic gradient descent, stochastic networks, signal processing and connections to quantum computation. However, starting from basics of negative dependence for the uninitiated reader, extensive references are provided to a broad swath of related developments which could not be covered within our limited scope. While existing works and reviews generally focus on specific negatively dependent models (e.g. DPPs), a notable feature of this article is that it addresses negative dependence as a machine learning methodology as a whole. In this vein, it covers within its span an array of negatively dependent models and their applications well beyond DPPs, thereby putting forward a very general and rather unique perspective.
Authors: Denan Li, Jiyuan Yang, Xiangkai Chen, Lintao Yu, Shi Liu
Abstract: Machine learning force fields have emerged as promising tools for molecular dynamics (MD) simulations, potentially offering quantum-mechanical accuracy with the efficiency of classical MD. Inspired by foundational large language models, recent years have seen considerable progress in developing foundational atomistic models, sometimes referred to as universal force fields, designed to cover most elements in the periodic table. This Perspective adopts a practitioner's viewpoint to ask a critical question: Are these foundational atomistic models reliable for one of their most compelling applications, in particular simulating finite-temperature dynamics? Instead of a broad benchmark, we use the canonical ferroelectric-paraelectric phase transition in PbTiO$_3$ as a focused case study to evaluate prominent foundational atomistic models. Our findings suggest a potential disconnect between static accuracy and dynamic reliability. While 0 K properties are often well-reproduced, we observed that the models can struggle to consistently capture the correct phase transition, sometimes exhibiting simulation instabilities. We believe these challenges may stem from inherent biases in training data and a limited description of anharmonicity. These observed shortcomings, though demonstrated on a single system, appear to point to broader, systemic challenges that can be addressed with targeted fine-tuning. This Perspective serves not to rank models, but to initiate a crucial discussion on the practical readiness of foundational atomistic models and to explore future directions for their improvement.
Authors: Tobias Christian Nauen, Brian Moser, Federico Raue, Stanislav Frolov, Andreas Dengel
Abstract: Transformers, particularly Vision Transformers (ViTs), have achieved state-of-the-art performance in large-scale image classification. However, they often require large amounts of data and can exhibit biases that limit their robustness and generalizability. This paper introduces ForAug, a novel data augmentation scheme that addresses these challenges and explicitly includes inductive biases, which commonly are part of the neural network architecture, into the training data. ForAug is constructed by using pretrained foundation models to separate and recombine foreground objects with different backgrounds, enabling fine-grained control over image composition during training. It thus increases the data diversity and effective number of training samples. We demonstrate that training on ForNet, the application of ForAug to ImageNet, significantly improves the accuracy of ViTs and other architectures by up to 4.5 percentage points (p.p.) on ImageNet and 7.3 p.p. on downstream tasks. Importantly, ForAug enables novel ways of analyzing model behavior and quantifying biases. Namely, we introduce metrics for background robustness, foreground focus, center bias, and size bias and show that training on ForNet substantially reduces these biases compared to training on ImageNet. In summary, ForAug provides a valuable tool for analyzing and mitigating biases, enabling the development of more robust and reliable computer vision models. Our code and dataset are publicly available at https://github.com/tobna/ForAug.
Authors: Shih-Chieh Dai, Jun Xu, Guanhong Tao
Abstract: Large language models (LLMs) are widely used in software development. However, the code generated by LLMs often contains vulnerabilities. Several secure code generation methods have been proposed to address this issue, but their current evaluation schemes leave several concerns unaddressed. Specifically, most existing studies evaluate security and functional correctness separately, using different datasets. That is, they assess vulnerabilities using security-related code datasets while validating functionality with general code datasets. In addition, prior research primarily relies on a single static analyzer, CodeQL, to detect vulnerabilities in generated code, which limits the scope of security evaluation. In this work, we conduct a comprehensive study to systematically assess the improvements introduced by four state-of-the-art secure code generation techniques. Specifically, we apply both security inspection and functionality validation to the same generated code and evaluate these two aspects together. We also employ three popular static analyzers and two LLMs to identify potential vulnerabilities in the generated code. Our study reveals that existing techniques often compromise the functionality of generated code to enhance security. Their overall performance remains limited when evaluating security and functionality together. In fact, many techniques even degrade the performance of the base LLM by more than 50%. Our further inspection reveals that these techniques often either remove vulnerable lines of code entirely or generate ``garbage code'' that is unrelated to the intended task. Moreover, the commonly used static analyzer CodeQL fails to detect several vulnerabilities, further obscuring the actual security improvements achieved by existing techniques.
Authors: Mihaela-Larisa Clement, M\'onika Farsang, Felix Resch, Mihai-Teodor Stanusoiu, Radu Grosu
Abstract: Autonomous agents that rely purely on perception to make real-time control decisions require efficient and robust architectures. In this work, we demonstrate that augmenting RGB input with depth information significantly enhances our agents' ability to predict steering commands compared to using RGB alone. We benchmark lightweight recurrent controllers that leverage the fused RGB-D features for sequential decision-making. To train our models, we collect high-quality data using a small-scale autonomous car controlled by an expert driver via a physical steering wheel, capturing varying levels of steering difficulty. Our models were successfully deployed on real hardware and inherently avoided dynamic and static obstacles, under out-of-distribution conditions. Specifically, our findings reveal that the early fusion of depth data results in a highly robust controller, which remains effective even with frame drops and increased noise levels, without compromising the network's focus on the task.
Authors: Ryu Tadokoro, Tsukasa Takagi, Shin-ichi Maeda
Abstract: In semantic segmentation, the accuracy of models heavily depends on the high-quality annotations. However, in many practical scenarios, such as medical imaging and remote sensing, obtaining true annotations is not straightforward and usually requires significant human labor. Relying on human labor often introduces annotation errors, including mislabeling, omissions, and inconsistency between annotators. In the case of remote sensing, differences in procurement time can lead to misaligned ground-truth annotations. These label errors are not independently distributed, and instead usually appear in spatially connected regions where adjacent pixels are more likely to share the same errors. To address these issues, we propose an approximate Bayesian estimation based on a probabilistic model that assumes training data include label errors, incorporating the tendency for these errors to occur with spatial correlations between adjacent pixels. However, Bayesian inference for such spatially correlated discrete variables is notoriously intractable. To overcome this fundamental challenge, we introduce a novel class of probabilistic models, which we term the ELBO-Computable Correlated Discrete Distribution (ECCD). By representing the discrete dependencies through a continuous latent Gaussian field with a Kac-Murdock-Szeg\"{o} (KMS) structured covariance, our framework enables scalable and efficient variational inference for problems previously considered computationally prohibitive. Through experiments on multiple segmentation tasks, we confirm that leveraging the spatial correlation of label errors significantly improves performance. Notably, in specific tasks such as lung segmentation, the proposed method achieves performance comparable to training with clean labels under moderate noise levels. Code is available at https://github.com/pfnet-research/Bayesian_SpatialCorr.
URLs: https://github.com/pfnet-research/Bayesian_SpatialCorr.
Authors: Ashirbad Mishra, Jinyu Zhao, Soumik Dey, Hansi Wu, Binbin Li, Kamesh Madduri
Abstract: In the domain of sponsored search advertising, the focus of Keyphrase recommendation has largely been on exact match types, which pose issues such as high management expenses, limited targeting scope, and evolving search query patterns. Alternatives like Broad match types can alleviate certain drawbacks of exact matches but present challenges like poor targeting accuracy and minimal supervisory signals owing to limited advertiser usage. This research defines the criteria for an ideal broad match, emphasizing on both efficiency and effectiveness, ensuring that a significant portion of matched queries are relevant. We propose BroadGen, an innovative framework that recommends efficient and effective broad match keyphrases by utilizing historical search query data. Additionally, we demonstrate that BroadGen, through token correspondence modeling, maintains better query stability over time. BroadGen's capabilities allow it to serve daily, millions of sellers at eBay with over 2.5 billion items.
Authors: Shenzhi Wang, Le Yu, Chang Gao, Chujie Zheng, Shixuan Liu, Rui Lu, Kai Dang, Xionghui Chen, Jianxin Yang, Zhenru Zhang, Yuqiong Liu, An Yang, Andrew Zhao, Yang Yue, Shiji Song, Bowen Yu, Gao Huang, Junyang Lin
Abstract: Reinforcement Learning with Verifiable Rewards (RLVR) has emerged as a powerful approach to enhancing the reasoning capabilities of Large Language Models (LLMs), while its mechanisms are not yet well understood. In this work, we undertake a pioneering exploration of RLVR through the novel perspective of token entropy patterns, comprehensively analyzing how different tokens influence reasoning performance. By examining token entropy patterns in Chain-of-Thought (CoT) reasoning, we observe that only a small fraction of tokens exhibit high entropy, and these tokens act as critical forks that steer the model toward diverse reasoning pathways. Furthermore, studying how entropy patterns evolve during RLVR training reveals that RLVR largely adheres to the base model's entropy patterns, primarily adjusting the entropy of high-entropy tokens. These findings highlight the significance of high-entropy tokens (i.e., forking tokens) to RLVR. We ultimately improve RLVR by restricting policy gradient updates to forking tokens and uncover a finding even beyond the 80/20 rule: utilizing only 20% of the tokens while maintaining performance comparable to full-gradient updates on the Qwen3-8B base model and significantly surpassing full-gradient updates on the Qwen3-32B (+11.04 on AIME'25 and +7.71 on AIME'24) and Qwen3-14B (+4.79 on AIME'25 and +5.21 on AIME'24) base models, highlighting a strong scaling trend. In contrast, training exclusively on the 80% lowest-entropy tokens leads to a marked decline in performance. These findings indicate that the efficacy of RLVR primarily arises from optimizing the high-entropy tokens that decide reasoning directions. Collectively, our results highlight the potential to understand RLVR through a token-entropy perspective and optimize RLVR by leveraging high-entropy minority tokens to further improve LLM reasoning.
Authors: Stella Huang, Qing Zhou
Abstract: Recent advances have established the identifiability of a directed acyclic graph (DAG) under additive noise models (ANMs), spurring the development of various causal discovery methods. However, most existing methods make restrictive model assumptions, rely heavily on general independence tests, or require substantial computational time. To address these limitations, we propose a sequential procedure to orient undirected edges in a completed partial DAG (CPDAG), representing an equivalence class of DAGs, by leveraging the pairwise additive noise model (PANM) to identify their causal directions. We prove that this procedure can recover the true causal DAG assuming a restricted ANM. Building on this result, we develop a novel constraint-based algorithm for learning causal DAGs under nonlinear ANMs. Given an estimated CPDAG, we develop a ranking procedure that sorts undirected edges by their adherence to the PANM, which defines an evaluation order of the edges. To determine the edge direction, we devise a statistical test that compares the log-likelihood values, evaluated with respect to the competing directions, of a sub-graph comprising just the candidate nodes and their identified parents in the partial DAG. We further establish the structural learning consistency of our algorithm in the large-sample limit. Extensive experiments on synthetic and real-world datasets demonstrate that our method is computationally efficient, robust to model misspecification, and consistently outperforms many existing nonlinear DAG learning methods.
Authors: Elim Kwan, Rehman Qureshi, Liam Fletcher, Colin Laganier, Victoria Nockles, Richard Walters
Abstract: Cooperative autonomous robotic systems have significant potential for executing complex multi-task missions across space, air, ground, and maritime domains. But they commonly operate in remote, dynamic and hazardous environments, requiring rapid in-mission adaptation without reliance on fragile or slow communication links to centralised compute. Fast, on-board replanning algorithms are therefore needed to enhance resilience. Reinforcement Learning shows strong promise for efficiently solving mission planning tasks when formulated as Travelling Salesperson Problems (TSPs), but existing methods: 1) are unsuitable for replanning, where agents do not start at a single location; 2) do not allow cooperation between agents; 3) are unable to model tasks with variable durations; or 4) lack practical considerations for on-board deployment. Here we define the Cooperative Mission Replanning Problem as a novel variant of multiple TSP with adaptations to overcome these issues, and develop a new encoder/decoder-based model using Graph Attention Networks and Attention Models to solve it effectively and efficiently. Using a simple example of cooperative drones, we show our replanner consistently (90% of the time) maintains performance within 10% of the state-of-the-art LKH3 heuristic solver, whilst running 85-370 times faster on a Raspberry Pi. This work paves the way for increased resilience in autonomous multi-agent systems.
Authors: Yuqi Zhu, Yi Zhong, Jintian Zhang, Ziheng Zhang, Shuofei Qiao, Yujie Luo, Lun Du, Da Zheng, Ningyu Zhang, Huajun Chen
Abstract: Large Language Models (LLMs) hold promise in automating data analysis tasks, yet open-source models face significant limitations in these kinds of reasoning-intensive scenarios. In this work, we investigate strategies to enhance the data analysis capabilities of open-source LLMs. By curating a seed dataset of diverse, realistic scenarios, we evaluate model behavior across three core dimensions: data understanding, code generation, and strategic planning. Our analysis reveals three key findings: (1) Strategic planning quality serves as the primary determinant of model performance; (2) Interaction design and task complexity significantly influence reasoning capabilities; (3) Data quality demonstrates a greater impact than diversity in achieving optimal performance. We leverage these insights to develop a data synthesis methodology, demonstrating significant improvements in open-source LLMs' analytical reasoning capabilities. Code is available at https://github.com/zjunlp/DataMind.
Authors: Jacky H. T. Yip, Charles Arnal, Fran\c{c}ois Charton, Gary Shiu
Abstract: Fine, regular, and star triangulations (FRSTs) of four-dimensional reflexive polytopes give rise to toric varieties, within which generic anticanonical hypersurfaces yield smooth Calabi-Yau threefolds. We introduce CYTransformer, a deep learning model based on the transformer architecture, to automate the generation of FRSTs. We demonstrate that CYTransformer efficiently and unbiasedly samples FRSTs for polytopes across a range of sizes, and can self-improve through retraining on its own output. These results lay the foundation for AICY: a community-driven platform designed to combine self-improving machine learning models with a continuously expanding database to explore and catalog the Calabi-Yau landscape.
Authors: Monica L\u{a}c\u{a}tu\c{s}, Matthias M\"oller
Abstract: This study introduces a framework for learning a low-depth surrogate quantum circuit (SQC) that approximates the nonlinear, dissipative, and hence non-unitary Bhatnagar-Gross-Krook (BGK) collision operator in the lattice Boltzmann method (LBM) for the D2Q9 lattice. By appropriately selecting the quantum state encoding, circuit architecture, and measurement protocol, non-unitary dynamics emerge naturally within the physical population space. This approach removes the need for probabilistic algorithms relying on any ancilla qubits and post-selection to reproduce dissipation, or for multiple state copies to capture nonlinearity. The SQC is designed to preserve key physical properties of the BGK operator, including mass conservation, scale equivariance, and D8 equivariance, while momentum conservation is encouraged through penalization in the training loss. When compiled to the IBM Heron quantum processor's native gate set, assuming all-to-all qubit connectivity, the circuit requires only 724 native gates and operates locally on the velocity register, making it independent of the lattice size. The learned SQC is validated on two benchmark cases, the Taylor-Green vortex decay and the lid-driven cavity, showing accurate reproduction of vortex decay and flow recirculation. While integration of the SQC into a quantum LBM framework presently requires measurement and re-initialization at each timestep, the necessary steps towards a measurement-free formulation are outlined.
Authors: Sarat Chandra Bobbili, Ujwal Dinesha, Dheeraj Narasimha, Srinivas Shakkottai
Abstract: Inference-time alignment enables large language models (LLMs) to generate outputs aligned with end-user preferences without further training. Recent post-training methods achieve this by using small guidance models to modify token generation during inference. These methods typically optimize a reward function KL-regularized by the original LLM taken as the reference policy. A critical limitation, however, is their dependence on a pre-trained reward model, which requires fitting to human preference feedback--a potentially unstable process. In contrast, we introduce PITA, a novel framework that integrates preference feedback directly into the LLM's token generation, eliminating the need for a reward model. PITA learns a small preference-based guidance policy to modify token probabilities at inference time without LLM fine-tuning, reducing computational cost and bypassing the pre-trained reward model dependency. The problem is framed as identifying an underlying preference distribution, solved through stochastic search and iterative refinement of the preference-based guidance model. We evaluate PITA across diverse tasks, including mathematical reasoning and sentiment classification, demonstrating its effectiveness in aligning LLM outputs with user preferences.
Authors: F. Gallavotti, A. Zaccone
Abstract: We introduce a three-dimensional vectorial extension of the Hopfield associative-memory model in which each neuron is a unit vector on $S^2$ and synaptic couplings are $3\times 3$ blocks generated through a vectorial Hebbian rule. The resulting block-structured operator is mathematically analogous to the Hessian of amorphous solids and induces a rigid energy landscape with deep minima for stored patterns. Simulations and spectral analysis show that the vectorial network substantially outperforms the classical binary Hopfield model. For moderate connectivity, the critical storage ratio $\gamma_c$ grows approximately linearly with the coordination number $Z$, while for $Z\gtrsim 40$ a high-connectivity regime emerges in which $\gamma_c$ systematically exceeds the extrapolated low-$Z$ linear fit. At the same time, a persistent spectral gap separates pattern modes from the bulk and basins of attraction enlarge, yielding enhanced robustness to initialization noise. Thus geometric constraints combined with amorphous-solid-inspired structure produce associative memories with superior storage and retrieval performance, especially in the high-connectivity ($Z \gtrsim 20$-$30$) regime.
Authors: Markus Pettersson, Connor T. Jerzak, Adel Daoud
Abstract: Machine learning models trained on Earth observation data, such as satellite imagery, have demonstrated significant promise in predicting household-level wealth indices, enabling the creation of high-resolution wealth maps that can be leveraged across multiple causal trials while addressing chronic data scarcity in global development research. However, because standard training objectives prioritize overall predictive accuracy, these predictions often suffer from shrinkage toward the mean, leading to attenuated estimates of causal treatment effects and limiting their utility in policy evaluations. Existing debiasing methods, such as Prediction-Powered Inference (PPI), can handle this attenuation bias but require additional fresh ground-truth data at the downstream stage of causal inference, which restricts their applicability in data-scarce environments. We introduce and evaluate two post-hoc correction methods -- Linear Calibration Correction (LCC) and a Tweedie's correction approach -- that substantially reduce shrinkage-induced prediction bias without relying on newly collected labeled data. LCC applies a simple linear transformation estimated on a held-out calibration split; Tweedie's method locally de-shrink predictions using density score estimates and a noise scale learned upstream. We provide practical diagnostics for when a correction is warranted and discuss practical limitations. Across analytical results, simulations, and experiments with Demographic and Health Surveys (DHS) data, both approaches reduce attenuation; Tweedie's correction yields nearly unbiased treatment-effect estimates, enabling a "one map, many trials" paradigm. Although we demonstrate on EO-ML wealth mapping, the methods are not geospatial-specific: they apply to any setting where imputed outcomes are reused downstream (e.g., pollution indices, population density, or LLM-derived indicators).
Authors: Sahar Salimpour, Lei Fu, Kajetan Rachwa{\l}, Pascal Bertrand, Kevin O'Sullivan, Robert Jakob, Farhad Keramat, Leonardo Militano, Giovanni Toffetti, Harry Edelman, Jorge Pe\~na Queralta
Abstract: Foundation models, including large language models (LLMs) and vision-language models (VLMs), have recently enabled novel approaches to robot autonomy and human-robot interfaces. In parallel, vision-language-action models (VLAs) or large behavior models (LBMs) are increasing the dexterity and capabilities of robotic systems. This survey paper reviews works that advance agentic applications and architectures, including initial efforts with GPT-style interfaces and more complex systems where AI agents function as coordinators, planners, perception actors, or generalist interfaces. Such agentic architectures allow robots to reason over natural language instructions, invoke APIs, plan task sequences, or assist in operations and diagnostics. In addition to peer-reviewed research, due to the fast-evolving nature of the field, we highlight and include community-driven projects, ROS packages, and industrial frameworks that show emerging trends. We propose a taxonomy for classifying model integration approaches and present a comparative analysis of the role that agents play in different solutions in today's literature.
Authors: Ratun Rahman, Atit Pokharel, Md Raihan Uddin, Dinh C. Nguyen
Abstract: Quantum federated learning (QFL) is an emerging field that has the potential to revolutionize computation by taking advantage of quantum physics concepts in a distributed machine learning (ML) environment. However, the majority of available quantum simulators are primarily built for general quantum circuit simulation and do not include integrated support for machine learning tasks such as training, evaluation, and iterative optimization. Furthermore, designing and assessing quantum learning algorithms is still a difficult and resource-intensive task. Real-time updates are essential for observing model convergence, debugging quantum circuits, and making conscious choices during training with the use of limited resources. Furthermore, most current simulators fail to support the integration of user-specific data for training purposes, undermining the main purpose of using a simulator. In this study, we introduce SimQFL, a customized simulator that simplifies and accelerates QFL experiments in quantum network applications. SimQFL supports real-time, epoch-wise output development and visualization, allowing researchers to monitor the process of learning across each training round. Furthermore, SimQFL offers an intuitive and visually appealing interface that facilitates ease of use and seamless execution. Users can customize key variables such as the number of epochs, learning rates, number of clients, and quantum hyperparameters such as qubits and quantum layers, making the simulator suitable for various QFL applications. The system gives immediate feedback following each epoch by showing intermediate outcomes and dynamically illustrating learning curves. SimQFL is a practical and interactive platform enabling academics and developers to prototype, analyze, and tune quantum neural networks with greater transparency and control in distributed quantum networks.
Authors: Sojun Ono, Kazuyuki Sugimura
Abstract: We present a neural-network emulator for the thermal and chemical evolution in Population III star formation. The emulator accurately reproduces the thermochemical evolution over a wide density range spanning 21 orders of magnitude (10$^{-3}$-10$^{18}$ cm$^{-3}$), tracking six primordial species: H, H$_2$, e$^{-}$, H$^{+}$, H$^{-}$, and H$_2^{+}$. To handle the broad dynamic range, we partition the density range into five subregions and train separate deep operator networks (DeepONets) in each region. When applied to randomly sampled thermochemical states, the emulator achieves relative errors below 10% in over 90% of cases for both temperature and chemical abundances (except for the rare species H$_2^{+}$). The emulator is roughly ten times faster on a CPU and more than 1000 times faster for batched predictions on a GPU, compared with conventional numerical integration. Furthermore, to ensure robust predictions under many iterations, we introduce a novel timescale-based update method, where a short-timestep update of each variable is computed by rescaling the predicted change over a longer timestep equal to its characteristic variation timescale. In one-zone collapse calculations, the results from the timescale-based method agree well with traditional numerical integration even with many iterations at a timestep as short as 10$^{-4}$ of the free-fall time. This proof-of-concept study suggests the potential for neural network-based chemical emulators to accelerate hydrodynamic simulations of star formation.
Authors: Kang-Hyun Lee, Faez Ahmed
Abstract: A major obstacle to establishing reliable structure-property (SP) linkages in materials engineering is the scarcity of diverse 3D microstructure datasets. Limited dataset availability and insufficient control over the analysis and design space restrict the variety of achievable microstructure morphologies, hindering progress in solving the inverse (property-to-structure) design problem. To address these challenges, we introduce MicroLad, a latent diffusion framework specifically designed for reconstructing 3D microstructures from 2D data. Trained on 2D images and employing multi-plane denoising diffusion sampling in the latent space, the framework reliably generates stable and coherent 3D volumes that remain statistically consistent with the original data. While this reconstruction capability enables dimensionality expansion (2D-to-3D) for generating statistically equivalent 3D samples from 2D data, effective exploration of microstructure design requires methods to guide the generation process toward specific objectives. To achieve this, MicroLad integrates score distillation sampling (SDS), which combines a differentiable score loss with microstructural descriptor-matching and property-alignment terms. This approach updates encoded 2D slices of the 3D volume in the latent space, enabling robust inverse-controlled 2D-to-3D microstructure generation. Consequently, the method facilitates exploration of an expanded 3D microstructure analysis and design space in terms of both microstructural descriptors and material properties.
Authors: Zichun Hao, Raghav Kansal, Abhijith Gandrakota, Chang Sun, Ngadiuba Jennifer, Javier Duarte, Maria Spiropulu
Abstract: A common challenge with supervised machine learning (ML) in high energy physics (HEP) is the reliance on simulations for labeled data, which can often mismodel the underlying collision or detector response. To help mitigate this problem of domain shift, we propose RINO (Renormalization Group Invariance with No Labels), a self-supervised learning approach that can instead pretrain models directly on collision data, learning embeddings invariant to renormalization group flow scales. In this work, we pretrain a transformer-based model on jets originating from quantum chromodynamic (QCD) interactions from the JetClass dataset, emulating real QCD-dominated experimental data, and then finetune on the JetNet dataset -- emulating simulations -- for the task of identifying jets originating from top quark decays. RINO demonstrates improved generalization from the JetNet training data to JetClass data compared to supervised training on JetNet from scratch, demonstrating the potential for RINO pretraining on real collision data followed by fine-tuning on small, high-quality MC datasets, to improve the robustness of ML models in HEP.
Authors: Zakaria El Kassimi, Fares Fourati, Mohamed-Slim Alouini
Abstract: We study question answering in the domain of radio regulations, a legally sensitive and high-stakes area. We propose a telecom-specific Retrieval-Augmented Generation (RAG) pipeline and introduce, to our knowledge, the first multiple-choice evaluation set for this domain, constructed from authoritative sources using automated filtering and human validation. To assess retrieval quality, we define a domain-specific retrieval metric, under which our retriever achieves approximately 97% accuracy. Beyond retrieval, our approach consistently improves generation accuracy across all tested models. In particular, while naively inserting documents without structured retrieval yields only marginal gains for GPT-4o (less than 1%), applying our pipeline results in nearly a 12% relative improvement. These findings demonstrate that carefully targeted grounding provides a simple yet strong baseline and an effective domain-specific solution for regulatory question answering. All code and evaluation scripts, along with our derived question-answer dataset, are available at https://github.com/Zakaria010/Radio-RAG.
Authors: Lanxiao Huang, Daksh Dave, Tyler Cody, Peter Beling, Ming Jin
Abstract: Large language models (LLMs) are increasingly used to automate or augment penetration testing, but their effectiveness and reliability across attack phases remain unclear. We present a comprehensive evaluation of multiple LLM-based agents, from single-agent to modular designs, across realistic penetration testing scenarios, measuring empirical performance and recurring failure patterns. We also isolate the impact of five core functional capabilities via targeted augmentations: Global Context Memory (GCM), Inter-Agent Messaging (IAM), Context-Conditioned Invocation (CCI), Adaptive Planning (AP), and Real-Time Monitoring (RTM). These interventions support, respectively: (i) context coherence and retention, (ii) inter-component coordination and state management, (iii) tool use accuracy and selective execution, (iv) multi-step strategic planning, error detection, and recovery, and (v) real-time dynamic responsiveness. Our results show that while some architectures natively exhibit subsets of these properties, targeted augmentations substantially improve modular agent performance, especially in complex, multi-step, and real-time penetration testing tasks.
Authors: Shaofeng Yin, Yanjie Ze, Hong-Xing Yu, C. Karen Liu, Jiajun Wu
Abstract: Humanoid loco-manipulation in unstructured environments demands tight integration of egocentric perception and whole-body control. However, existing approaches either depend on external motion capture systems or fail to generalize across diverse tasks. We introduce VisualMimic, a visual sim-to-real framework that unifies egocentric vision with hierarchical whole-body control for humanoid robots. VisualMimic combines a task-agnostic low-level keypoint tracker -- trained from human motion data via a teacher-student scheme -- with a task-specific high-level policy that generates keypoint commands from visual and proprioceptive input. To ensure stable training, we inject noise into the low-level policy and clip high-level actions using human motion statistics. VisualMimic enables zero-shot transfer of visuomotor policies trained in simulation to real humanoid robots, accomplishing a wide range of loco-manipulation tasks such as box lifting, pushing, football dribbling, and kicking. Beyond controlled laboratory settings, our policies also generalize robustly to outdoor environments. Videos are available at: https://visualmimic.github.io .
Authors: Fr\'ed\'eric Berdoz, Luca A. Lanzend\"orfer, Nick Tuninga, Roger Wattenhofer
Abstract: Prompt-driven scene synthesis allows users to generate complete 3D environments from textual descriptions. Current text-to-scene methods often struggle with complex geometries and object transformations, and tend to show weak adherence to complex instructions. We address these limitations by introducing Reason-3D, a text-to-scene model powered by large reasoning models (LRMs). Reason-3D integrates object retrieval using captions covering physical, functional, and contextual attributes. Reason-3D then places the selected objects based on implicit and explicit layout constraints, and refines their positions with collision-aware spatial reasoning. Evaluated on instructions ranging from simple to complex indoor configurations, Reason-3D significantly outperforms previous methods in human-rated visual fidelity, adherence to constraints, and asset retrieval quality. Beyond its contribution to the field of text-to-scene generation, our work showcases the advanced spatial reasoning abilities of modern LRMs. Additionally, we release the codebase to further the research in object retrieval and placement with LRMs.
Authors: Mariam Mahran, Katharina Simbeck
Abstract: Large Language Models (LLMs) are trained on massive, unstructured corpora, making it unclear which social patterns and biases they absorb and later reproduce. Existing evaluations typically examine outputs or activations, but rarely connect them back to the pre-training data. We introduce a pipeline that couples LLMs with sparse autoencoders (SAEs) to trace how different themes are encoded during training. As a controlled case study, we trained a GPT-style model on 37 nineteenth-century novels by ten female authors, a corpus centered on themes such as gender, marriage, class, and morality. By applying SAEs across layers and probing with eleven social and moral categories, we mapped sparse features to human-interpretable concepts. The analysis revealed stable thematic backbones (most prominently around gender and kinship) and showed how associations expand and entangle with depth. More broadly, we argue that the LLM+SAEs pipeline offers a scalable framework for auditing how cultural assumptions from the data are embedded in model representations.
Authors: Eduardo Fernandes Montesuma, Yassir Bendou, Mike Gartrell
Abstract: Wasserstein barycenters provide a powerful tool for aggregating probability measures, while leveraging the geometry of their ambient space. Existing discrete methods suffer from poor scalability, as they require access to the complete set of samples from input measures. We address this issue by recasting the original barycenter problem as a gradient flow in the Wasserstein space. Our approach offers two advantages. First, we achieve scalability by sampling mini-batches from the input measures. Second, we incorporate functionals over probability measures, which regularize the barycenter problem through internal, potential, and interaction energies. We present two algorithms for empirical and Gaussian mixture measures, providing convergence guarantees under the Polyak-{\L}ojasiewicz inequality. Experimental validation on toy datasets and domain adaptation benchmarks show that our methods outperform previous discrete and neural net-based methods for computing Wasserstein barycenters.
Authors: Aditi Dutta
Abstract: Addressing climate change effectively requires more than cataloguing the number of policies in place; it calls for tools that can reveal their thematic priorities and their tangible impacts on development outcomes. Existing assessments often rely on qualitative descriptions or composite indices, which can mask crucial differences between key domains such as mitigation, adaptation, disaster risk management, and loss and damage. To bridge this gap, we develop a quantitative indicator of climate policy orientation by applying a multilingual transformer-based language model to official national policy documents, achieving a classification accuracy of 0.90 (F1-score). Linking these indicators with World Bank development data in panel regressions reveals that mitigation policies are associated with higher GDP and GNI; disaster risk management correlates with greater GNI and debt but reduced foreign direct investment; adaptation and loss and damage show limited measurable effects. This integrated NLP-econometric framework enables comparable, theme-specific analysis of climate governance, offering a scalable method to monitor progress, evaluate trade-offs, and align policy emphasis with development goals.
Authors: Nguyen Linh Bao Nguyen, Alsharif Abuadbba, Kristen Moore, Tingmin Wu
Abstract: The rapid advancement of generative models has enabled the creation of increasingly stealthy synthetic voices, commonly referred to as audio deepfakes. A recent technique, FOICE [USENIX'24], demonstrates a particularly alarming capability: generating a victim's voice from a single facial image, without requiring any voice sample. By exploiting correlations between facial and vocal features, FOICE produces synthetic voices realistic enough to bypass industry-standard authentication systems, including WeChat Voiceprint and Microsoft Azure. This raises serious security concerns, as facial images are far easier for adversaries to obtain than voice samples, dramatically lowering the barrier to large-scale attacks. In this work, we investigate two core research questions: (RQ1) can state-of-the-art audio deepfake detectors reliably detect FOICE-generated speech under clean and noisy conditions, and (RQ2) whether fine-tuning these detectors on FOICE data improves detection without overfitting, thereby preserving robustness to unseen voice generators such as SpeechT5. Our study makes three contributions. First, we present the first systematic evaluation of FOICE detection, showing that leading detectors consistently fail under both standard and noisy conditions. Second, we introduce targeted fine-tuning strategies that capture FOICE-specific artifacts, yielding significant accuracy improvements. Third, we assess generalization after fine-tuning, revealing trade-offs between specialization to FOICE and robustness to unseen synthesis pipelines. These findings expose fundamental weaknesses in today's defenses and motivate new architectures and training protocols for next-generation audio deepfake detection.
Authors: Ayanava Dasgupta, Naqueeb Ahmad Warsi, Masahito Hayashi
Abstract: We study quantum differential privacy (QDP) by defining a notion of the order of informativeness between two pairs of quantum states. In particular, we show that if the hypothesis testing divergence of the one pair dominates over that of the other pair, then this dominance holds for every $f$-divergence. This approach completely characterizes $(\varepsilon,\delta)$-QDP mechanisms by identifying the most informative $(\varepsilon,\delta)$-DP quantum state pairs. We apply this to analyze the stability of quantum differentially private learning algorithms, generalizing classical results to the case $\delta>0$. Additionally, we study precise limits for privatized hypothesis testing and privatized quantum parameter estimation, including tight upper-bounds on the quantum Fisher information under QDP. Finally, we establish near-optimal contraction bounds for differentially private quantum channels with respect to the hockey-stick divergence.
Authors: Zhangsong Li
Abstract: We study the computational task of detecting and estimating correlated signals in a pair of spiked Wigner matrices. Our model consists of observations $$ X = \tfrac{\lambda}{\sqrt{n}} xx^{\top} + W \,, \quad Y = \tfrac{\mu}{\sqrt{n}} yy^{\top} + Z \,. $$ where $x,y \in \mathbb R^n$ are signal vectors with norm $\|x\|,\|y\| \approx\sqrt{n}$ and correlation $\langle x,y \rangle \approx \rho\|x\|\|y\|$, while $W,Z$ are independent Gaussian Wigner matrices. We propose an efficient algorithm that succeeds whenever $F(\lambda,\mu,\rho)>1$, where $$ F(\lambda,\mu,\rho)=\max\Big\{ \lambda,\mu, \frac{ \lambda^2 \rho^2 }{ 1-\lambda^2+\lambda^2 \rho^2 } + \frac{ \mu^2 \rho^2 }{ 1-\mu^2+\mu^2 \rho^2 } \Big\} \,. $$ Our result shows that an algorithm can leverage the correlation between the spikes to detect and estimate the signals even in regimes where efficiently recovering either $x$ from $X$ alone or $y$ from $Y$ alone is believed to be computationally infeasible. We complement our algorithmic result with evidence for a matching computational lower bound. In particular, we prove that when $F(\lambda,\mu,\rho)<1$, all algorithms based on {\em low-degree polynomials} fails to distinguish $(X,Y)$ with two independent Wigner matrices. This low-degree analysis strongly suggests that $F(\lambda,\mu,\rho)=1$ is the precise computation threshold for this problem.
Authors: Yifei Zhang, Jiashuo Zhang, Mojtaba Safari, Xiaofeng Yang, Liang Zhao
Abstract: Low-dose chest computed tomography (LDCT) inherently captures both pulmonary and cardiac structures, offering a unique opportunity for joint assessment of lung and cardiovascular health. However, most existing approaches treat these domains as independent tasks, overlooking their physiological interplay and shared imaging biomarkers. We propose an Explainable Cross-Disease Reasoning Framework that enables interpretable cardiopulmonary risk assessment from a single LDCT scan. The framework introduces an agentic reasoning process that emulates clinical diagnostic thinking-first perceiving pulmonary findings, then reasoning through established medical knowledge, and finally deriving a cardiovascular judgment with explanatory rationale. It integrates three synergistic components: a pulmonary perception module that summarizes lung abnormalities, a knowledge-guided reasoning module that infers their cardiovascular implications, and a cardiac representation module that encodes structural biomarkers. Their outputs are fused to produce a holistic cardiovascular risk prediction that is both accurate and physiologically grounded. Experiments on the NLST cohort demonstrate that the proposed framework achieves state-of-the-art performance for CVD screening and mortality prediction, outperforming single-disease and purely image-based baselines. Beyond quantitative gains, the framework provides human-verifiable reasoning that aligns with cardiological understanding, revealing coherent links between pulmonary abnormalities and cardiac stress mechanisms. Overall, this work establishes a unified and explainable paradigm for cardiovascular analysis from LDCT, bridging the gap between image-based prediction and mechanism-based medical interpretation.
Authors: Jiaqi Wang, Liutao Yu, Xiongri Shen, Sihang Guo, Chenlin Zhou, Leilei Zhao, Yi Zhong, Zhiguo Zhang, Zhengyu Ma
Abstract: Spiking neural networks (SNNs) offer a promising path toward energy-efficient speech command recognition (SCR) by leveraging their event-driven processing paradigm. However, existing SNN-based SCR methods often struggle to capture rich temporal dependencies and contextual information from speech due to limited temporal modeling and binary spike-based representations. To address these challenges, we first introduce the multi-view spiking temporal-aware self-attention (MSTASA) module, which combines effective spiking temporal-aware attention with a multi-view learning framework to model complementary temporal dependencies in speech commands. Building on MSTASA, we further propose SpikCommander, a fully spike-driven transformer architecture that integrates MSTASA with a spiking contextual refinement channel MLP (SCR-MLP) to jointly enhance temporal context modeling and channel-wise feature integration. We evaluate our method on three benchmark datasets: the Spiking Heidelberg Dataset (SHD), the Spiking Speech Commands (SSC), and the Google Speech Commands V2 (GSC). Extensive experiments demonstrate that SpikCommander consistently outperforms state-of-the-art (SOTA) SNN approaches with fewer parameters under comparable time steps, highlighting its effectiveness and efficiency for robust speech command recognition.
Authors: Giorgio Piras, Raffaele Mura, Fabio Brau, Luca Oneto, Fabio Roli, Battista Biggio
Abstract: Refusal refers to the functional behavior enabling safety-aligned language models to reject harmful or unethical prompts. Following the growing scientific interest in mechanistic interpretability, recent work encoded refusal behavior as a single direction in the model's latent space; e.g., computed as the difference between the centroids of harmful and harmless prompt representations. However, emerging evidence suggests that concepts in LLMs often appear to be encoded as a low-dimensional manifold embedded in the high-dimensional latent space. Motivated by these findings, we propose a novel method leveraging Self-Organizing Maps (SOMs) to extract multiple refusal directions. To this end, we first prove that SOMs generalize the prior work's difference-in-means technique. We then train SOMs on harmful prompt representations to identify multiple neurons. By subtracting the centroid of harmless representations from each neuron, we derive a set of multiple directions expressing the refusal concept. We validate our method on an extensive experimental setup, demonstrating that ablating multiple directions from models' internals outperforms not only the single-direction baseline but also specialized jailbreak algorithms, leading to an effective suppression of refusal. Finally, we conclude by analyzing the mechanistic implications of our approach.
Authors: Po-Heng Chou, Da-Chih Lin, Hung-Yu Wei, Walid Saad, Yu Tsao
Abstract: In this paper, a measurement-driven framework is proposed for early radio link failure (RLF) prediction in 5G non-standalone (NSA) railway environments. Using 10 Hz metro-train traces with serving and neighbor-cell indicators, we benchmark six models, namely CNN, LSTM, XGBoost, Anomaly Transformer, PatchTST, and TimesNet, under varied observation windows and prediction horizons. When the observation window is three seconds, TimesNet attains the highest F1 score with a three-second prediction horizon, while CNN provides a favorable accuracy-latency tradeoff with a two-second horizon, enabling proactive actions such as redundancy and adaptive handovers. The results indicate that deep temporal models can anticipate reliability degradations several seconds in advance using lightweight features available on commercial devices, offering a practical path to early-warning control in 5G-based railway systems.
Authors: Yipeng Zou, Qin Liu, Jie Wu, Yu Peng, Guo Chen, Hui Zhou, Guanghui Ye
Abstract: Ensemble attacks integrate the outputs of surrogate models with diverse architectures, which can be combined with various gradient-based attacks to improve adversarial transferability. However, previous work shows unsatisfactory attack performance when transferring across heterogeneous model architectures. The main reason is that the gradient update directions of heterogeneous surrogate models differ widely, making it hard to reduce the gradient variance of ensemble models while making the best of individual model. To tackle this challenge, we design a novel ensemble attack, NAMEA, which for the first time integrates the gradients from the non-attention areas of ensemble models into the iterative gradient optimization process. Our design is inspired by the observation that the attention areas of heterogeneous models vary sharply, thus the non-attention areas of ViTs are likely to be the focus of CNNs and vice versa. Therefore, we merge the gradients respectively from the attention and non-attention areas of ensemble models so as to fuse the transfer information of CNNs and ViTs. Specifically, we pioneer a new way of decoupling the gradients of non-attention areas from those of attention areas, while merging gradients by meta-learning. Empirical evaluations on ImageNet dataset indicate that NAMEA outperforms AdaEA and SMER, the state-of-the-art ensemble attacks by an average of 15.0% and 9.6%, respectively. This work is the first attempt to explore the power of ensemble non-attention in boosting cross-architecture transferability, providing new insights into launching ensemble attacks.
Authors: Hongru Zhao, Jinwen Fu, Tuan Pham
Abstract: Self-consuming generative models have received significant attention over the last few years. In this paper, we study a self-consuming generative model with heterogeneous preferences that is a generalization of the model in Ferbach et al. (2024). The model is retrained round by round using real data and its previous-round synthetic outputs. The asymptotic behavior of the retraining dynamics is investigated across four regimes using different techniques including the nonlinear Perron--Frobenius theory. Our analyses improve upon that of Ferbach et al. (2024) and provide convergence results in settings where the well-known Banach contraction mapping arguments do not apply. Stability and non-stability results regarding the retraining dynamics are also given.
Authors: PAN Team, Jiannan Xiang, Yi Gu, Zihan Liu, Zeyu Feng, Qiyue Gao, Yiyan Hu, Benhao Huang, Guangyi Liu, Yichi Yang, Kun Zhou, Davit Abrahamyan, Arif Ahmad, Ganesh Bannur, Junrong Chen, Kimi Chen, Mingkai Deng, Ruobing Han, Xinqi Huang, Haoqiang Kang, Zheqi Li, Enze Ma, Hector Ren, Yashowardhan Shinde, Rohan Shingre, Ramsundar Tanikella, Kaiming Tao, Dequan Yang, Xinle Yu, Cong Zeng, Binglin Zhou, Zhengzhong Liu, Zhiting Hu, Eric P. Xing
Abstract: A world model enables an intelligent agent to imagine, predict, and reason about how the world evolves in response to its actions, and accordingly to plan and strategize. While recent video generation models produce realistic visual sequences, they typically operate in the prompt-to-full-video manner without causal control, interactivity, or long-horizon consistency required for purposeful reasoning. Existing world modeling efforts, on the other hand, often focus on restricted domains (e.g., physical, game, or 3D-scene dynamics) with limited depth and controllability, and struggle to generalize across diverse environments and interaction formats. In this work, we introduce PAN, a general, interactable, and long-horizon world model that predicts future world states through high-quality video simulation conditioned on history and natural language actions. PAN employs the Generative Latent Prediction (GLP) architecture that combines an autoregressive latent dynamics backbone based on a large language model (LLM), which grounds simulation in extensive text-based knowledge and enables conditioning on language-specified actions, with a video diffusion decoder that reconstructs perceptually detailed and temporally coherent visual observations, to achieve a unification between latent space reasoning (imagination) and realizable world dynamics (reality). Trained on large-scale video-action pairs spanning diverse domains, PAN supports open-domain, action-conditioned simulation with coherent, long-term dynamics. Extensive experiments show that PAN achieves strong performance in action-conditioned world simulation, long-horizon forecasting, and simulative reasoning compared to other video generators and world models, taking a step towards general world models that enable predictive simulation of future world states for reasoning and acting.
Authors: Prajwal Singhania, Siddharth Singh, Lannie Dalton Hough, Akarsh Srivastava, Harshitha Menon, Charles Fredrick Jekel, Abhinav Bhatele
Abstract: As large language models (LLMs) continue to grow in size, distributed inference has become increasingly important. Model-parallel strategies must now efficiently scale not only across multiple GPUs but also across multiple nodes. In this work, we present a detailed performance study of multi-node distributed inference using LLMs on GPU-based supercomputers. We conduct experiments with several state-of-the-art inference engines alongside YALIS, a research-oriented prototype engine designed for controlled experimentation. We analyze the strong-scaling behavior of different model-parallel schemes and identify key bottlenecks. Since all-reduce operations are a common performance bottleneck, we develop NVRAR, a hierarchical all-reduce algorithm based on recursive doubling with NVSHMEM. NVRAR achieves up to 1.9x-3.6x lower latency than NCCL for message sizes between 128 KB and 2 MB on HPE Slingshot and InfiniBand interconnects. Integrated into YALIS, NVRAR achieves up to a 1.72x reduction in end-to-end batch latency for the Llama 3.1 405B model in multi-node decode-heavy workloads using tensor parallelism.