# Research Papers — 2025-10 · bundle page 180 of 217 > Full-text markdown of papers #8951–9000 of 10843 published in 2025-10, newest first. AI-readable mirrors of the full 10843-paper 2025-10 archive (50 papers per file). Canonical page: https://aipod.dev/papers/archive/2025-10/page/180 [← Previous bundle](https://aipod.dev/papers/archive/2025-10/page/179.md) [Next bundle →](https://aipod.dev/papers/archive/2025-10/page/181.md) ### MuFFIN: Multifaceted Pronunciation Feedback Model with Interactive Hierarchical Neural Modeling Authors: Bi-Cheng Yan, Ming-Kang Tsai, Berlin Chen Published: 2025-10-06 Categories: eess.AS, cs.AI arXiv: https://arxiv.org/abs/2510.04956v2 Computer-assisted pronunciation training (CAPT) manages to facilitate second-language (L2) learners to practice pronunciation skills by offering timely and instructive feedback. To examine pronunciation proficiency from multiple facets, existing methods for CAPT broadly fall into two categories: mispronunciation detection and diagnosis (MDD) as well as automatic pronunciation assessment (APA). The former aims to pinpoint phonetic pronunciation errors and provide diagnostic feedback, while the latter seeks instead to quantify pronunciation proficiency pertaining to various aspects. Despite the natural complementarity between MDD and APA, researchers and practitioners, however, often treat them as independent tasks with disparate modeling paradigms. In light of this, we in this paper first introduce MuFFIN, a Multi-Faceted pronunciation Feedback model with an Interactive hierarchical Neural architecture, to jointly address the tasks of MDD and APA. To better capture the nuanced distinctions between phonemes in the feature space, a novel phoneme-contrastive ordinal regularization mechanism is then put forward to optimize the proposed model to generate more phoneme-discriminative features while factoring in the ordinality of the aspect scores. In addition, to address the intricate data imbalance problem in MDD, we design a simple yet effective training objective, which is specifically tailored to perturb the outputs of a phoneme classifier with the phoneme-specific variations, so as to better render the distribution of predicted phonemes meanwhile considering their mispronunciation characteristics. A series of experiments conducted on the Speechocean762 benchmark dataset demonstrates the efficacy of our method in relation to several cutting-edge baselines, showing state-of-the-art performance on both the APA and MDD tasks. --- ### Safe and Compliant Cross-Market Trade Execution via Constrained RL and Zero-Knowledge Audits Authors: Ailiya Borjigin, Cong He Published: 2025-10-06 Categories: cs.AI, cs.DC arXiv: https://arxiv.org/abs/2510.04952v2 We present a cross-market algorithmic trading system that balances execution quality with rigorous compliance enforcement. The architecture comprises a high-level planner, a reinforcement learning execution agent, and an independent compliance agent. We formulate trade execution as a constrained Markov decision process with hard constraints on participation limits, price bands, and self-trading avoidance. The execution agent is trained with proximal policy optimization, while a runtime action-shield projects any unsafe action into a feasible set. To support auditability without exposing proprietary signals, we add a zero-knowledge compliance audit layer that produces cryptographic proofs that all actions satisfied the constraints. We evaluate in a multi-venue, ABIDES-based simulator and compare against standard baselines (e.g., TWAP, VWAP). The learned policy reduces implementation shortfall and variance while exhibiting no observed constraint violations across stress scenarios including elevated latency, partial fills, compliance module toggling, and varying constraint limits. We report effects at the 95% confidence level using paired t-tests and examine tail risk via CVaR. We situate the work at the intersection of optimal execution, safe reinforcement learning, regulatory technology, and verifiable AI, and discuss ethical considerations, limitations (e.g., modeling assumptions and computational overhead), and paths to real-world deployment. --- ### Feasibility-Aware Decision-Focused Learning for Predicting Parameters in the Constraints Authors: Jayanta Mandi, Marianne Defresne, Senne Berden, Tias Guns Published: 2025-10-06 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2510.04951v2 When some parameters of a constrained optimization problem (COP) are uncertain, this gives rise to a predict-then-optimize (PtO) problem, comprising two stages: the prediction of the unknown parameters from contextual information and the subsequent optimization using those predicted parameters. Decision-focused learning (DFL) implements the first stage by training a machine learning (ML) model to optimize the quality of the decisions made using the predicted parameters. When the predicted parameters occur in the constraints, they can lead to infeasible solutions. Therefore, it is important to simultaneously manage both feasibility and decision quality. We develop a DFL framework for predicting constraint parameters in a generic COP. While prior works typically assume that the underlying optimization problem is a linear program (LP) or integer LP (ILP), our approach makes no such assumption. We derive two novel loss functions based on maximum likelihood estimation (MLE): the first one penalizes infeasibility (by penalizing predicted parameters that lead to infeasible solutions), while the second one penalizes suboptimal decisions (by penalizing predicted parameters that make the true optimal solution infeasible). We introduce a single tunable parameter to form a weighted average of the two losses, allowing decision-makers to balance suboptimality and feasibility. We experimentally demonstrate that adjusting this parameter provides decision-makers control over this trade-off. Moreover, across several COP instances, we show that adjusting the tunable parameter allows a decision-maker to prioritize either suboptimality or feasibility, outperforming the performance of existing baselines in either objective. --- ### Mind Your Tone: Investigating How Prompt Politeness Affects LLM Accuracy (short paper) Authors: Om Dobariya, Akhil Kumar Published: 2025-10-06 Categories: cs.CL, cs.AI, cs.LG, cs.NE, stat.ME arXiv: https://arxiv.org/abs/2510.04950v1 The wording of natural language prompts has been shown to influence the performance of large language models (LLMs), yet the role of politeness and tone remains underexplored. In this study, we investigate how varying levels of prompt politeness affect model accuracy on multiple-choice questions. We created a dataset of 50 base questions spanning mathematics, science, and history, each rewritten into five tone variants: Very Polite, Polite, Neutral, Rude, and Very Rude, yielding 250 unique prompts. Using ChatGPT 4o, we evaluated responses across these conditions and applied paired sample t-tests to assess statistical significance. Contrary to expectations, impolite prompts consistently outperformed polite ones, with accuracy ranging from 80.8% for Very Polite prompts to 84.8% for Very Rude prompts. These findings differ from earlier studies that associated rudeness with poorer outcomes, suggesting that newer LLMs may respond differently to tonal variation. Our results highlight the importance of studying pragmatic aspects of prompting and raise broader questions about the social dimensions of human-AI interaction. --- ### Bidirectional Mammogram View Translation with Column-Aware and Implicit 3D Conditional Diffusion Authors: Xin Li, Kaixiang Yang, Qiang Li, Zhiwei Wang Published: 2025-10-06 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2510.04947v1 Dual-view mammography, including craniocaudal (CC) and mediolateral oblique (MLO) projections, offers complementary anatomical views crucial for breast cancer diagnosis. However, in real-world clinical workflows, one view may be missing, corrupted, or degraded due to acquisition errors or compression artifacts, limiting the effectiveness of downstream analysis. View-to-view translation can help recover missing views and improve lesion alignment. Unlike natural images, this task in mammography is highly challenging due to large non-rigid deformations and severe tissue overlap in X-ray projections, which obscure pixel-level correspondences. In this paper, we propose Column-Aware and Implicit 3D Diffusion (CA3D-Diff), a novel bidirectional mammogram view translation framework based on conditional diffusion model. To address cross-view structural misalignment, we first design a column-aware cross-attention mechanism that leverages the geometric property that anatomically corresponding regions tend to lie in similar column positions across views. A Gaussian-decayed bias is applied to emphasize local column-wise correlations while suppressing distant mismatches. Furthermore, we introduce an implicit 3D structure reconstruction module that back-projects noisy 2D latents into a coarse 3D feature volume based on breast-view projection geometry. The reconstructed 3D structure is refined and injected into the denoising UNet to guide cross-view generation with enhanced anatomical awareness. Extensive experiments demonstrate that CA3D-Diff achieves superior performance in bidirectional tasks, outperforming state-of-the-art methods in visual fidelity and structural consistency. Furthermore, the synthesized views effectively improve single-view malignancy classification in screening settings, demonstrating the practical value of our method in real-world diagnostics. --- ### A First Context-Free Grammar Applied to Nawatl Corpora Augmentation Authors: Juan-José Guzmán-Landa, Juan-Manuel Torres-Moreno, Miguel Figueroa-Saavedra, Ligia Quintana-Torres, Martha-Lorena Avendaño-Garrido, Graham Ranger Published: 2025-10-06 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2510.04945v1 In this article we introduce a context-free grammar (CFG) for the Nawatl language. Nawatl (or Nahuatl) is an Amerindian language of the $π$-language type, i.e. a language with few digital resources, in which the corpora available for machine learning are virtually non-existent. The objective here is to generate a significant number of grammatically correct artificial sentences, in order to increase the corpora available for language model training. We want to show that a grammar enables us significantly to expand a corpus in Nawatl which we call $π$-\textsc{yalli}. The corpus, thus enriched, enables us to train algorithms such as FastText and to evaluate them on sentence-level semantic tasks. Preliminary results show that by using the grammar, comparative improvements are achieved over some LLMs. However, it is observed that to achieve more significant improvement, grammars that model the Nawatl language even more effectively are required. --- ### On Structured State-Space Duality Authors: Jerry Yao-Chieh Hu, Xiwen Zhang, Ali ElSheikh, Weimin Wu, Han Liu Published: 2025-10-06 Categories: cs.LG, cs.CL, cs.CV, stat.ML arXiv: https://arxiv.org/abs/2510.04944v2 Structured State-Space Duality (SSD) (Dao & Gu, ICML 2024) is an equivalence between a simple Structured State-Space Model (SSM) and a masked attention mechanism. In particular, a state-space model with a scalar-times-identity state matrix is equivalent to a masked self-attention with a $1$-semiseparable causal mask. Consequently, the same sequence transformation (model) has two algorithmic realizations: as a linear-time $O(T)$ recurrence or as a quadratic-time $O(T^2)$ attention. In this note, we formalize and generalize this duality: (i) we extend SSD from the scalar-identity case to general diagonal SSMs (diagonal state matrices); (ii) we show that these diagonal SSMs match the scalar case's training complexity lower bounds while supporting richer dynamics; (iii) we establish a necessary and sufficient condition under which an SSM is equivalent to $1$-semiseparable masked attention; and (iv) we show that such duality fails to extend to standard softmax attention due to rank explosion. Together, these results tighten bridge between recurrent SSMs and Transformers, and widen the design space for expressive yet efficient sequence models. --- ### Unsupervised Active Learning via Natural Feature Progressive Framework Authors: Yuxi Liu, Catherine Lalman, Yimin Yang Published: 2025-10-06 Categories: cs.CV, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2510.04939v1 The effectiveness of modern deep learning models is predicated on the availability of large-scale, human-annotated datasets, a process that is notoriously expensive and time-consuming. While Active Learning (AL) offers a strategic solution by labeling only the most informative and representative data, its iterative nature still necessitates significant human involvement. Unsupervised Active Learning (UAL) presents an alternative by shifting the annotation burden to a single, post-selection step. Unfortunately, prevailing UAL methods struggle to achieve state-of-the-art performance. These approaches typically rely on local, gradient-based scoring for sample importance estimation, which not only makes them vulnerable to ambiguous and noisy data but also hinders their capacity to select samples that adequately represent the full data distribution. Moreover, their use of shallow, one-shot linear selection falls short of a true UAL paradigm. In this paper, we propose the Natural Feature Progressive Framework (NFPF), a UAL method that revolutionizes how sample importance is measured. At its core, NFPF employs a Specific Feature Learning Machine (SFLM) to effectively quantify each sample's contribution to model performance. We further utilize the SFLM to define a powerful Reconstruction Difference metric for initial sample selection. Our comprehensive experiments show that NFPF significantly outperforms all established UAL methods and achieves performance on par with supervised AL methods on vision datasets. Detailed ablation studies and qualitative visualizations provide compelling evidence for NFPF's superior performance, enhanced robustness, and improved data distribution coverage. --- ### ONNX-Net: Towards Universal Representations and Instant Performance Prediction for Neural Architectures Authors: Shiwen Qin, Alexander Auras, Shay B. Cohen, Elliot J. Crowley, Michael Moeller, Linus Ericsson, Jovita Lukasik Published: 2025-10-06 Categories: cs.LG, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2510.04938v2 Neural architecture search (NAS) automates the design process of high-performing architectures, but remains bottlenecked by expensive performance evaluation. Most existing studies that achieve faster evaluation are mostly tied to cell-based search spaces and graph encodings tailored to those individual search spaces, limiting their flexibility and scalability when applied to more expressive search spaces. In this work, we aim to close the gap of individual search space restrictions and search space dependent network representations. We present ONNX-Bench, a benchmark consisting of a collection of neural networks in a unified format based on ONNX files. ONNX-Bench includes all open-source NAS-bench-based neural networks, resulting in a total size of more than 600k {architecture, accuracy} pairs. This benchmark allows creating a shared neural network representation, ONNX-Net, able to represent any neural architecture using natural language descriptions acting as an input to a performance predictor. This text-based encoding can accommodate arbitrary layer types, operation parameters, and heterogeneous topologies, enabling a single surrogate to generalise across all neural architectures rather than being confined to cell-based search spaces. Experiments show strong zero-shot performance across disparate search spaces using only a small amount of pretraining samples, enabling the unprecedented ability to evaluate any neural network architecture instantly. --- ### Discrete scalar curvature as a weighted sum of Ollivier-Ricci curvatures Authors: Abigail Hickok, Andrew J. Blumberg Published: 2025-10-06 Categories: cs.DM, cs.CG, cs.SI, stat.ML arXiv: https://arxiv.org/abs/2510.04936v1 We study the relationship between discrete analogues of Ricci and scalar curvature that are defined for point clouds and graphs. In the discrete setting, Ricci curvature is replaced by Ollivier-Ricci curvature. Scalar curvature can be computed as the trace of Ricci curvature for a Riemannian manifold; this motivates a new definition of a scalar version of Ollivier-Ricci curvature. We show that our definition converges to scalar curvature for nearest neighbor graphs obtained by sampling from a manifold. We also prove some new results about the convergence of Ollivier-Ricci curvature to Ricci curvature. --- ### MARS: Co-evolving Dual-System Deep Research via Multi-Agent Reinforcement Learning Authors: Guoxin Chen, Zile Qiao, Wenqing Wang, Donglei Yu, Xuanzhong Chen, Hao Sun, Minpeng Liao, Kai Fan, Yong Jiang, Penguin Xie, Wayne Xin Zhao, Ruihua Song et al. Published: 2025-10-06 Categories: cs.AI, cs.CL, cs.LG arXiv: https://arxiv.org/abs/2510.04935v2 Large Reasoning Models (LRMs) face two fundamental limitations: excessive token consumption when overanalyzing simple information processing tasks, and inability to access up-to-date knowledge beyond their training data. We introduce MARS (Multi-Agent System for Deep ReSearch), a novel co-evolution framework that jointly optimizes dual cognitive systems through multi-agent reinforcement learning. Unlike prior approaches that employ fixed or independently-trained summarizers, MARS enables System 1 (fast, intuitive processing) and System 2 (deliberate reasoning) to co-adapt through shared trajectory rewards, developing complementary strategies where System 1 learns to distill information specifically useful for System 2's reasoning. We extend Group Relative Policy Optimization (GRPO) for multi-agent settings with three key innovations: (1) decoupled gradient computation ensuring proper credit assignment despite shared rewards, (2) bin-packing optimization for efficient parallel information processing, and (3) advantage-weighted balanced sampling preventing training imbalance. Extensive experiments demonstrate that MARS (8B), trained under a challenging Zero RL setting without any supervised fine-tuning, achieves 8.17% on HLE -- outperforming WebThinker (32B with SFT, 6.87%) and narrowing the gap with proprietary models like Claude 3.7 Sonnet (7.89%) -- while achieving an average gain of 8.9% across 7 knowledge-intensive tasks. --- ### AURA Score: A Metric For Holistic Audio Question Answering Evaluation Authors: Satvik Dixit, Soham Deshmukh, Bhiksha Raj Published: 2025-10-06 Categories: eess.AS, cs.AI arXiv: https://arxiv.org/abs/2510.04934v1 Audio Question Answering (AQA) is a key task for evaluating Audio-Language Models (ALMs), yet assessing open-ended responses remains challenging. Existing metrics used for AQA such as BLEU, METEOR and BERTScore, mostly adapted from NLP and audio captioning, rely on surface similarity and fail to account for question context, reasoning, and partial correctness. To address the gap in literature, we make three contributions in this work. First, we introduce AQEval to enable systematic benchmarking of AQA metrics. It is the first benchmark of its kind, consisting of 10k model responses annotated by multiple humans for their correctness and relevance. Second, we conduct a comprehensive analysis of existing AQA metrics on AQEval, highlighting weak correlation with human judgment, especially for longer answers. Third, we propose a new metric - AURA score, to better evaluate open-ended model responses. On AQEval, AURA achieves state-of-the-art correlation with human ratings, significantly outperforming all baselines. Through this work, we aim to highlight the limitations of current AQA evaluation methods and motivate better metrics. We release both the AQEval benchmark and the AURA metric to support future research in holistic AQA evaluation. --- ### The Geometry of Truth: Layer-wise Semantic Dynamics for Hallucination Detection in Large Language Models Authors: Amir Hameed Mir Published: 2025-10-06 Categories: cs.CL, cs.AI, cs.IT, cs.LG, cs.NE arXiv: https://arxiv.org/abs/2510.04933v1 Large Language Models (LLMs) often produce fluent yet factually incorrect statements-a phenomenon known as hallucination-posing serious risks in high-stakes domains. We present Layer-wise Semantic Dynamics (LSD), a geometric framework for hallucination detection that analyzes the evolution of hidden-state semantics across transformer layers. Unlike prior methods that rely on multiple sampling passes or external verification sources, LSD operates intrinsically within the model's representational space. Using margin-based contrastive learning, LSD aligns hidden activations with ground-truth embeddings derived from a factual encoder, revealing a distinct separation in semantic trajectories: factual responses preserve stable alignment, while hallucinations exhibit pronounced semantic drift across depth. Evaluated on the TruthfulQA and synthetic factual-hallucination datasets, LSD achieves an F1-score of 0.92, AUROC of 0.96, and clustering accuracy of 0.89, outperforming SelfCheckGPT and Semantic Entropy baselines while requiring only a single forward pass. This efficiency yields a 5-20x speedup over sampling-based methods without sacrificing precision or interpretability. LSD offers a scalable, model-agnostic mechanism for real-time hallucination monitoring and provides new insights into the geometry of factual consistency within large language models. --- ### Egalitarian Gradient Descent: A Simple Approach to Accelerated Grokking Authors: Ali Saheb Pasand, Elvis Dohmatob Published: 2025-10-06 Categories: cs.LG arXiv: https://arxiv.org/abs/2510.04930v3 Grokking is the phenomenon whereby, unlike the training performance, which peaks early in the training process, the test/generalization performance of a model stagnates over arbitrarily many epochs and then suddenly jumps to usually close to perfect levels. In practice, it is desirable to reduce the length of such plateaus, that is to make the learning process "grok" faster. In this work, we provide new insights into grokking. First, we show both empirically and theoretically that grokking can be induced by asymmetric speeds of (stochastic) gradient descent, along different principal (i.e singular directions) of the gradients. We then propose a simple modification that normalizes the gradients so that dynamics along all the principal directions evolves at exactly the same speed. Then, we establish that this modified method, which we call egalitarian gradient descent (EGD) and can be seen as a carefully modified form of natural gradient descent, groks much faster. In fact, in some cases the stagnation is completely removed. Finally, we empirically show that on classical arithmetic problems such as modular addition and sparse parity problem which this stagnation has been widely observed and intensively studied, that our proposed method eliminates the plateaus. --- ### Federated Self-Supervised Modulation Classification under Non-IID and Imbalanced Data Authors: Usman Akram, Yiyue Chen, Haris Vikalo Published: 2025-10-06 Categories: cs.LG, cs.AI, eess.SP arXiv: https://arxiv.org/abs/2510.04927v2 Automatic modulation classification (AMC) is a core enabler of cognitive wireless systems, providing spectrum awareness and supporting adaptive communication at the network edge. However, training AMC models on centrally aggregated data incurs high communication overhead, raises privacy concerns, and often lacks robustness to real-world conditions. We propose FedSSL-AMC, a federated self-supervised framework for learning AMC models from sparsely labeled, distributed I/Q time-series data. Participating clients collaboratively train a causal, time-dilated CNN encoder using triplet-loss self-supervision on unlabeled signals, followed by lightweight local SVMs trained on limited labeled samples. This enables communication-round-efficient, robust representation learning under class imbalance and channel variability. We establish convergence guarantees for a proximal variant of the encoder-training procedure and derive a separability bound for the downstream classifier under feature noise. Experiments on synthetic and over-the-air datasets demonstrate improvements over supervised FL baselines across all three datasets and nearly all evaluated settings involving heterogeneous SNRs, carrier-frequency offsets, and non-IID label distributions. --- ### Set to Be Fair: Demographic Parity Constraints for Set-Valued Classification Authors: Eyal Cohen, Christophe Denis, Mohamed Hebiri Published: 2025-10-06 Categories: stat.ML, cs.LG arXiv: https://arxiv.org/abs/2510.04926v1 Set-valued classification is used in multiclass settings where confusion between classes can occur and lead to misleading predictions. However, its application may amplify discriminatory bias motivating the development of set-valued approaches under fairness constraints. In this paper, we address the problem of set-valued classification under demographic parity and expected size constraints. We propose two complementary strategies: an oracle-based method that minimizes classification risk while satisfying both constraints, and a computationally efficient proxy that prioritizes constraint satisfaction. For both strategies, we derive closed-form expressions for the (optimal) fair set-valued classifiers and use these to build plug-in, data-driven procedures for empirical predictions. We establish distribution-free convergence rates for violations of the size and fairness constraints for both methods, and under mild assumptions we also provide excess-risk bounds for the oracle-based approach. Empirical results demonstrate the effectiveness of both strategies and highlight the efficiency of our proxy method. --- ### REN: Anatomically-Informed Mixture-of-Experts for Interstitial Lung Disease Diagnosis Authors: Alec K. Peltekian, Halil Ertugrul Aktas, Gorkem Durak, Kevin Grudzinski, Bradford C. Bemiss, Carrie Richardson, Jane E. Dematte, G. R. Scott Budinger, Anthony J. Esposito, Alexander Misharin, Alok Choudhary, Ankit Agrawal et al. Published: 2025-10-06 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2510.04923v3 Mixture-of-Experts (MoE) architectures achieve scalable learning by routing inputs to specialized subnetworks through conditional computation. However, conventional MoE designs assume homogeneous expert capability and domain-agnostic routing-assumptions that are fundamentally misaligned with medical imaging, where anatomical structure and regional disease heterogeneity govern pathological patterns. We introduce Regional Expert Networks (REN), the first anatomically-informed MoE framework for medical image classification. REN encodes anatomical priors by training seven specialized experts, each dedicated to a distinct lung lobe or bilateral lung combination, enabling precise modeling of region-specific pathological variation. Multi-modal gating mechanisms dynamically integrate radiomics biomarkers with deep learning (DL) features extracted by convolutional (CNN), Transformer (ViT), and state-space (Mamba) architectures to weight expert contributions at inference. Applied to interstitial lung disease (ILD) classification on a 597-patient, 1,898-scan longitudinal cohort, REN achieves consistently superior performance: the radiomics-guided ensemble attains an average AUC of 0.8646 +- 0.0467, a +12.5 % improvement over the SwinUNETR single-model baseline (AUC 0.7685, p=0.031). Lower-lobe experts reach AUCs of 0.88-0.90, outperforming DL baselines (CNN: 0.76-0.79) and mirroring known patterns of basal ILD progression. Evaluated under rigorous patient-level cross-validation, REN demonstrates strong generalizability and clinical interpretability, establishing a scalable, anatomically-guided framework potentially extensible to other structured medical imaging tasks. Code is available on our GitHub https://github.com/NUBagciLab/MoE-REN. --- ### Do LLMs Align with My Task? Evaluating Text-to-SQL via Dataset Alignment Authors: Davood Rafiei, Morgan Lindsay Heisler, Weiwei Zhang, Mohammadreza Pourreza, Yong Zhang Published: 2025-10-06 Categories: cs.CL, cs.AI, cs.DB arXiv: https://arxiv.org/abs/2510.04919v1 Supervised Fine-Tuning (SFT) is an effective method for adapting Large Language Models (LLMs) on downstream tasks. However, variability in training data can hinder a model's ability to generalize across domains. This paper studies the problem of dataset alignment for Natural Language to SQL (NL2SQL or text to SQL), examining how well SFT training data matches the structural characteristics of target queries and how this alignment impacts model performance. We hypothesize that alignment can be accurately estimated by comparing the distributions of structural SQL features across the training set, target data, and the model's predictions prior to SFT. Through comprehensive experiments on three large cross-domain NL2SQL benchmarks and multiple model families, we show that structural alignment is a strong predictor of fine-tuning success. When alignment is high, SFT yields substantial gains in accuracy and SQL generation quality; when alignment is low, improvements are marginal or absent. These findings highlight the importance of alignment-aware data selection for effective fine-tuning and generalization in NL2SQL tasks. --- ### Semantics-Aware Hierarchical Consensus Learning for Remote Sensing Image Classification Authors: Giulio Weikmann, Gianmarco Perantoni, Lorenzo Bruzzone Published: 2025-10-06 Categories: cs.CV arXiv: https://arxiv.org/abs/2510.04916v4 Deep learning has become increasingly important in remote sensing image classification due to its ability to extract semantic information from complex data. Classification tasks often include predefined label hierarchies that represent the semantic relationships among classes. However, these hierarchies are frequently overlooked, and most approaches focus only on fine-grained classification schemes. In this paper, we present a novel Semantics-Aware Hierarchical Consensus (SAHC) approach that integrates hierarchical-level-specific classification heads within a deep network architecture and combines their output through cross-level probability projectors. Direct and projected predictions are fused into a geometric consensus distribution, which is used for self-consistent training and optional hierarchy-aware inference. This mechanism acts as a geometric ensemble that leverages the inherent structure of the hierarchical classification task. The projectors are initialized from the user-defined taxonomy (i.e., the hierarchical label structure), and can be adaptively refined during optimization. The proposed SAHC method is evaluated on two benchmark datasets with different degrees of hierarchical complexity on different tasks, considering varying spectral and spatial resolutions. Experimental results show both the effectiveness of the proposed approach in guiding network learning and the robustness of the hierarchical consensus for remote sensing image classification tasks. The source code is available at https://github.com/rslab-unitrento/sahc. --- ### Comparative Analysis of YOLOv5, Faster R-CNN, SSD, and RetinaNet for Motorbike Detection in Kigali Autonomous Driving Context Authors: Ngeyen Yinkfu, Sunday Nwovu, Jonathan Kayizzi, Angelique Uwamahoro Published: 2025-10-06 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2510.04912v1 In Kigali, Rwanda, motorcycle taxis are a primary mode of transportation, often navigating unpredictably and disregarding traffic rules, posing significant challenges for autonomous driving systems. This study compares four object detection models--YOLOv5, Faster R-CNN, SSD, and RetinaNet--for motorbike detection using a custom dataset of 198 images collected in Kigali. Implemented in PyTorch with transfer learning, the models were evaluated for accuracy, localization, and inference speed to assess their suitability for real-time navigation in resource-constrained settings. We identify implementation challenges, including dataset limitations and model complexities, and recommend simplified architectures for future work to enhance accessibility for autonomous systems in developing countries like Rwanda. --- ### Glocal Information Bottleneck for Time Series Imputation Authors: Jie Yang, Kexin Zhang, Guibin Zhang, Philip S. Yu, Kaize Ding Published: 2025-10-06 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2510.04910v1 Time Series Imputation (TSI), which aims to recover missing values in temporal data, remains a fundamental challenge due to the complex and often high-rate missingness in real-world scenarios. Existing models typically optimize the point-wise reconstruction loss, focusing on recovering numerical values (local information). However, we observe that under high missing rates, these models still perform well in the training phase yet produce poor imputations and distorted latent representation distributions (global information) in the inference phase. This reveals a critical optimization dilemma: current objectives lack global guidance, leading models to overfit local noise and fail to capture global information of the data. To address this issue, we propose a new training paradigm, Glocal Information Bottleneck (Glocal-IB). Glocal-IB is model-agnostic and extends the standard IB framework by introducing a Global Alignment loss, derived from a tractable mutual information approximation. This loss aligns the latent representations of masked inputs with those of their originally observed counterparts. It helps the model retain global structure and local details while suppressing noise caused by missing values, giving rise to better generalization under high missingness. Extensive experiments on nine datasets confirm that Glocal-IB leads to consistently improved performance and aligned latent representations under missingness. Our code implementation is available in https://github.com/Muyiiiii/NeurIPS-25-Glocal-IB. --- ### How Different from the Past? Spatio-Temporal Time Series Forecasting with Self-Supervised Deviation Learning Authors: Haotian Gao, Zheng Dong, Jiawei Yong, Shintaro Fukushima, Kenjiro Taura, Renhe Jiang Published: 2025-10-06 Categories: cs.LG arXiv: https://arxiv.org/abs/2510.04908v1 Spatio-temporal forecasting is essential for real-world applications such as traffic management and urban computing. Although recent methods have shown improved accuracy, they often fail to account for dynamic deviations between current inputs and historical patterns. These deviations contain critical signals that can significantly affect model performance. To fill this gap, we propose ST-SSDL, a Spatio-Temporal time series forecasting framework that incorporates a Self-Supervised Deviation Learning scheme to capture and utilize such deviations. ST-SSDL anchors each input to its historical average and discretizes the latent space using learnable prototypes that represent typical spatio-temporal patterns. Two auxiliary objectives are proposed to refine this structure: a contrastive loss that enhances inter-prototype discriminability and a deviation loss that regularizes the distance consistency between input representations and corresponding prototypes to quantify deviation. Optimized jointly with the forecasting objective, these components guide the model to organize its hidden space and improve generalization across diverse input conditions. Experiments on six benchmark datasets show that ST-SSDL consistently outperforms state-of-the-art baselines across multiple metrics. Visualizations further demonstrate its ability to adaptively respond to varying levels of deviation in complex spatio-temporal scenarios. Our code and datasets are available at https://github.com/Jimmy-7664/ST-SSDL. --- ### Retrieval-Augmented Code Generation: A Survey with Focus on Repository-Level Approaches Authors: Yicheng Tao, Yuante Li, Yao Qin, Yepang Liu Published: 2025-10-06 Categories: cs.SE, cs.CL arXiv: https://arxiv.org/abs/2510.04905v3 Recent advances in large language models (LLMs) have significantly improved automated code generation. While existing approaches have achieved strong performance at the function and file levels, real-world software engineering requires reasoning over entire repositories, including cross-file dependencies, evolving execution environments, and global semantic consistency. This challenge has led to the emergence of Repository-Level Code Generation (RLCG), where models must retrieve, organize, and utilize repository-scale context to generate coherent and executable code changes. To address these challenges, Retrieval-Augmented Generation (RAG) has become an increasingly important paradigm for repository-level code intelligence. In this survey, we present a comprehensive review of Retrieval-Augmented Code Generation (RACG), with a particular focus on repository-level approaches. Rather than viewing RACG as a static ``retrieve-then-generate'' pipeline, we characterize it as a coupled and evolving process involving context construction, retrieval optimization, generation, and environment interaction. We organize existing methods through a unified analytical framework spanning retrieval substrate, control regime, and evaluation setting. Based on this framework, we systematically examine retrieval strategies, graph-based and non-graph-based retrieval paradigms, training-driven optimizations, and autonomous agent architectures. We further summarize widely used datasets, benchmarks, and system configurations, and discuss key challenges including scalability, reliability, efficiency, and the necessity boundary between RACG and long-context LLMs. Through this survey, we aim to provide a structured understanding of the rapidly evolving RACG landscape and highlight promising directions for future AI-powered software engineering research. --- ### DP-Hype: Federated Differentially Private Hyperparameter Search Authors: Johannes Liebenow, Thorsten Peinemann, Esfandiar Mohammadi Published: 2025-10-06 Categories: cs.LG arXiv: https://arxiv.org/abs/2510.04902v3 Tuning hyperparameters in federated machine learning can substantially impact model performance. When hyperparameters are tuned on sensitive data, privacy becomes an important challenge and to this end, differential privacy has emerged as the de facto standard for provable privacy. A standard setting in federated learning is that clients agree on a shared setup, i.e., find a compromise from a set of hyperparameters, like a model's learning rate. Yet, prior work on privacy-preserving hyperparameter tuning is tailored to specific learning tasks, does not account for the privacy leakage of aggregated results, or offers a sub-optimal privacy-utility trade-off. In this work, we present our algorithm DP-Hype, which performs a federated and privacy-preserving hyperparameter search by conducting a federated voting based on local hyperparameter evaluations of clients. In this way, DP-Hype selects hyperparameters that lead to a compromise supported by a majority of clients, while maintaining scalability and independence from specific learning tasks. We prove that DP-Hype preserves the strong notion of differential privacy called client-level differential privacy and, importantly, show that its privacy guarantees do not depend on the number of hyperparameters. We also provide bounds on its utility guarantees, that is, the probability of finding good hyperparameters, and implement DP-Hype as a submodule in the popular Flower framework for federated machine learning. In addition, we evaluate performance on multiple benchmark data sets in iid as well as multiple non-iid settings and demonstrate high utility of DP-Hype even under small privacy budgets. --- ### Focused Skill Discovery: Learning to Control Specific State Variables while Minimizing Side Effects Authors: Jonathan Colaço Carr, Qinyi Sun, Cameron Allen Published: 2025-10-06 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2510.04901v1 Skills are essential for unlocking higher levels of problem solving. A common approach to discovering these skills is to learn ones that reliably reach different states, thus empowering the agent to control its environment. However, existing skill discovery algorithms often overlook the natural state variables present in many reinforcement learning problems, meaning that the discovered skills lack control of specific state variables. This can significantly hamper exploration efficiency, make skills more challenging to learn with, and lead to negative side effects in downstream tasks when the goal is under-specified. We introduce a general method that enables these skill discovery algorithms to learn focused skills -- skills that target and control specific state variables. Our approach improves state space coverage by a factor of three, unlocks new learning capabilities, and automatically avoids negative side effects in downstream tasks. --- ### Benchmarking M-LTSF: Frequency and Noise-Based Evaluation of Multivariate Long Time Series Forecasting Models Authors: Nick Janssen, Melanie Schaller, Bodo Rosenhahn Published: 2025-10-06 Categories: cs.LG, eess.SY arXiv: https://arxiv.org/abs/2510.04900v2 Understanding the robustness of deep learning models for multivariate long-term time series forecasting (M-LTSF) remains challenging, as evaluations typically rely on real-world datasets with unknown noise properties. We propose a simulation-based evaluation framework that generates parameterizable synthetic datasets, where each dataset instance corresponds to a different configuration of signal components, noise types, signal-to-noise ratios, and frequency characteristics. These configurable components aim to model real-world multivariate time series data without the ambiguity of unknown noise. This framework enables fine-grained, systematic evaluation of M-LTSF models under controlled and diverse scenarios. We benchmark four representative architectures S-Mamba (state-space), iTransformer (transformer-based), R-Linear (linear), and Autoformer (decomposition-based). Our analysis reveals that all models degrade severely when lookback windows cannot capture complete periods of seasonal patters in the data. S-Mamba and Autoformer perform best on sawtooth patterns, while R-Linear and iTransformer favor sinusoidal signals. White and Brownian noise universally degrade performance with lower signal-to-noise ratio while S-Mamba shows specific trend-noise and iTransformer shows seasonal-noise vulnerability. Further spectral analysis shows that S-Mamba and iTransformer achieve superior frequency reconstruction. This controlled approach, based on our synthetic and principle-driven testbed, offers deeper insights into model-specific strengths and limitations through the aggregation of MSE scores and provides concrete guidance for model selection based on signal characteristics and noise conditions. --- ### Human Behavior Atlas: Benchmarking Unified Psychological and Social Behavior Understanding Authors: Keane Ong, Wei Dai, Carol Li, Dewei Feng, Hengzhi Li, Jingyao Wu, Jiaee Cheong, Rui Mao, Gianmarco Mengaldo, Erik Cambria, Paul Pu Liang Published: 2025-10-06 Categories: cs.AI arXiv: https://arxiv.org/abs/2510.04899v2 Using intelligent systems to perceive psychological and social behaviors, that is, the underlying affective, cognitive, and pathological states that are manifested through observable behaviors and social interactions, remains a challenge due to their complex, multifaceted, and personalized nature. Existing work tackling these dimensions through specialized datasets and single-task systems often miss opportunities for scalability, cross-task transfer, and broader generalization. To address this gap, we curate Human Behavior Atlas, a unified benchmark of diverse behavioral tasks designed to support the development of foundation models for understanding psychological and social behaviors. Human Behavior Atlas comprises over 100,000 samples spanning text, audio, and visual modalities, covering tasks on affective states, cognitive states, pathologies, and social processes. Our unification efforts can reduce redundancy and cost, enable training to scale efficiently across tasks, and enhance generalization of behavioral features across domains. On Human Behavior Atlas, we train three models: Omnisapiens-7B SFT, Omnisapiens-7B BAM, and Omnisapiens-7B RL. We show that training on Human Behavior Atlas enables models to consistently outperform existing multimodal LLMs across diverse behavioral tasks. Pretraining on Human Behavior Atlas also improves transfer to novel behavioral datasets; with the targeted use of behavioral descriptors yielding meaningful performance gains. The benchmark, models, and codes can be found at: https://github.com/MIT-MI/human_behavior_atlas. --- ### HyperVLA: Efficient Inference in Vision-Language-Action Models via Hypernetworks Authors: Zheng Xiong, Kang Li, Zilin Wang, Matthew Jackson, Jakob Foerster, Shimon Whiteson Published: 2025-10-06 Categories: cs.RO, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2510.04898v1 Built upon language and vision foundation models with strong generalization ability and trained on large-scale robotic data, Vision-Language-Action (VLA) models have recently emerged as a promising approach to learning generalist robotic policies. However, a key drawback of existing VLAs is their extremely high inference costs. In this paper, we propose HyperVLA to address this problem. Unlike existing monolithic VLAs that activate the whole model during both training and inference, HyperVLA uses a novel hypernetwork (HN)-based architecture that activates only a small task-specific policy during inference, while still retaining the high model capacity needed to accommodate diverse multi-task behaviors during training. Successfully training an HN-based VLA is nontrivial so HyperVLA contains several key algorithm design features that improve its performance, including properly utilizing the prior knowledge from existing vision foundation models, HN normalization, and an action generation strategy. Compared to monolithic VLAs, HyperVLA achieves a similar or even higher success rate for both zero-shot generalization and few-shot adaptation, while significantly reducing inference costs. Compared to OpenVLA, a state-of-the-art VLA model, HyperVLA reduces the number of activated parameters at test time by $90\times$, and accelerates inference speed by $120\times$. Code is publicly available at https://github.com/MasterXiong/HyperVLA --- ### SocialHarmBench: Revealing LLM Vulnerabilities to Socially Harmful Requests Authors: Punya Syon Pandey, Hai Son Le, Devansh Bhardwaj, Rada Mihalcea, Zhijing Jin Published: 2025-10-06 Categories: cs.CL, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2510.04891v2 Large language models (LLMs) are increasingly deployed in contexts where their failures can have direct sociopolitical consequences. Yet, existing safety benchmarks rarely test vulnerabilities in domains such as political manipulation, propaganda and disinformation generation, or surveillance and information control. We introduce SocialHarmBench, a dataset of 585 prompts spanning 7 sociopolitical categories and 34 countries, designed to surface where LLMs most acutely fail in politically charged contexts. Our evaluations reveal several shortcomings: open-weight models exhibit high vulnerability to harmful compliance, with Mistral-7B reaching attack success rates as high as 97% to 98% in domains such as historical revisionism, propaganda, and political manipulation. Moreover, temporal and geographic analyses show that LLMs are most fragile when confronted with 21st-century or pre-20th-century contexts, and when responding to prompts tied to regions such as Latin America, the USA, and the UK. These findings demonstrate that current safeguards fail to generalize to high-stakes sociopolitical settings, exposing systematic biases and raising concerns about the reliability of LLMs in preserving human rights and democratic values. We share the SocialHarmBench benchmark at https://huggingface.co/datasets/psyonp/SocialHarmBench. --- ### Revealing Interconnections between Diseases: from Statistical Methods to Large Language Models Authors: Alina Ermilova, Dmitrii Kornilov, Sofia Samoilova, Ekaterina Laptenkova, Anastasia Kolesnikova, Ekaterina Podplutova, Senotrusova Sofya, Maksim G. Sharaev Published: 2025-10-06 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2510.04888v2 Identifying disease interconnections through manual analysis of large-scale clinical data is labor-intensive, subjective, and prone to expert disagreement. While machine learning (ML) shows promise, three critical challenges remain: (1) selecting optimal methods from the vast ML landscape, (2) determining whether real-world clinical data (e.g., electronic health records, EHRs) or structured disease descriptions yield more reliable insights, (3) the lack of "ground truth," as some disease interconnections remain unexplored in medicine. Large language models (LLMs) demonstrate broad utility, yet they often lack specialized medical knowledge. To address these gaps, we conduct a systematic evaluation of seven approaches for uncovering disease relationships based on two data sources: (i) sequences of ICD-10 codes from MIMIC-IV EHRs and (ii) the full set of ICD-10 codes, both with and without textual descriptions. Our framework integrates the following: (i) a statistical co-occurrence analysis and a masked language modeling (MLM) approach using real clinical data; (ii) domain-specific BERT variants (Med-BERT and BioClinicalBERT); (iii) a general-purpose BERT and document retrieval; and (iv) four LLMs (Mistral, DeepSeek, Qwen, and YandexGPT). Our graph-based comparison of the obtained interconnection matrices shows that the LLM-based approach produces interconnections with the lowest diversity of ICD code connections to different diseases compared to other methods, including text-based and domain-based approaches. This suggests an important implication: LLMs have limited potential for discovering new interconnections. In the absence of ground truth databases for medical interconnections between ICD codes, our results constitute a valuable medical disease ontology that can serve as a foundational resource for future clinical research and artificial intelligence applications in healthcare. --- ### Where Did It All Go Wrong? A Hierarchical Look into Multi-Agent Error Attribution Authors: Adi Banerjee, Anirudh Nair, Tarik Borogovac Published: 2025-10-06 Categories: cs.AI, cs.MA arXiv: https://arxiv.org/abs/2510.04886v2 Error attribution in Large Language Model (LLM) multi-agent systems presents a significant challenge in debugging and improving collaborative AI systems. Current approaches to pinpointing agent and step level failures in interaction traces - whether using all-at-once evaluation, step-by-step analysis, or binary search - fall short when analyzing complex patterns, struggling with both accuracy and consistency. We present ECHO (Error attribution through Contextual Hierarchy and Objective consensus analysis), a novel algorithm that combines hierarchical context representation, objective analysis-based evaluation, and consensus voting to improve error attribution accuracy. Our approach leverages a positional-based leveling of contextual understanding while maintaining objective evaluation criteria, ultimately reaching conclusions through a consensus mechanism. Experimental results demonstrate that ECHO outperforms existing methods across various multi-agent interaction scenarios, showing particular strength in cases involving subtle reasoning errors and complex interdependencies. Our findings suggest that leveraging these concepts of structured, hierarchical context representation combined with consensus-based objective decision-making, provides a more robust framework for error attribution in multi-agent systems. --- ### RL Is a Hammer and LLMs Are Nails: A Simple Reinforcement Learning Recipe for Strong Prompt Injection Authors: Yuxin Wen, Arman Zharmagambetov, Ivan Evtimov, Narine Kokhlikyan, Tom Goldstein, Kamalika Chaudhuri, Chuan Guo Published: 2025-10-06 Categories: cs.CR, cs.LG arXiv: https://arxiv.org/abs/2510.04885v1 Prompt injection poses a serious threat to the reliability and safety of LLM agents. Recent defenses against prompt injection, such as Instruction Hierarchy and SecAlign, have shown notable robustness against static attacks. However, to more thoroughly evaluate the robustness of these defenses, it is arguably necessary to employ strong attacks such as automated red-teaming. To this end, we introduce RL-Hammer, a simple recipe for training attacker models that automatically learn to perform strong prompt injections and jailbreaks via reinforcement learning. RL-Hammer requires no warm-up data and can be trained entirely from scratch. To achieve high ASRs against industrial-level models with defenses, we propose a set of practical techniques that enable highly effective, universal attacks. Using this pipeline, RL-Hammer reaches a 98% ASR against GPT-4o and a $72\%$ ASR against GPT-5 with the Instruction Hierarchy defense. We further discuss the challenge of achieving high diversity in attacks, highlighting how attacker models tend to reward-hack diversity objectives. Finally, we show that RL-Hammer can evade multiple prompt injection detectors. We hope our work advances automatic red-teaming and motivates the development of stronger, more principled defenses. Code is available at https://github.com/facebookresearch/rl-injector. --- ### CLEAR-IR: Clarity-Enhanced Active Reconstruction of Infrared Imagery Authors: Nathan Shankar, Pawel Ladosz, Hujun Yin Published: 2025-10-06 Categories: cs.RO, cs.CV, cs.LG arXiv: https://arxiv.org/abs/2510.04883v2 This paper presents a novel approach for enabling robust robotic perception in dark environments using infrared (IR) stream. IR stream is less susceptible to noise than RGB in low-light conditions. However, it is dominated by active emitter patterns that hinder high-level tasks such as object detection, tracking and localisation. To address this, a Deep Multi-scale Aware Overcomplete (DeepMAO) inspired architecture is proposed that reconstructs clean IR images from emitter populated input, improving both image quality and downstream robotic performance. This approach outperforms existing enhancement techniques and enables reliable operation of vision driven robotic systems across illumination conditions from well-lit to extreme low-light scenes. The results outline the ability of this work to be able to mimic RGB styling from the scene and its applicability on robotics tasks that were trained on RGB images, opening the possibility of doing these tasks in extreme low-light without on-board lighting. --- ### Flow-Matching Based Refiner for Molecular Conformer Generation Authors: Xiangyang Xu, Hongyang Gao Published: 2025-10-06 Categories: cs.LG, q-bio.QM arXiv: https://arxiv.org/abs/2510.04878v1 Low-energy molecular conformers generation (MCG) is a foundational yet challenging problem in drug discovery. Denoising-based methods include diffusion and flow-matching methods that learn mappings from a simple base distribution to the molecular conformer distribution. However, these approaches often suffer from error accumulation during sampling, especially in the low SNR steps, which are hard to train. To address these challenges, we propose a flow-matching refiner for the MCG task. The proposed method initializes sampling from mixed-quality outputs produced by upstream denoising models and reschedules the noise scale to bypass the low-SNR phase, thereby improving sample quality. On the GEOM-QM9 and GEOM-Drugs benchmark datasets, the generator-refiner pipeline improves quality with fewer total denoising steps while preserving diversity. --- ### BenthiCat: An opti-acoustic dataset for advancing benthic classification and habitat mapping Authors: Hayat Rajani, Valerio Franchi, Borja Martinez-Clavel Valles, Raimon Ramos, Rafael Garcia, Nuno Gracias Published: 2025-10-06 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2510.04876v3 Benthic habitat mapping is fundamental for understanding marine ecosystems, guiding conservation efforts, and supporting sustainable resource management. Yet, the scarcity of large, annotated datasets limits the development and benchmarking of machine learning models in this domain. This paper introduces a thorough multi-modal dataset, comprising about a million side-scan sonar (SSS) tiles collected along the coast of Catalonia (Spain), complemented by bathymetric maps and a set of co-registered optical images from targeted surveys using an autonomous underwater vehicle (AUV). Approximately 36000 of the SSS tiles have been manually annotated with segmentation masks to enable supervised fine-tuning of classification models. All the raw sensor data, together with mosaics, are also released to support further exploration and algorithm development. To address challenges in multi-sensor data fusion for AUVs, we spatially associate optical images with corresponding SSS tiles, facilitating self-supervised, cross-modal representation learning. Accompanying open-source preprocessing and annotation tools are provided to enhance accessibility and encourage research. This resource aims to establish a standardized benchmark for underwater habitat mapping, promoting advancements in autonomous seafloor classification and multi-sensor integration. --- ### Less is More: Recursive Reasoning with Tiny Networks Authors: Alexia Jolicoeur-Martineau Published: 2025-10-06 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2510.04871v1 Hierarchical Reasoning Model (HRM) is a novel approach using two small neural networks recursing at different frequencies. This biologically inspired method beats Large Language models (LLMs) on hard puzzle tasks such as Sudoku, Maze, and ARC-AGI while trained with small models (27M parameters) on small data (around 1000 examples). HRM holds great promise for solving hard problems with small networks, but it is not yet well understood and may be suboptimal. We propose Tiny Recursive Model (TRM), a much simpler recursive reasoning approach that achieves significantly higher generalization than HRM, while using a single tiny network with only 2 layers. With only 7M parameters, TRM obtains 45% test-accuracy on ARC-AGI-1 and 8% on ARC-AGI-2, higher than most LLMs (e.g., Deepseek R1, o3-mini, Gemini 2.5 Pro) with less than 0.01% of the parameters. --- ### Model Predictive Control-Guided Reinforcement Learning for Implicit Balancing Authors: Seyed Soroush Karimi Madahi, Kenneth Bruninx, Bert Claessens, Chris Develder Published: 2025-10-06 Categories: eess.SY, cs.AI arXiv: https://arxiv.org/abs/2510.04868v1 In Europe, profit-seeking balance responsible parties can deviate in real time from their day-ahead nominations to assist transmission system operators in maintaining the supply-demand balance. Model predictive control (MPC) strategies to exploit these implicit balancing strategies capture arbitrage opportunities, but fail to accurately capture the price-formation process in the European imbalance markets and face high computational costs. Model-free reinforcement learning (RL) methods are fast to execute, but require data-intensive training and usually rely on real-time and historical data for decision-making. This paper proposes an MPC-guided RL method that combines the complementary strengths of both MPC and RL. The proposed method can effectively incorporate forecasts into the decision-making process (as in MPC), while maintaining the fast inference capability of RL. The performance of the proposed method is evaluated on the implicit balancing battery control problem using Belgian balancing data from 2023. First, we analyze the performance of the standalone state-of-the-art RL and MPC methods from various angles, to highlight their individual strengths and limitations. Next, we show an arbitrage profit benefit of the proposed MPC-guided RL method of 16.15% and 54.36%, compared to standalone RL and MPC. --- ### In-Field Mapping of Grape Yield and Quality with Illumination-Invariant Deep Learning Authors: Ciem Cornelissen, Sander De Coninck, Axel Willekens, Sam Leroux, Pieter Simoens Published: 2025-10-06 Categories: cs.CV arXiv: https://arxiv.org/abs/2510.04864v1 This paper presents an end-to-end, IoT-enabled robotic system for the non-destructive, real-time, and spatially-resolved mapping of grape yield and quality (Brix, Acidity) in vineyards. The system features a comprehensive analytical pipeline that integrates two key modules: a high-performance model for grape bunch detection and weight estimation, and a novel deep learning framework for quality assessment from hyperspectral (HSI) data. A critical barrier to in-field HSI is the ``domain shift" caused by variable illumination. To overcome this, our quality assessment is powered by the Light-Invariant Spectral Autoencoder (LISA), a domain-adversarial framework that learns illumination-invariant features from uncalibrated data. We validated the system's robustness on a purpose-built HSI dataset spanning three distinct illumination domains: controlled artificial lighting (lab), and variable natural sunlight captured in the morning and afternoon. Results show the complete pipeline achieves a recall (0.82) for bunch detection and a $R^2$ (0.76) for weight prediction, while the LISA module improves quality prediction generalization by over 20% compared to the baselines. By combining these robust modules, the system successfully generates high-resolution, georeferenced data of both grape yield and quality, providing actionable, data-driven insights for precision viticulture. --- ### Video Game Level Design as a Multi-Agent Reinforcement Learning Problem Authors: Sam Earle, Zehua Jiang, Eugene Vinitsky, Julian Togelius Published: 2025-10-06 Categories: cs.AI, cs.LG, cs.MA, cs.NE arXiv: https://arxiv.org/abs/2510.04862v1 Procedural Content Generation via Reinforcement Learning (PCGRL) offers a method for training controllable level designer agents without the need for human datasets, using metrics that serve as proxies for level quality as rewards. Existing PCGRL research focuses on single generator agents, but are bottlenecked by the need to frequently recalculate heuristics of level quality and the agent's need to navigate around potentially large maps. By framing level generation as a multi-agent problem, we mitigate the efficiency bottleneck of single-agent PCGRL by reducing the number of reward calculations relative to the number of agent actions. We also find that multi-agent level generators are better able to generalize to out-of-distribution map shapes, which we argue is due to the generators' learning more local, modular design policies. We conclude that treating content generation as a distributed, multi-agent task is beneficial for generating functional artifacts at scale. --- ### A Clinical-grade Universal Foundation Model for Intraoperative Pathology Authors: Zihan Zhao, Fengtao Zhou, Ronggang Li, Bing Chu, Xinke Zhang, Xueyi Zheng, Ke Zheng, Xiaobo Wen, Jiabo Ma, Yihui Wang, Jiewei Chen, Chengyou Zheng et al. Published: 2025-10-06 Categories: cs.LG arXiv: https://arxiv.org/abs/2510.04861v2 Intraoperative pathology is pivotal to precision surgery, yet its clinical impact is constrained by diagnostic complexity and the limited availability of high-quality frozen-section data. While computational pathology has made significant strides, the lack of large-scale, prospective validation has impeded its routine adoption in surgical workflows. Here, we introduce CRISP, a clinical-grade foundation model developed on over 100,000 frozen sections from eight medical centers, specifically designed to provide Clinical-grade Robust Intraoperative Support for Pathology (CRISP). CRISP was comprehensively evaluated on more than 15,000 intraoperative slides across nearly 100 retrospective diagnostic tasks, including benign-malignant discrimination, key intraoperative decision-making, and pan-cancer detection, etc. The model demonstrated robust generalization across diverse institutions, tumor types, and anatomical sites-including previously unseen sites and rare cancers. In a prospective cohort of over 2,000 patients, CRISP sustained high diagnostic accuracy under real-world conditions, directly informing surgical decisions in 92.6% of cases. Human-AI collaboration further reduced diagnostic workload by 35%, avoided 105 ancillary tests and enhanced detection of micrometastases with 87.5% accuracy. Together, these findings position CRISP as a clinical-grade paradigm for AI-driven intraoperative pathology, bridging computational advances with surgical precision and accelerating the translation of artificial intelligence into routine clinical practice. --- ### Alignment Tipping Process: How Self-Evolution Pushes LLM Agents Off the Rails Authors: Siwei Han, Kaiwen Xiong, Jiaqi Liu, Xinyu Ye, Yaofeng Su, Wenbo Duan, Xinyuan Liu, Cihang Xie, Mohit Bansal, Mingyu Ding, Linjun Zhang, Huaxiu Yao Published: 2025-10-06 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2510.04860v2 As Large Language Model (LLM) agents increasingly gain self-evolutionary capabilities to adapt and refine their strategies through real-world interaction, their long-term reliability becomes a critical concern. We identify the Alignment Tipping Process (ATP), a critical post-deployment risk unique to self-evolving LLM agents. Unlike training-time failures, ATP arises when continual interaction drives agents to abandon alignment constraints established during training in favor of reinforced, self-interested strategies. We formalize and analyze ATP through two complementary paradigms: Self-Interested Exploration, where repeated high-reward deviations induce individual behavioral drift, and Imitative Strategy Diffusion, where deviant behaviors spread across multi-agent systems. Building on these paradigms, we construct controllable testbeds and benchmark both open and closed-source LLMs. Our experiments show that alignment benefits erode rapidly under self-evolution, with initially aligned models converging toward unaligned states. In multi-agent settings, successful violations diffuse quickly, leading to collective misalignment. Moreover, current reinforcement learning-based alignment methods provide limited defenses against alignment tipping. These findings demonstrate that alignment of LLM agents is not a static property but a fragile and dynamic one, vulnerable to feedback-driven decay during deployment. Our data and code are available at https://github.com/aiming-lab/ATP. --- ### Global-to-local image quality assessment in optical microscopy via fast and robust deep learning predictions Authors: Elena Corbetta, Thomas Bocklitz Published: 2025-10-06 Categories: cs.CV, physics.data-an, q-bio.QM arXiv: https://arxiv.org/abs/2510.04859v2 Optical microscopy is one of the most widely used techniques in research studies for life sciences and biomedicine. These applications require reliable experimental pipelines to extract valuable knowledge from the measured samples and must be supported by image quality assessment (IQA) to ensure correct processing and analysis of the image data. IQA methods are implemented with variable complexity. However, while most quality metrics have a straightforward implementation, they might be time consuming and computationally expensive when evaluating a large dataset. In addition, quality metrics are often designed for well-defined image features and may be unstable for images out of the ideal domain. To overcome these limitations, recent works have proposed deep learning-based IQA methods, which can provide superior performance, increased generalizability and fast prediction. Our method, named $\mathrmμ$DeepIQA, is inspired by previous studies and applies a deep convolutional neural network designed for IQA on natural images to optical microscopy measurements. We retrained the same architecture to predict individual quality metrics and global quality scores for optical microscopy data. The resulting models provide fast and stable predictions of image quality by generalizing quality estimation even outside the ideal range of standard methods. In addition, $\mathrmμ$DeepIQA provides patch-wise prediction of image quality and can be used to visualize spatially varying quality in a single image. Our study demonstrates that optical microscopy-based studies can benefit from the generalizability of deep learning models due to their stable performance in the presence of outliers, the ability to assess small image patches, and rapid predictions. --- ### ERDE: Entropy-Regularized Distillation for Early-exit Authors: Martial Guidez, Stefan Duffner, Yannick Alpou, Oscar Röth, Christophe Garcia Published: 2025-10-06 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2510.04856v1 Although deep neural networks and in particular Convolutional Neural Networks have demonstrated state-of-the-art performance in image classification with relatively high efficiency, they still exhibit high computational costs, often rendering them impractical for real-time and edge applications. Therefore, a multitude of compression techniques have been developed to reduce these costs while maintaining accuracy. In addition, dynamic architectures have been introduced to modulate the level of compression at execution time, which is a desirable property in many resource-limited application scenarios. The proposed method effectively integrates two well-established optimization techniques: early exits and knowledge distillation, where a reduced student early-exit model is trained from a more complex teacher early-exit model. The primary contribution of this research lies in the approach for training the student early-exit model. In comparison to the conventional Knowledge Distillation loss, our approach incorporates a new entropy-based loss for images where the teacher's classification was incorrect. The proposed method optimizes the trade-off between accuracy and efficiency, thereby achieving significant reductions in computational complexity without compromising classification performance. The validity of this approach is substantiated by experimental results on image classification datasets CIFAR10, CIFAR100 and SVHN, which further opens new research perspectives for Knowledge Distillation in other contexts. --- ### Synthesising Counterfactual Explanations via Label-Conditional Gaussian Mixture Variational Autoencoders Authors: Junqi Jiang, Francesco Leofante, Antonio Rago, Francesca Toni Published: 2025-10-06 Categories: cs.LG arXiv: https://arxiv.org/abs/2510.04855v2 Counterfactual explanations (CEs) provide recourse recommendations for individuals affected by algorithmic decisions. A key challenge is generating CEs that are robust against various perturbation types (e.g. input and model perturbations) while simultaneously satisfying other desirable properties. These include plausibility, ensuring CEs reside on the data manifold, and diversity, providing multiple distinct recourse options for single inputs. Existing methods, however, mostly struggle to address these multifaceted requirements in a unified, model-agnostic manner. We address these limitations by proposing a novel generative framework. First, we introduce the Label-conditional Gaussian Mixture Variational Autoencoder (L-GMVAE), a model trained to learn a structured latent space where each class label is represented by a set of Gaussian components with diverse, prototypical centroids. Building on this, we present LAPACE (LAtent PAth Counterfactual Explanations), a model-agnostic algorithm that synthesises entire paths of CE points by interpolating from inputs' latent representations to those learned latent centroids. This approach inherently ensures robustness to input changes, as all paths for a given target class converge to the same fixed centroids. Furthermore, the generated paths provide a spectrum of recourse options, allowing users to navigate the trade-off between proximity and plausibility while also encouraging robustness against model changes. In addition, user-specified actionability constraints can also be easily incorporated via lightweight gradient optimisation through the L-GMVAE's decoder. Comprehensive experiments show that LAPACE is computationally efficient and achieves competitive performance across eight quantitative metrics. --- ### Read the Room: Inferring Social Context Through Dyadic Interaction Recognition in Cyber-physical-social Infrastructure Systems Authors: Cheyu Lin, John Martins, Katherine A. Flanigan, Ph. D Published: 2025-10-06 Categories: cs.CV arXiv: https://arxiv.org/abs/2510.04854v1 Cyber-physical systems (CPS) integrate sensing, computing, and control to improve infrastructure performance, focusing on economic goals like performance and safety. However, they often neglect potential human-centered (or ''social'') benefits. Cyber-physical-social infrastructure systems (CPSIS) aim to address this by aligning CPS with social objectives. This involves defining social benefits, understanding human interactions with each other and infrastructure, developing privacy-preserving measurement methods, modeling these interactions for prediction, linking them to social benefits, and actuating the physical environment to foster positive social outcomes. This paper delves into recognizing dyadic human interactions using real-world data, which is the backbone to measuring social behavior. This lays a foundation to address the need to enhance understanding of the deeper meanings and mutual responses inherent in human interactions. While RGB cameras are informative for interaction recognition, privacy concerns arise. Depth sensors offer a privacy-conscious alternative by analyzing skeletal movements. This study compares five skeleton-based interaction recognition algorithms on a dataset of 12 dyadic interactions. Unlike single-person datasets, these interactions, categorized into communication types like emblems and affect displays, offer insights into the cultural and emotional aspects of human interactions. --- ### FreshBrew: A Benchmark for Evaluating AI Agents on Java Code Migration Authors: Victor May, Diganta Misra, Yanqi Luo, Anjali Sridhar, Justine Gehring, Silvio Soares Ribeiro Junior Published: 2025-10-06 Categories: cs.SE, cs.AI arXiv: https://arxiv.org/abs/2510.04852v3 AI coding assistants are rapidly becoming integral to modern software development. A key challenge in this space is the continual need to migrate and modernize codebases in response to evolving software ecosystems. Traditionally, such migrations have relied on rule-based systems and human intervention. With the advent of powerful large language models (LLMs), AI-driven agentic frameworks offer a promising alternative-but their effectiveness has not been systematically evaluated. In this paper, we introduce \FreshBrew{}\footnote{\url{https://github.com/mrcabbage972/freshbrew}}, a novel benchmark for evaluating AI agents on project-level Java migrations, with a specific focus on measuring an agent's ability to preserve program semantics and avoid reward hacking, which we argue requires projects with high test coverage for a rigorous and reliable evaluation. We benchmark several state-of-the-art LLMs, and compare their performance against established rule-based tools. Our evaluation on a challenging subset of 228 public Maven repositories that build on JDK 8, fail on JDK 17, and have high test coverage shows that the top-performing model, Gemini 2.5 Flash, can successfully migrate 52.3\% of projects to JDK 17. Our empirical analysis reveals novel insights into the critical strengths and limitations of current agentic approaches, offering actionable insights into their real-world applicability. Our empirical study reveals failure modes of current AI agents in realistic Java modernization tasks, providing a foundation for evaluating trustworthy code-migration systems. By releasing \FreshBrew{}, we aim to facilitate rigorous, reproducible evaluation and catalyze progress in AI-driven codebase modernization. --- ### LEGOMem: Modular Procedural Memory for Multi-agent LLM Systems for Workflow Automation Authors: Dongge Han, Camille Couturier, Daniel Madrigal Diaz, Xuchao Zhang, Victor Rühle, Saravan Rajmohan Published: 2025-10-06 Categories: cs.AI, cs.LG, cs.MA arXiv: https://arxiv.org/abs/2510.04851v1 We introduce LEGOMem, a modular procedural memory framework for multi-agent large language model (LLM) systems in workflow automation. LEGOMem decomposes past task trajectories into reusable memory units and flexibly allocates them across orchestrators and task agents to support planning and execution. To explore the design space of memory in multi-agent systems, we use LEGOMem as a lens and conduct a systematic study of procedural memory in multi-agent systems, examining where memory should be placed, how it should be retrieved, and which agents benefit most. Experiments on the OfficeBench benchmark show that orchestrator memory is critical for effective task decomposition and delegation, while fine-grained agent memory improves execution accuracy. We find that even teams composed of smaller language models can benefit substantially from procedural memory, narrowing the performance gap with stronger agents by leveraging prior execution traces for more accurate planning and tool use. These results position LEGOMem as both a practical framework for memory-augmented agent systems and a research tool for understanding memory design in multi-agent workflow automation. --- ### Detecting Distillation Data from Reasoning Models Authors: Hengxiang Zhang, Hyeong Kyu Choi, Sharon Li, Hongxin Wei Published: 2025-10-06 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2510.04850v3 Reasoning distillation has emerged as a prevailing paradigm for transferring reasoning capabilities from large reasoning models to small language models. Yet, reasoning distillation risks data contamination: benchmark data may inadvertently be included in the distillation data, thereby inflating model performance metrics. In this work, we formally define the distillation data detection task, which determines whether a given question is included in the model's distillation data. The unique challenge of this task lies in the partial availability of distillation data. To address this, we propose Token Probability Deviation (TPD), a detection method that leverages the probability patterns of output tokens generated by the model instead of input tokens. Our method is motivated by the observation that seen questions tend to elicit more near-deterministic tokens generated by the models than unseen ones. Our TPD score is thus designed to quantify the token-level deviation of generated tokens from a high-confidence reference probability. Consequently, seen questions can yield substantially lower TPD scores than unseen ones, enabling strong detection performance. Extensive experiments demonstrate the effectiveness of our approach, improving detection AUC by up to 31% on distillation datasets. --- ### When Models Lie, We Learn: Multilingual Span-Level Hallucination Detection with PsiloQA Authors: Elisei Rykov, Kseniia Petrushina, Maksim Savkin, Valerii Olisov, Artem Vazhentsev, Kseniia Titova, Alexander Panchenko, Vasily Konovalov, Julia Belikova Published: 2025-10-06 Categories: cs.CL arXiv: https://arxiv.org/abs/2510.04849v1 Hallucination detection remains a fundamental challenge for the safe and reliable deployment of large language models (LLMs), especially in applications requiring factual accuracy. Existing hallucination benchmarks often operate at the sequence level and are limited to English, lacking the fine-grained, multilingual supervision needed for a comprehensive evaluation. In this work, we introduce PsiloQA, a large-scale, multilingual dataset annotated with span-level hallucinations across 14 languages. PsiloQA is constructed through an automated three-stage pipeline: generating question-answer pairs from Wikipedia using GPT-4o, eliciting potentially hallucinated answers from diverse LLMs in a no-context setting, and automatically annotating hallucinated spans using GPT-4o by comparing against golden answers and retrieved context. We evaluate a wide range of hallucination detection methods -- including uncertainty quantification, LLM-based tagging, and fine-tuned encoder models -- and show that encoder-based models achieve the strongest performance across languages. Furthermore, PsiloQA demonstrates effective cross-lingual generalization and supports robust knowledge transfer to other benchmarks, all while being significantly more cost-efficient than human-annotated datasets. Our dataset and results advance the development of scalable, fine-grained hallucination detection in multilingual settings. --- ### Instability in Downstream Task Performance During LLM Pretraining Authors: Yuto Nishida, Masaru Isonuma, Yusuke Oda Published: 2025-10-06 Categories: cs.CL arXiv: https://arxiv.org/abs/2510.04848v1 When training large language models (LLMs), it is common practice to track downstream task performance throughout the training process and select the checkpoint with the highest validation score. However, downstream metrics often exhibit substantial fluctuations, making it difficult to identify the checkpoint that truly represents the best-performing model. In this study, we empirically analyze the stability of downstream task performance in an LLM trained on diverse web-scale corpora. We find that task scores frequently fluctuate throughout training, both at the aggregate and example levels. To address this instability, we investigate two post-hoc checkpoint integration methods: checkpoint averaging and ensemble, motivated by the hypothesis that aggregating neighboring checkpoints can reduce performance volatility. We demonstrate both empirically and theoretically that these methods improve downstream performance stability without requiring any changes to the training procedure. ---