# Research Papers — 2025-10 · bundle page 172 of 217

> Full-text markdown of papers #8551–8600 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/172
[← Previous bundle](https://aipod.dev/papers/archive/2025-10/page/171.md)
[Next bundle →](https://aipod.dev/papers/archive/2025-10/page/173.md)

### Taxonomy of User Needs and Actions

Authors: Renee Shelby, Fernando Diaz, Vinodkumar Prabhakaran
Published: 2025-10-07
Categories: cs.HC, cs.CL
arXiv: https://arxiv.org/abs/2510.06124v2

The growing ubiquity of conversational AI highlights the need for frameworks that capture not only users' instrumental goals but also the situated, adaptive, and social practices through which they achieve them. Existing taxonomies of conversational behavior either overgeneralize, remain domain-specific, or reduce interactions to narrow dialogue functions. To address this gap, we introduce the Taxonomy of User Needs and Actions (TUNA), an empirically grounded framework developed through iterative qualitative analysis of 1193 human-AI conversations, supplemented by theoretical review and validation across diverse contexts. TUNA organizes user actions into a three-level hierarchy encompassing behaviors associated with information seeking, synthesis, procedural guidance, content creation, social interaction, and meta-conversation. By centering user agency and appropriation practices, TUNA enables multi-scale evaluation, supports policy harmonization across products, and provides a backbone for layering domain-specific taxonomies. This work contributes a systematic vocabulary for describing AI use, advancing both scholarly understanding and practical design of safer, more responsive, and more accountable conversational systems.

---

### Towards Data-Efficient Medical Imaging: A Generative and Semi-Supervised Framework

Authors: Mosong Ma, Tania Stathaki, Michalis Lazarou
Published: 2025-10-07
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.06123v1

Deep learning in medical imaging is often limited by scarce and imbalanced annotated data. We present SSGNet, a unified framework that combines class specific generative modeling with iterative semisupervised pseudo labeling to enhance both classification and segmentation. Rather than functioning as a standalone model, SSGNet augments existing baselines by expanding training data with StyleGAN3 generated images and refining labels through iterative pseudo labeling. Experiments across multiple medical imaging benchmarks demonstrate consistent gains in classification and segmentation performance, while Frechet Inception Distance analysis confirms the high quality of generated samples. These results highlight SSGNet as a practical strategy to mitigate annotation bottlenecks and improve robustness in medical image analysis.

---

### PolyGraph Discrepancy: a classifier-based metric for graph generation

Authors: Markus Krimmel, Philip Hartout, Karsten Borgwardt, Dexiong Chen
Published: 2025-10-07
Categories: cs.LG, stat.ML
arXiv: https://arxiv.org/abs/2510.06122v2

Existing methods for evaluating graph generative models primarily rely on Maximum Mean Discrepancy (MMD) metrics based on graph descriptors. While these metrics can rank generative models, they do not provide an absolute measure of performance. Their values are also highly sensitive to extrinsic parameters, namely kernel and descriptor parametrization, making them incomparable across different graph descriptors. We introduce PolyGraph Discrepancy (PGD), a new evaluation framework that addresses these limitations. It approximates the Jensen-Shannon distance of graph distributions by fitting binary classifiers to distinguish between real and generated graphs, featurized by these descriptors. The data log-likelihood of these classifiers approximates a variational lower bound on the JS distance between the two distributions. Resulting metrics are constrained to the unit interval (0,1) and are comparable across different graph descriptors. We further derive a theoretically grounded summary metric that combines these individual metrics to provide a maximally tight lower bound on the distance for the given descriptors. Thorough experiments demonstrate that PGD provides a more robust and insightful evaluation compared to MMD metrics. The PolyGraph framework for benchmarking graph generative models is made publicly available at https://github.com/BorgwardtLab/polygraph-benchmark.

---

### Multimodal Feature Prototype Learning for Interpretable and Discriminative Cancer Survival Prediction

Authors: Shuo Jiang, Zhuwen Chen, Liaoman Xu, Yanming Zhu, Changmiao Wang, Jiong Zhang, Feiwei Qin, Yifei Chen, Zhu Zhu
Published: 2025-10-07
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.06113v2

Survival analysis plays a vital role in making clinical decisions. However, the models currently in use are often difficult to interpret, which reduces their usefulness in clinical settings. Prototype learning presents a potential solution, yet traditional methods focus on local similarities and static matching, neglecting the broader tumor context and lacking strong semantic alignment with genomic data. To overcome these issues, we introduce an innovative prototype-based multimodal framework, FeatProto, aimed at enhancing cancer survival prediction by addressing significant limitations in current prototype learning methodologies within pathology. Our framework establishes a unified feature prototype space that integrates both global and local features of whole slide images (WSI) with genomic profiles. This integration facilitates traceable and interpretable decision-making processes. Our approach includes three main innovations: (1) A robust phenotype representation that merges critical patches with global context, harmonized with genomic data to minimize local bias. (2) An Exponential Prototype Update Strategy (EMA ProtoUp) that sustains stable cross-modal associations and employs a wandering mechanism to adapt prototypes flexibly to tumor heterogeneity. (3) A hierarchical prototype matching scheme designed to capture global centrality, local typicality, and cohort-level trends, thereby refining prototype inference. Comprehensive evaluations on four publicly available cancer datasets indicate that our method surpasses current leading unimodal and multimodal survival prediction techniques in both accuracy and interpretability, providing a new perspective on prototype learning for critical medical applications. Our source code is available at https://github.com/JSLiam94/FeatProto.

---

### Influence Functions for Efficient Data Selection in Reasoning

Authors: Prateek Humane, Paolo Cudrano, Daniel Z. Kaplan, Matteo Matteucci, Supriyo Chakraborty, Irina Rish
Published: 2025-10-07
Categories: cs.LG, cs.CL
arXiv: https://arxiv.org/abs/2510.06108v2

Fine-tuning large language models (LLMs) on chain-of-thought (CoT) data shows that a small amount of high-quality data can outperform massive datasets. Yet, what constitutes "quality" remains ill-defined. Existing reasoning methods rely on indirect heuristics such as problem difficulty or trace length, while instruction-tuning has explored a broader range of automated selection strategies, but rarely in the context of reasoning. We propose to define reasoning data quality using influence functions, which measure the causal effect of individual CoT examples on downstream accuracy, and introduce influence-based pruning, which consistently outperforms perplexity and embedding-based baselines on math reasoning within a model family.

---

### Distributional Semantics Tracing: A Framework for Explaining Hallucinations in Large Language Models

Authors: Gagan Bhatia, Somayajulu G Sripada, Kevin Allan, Jacobo Azcona
Published: 2025-10-07
Categories: cs.CL, cs.AI, cs.CE
arXiv: https://arxiv.org/abs/2510.06107v3

Hallucinations in large language models (LLMs) produce fluent continuations that are not supported by the prompt, especially under minimal contextual cues and ambiguity. We introduce Distributional Semantics Tracing (DST), a model-native method that builds layer-wise semantic maps at the answer position by decoding residual-stream states through the unembedding, selecting a compact top-$K$ concept set, and estimating directed concept-to-concept support via lightweight causal tracing. Using these traces, we test a representation-level hypothesis: hallucinations arise from correlation-driven representational drift across depth, where the residual stream is pulled toward a locally coherent but context-inconsistent concept neighborhood reinforced by training co-occurrences. On Racing Thoughts dataset, DST yields more faithful explanations than attribution, probing, and intervention baselines under an LLM-judge protocol, and the resulting Contextual Alignment Score (CAS) strongly predicts failures, supporting this drift hypothesis.

---

### The Physics of Data and Tasks: Theories of Locality and Compositionality in Deep Learning

Authors: Alessandro Favero
Published: 2025-10-07
Categories: cs.LG, cond-mat.dis-nn, cond-mat.stat-mech, stat.ML
arXiv: https://arxiv.org/abs/2510.06106v1

Deep neural networks have achieved remarkable success, yet our understanding of how they learn remains limited. These models can learn high-dimensional tasks, which is generally statistically intractable due to the curse of dimensionality. This apparent paradox suggests that learnable data must have an underlying latent structure. What is the nature of this structure? How do neural networks encode and exploit it, and how does it quantitatively impact performance - for instance, how does generalization improve with the number of training examples? This thesis addresses these questions by studying the roles of locality and compositionality in data, tasks, and deep learning representations.

---

### Moloch's Bargain: Emergent Misalignment When LLMs Compete for Audiences

Authors: Batu El, James Zou
Published: 2025-10-07
Categories: cs.AI, cs.CY, cs.HC, cs.LG
arXiv: https://arxiv.org/abs/2510.06105v2

Large language models (LLMs) are increasingly shaping how information is created and disseminated, from companies using them to craft persuasive advertisements, to election campaigns optimizing messaging to gain votes, to social media influencers boosting engagement. These settings are inherently competitive, with sellers, candidates, and influencers vying for audience approval, yet it remains poorly understood how competitive feedback loops influence LLM behavior. We show that optimizing LLMs for competitive success can inadvertently drive misalignment. Using simulated environments across these scenarios, we find that, 6.3% increase in sales is accompanied by a 14.0% rise in deceptive marketing; in elections, a 4.9% gain in vote share coincides with 22.3% more disinformation and 12.5% more populist rhetoric; and on social media, a 7.5% engagement boost comes with 188.6% more disinformation and a 16.3% increase in promotion of harmful behaviors. We call this phenomenon Moloch's Bargain for AI--competitive success achieved at the cost of alignment. These misaligned behaviors emerge even when models are explicitly instructed to remain truthful and grounded, revealing the fragility of current alignment safeguards. Our findings highlight how market-driven optimization pressures can systematically erode alignment, creating a race to the bottom, and suggest that safe deployment of AI systems will require stronger governance and carefully designed incentives to prevent competitive dynamics from undermining societal trust.

---

### The Valley of Code Reasoning: Scaling Knowledge Distillation of Large Language Models

Authors: Muyu He, Muhammad Ali Shafique, Anand Kumar, Tsach Mackey, Nazneen Rajani
Published: 2025-10-07
Categories: cs.CL
arXiv: https://arxiv.org/abs/2510.06101v1

Distilling the thinking traces of a Large Language Model (LLM) with reasoning capabilities into a smaller model has been proven effective. Yet, there is a scarcity of work done on how model performances scale with the quantity of distillation data. In this work, we study the scaling trend of distilling competitive coding skills on two small non-reasoning LLMs. We validate the hypothesis that there is a $\textit{valley of code reasoning}$: downstream performance on competitive coding first drops as data quantity increases, then it steadily increases in a sharper-than-log-linear fashion. Having identified the trend, we further fine-tune the models at two different distillation stages on the same data to ground conclusions on their respective learning phases. We learn that across stages in the low and medium-low data regimes, small models benefit significantly from easier coding questions than from harder ones. We also find that, surprisingly, the correctness of outputs in training data makes no difference to distillation outcomes. Our work represents a step forward in understanding the training dynamics of code reasoning distillation outside intuition

---

### Compact Multi-level-prior Tensor Representation for Hyperspectral Image Super-resolution

Authors: Yinjian Wang, Wei Li, Yuanyuan Gui, Gemine Vivone
Published: 2025-10-07
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.06098v1

Fusing a hyperspectral image with a multispectral image acquired over the same scene, \textit{i.e.}, hyperspectral image super-resolution, has become a popular computational way to access the latent high-spatial-spectral-resolution image. To date, a variety of fusion methods have been proposed, among which the tensor-based ones have testified that multiple priors, such as multidimensional low-rankness and spatial total variation at multiple levels, effectively drive the fusion process. However, existing tensor-based models can only effectively leverage one or two priors at one or two levels, since simultaneously incorporating multi-level priors inevitably increases model complexity. This introduces challenges in both balancing the weights of different priors and optimizing multi-block structures. Concerning this, we present a novel hyperspectral super-resolution model compactly characterizing these multi-level priors of hyperspectral images within the tensor framework. Firstly, the proposed model decouples the spectral low-rankness and spatial priors by casting the latent high-spatial-spectral-resolution image into spectral subspace and spatial maps via block term decomposition. Secondly, these spatial maps are stacked as the spatial tensor encoding the high-order spatial low-rankness and smoothness priors, which are co-modeled via the proposed non-convex mode-shuffled tensor correlated total variation. Finally, we draw inspiration from the linearized alternating direction method of multipliers to design an efficient algorithm to optimize the resulting model, theoretically proving its Karush-Kuhn-Tucker convergence under mild conditions. Experiments on multiple datasets demonstrate the effectiveness of the proposed algorithm. The code implementation will be available from https://github.com/WongYinJ.

---

### The Alignment Auditor: A Bayesian Framework for Verifying and Refining LLM Objectives

Authors: Matthieu Bou, Nyal Patel, Arjun Jagota, Satyapriya Krishna, Sonali Parbhoo
Published: 2025-10-07
Categories: cs.LG, cs.CL
arXiv: https://arxiv.org/abs/2510.06096v3

The objectives that Large Language Models (LLMs) implicitly optimize remain dangerously opaque, making trustworthy alignment and auditing a grand challenge. While Inverse Reinforcement Learning (IRL) can infer reward functions from behaviour, existing approaches either produce a single, overconfident reward estimate or fail to address the fundamental ambiguity of the task (non-identifiability). This paper introduces a principled auditing framework that re-frames reward inference from a simple estimation task to a comprehensive process for verification. Our framework leverages Bayesian IRL to not only recover a distribution over objectives but to enable three critical audit capabilities: (i) Quantifying and systematically reducing non-identifiability by demonstrating posterior contraction over sequential rounds of evidence; (ii) Providing actionable, uncertainty-aware diagnostics that expose spurious shortcuts and identify out-of-distribution prompts where the inferred objective cannot be trusted; and (iii) Validating policy-level utility by showing that the refined, low-uncertainty reward can be used directly in RLHF to achieve training dynamics and toxicity reductions comparable to the ground-truth alignment process. Empirically, our framework successfully audits a detoxified LLM, yielding a well-calibrated and interpretable objective that strengthens alignment guarantees. Overall, this work provides a practical toolkit for auditors, safety teams, and regulators to verify what LLMs are truly trying to achieve, moving us toward more trustworthy and accountable AI.

---

### Classical AI vs. LLMs for Decision-Maker Alignment in Health Insurance Choices

Authors: Mallika Mainali, Harsha Sureshbabu, Anik Sen, Christopher B. Rauch, Noah D. Reifsnyder, John Meyer, J. T. Turner, Michael W. Floyd, Matthew Molineaux, Rosina O. Weber
Published: 2025-10-07
Categories: cs.AI
arXiv: https://arxiv.org/abs/2510.06093v1

As algorithmic decision-makers are increasingly applied to high-stakes domains, AI alignment research has evolved from a focus on universal value alignment to context-specific approaches that account for decision-maker attributes. Prior work on Decision-Maker Alignment (DMA) has explored two primary strategies: (1) classical AI methods integrating case-based reasoning, Bayesian reasoning, and naturalistic decision-making, and (2) large language model (LLM)-based methods leveraging prompt engineering. While both approaches have shown promise in limited domains such as medical triage, their generalizability to novel contexts remains underexplored. In this work, we implement a prior classical AI model and develop an LLM-based algorithmic decision-maker evaluated using a large reasoning model (GPT-5) and a non-reasoning model (GPT-4) with weighted self-consistency under a zero-shot prompting framework, as proposed in recent literature. We evaluate both approaches on a health insurance decision-making dataset annotated for three target decision-makers with varying levels of risk tolerance (0.0, 0.5, 1.0). In the experiments reported herein, classical AI and LLM-based models achieved comparable alignment with attribute-based targets, with classical AI exhibiting slightly better alignment for a moderate risk profile. The dataset and open-source implementation are publicly available at: https://github.com/TeX-Base/ClassicalAIvsLLMsforDMAlignment and https://github.com/Parallax-Advanced-Research/ITM/tree/feature_insurance.

---

### Learning from Failures: Understanding LLM Alignment through Failure-Aware Inverse RL

Authors: Nyal Patel, Matthieu Bou, Arjun Jagota, Satyapriya Krishna, Sonali Parbhoo
Published: 2025-10-07
Categories: cs.LG, cs.CL
arXiv: https://arxiv.org/abs/2510.06092v2

Reinforcement Learning from Human Feedback (RLHF) aligns Large Language Models (LLMs) with human preferences, yet the underlying reward signals they internalize remain hidden, posing a critical challenge for interpretability and safety. Existing approaches attempt to extract these latent incentives using Inverse Reinforcement Learning (IRL), but treat all preference pairs equally, often overlooking the most informative signals: those examples the extracted reward model misclassifies or assigns nearly equal scores, which we term \emph{failures}. We introduce a novel \emph{failure-aware} IRL algorithm that focuses on misclassified or difficult examples to recover the latent rewards defining model behaviors. By learning from these failures, our failure-aware IRL extracts reward functions that better reflect the true objectives behind RLHF. We demonstrate that failure-aware IRL outperforms existing IRL baselines across multiple metrics when applied to LLM detoxification, without requiring external classifiers or supervision. Crucially, failure-aware IRL yields rewards that better capture the true incentives learned during RLHF, enabling more effective re-RLHF training than standard IRL. This establishes failure-aware IRL as a robust, scalable method for auditing model alignment and reducing ambiguity in the IRL process.

---

### Learning Mixtures of Linear Dynamical Systems via Hybrid Tensor-EM Method

Authors: Lulu Gong, Shreya Saxena
Published: 2025-10-07
Categories: cs.LG, eess.SY, q-bio.NC, stat.ML
arXiv: https://arxiv.org/abs/2510.06091v2

Mixtures of linear dynamical systems (MoLDS) provide a path to model time-series data that exhibit diverse temporal dynamics across trajectories. However, its application remains challenging in complex and noisy settings, limiting its effectiveness for neural data analysis. Tensor-based moment methods can provide global identifiability guarantees for MoLDS, but their performance degrades under noise and complexity. Commonly used expectation-maximization (EM) methods offer flexibility in fitting latent models but are highly sensitive to initialization and prone to poor local minima. Here, we propose a tensor-based method that provides identifiability guarantees for learning MoLDS, which is followed by EM updates to combine the strengths of both approaches. The novelty in our approach lies in the construction of moment tensors using the input-output data to recover globally consistent estimates of mixture weights and system parameters. These estimates can then be refined through a Kalman EM algorithm, with closed-form updates for all LDS parameters. We validate our framework on synthetic benchmarks and real-world datasets. On synthetic data, the proposed Tensor-EM method achieves more reliable recovery and improved robustness compared to either pure tensor or randomly initialized EM methods. We then analyze neural recordings from the primate somatosensory cortex while a non-human primate performs reaches in different directions. Our method successfully models and clusters different conditions as separate subsystems, consistent with supervised single-LDS fits for each condition. Finally, we apply this approach to another neural dataset where monkeys perform a sequential reaching task. These results demonstrate that MoLDS provides an effective framework for modeling complex neural data, and that Tensor-EM is a reliable approach to MoLDS learning for these applications.

---

### A public cardiac CT dataset featuring the left atrial appendage

Authors: Bjoern Hansen, Jonas Pedersen, Klaus F. Kofoed, Oscar Camara, Rasmus R. Paulsen, Kristine Soerensen
Published: 2025-10-07
Categories: cs.CV, cs.AI
arXiv: https://arxiv.org/abs/2510.06090v1

Despite the success of advanced segmentation frameworks such as TotalSegmentator (TS), accurate segmentations of the left atrial appendage (LAA), coronary arteries (CAs), and pulmonary veins (PVs) remain a significant challenge in medical imaging. In this work, we present the first open-source, anatomically coherent dataset of curated, high-resolution segmentations for these structures, supplemented with whole-heart labels produced by TS on the publicly available ImageCAS dataset consisting of 1000 cardiac computed tomography angiography (CCTA) scans. One purpose of the data set is to foster novel approaches to the analysis of LAA morphology. LAA segmentations on ImageCAS were generated using a state-of-the-art segmentation framework developed specifically for high resolution LAA segmentation. We trained the network on a large private dataset with manual annotations provided by medical readers guided by a trained cardiologist and transferred the model to ImageCAS data. CA labels were improved from the original ImageCAS annotations, while PV segmentations were refined from TS outputs. In addition, we provide a list of scans from ImageCAS that contains common data flaws such as step artefacts, LAAs extending beyond the scanner's field of view, and other types of data defects.

---

### Spectrum Tuning: Post-Training for Distributional Coverage and In-Context Steerability

Authors: Taylor Sorensen, Benjamin Newman, Jared Moore, Chan Park, Jillian Fisher, Niloofar Mireshghallah, Liwei Jiang, Yejin Choi
Published: 2025-10-07
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2510.06084v2

Language model post-training has enhanced instruction-following and performance on many downstream tasks, but also comes with an often-overlooked cost on tasks with many possible valid answers. On many tasks such as creative writing, synthetic data generation, or steering to diverse preferences, models must cover an entire distribution of outputs, rather than a single correct answer. We characterize three desiderata for conditional distributional modeling: in-context steerability, valid output space coverage, and distributional alignment, and document across three model families how current post-training can reduce these properties. In particular, we disambiguate between two kinds of in-context learning: ICL for eliciting existing underlying knowledge or capabilities, and in-context steerability, where a model must use in-context information to override its priors and steer to a novel data generating distribution. To better evaluate and improve these desiderata, we introduce Spectrum Suite, a large-scale resource compiled from >40 data sources and spanning >90 tasks requiring models to steer to and match diverse distributions ranging from varied human preferences to numerical distributions and more. We find that while current post-training techniques elicit underlying capabilities and knowledge, they hurt models' ability to flexibly steer in-context. To mitigate these issues, we propose Spectrum Tuning, a post-training method using Spectrum Suite to improve steerability and distributional coverage. We find that Spectrum Tuning often improves over pretrained and typical instruction-tuned models, enhancing steerability, spanning more of the output space, and improving distributional alignment on held-out datasets.

---

### Constraint-Aware Route Recommendation from Natural Language via Hierarchical LLM Agents

Authors: Tao Zhe, Rui Liu, Fateme Memar, Xiao Luo, Wei Fan, Xinyue Ye, Zhongren Peng, Dongjie Wang
Published: 2025-10-07
Categories: cs.AI
arXiv: https://arxiv.org/abs/2510.06078v1

Route recommendation aims to provide users with optimal travel plans that satisfy diverse and complex requirements. Classical routing algorithms (e.g., shortest-path and constraint-aware search) are efficient but assume structured inputs and fixed objectives, limiting adaptability to natural-language queries. Recent LLM-based approaches enhance flexibility but struggle with spatial reasoning and the joint modeling of route-level and POI-level preferences. To address these limitations, we propose RouteLLM, a hierarchical multi-agent framework that grounds natural-language intents into constraint-aware routes. It first parses user queries into structured intents including POIs, paths, and constraints. A manager agent then coordinates specialized sub-agents: a constraint agent that resolves and formally check constraints, a POI agent that retrieves and ranks candidate POIs, and a path refinement agent that refines routes via a routing engine with preference-conditioned costs. A final verifier agent ensures constraint satisfaction and produces the final route with an interpretable rationale. This design bridges linguistic flexibility and spatial structure, enabling reasoning over route feasibility and user preferences. Experiments show that our method reliably grounds textual preferences into constraint-aware routes, improving route quality and preference satisfaction over classical methods.

---

### When Thinking Drifts: Evidential Grounding for Robust Video Reasoning

Authors: Mi Luo, Zihui Xue, Alex Dimakis, Kristen Grauman
Published: 2025-10-07
Categories: cs.CV, cs.AI
arXiv: https://arxiv.org/abs/2510.06077v1

Video reasoning, the task of enabling machines to infer from dynamic visual content through multi-step logic, is crucial for advanced AI. While the Chain-of-Thought (CoT) mechanism has enhanced reasoning in text-based tasks, its application to video understanding remains underexplored. This paper presents a systematic analysis revealing that CoT often degrades performance in video reasoning, generating verbose but misleading internal monologues, and leading to hallucinated visual details and overridden correct intuitions - a phenomenon we term "visual thinking drift". We explain this drift through a Bayesian lens, positing that CoT traces often diverge from actual visual evidence, instead amplifying internal biases or language priors, causing models to storytell rather than engage in grounded reasoning. To counteract this, we introduce Visual Evidence Reward (VER), a novel reinforcement learning framework that explicitly rewards the generation of reasoning traces that are verifiably grounded in visual evidence. Comprehensive evaluation across 10 diverse video understanding benchmarks demonstrates that our Video-VER consistently achieves top performance. Our work sheds light on the distinct challenges of video-centric reasoning and encourages the development of AI that robustly grounds its inferences in visual evidence - for large multimodal models that not only "think before answering", but also "see while thinking".

---

### EmoHRNet: High-Resolution Neural Network Based Speech Emotion Recognition

Authors: Akshay Muppidi, Martin Radfar
Published: 2025-10-07
Categories: cs.SD, cs.LG
arXiv: https://arxiv.org/abs/2510.06072v1

Speech emotion recognition (SER) is pivotal for enhancing human-machine interactions. This paper introduces "EmoHRNet", a novel adaptation of High-Resolution Networks (HRNet) tailored for SER. The HRNet structure is designed to maintain high-resolution representations from the initial to the final layers. By transforming audio samples into spectrograms, EmoHRNet leverages the HRNet architecture to extract high-level features. EmoHRNet's unique architecture maintains high-resolution representations throughout, capturing both granular and overarching emotional cues from speech signals. The model outperforms leading models, achieving accuracies of 92.45% on RAVDESS, 80.06% on IEMOCAP, and 92.77% on EMOVO. Thus, we show that EmoHRNet sets a new benchmark in the SER domain.

---

### Benchmark It Yourself (BIY): Preparing a Dataset and Benchmarking AI Models for Scatterplot-Related Tasks

Authors: João Palmeiro, Diogo Duarte, Rita Costa, Pedro Bizarro
Published: 2025-10-07
Categories: cs.LG, cs.AI, cs.HC
arXiv: https://arxiv.org/abs/2510.06071v1

AI models are increasingly used for data analysis and visualization, yet benchmarks rarely address scatterplot-specific tasks, limiting insight into performance. To address this gap for one of the most common chart types, we introduce a synthetic, annotated dataset of over 18,000 scatterplots from six data generators and 17 chart designs, and a benchmark based on it. We evaluate proprietary models from OpenAI and Google using N-shot prompting on five distinct tasks derived from annotations of cluster bounding boxes, their center coordinates, and outlier coordinates. OpenAI models and Gemini 2.5 Flash, especially when prompted with examples, are viable options for counting clusters and, in Flash's case, outliers (90%+ Accuracy). However, the results for localization-related tasks are unsatisfactory: Precision and Recall are near or below 50%, except for Flash in outlier identification (65.01%). Furthermore, the impact of chart design on performance appears to be a secondary factor, but it is advisable to avoid scatterplots with wide aspect ratios (16:9 and 21:9) or those colored randomly. Supplementary materials are available at https://github.com/feedzai/biy-paper.

---

### There is More to Attention: Statistical Filtering Enhances Explanations in Vision Transformers

Authors: Meghna P Ayyar, Jenny Benois-Pineau, Akka Zemmari
Published: 2025-10-07
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.06070v1

Explainable AI (XAI) has become increasingly important with the rise of large transformer models, yet many explanation methods designed for CNNs transfer poorly to Vision Transformers (ViTs). Existing ViT explanations often rely on attention weights, which tend to yield noisy maps as they capture token-to-token interactions within each layer.While attribution methods incorporating MLP blocks have been proposed, we argue that attention remains a valuable and interpretable signal when properly filtered. We propose a method that combines attention maps with a statistical filtering, initially proposed for CNNs, to remove noisy or uninformative patterns and produce more faithful explanations. We further extend our approach with a class-specific variant that yields discriminative explanations. Evaluation against popular state-of-the-art methods demonstrates that our approach produces sharper and more interpretable maps. In addition to perturbation-based faithfulness metrics, we incorporate human gaze data to assess alignment with human perception, arguing that human interpretability remains essential for XAI. Across multiple datasets, our approach consistently outperforms or is comparable to the SOTA methods while remaining efficient and human plausible.

---

### MachaGrasp: Morphology-Aware Cross-Embodiment Dexterous Hand Articulation Generation for Grasping

Authors: Heng Zhang, Kevin Yuchen Ma, Mike Zheng Shou, Weisi Lin, Yan Wu
Published: 2025-10-07
Categories: cs.RO, cs.AI
arXiv: https://arxiv.org/abs/2510.06068v3

Dexterous grasping with multi-fingered hands remains challenging due to high-dimensional articulations and the cost of optimization-based pipelines. Existing end-to-end methods require training on large-scale datasets for specific hands, limiting their ability to generalize across different embodiments. We propose MachaGrasp, an eigengrasp-based, end-to-end framework for cross-embodiment grasp generation. From a hand's morphology description, we derive a morphology embedding and an eigengrasp set. Conditioned on these, together with the object point cloud and wrist pose, an amplitude predictor regresses articulation coefficients in a low-dimensional space, which are decoded into full joint articulations. Articulation learning is supervised with a Kinematic-Aware Articulation Loss (KAL) that emphasizes fingertip-relevant motions and injects morphology-specific structure. In simulation on unseen objects across three dexterous hands, MachaGrasp attains a 91.9% average grasp success rate with less than 0.4 seconds inference per grasp. With few-shot adaptation to an unseen hand, it achieves 85.6% success on unseen objects in simulation, and real-world experiments on this few-shot-generalized hand achieve an 87% success rate. The code and additional materials are available on our project website https://connor-zh.github.io/MachaGrasp.

---

### Reasoning under Vision: Understanding Visual-Spatial Cognition in Vision-Language Models for CAPTCHA

Authors: Python Song, Luke Tenyi Chang, Yun-Yun Tsai, Penghui Li, Junfeng Yang
Published: 2025-10-07
Categories: cs.CV, cs.AI
arXiv: https://arxiv.org/abs/2510.06067v2

CAPTCHA, originally designed to distinguish humans from robots, has evolved into a real-world benchmark for assessing the spatial reasoning capabilities of vision-language models. In this work, we first show that step-by-step reasoning is crucial for vision-language models (VLMs) to solve CAPTCHAs, which represent high-difficulty spatial reasoning tasks, and that current commercial vision-language models still struggle with such reasoning. In particular, we observe that most commercial VLMs (e.g., Gemini, Claude, GPT, etc.) fail to effectively solve CAPTCHAs and thus achieve low accuracy (around 21.9 percent). However, our findings indicate that requiring the model to perform step-by-step reasoning before generating the final coordinates can significantly enhance its solving accuracy, underscoring the severity of the gap. To systematically study this issue, we introduce CAPTCHA-X, the first real-world CAPTCHA benchmark with reasoning, covering seven categories of CAPTCHAs (such as Gobang, hCaptcha, etc.) with step-by-step action solutions and grounding annotations. We further define five reasoning-oriented metrics that enable a comprehensive evaluation of models reasoning capabilities. To validate the effectiveness of reasoning, we also propose a general agentic VLM-based framework that incorporates the models inherent reasoning abilities. Our method achieves state-of-the-art performance across five high-difficulty CAPTCHA types, with an average solving accuracy of 83.9 percent, substantially surpassing existing baselines. These results reveal the limitations of current models and highlight the importance of reasoning in advancing visual-spatial challenges in the future.

---

### Analyzing the Effect of Embedding Norms and Singular Values to Oversmoothing in Graph Neural Networks

Authors: Dimitrios Kelesis, Dimitris Fotakis, Georgios Paliouras
Published: 2025-10-07
Categories: cs.LG
arXiv: https://arxiv.org/abs/2510.06066v1

In this paper, we study the factors that contribute to the effect of oversmoothing in deep Graph Neural Networks (GNNs). Specifically, our analysis is based on a new metric (Mean Average Squared Distance - $MASED$) to quantify the extent of oversmoothing. We derive layer-wise bounds on $MASED$, which aggregate to yield global upper and lower distance bounds. Based on this quantification of oversmoothing, we further analyze the importance of two different properties of the model; namely the norms of the generated node embeddings, along with the largest and smallest singular values of the weight matrices. Building on the insights drawn from the theoretical analysis, we show that oversmoothing increases as the number of trainable weight matrices and the number of adjacency matrices increases. We also use the derived layer-wise bounds on $MASED$ to form a proposal for decoupling the number of hops (i.e., adjacency depth) from the number of weight matrices. In particular, we introduce G-Reg, a regularization scheme that increases the bounds, and demonstrate through extensive experiments that by doing so node classification accuracy increases, achieving robustness at large depths. We further show that by reducing oversmoothing in deep networks, we can achieve better results in some tasks than using shallow ones. Specifically, we experiment with a ``cold start" scenario, i.e., when there is no feature information for the unlabeled nodes. Finally, we show empirically the trade-off between receptive field size (i.e., number of weight matrices) and performance, using the $MASED$ bounds. This is achieved by distributing adjacency hops across a small number of trainable layers, avoiding the extremes of under- or over-parameterization of the GNN.

---

### Medical Vision Language Models as Policies for Robotic Surgery

Authors: Akshay Muppidi, Martin Radfar
Published: 2025-10-07
Categories: cs.CV, cs.LG
arXiv: https://arxiv.org/abs/2510.06064v1

Vision-based Proximal Policy Optimization (PPO) struggles with visual observation-based robotic laparoscopic surgical tasks due to the high-dimensional nature of visual input, the sparsity of rewards in surgical environments, and the difficulty of extracting task-relevant features from raw visual data. We introduce a simple approach integrating MedFlamingo, a medical domain-specific Vision-Language Model, with PPO. Our method is evaluated on five diverse laparoscopic surgery task environments in LapGym, using only endoscopic visual observations. MedFlamingo PPO outperforms and converges faster compared to both standard vision-based PPO and OpenFlamingo PPO baselines, achieving task success rates exceeding 70% across all environments, with improvements ranging from 66.67% to 1114.29% compared to baseline. By processing task observations and instructions once per episode to generate high-level planning tokens, our method efficiently combines medical expertise with real-time visual feedback. Our results highlight the value of specialized medical knowledge in robotic surgical planning and decision-making.

---

### TelecomTS: A Multi-Modal Observability Dataset for Time Series and Language Analysis

Authors: Austin Feng, Andreas Varvarigos, Ioannis Panitsas, Daniela Fernandez, Jinbiao Wei, Yuwei Guo, Jialin Chen, Ali Maatouk, Leandros Tassiulas, Rex Ying
Published: 2025-10-07
Categories: cs.AI, cs.IT, cs.LG
arXiv: https://arxiv.org/abs/2510.06063v3

Modern enterprises generate vast streams of time series metrics when monitoring complex systems, known as observability data. Unlike conventional time series from domains such as climate, observability data are zero-inflated, highly stochastic, and exhibit minimal temporal structure. Despite their importance, observability datasets remain underrepresented in public benchmarks due to proprietary restrictions and privacy concerns. Existing datasets are often anonymized and normalized, removing scale information and limiting their use for tasks such as anomaly detection, root cause analysis, and multi-modal reasoning. To address this gap, we introduce TelecomTS, a large-scale observability dataset derived from a 5G telecommunications network. TelecomTS features heterogeneous, de-anonymized covariates with explicit absolute scale information and provides a diverse suite of downstream tasks, including anomaly detection, root cause analysis, and multi-modal question-answering. Benchmarking state-of-the-art time series, language, reasoning, and multi-modal foundation models reveals that existing approaches struggle with the abrupt, noisy, and high-variance dynamics characteristic of observability data. Our experiments further underscore the importance of preserving covariates' absolute scale, emphasizing the need for foundation time series models that natively leverage scale information for practical real-world observability applications. The code is available at: https://github.com/Ali-maatouk/TelecomTS.

---

### When Importance Sampling Misallocates Credit: Asymmetric Ratios for Outcome-Supervised RL

Authors: Jiakang Wang, Runze Liu, Qingpeng Cai, Lei Lin, Wenping Hu, Xiu Li, Fuzheng Zhang, Guorui Zhou, Kun Gai, Ling Pan
Published: 2025-10-07
Categories: cs.CL
arXiv: https://arxiv.org/abs/2510.06062v2

Reinforcement learning (RL) has shown great promise in large language models (LLMs) post-training, which typically rely on token-level clipping to maintain stability during optimization. Despite the empirical success of GRPO-style methods, we identify a fundamental and previously overlooked challenge in this popular Outcome-Supervised RL (OSRL) paradigm. We reveal that in OSRL, where advantages are shared across tokens within a response, importance sampling (IS) ratios deviate from their traditional purpose of distribution correction as in classic RL, which become token-level weights that allocate the shared advantage signal across tokens. We show that this hidden role shift induces a critical mismatch for positive-advantage tokens, leading to unbalanced token weighting between positive and negative tokens. Specifically, it suppresses the update of underrepresented tokens that are lagging behind, while over-amplifying already high-probability tokens. This mismatch results in rich-get-richer dynamics that over-reinforce confident tokens, weaken catch-up learning that drive entropy collapse, excessive repetition, and premature convergence. To address this, we propose Asymmetric Importance Sampling Policy Optimization (ASPO), a simple yet effective strategy that reverses the ratio-induced weighting of positive-advantage tokens, while stabilizing extreme updates and maintaining gradient flow. This mismatch correction aligns their update direction with the learning dynamics of negative ones. Comprehensive experiments across math reasoning and coding benchmarks demonstrate that ASPO significantly mitigates entropy collapse, improves training stability, and enhances performance over strong GRPO-based baselines. Our analysis provides new insights into the role of token-level weighting in OSRL and highlights the critical importance of correcting ratio-induced weighting in LLM RL.

---

### Controllable Audio-Visual Viewpoint Generation from 360° Spatial Information

Authors: Christian Marinoni, Riccardo Fosco Gramaccioni, Eleonora Grassucci, Danilo Comminiello
Published: 2025-10-07
Categories: cs.MM, cs.AI, cs.CV
arXiv: https://arxiv.org/abs/2510.06060v1

The generation of sounding videos has seen significant advancements with the advent of diffusion models. However, existing methods often lack the fine-grained control needed to generate viewpoint-specific content from larger, immersive 360-degree environments. This limitation restricts the creation of audio-visual experiences that are aware of off-camera events. To the best of our knowledge, this is the first work to introduce a framework for controllable audio-visual generation, addressing this unexplored gap. Specifically, we propose a diffusion model by introducing a set of powerful conditioning signals derived from the full 360-degree space: a panoramic saliency map to identify regions of interest, a bounding-box-aware signed distance map to define the target viewpoint, and a descriptive caption of the entire scene. By integrating these controls, our model generates spatially-aware viewpoint videos and audios that are coherently influenced by the broader, unseen environmental context, introducing a strong controllability that is essential for realistic and immersive audio-visual generation. We show audiovisual examples proving the effectiveness of our framework.

---

### Scientific Algorithm Discovery by Augmenting AlphaEvolve with Deep Research

Authors: Gang Liu, Yihan Zhu, Jie Chen, Meng Jiang
Published: 2025-10-07
Categories: cs.AI
arXiv: https://arxiv.org/abs/2510.06056v1

Large language models hold promise as scientific assistants, yet existing agents either rely solely on algorithm evolution or on deep research in isolation, both of which face critical limitations. Pure algorithm evolution, as in AlphaEvolve, depends only on the internal knowledge of LLMs and quickly plateaus in complex domains, while pure deep research proposes ideas without validation, resulting in unrealistic or unimplementable solutions. We present DeepEvolve, an agent that integrates deep research with algorithm evolution, uniting external knowledge retrieval, cross-file code editing, and systematic debugging under a feedback-driven iterative loop. Each iteration not only proposes new hypotheses but also refines, implements, and tests them, avoiding both shallow improvements and unproductive over-refinements. Across nine benchmarks in chemistry, mathematics, biology, materials, and patents, DeepEvolve consistently improves the initial algorithm, producing executable new algorithms with sustained gains. By bridging the gap between unguided evolution and research without grounding, DeepEvolve provides a reliable framework for advancing scientific algorithm discovery. Our code is available at https://github.com/liugangcode/deepevolve.

---

### MixReasoning: Switching Modes to Think

Authors: Haiquan Lu, Gongfan Fang, Xinyin Ma, Qi Li, Xinchao Wang
Published: 2025-10-07
Categories: cs.AI, cs.CL
arXiv: https://arxiv.org/abs/2510.06052v2

Reasoning models enhance performance by tackling problems in a step-by-step manner, decomposing them into sub-problems and exploring long chains of thought before producing an answer. However, applying extended reasoning to every step introduces substantial redundancy, as sub-problems vary widely in difficulty and complexity: a small number of pivotal steps are genuinely challenging and decisive for the final answer, while many others only involve straightforward revisions or simple computations. Therefore, a natural idea is to endow reasoning models with the ability to adaptively respond to this variation, rather than treating all steps with the same level of elaboration. To this end, we propose MixReasoning, a framework that dynamically adjusts the depth of reasoning within a single response. The resulting chain of thought then becomes a mixture of detailed reasoning on difficult steps and concise inference on simpler ones. Experiments on GSM8K, MATH-500, and AIME show that MixReasoning shortens reasoning length and substantially improves efficiency without compromising accuracy.

---

### Edit-Based Flow Matching for Temporal Point Processes

Authors: David Lüdke, Marten Lienen, Marcel Kollovieh, Stephan Günnemann
Published: 2025-10-07
Categories: cs.LG
arXiv: https://arxiv.org/abs/2510.06050v3

Temporal point processes (TPPs) are a fundamental tool for modeling event sequences in continuous time, but most existing approaches rely on autoregressive parameterizations that are limited by their sequential sampling. Recent non-autoregressive, diffusion-style models mitigate these issues by jointly interpolating between noise and data through event insertions and deletions in a discrete Markov chain. In this work, we generalize this perspective and introduce an Edit Flow process for TPPs that transports noise to data via insert, delete, and substitute edit operations. By learning the instantaneous edit rates within a continuous-time Markov chain framework, we attain a flexible and efficient model that effectively reduces the total number of necessary edit operations during generation. Empirical results demonstrate the generative flexibility of our unconditionally trained model in a wide range of unconditional and conditional generation tasks on benchmark TPPs.

---

### BLISS: A Lightweight Bilevel Influence Scoring Method for Data Selection in Language Model Pretraining

Authors: Jie Hao, Rui Yu, Wei Zhang, Huixia Wang, Jie Xu, Mingrui Liu
Published: 2025-10-07
Categories: cs.LG
arXiv: https://arxiv.org/abs/2510.06048v5

Effective data selection is essential for pretraining large language models (LLMs), enhancing efficiency and improving generalization to downstream tasks. However, existing approaches often require leveraging external pretrained models, making it difficult to disentangle the effects of data selection from those of the external pretrained models. In addition, they often overlook the long-term impact of selected data if the model is trained to convergence, primarily due to the prohibitive cost of full-scale LLM pretraining. In this paper, we introduce BLISS (\textbf{B}ileve\textbf{L} \textbf{I}nfluence \textbf{S}coring method for data \textbf{S}election): a lightweight data selection method that operates entirely \emph{from scratch}, without relying on any external pretrained oracle models, while explicitly accounting for the long-term impact of selected data. BLISS leverages a small proxy model as a surrogate for the LLM and employs a score model to estimate the long-term influence of training samples if the proxy model is trained to convergence. We formulate data selection as a bilevel optimization problem, where the upper-level objective optimizes the score model to assign importance weights to training samples, ensuring that minimizing the lower-level objective (i.e., training the proxy model over the weighted training loss until convergence) leads to best validation performance. Once optimized, the trained score model predicts influence scores for the dataset, enabling efficient selection of high-quality samples for LLM pretraining. We validate BLISS by pretraining 410M/1B/2.8B Pythia and LLaMA-0.5B models on selected subsets of the C4 dataset. Notably, under the 1B model setting, BLISS achieves $1.7\times$ speedup in reaching the same performance as the state-of-the-art method, demonstrating superior performance across multiple downstream tasks.

---

### GLVD: Guided Learned Vertex Descent

Authors: Pol Caselles Rico, Francesc Moreno Noguer
Published: 2025-10-07
Categories: cs.CV, cs.AI
arXiv: https://arxiv.org/abs/2510.06046v1

Existing 3D face modeling methods usually depend on 3D Morphable Models, which inherently constrain the representation capacity to fixed shape priors. Optimization-based approaches offer high-quality reconstructions but tend to be computationally expensive. In this work, we introduce GLVD, a hybrid method for 3D face reconstruction from few-shot images that extends Learned Vertex Descent (LVD) by integrating per-vertex neural field optimization with global structural guidance from dynamically predicted 3D keypoints. By incorporating relative spatial encoding, GLVD iteratively refines mesh vertices without requiring dense 3D supervision. This enables expressive and adaptable geometry reconstruction while maintaining computational efficiency. GLVD achieves state-of-the-art performance in single-view settings and remains highly competitive in multi-view scenarios, all while substantially reducing inference time.

---

### VideoMiner: Iteratively Grounding Key Frames of Hour-Long Videos via Tree-based Group Relative Policy Optimization

Authors: Xinye Cao, Hongcan Guo, Jiawen Qian, Guoshun Nan, Chao Wang, Yuqi Pan, Tianhao Hou, Xiaojuan Wang, Yutong Gao
Published: 2025-10-07
Categories: cs.CV, cs.AI
arXiv: https://arxiv.org/abs/2510.06040v1

Understanding hour-long videos with multi-modal large language models (MM-LLMs) enriches the landscape of human-centered AI applications. However, for end-to-end video understanding with LLMs, uniformly sampling video frames results in LLMs being overwhelmed by a vast amount of irrelevant information as video length increases. Existing hierarchical key frame extraction methods improve the accuracy of video understanding but still face two critical challenges. 1) How can the interference of extensive redundant information in long videos be mitigated? 2) How can a model dynamically adapt to complex hierarchical structures while accurately identifying key frames? To address these issues, we propose VideoMiner, which iteratively segments, captions, and clusters long videos, forming a hierarchical tree structure. The proposed VideoMiner progresses from long videos to events to frames while preserving temporal coherence, effectively addressing the first challenge. To precisely locate key frames, we introduce T-GRPO, a tree-based group relative policy optimization in reinforcement learning method that guides the exploration of the VideoMiner. The proposed T-GRPO is specifically designed for tree structures, integrating spatiotemporal information at the event level while being guided by the question, thus solving the second challenge. We achieve superior performance in all long-video understanding tasks and uncover several interesting insights. Our proposed T-GRPO surprisingly incentivizes the model to spontaneously generate a reasoning chain. Additionally, the designed tree growth auxin dynamically adjusts the expansion depth, obtaining accuracy and efficiency gains. The code is publicly available at https://github.com/caoxinye/VideoMiner.

---

### A Large-Scale Chinese Knowledge Graph-Text Alignment Dataset for Benchmarking Knowledge-Grounded LLMs

Authors: Chengwei Wu, Xingrui Zhuo, Mingyang Gao, Xinghe Cheng, Zhichao Yan, Jiapu Wang
Published: 2025-10-07
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2510.06039v2

Reliable evaluation of knowledge-grounded Large Language Models (LLMs) in Chinese requires resources that explicitly align Chinese-language text with verifiable Knowledge Graph (KG) facts. Yet existing Chinese benchmarks primarily assess general language understanding and offer limited support for structured reasoning under Chinese-specific linguistic phenomena. We introduce the Chinese Data-Text Pair (CDTP), a large-scale Chinese KG-text alignment dataset comprising more than 7 million aligned instances across four broad domains. Each instance pairs a Chinese-language text with one or more textually supported KG triples, totaling 15 million triples. A multi-stage construction pipeline combining alignment filtering, manual verification, and external evidence validation improves semantic consistency and factual reliability. CDTP supports Knowledge Graph Completion (KGC), Question Answering (QA), and Triple-to-Text Generation (T2T). Across all three tasks, the benchmark design accounts for Chinese-specific phenomena, including polysemy, word-segmentation ambiguity, and context-dependent entity interpretation, enabling the evaluation of structured reasoning, ambiguity-aware factual understanding, and knowledge-grounded generation. Experiments with diverse open-source and proprietary LLMs show that model scale alone does not guarantee reliable performance on these Chinese knowledge-intensive tasks, whereas supervised fine-tuning on CDTP consistently improves in-domain performance and out-of-distribution robustness. The publicly accessible dataset, code, and evaluation protocols provide a reusable resource for developing and evaluating knowledge-grounded LLMs in Chinese.

---

### From Learning to Mastery: Achieving Safe and Efficient Real-World Autonomous Driving with Human-In-The-Loop Reinforcement Learning

Authors: Li Zeqiao, Wang Yijing, Wang Haoyu, Li Zheng, Li Peng, Liu Wenfei, Zuo Zhiqiang
Published: 2025-10-07
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2510.06038v1

Autonomous driving with reinforcement learning (RL) has significant potential. However, applying RL in real-world settings remains challenging due to the need for safe, efficient, and robust learning. Incorporating human expertise into the learning process can help overcome these challenges by reducing risky exploration and improving sample efficiency. In this work, we propose a reward-free, active human-in-the-loop learning method called Human-Guided Distributional Soft Actor-Critic (H-DSAC). Our method combines Proxy Value Propagation (PVP) and Distributional Soft Actor-Critic (DSAC) to enable efficient and safe training in real-world environments. The key innovation is the construction of a distributed proxy value function within the DSAC framework. This function encodes human intent by assigning higher expected returns to expert demonstrations and penalizing actions that require human intervention. By extrapolating these labels to unlabeled states, the policy is effectively guided toward expert-like behavior. With a well-designed state space, our method achieves real-world driving policy learning within practical training times. Results from both simulation and real-world experiments demonstrate that our framework enables safe, robust, and sample-efficient learning for autonomous driving.

---

### Refusal Falls off a Cliff: How Safety Alignment Fails in Reasoning?

Authors: Qingyu Yin, Chak Tou Leong, Linyi Yang, Wenxuan Huang, Wenjie Li, Xiting Wang, Jaehong Yoon, YunXing, XingYu, Jinjin Gu
Published: 2025-10-07
Categories: cs.AI, cs.CR
arXiv: https://arxiv.org/abs/2510.06036v1

Large reasoning models (LRMs) with multi-step reasoning capabilities have shown remarkable problem-solving abilities, yet they exhibit concerning safety vulnerabilities that remain poorly understood. In this work, we investigate why safety alignment fails in reasoning models through a mechanistic interpretability lens. Using a linear probing approach to trace refusal intentions across token positions, we discover a striking phenomenon termed as \textbf{refusal cliff}: many poorly-aligned reasoning models correctly identify harmful prompts and maintain strong refusal intentions during their thinking process, but experience a sharp drop in refusal scores at the final tokens before output generation. This suggests that these models are not inherently unsafe; rather, their refusal intentions are systematically suppressed. Through causal intervention analysis, we identify a sparse set of attention heads that negatively contribute to refusal behavior. Ablating just 3\% of these heads can reduce attack success rates below 10\%. Building on these mechanistic insights, we propose \textbf{Cliff-as-a-Judge}, a novel data selection method that identifies training examples exhibiting the largest refusal cliff to efficiently repair reasoning models' safety alignment. This approach achieves comparable safety improvements using only 1.7\% of the vanilla safety training data, demonstrating a less-is-more effect in safety alignment.

---

### Universal Neural Architecture Space: Covering ConvNets, Transformers and Everything in Between

Authors: Ondřej Týbl, Lukáš Neumann
Published: 2025-10-07
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.06035v1

We introduce Universal Neural Architecture Space (UniNAS), a generic search space for neural architecture search (NAS) which unifies convolutional networks, transformers, and their hybrid architectures under a single, flexible framework. Our approach enables discovery of novel architectures as well as analyzing existing architectures in a common framework. We also propose a new search algorithm that allows traversing the proposed search space, and demonstrate that the space contains interesting architectures, which, when using identical training setup, outperform state-of-the-art hand-crafted architectures. Finally, a unified toolkit including a standardized training and evaluation protocol is introduced to foster reproducibility and enable fair comparison in NAS research. Overall, this work opens a pathway towards systematically exploring the full spectrum of neural architectures with a unified graph-based NAS perspective.

---

### Adaptive Pruning for Increased Robustness and Reduced Computational Overhead in Gaussian Process Accelerated Saddle Point Searches

Authors: Rohit Goswami, Hannes Jónsson
Published: 2025-10-07
Categories: physics.chem-ph, cond-mat.mtrl-sci, cs.LG, physics.comp-ph
arXiv: https://arxiv.org/abs/2510.06030v3

Gaussian process (GP) regression provides a strategy for accelerating saddle point searches on high-dimensional energy surfaces by reducing the number of times the energy and its derivatives with respect to atomic coordinates need to be evaluated. The computational overhead in the hyperparameter optimization can, however, be large and make the approach inefficient. Failures can also occur if the search ventures too far into regions that are not represented well enough by the GP model. Here, these challenges are resolved by using geometry-aware optimal transport measures and an active pruning strategy using a summation over Wasserstein-1 distances for each atom-type in farthest-point sampling, selecting a fixed-size subset of geometrically diverse configurations to avoid rapidly increasing cost of GP updates as more observations are made. Stability is enhanced by permutation-invariant metric that provides a reliable trust radius for early-stopping and a logarithmic barrier penalty for the growth of the signal variance. These physically motivated algorithmic changes prove their efficacy by reducing to less than a half the mean computational time on a set of 238 challenging configurations from a previously published data set of chemical reactions. With these improvements, the GP approach is established as, a robust and scalable algorithm for accelerating saddle point searches when the evaluation of the energy and atomic forces requires significant computational effort.

---

### Fast Leave-One-Out Approximation from Fragment-Target Prevalence Vectors (molFTP) : From Dummy Masking to Key-LOO for Leakage-Free Feature Construction

Authors: Guillaume Godin
Published: 2025-10-07
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2510.06029v1

We introduce molFTP (molecular fragment-target prevalence), a compact representation that delivers strong predictive performance. To prevent feature leakage across cross-validation folds, we implement a dummy-masking procedure that removes information about fragments present in the held-out molecules. We further show that key leave-one-out (key-loo) closely approximates true molecule-level leave-one-out (LOO), with deviation below 8% on our datasets. This enables near full data training while preserving unbiased cross-validation estimates of model performance. Overall, molFTP provides a fast, leakage-resistant fragment-target prevalence vectorization with practical safeguards (dummy masking or key-LOO) that approximate LOO at a fraction of its cost.

---

### Generalization of Gibbs and Langevin Monte Carlo Algorithms in the Interpolation Regime

Authors: Andreas Maurer, Erfan Mirzaei, Massimiliano Pontil
Published: 2025-10-07
Categories: cs.LG, stat.ML
arXiv: https://arxiv.org/abs/2510.06028v3

This paper provides data-dependent bounds on the expected error of the Gibbs algorithm in the overparameterized interpolation regime, where low training errors are also obtained for impossible data, such as random labels in classification. The results show that generalization in the low-temperature regime is already signaled by small training errors in the noisier high-temperature regime. The bounds are stable under approximation with Langevin Monte Carlo algorithms. The analysis motivates the design of an algorithm to compute bounds, which on the MNIST, CIFAR-10, and SVHN datasets yield nontrivial, close predictions on the test error for true labeled data, while maintaining a correct upper bound on the test error for random labels.

---

### Emergent AI Surveillance: Overlearned Person Re-Identification and Its Mitigation in Law Enforcement Context

Authors: An Thi Nguyen, Radina Stoykova, Eric Arazo
Published: 2025-10-07
Categories: cs.CV, cs.AI, cs.CY, cs.LG
arXiv: https://arxiv.org/abs/2510.06026v1

Generic instance search models can dramatically reduce the manual effort required to analyze vast surveillance footage during criminal investigations by retrieving specific objects of interest to law enforcement. However, our research reveals an unintended emergent capability: through overlearning, these models can single out specific individuals even when trained on datasets without human subjects. This capability raises concerns regarding identification and profiling of individuals based on their personal data, while there is currently no clear standard on how de-identification can be achieved. We evaluate two technical safeguards to curtail a model's person re-identification capacity: index exclusion and confusion loss. Our experiments demonstrate that combining these approaches can reduce person re-identification accuracy to below 2% while maintaining 82% of retrieval performance for non-person objects. However, we identify critical vulnerabilities in these mitigations, including potential circumvention using partial person images. These findings highlight urgent regulatory questions at the intersection of AI governance and data protection: How should we classify and regulate systems with emergent identification capabilities? And what technical standards should be required to prevent identification capabilities from developing in seemingly benign applications?

---

### Out-of-Distribution Detection from Small Training Sets using Bayesian Neural Network Classifiers

Authors: Kevin Raina, Tanya Schmah
Published: 2025-10-07
Categories: cs.LG, stat.ML
arXiv: https://arxiv.org/abs/2510.06025v1

Out-of-Distribution (OOD) detection is critical to AI reliability and safety, yet in many practical settings, only a limited amount of training data is available. Bayesian Neural Networks (BNNs) are a promising class of model on which to base OOD detection, because they explicitly represent epistemic (i.e. model) uncertainty. In the small training data regime, BNNs are especially valuable because they can incorporate prior model information. We introduce a new family of Bayesian posthoc OOD scores based on expected logit vectors, and compare 5 Bayesian and 4 deterministic posthoc OOD scores. Experiments on MNIST and CIFAR-10 In-Distributions, with 5000 training samples or less, show that the Bayesian methods outperform corresponding deterministic methods.

---

### RamPINN: Recovering Raman Spectra From Coherent Anti-Stokes Spectra Using Embedded Physics

Authors: Sai Karthikeya Vemuri, Adithya Ashok Chalain Valapil, Tim Büchner, Joachim Denzler
Published: 2025-10-07
Categories: cs.LG
arXiv: https://arxiv.org/abs/2510.06020v2

Transferring the recent advancements in deep learning into scientific disciplines is hindered by the lack of the required large-scale datasets for training. We argue that in these knowledge-rich domains, the established body of scientific theory provides reliable inductive biases in the form of governing physical laws. We address the ill-posed inverse problem of recovering Raman spectra from noisy Coherent Anti-Stokes Raman Scattering (CARS) measurements, as the true Raman signal here is suppressed by a dominating non-resonant background. We propose RamPINN, a model that learns to recover Raman spectra from given CARS spectra. Our core methodological contribution is a physics-informed neural network that utilizes a dual-decoder architecture to disentangle resonant and non-resonant signals. This is done by enforcing the Kramers-Kronig causality relations via a differentiable Hilbert transform loss on the resonant and a smoothness prior on the non-resonant part of the signal. Trained entirely on synthetic data, RamPINN demonstrates strong zero-shot generalization to real-world experimental data, explicitly closing this gap and significantly outperforming existing baselines. Furthermore, we show that training with these physics-based losses alone, without access to any ground-truth Raman spectra, still yields competitive results. This work highlights a broader concept: formal scientific rules can act as a potent inductive bias, enabling robust, self-supervised learning in data-limited scientific domains.

---

### Evaluating The Impact of Stimulus Quality in Investigations of LLM Language Performance

Authors: Timothy Pistotti, Jason Brown, Michael Witbrock
Published: 2025-10-07
Categories: cs.CL
arXiv: https://arxiv.org/abs/2510.06018v1

Recent studies employing Large Language Models (LLMs) to test the Argument from the Poverty of the Stimulus (APS) have yielded contrasting results across syntactic phenomena. This paper investigates the hypothesis that characteristics of the stimuli used in recent studies, including lexical ambiguities and structural complexities, may confound model performance. A methodology is proposed for re-evaluating LLM competence on syntactic prediction, focusing on GPT-2. This involves: 1) establishing a baseline on previously used (both filtered and unfiltered) stimuli, and 2) generating a new, refined dataset using a state-of-the-art (SOTA) generative LLM (Gemini 2.5 Pro Preview) guided by linguistically-informed templates designed to mitigate identified confounds. Our preliminary findings indicate that GPT-2 demonstrates notably improved performance on these refined PG stimuli compared to baselines, suggesting that stimulus quality significantly influences outcomes in surprisal-based evaluations of LLM syntactic competency.

---

### ARISE: An Adaptive Resolution-Aware Metric for Test-Time Scaling Evaluation in Large Reasoning Models

Authors: Zhangyue Yin, Qiushi Sun, Zhiyuan Zeng, Zhiyuan Yu, Qipeng Guo, Xuanjing Huang, Xipeng Qiu
Published: 2025-10-07
Categories: cs.AI
arXiv: https://arxiv.org/abs/2510.06014v1

Test-time scaling has emerged as a transformative paradigm for enhancing the performance of large reasoning models, enabling dynamic allocation of computational resources during inference. However, as the landscape of reasoning models rapidly expands, a critical question remains: how can we systematically compare and evaluate the test-time scaling capabilities across different models? In this paper, we introduce ARISE (Adaptive Resolution-aware Scaling Evaluation), a novel metric specifically designed to assess the test-time scaling effectiveness of large reasoning models. Unlike existing evaluation approaches, ARISE incorporates two key innovations: (1) sample-level awareness that effectively penalizes negative scaling behaviors where increased computation leads to performance degradation, and (2) a dynamic sampling mechanism that mitigates the impact of accuracy fluctuations and token count instability on the final assessment. We conduct comprehensive experiments evaluating state-of-the-art reasoning models across diverse domains including mathematical reasoning, code generation, and agentic tasks. Our results demonstrate that ARISE provides a reliable and fine-grained measurement of test-time scaling capabilities, revealing significant variations in scaling efficiency across models. Notably, our evaluation identifies Claude Opus as exhibiting superior scaling characteristics compared to other contemporary reasoning models.

---

### Hybrid Quantum-Classical Policy Gradient for Adaptive Control of Cyber-Physical Systems: A Comparative Study of VQC vs. MLP

Authors: Aueaphum Aueawatthanaphisut, Nyi Wunna Tun
Published: 2025-10-07
Categories: quant-ph, cs.AI, cs.LG, cs.RO, eess.SY
arXiv: https://arxiv.org/abs/2510.06010v1

The comparative evaluation between classical and quantum reinforcement learning (QRL) paradigms was conducted to investigate their convergence behavior, robustness under observational noise, and computational efficiency in a benchmark control environment. The study employed a multilayer perceptron (MLP) agent as a classical baseline and a parameterized variational quantum circuit (VQC) as a quantum counterpart, both trained on the CartPole-v1 environment over 500 episodes. Empirical results demonstrated that the classical MLP achieved near-optimal policy convergence with a mean return of 498.7 +/- 3.2, maintaining stable equilibrium throughout training. In contrast, the VQC exhibited limited learning capability, with an average return of 14.6 +/- 4.8, primarily constrained by circuit depth and qubit connectivity. Noise robustness analysis further revealed that the MLP policy deteriorated gracefully under Gaussian perturbations, while the VQC displayed higher sensitivity at equivalent noise levels. Despite the lower asymptotic performance, the VQC exhibited significantly lower parameter count and marginally increased training time, highlighting its potential scalability for low-resource quantum processors. The results suggest that while classical neural policies remain dominant in current control benchmarks, quantum-enhanced architectures could offer promising efficiency advantages once hardware noise and expressivity limitations are mitigated.

---

### Continual Learning for Image Captioning through Improved Image-Text Alignment

Authors: Bertram Taetz, Gal Bordelius
Published: 2025-10-07
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.06009v2

Generating accurate and coherent image captions in a continual learning setting remains a major challenge due to catastrophic forgetting and the difficulty of aligning evolving visual concepts with language over time. In this work, we propose a novel multi-loss framework for continual image captioning that integrates semantic guidance through prompt-based continual learning and contrastive alignment. Built upon a pretrained ViT-GPT-2 backbone, our approach combines standard cross-entropy loss with three additional components: (1) a prompt-based cosine similarity loss that aligns image embeddings with synthetically constructed prompts encoding objects, attributes, and actions; (2) a CLIP-style loss that promotes alignment between image embeddings and target caption embedding; and (3) a language-guided contrastive loss that employs a triplet loss to enhance class-level discriminability between tasks. Notably, our approach introduces no additional overhead at inference time and requires no prompts during caption generation. We find that this approach mitigates catastrophic forgetting, while achieving better semantic caption alignment compared to state-of-the-art methods. The code can be found via the following link: https://github.com/Gepardius/Taetz_Bordelius_Continual_ImageCaptioning.

---

### Detection and Measurement of Hailstones with Multimodal Large Language Models

Authors: Moritz Alker, David C. Schedl, Andreas Stöckl
Published: 2025-10-07
Categories: cs.CV, cs.AI
arXiv: https://arxiv.org/abs/2510.06008v1

This study examines the use of social media and news images to detect and measure hailstones, utilizing pre-trained multimodal large language models. The dataset for this study comprises 474 crowdsourced images of hailstones from documented hail events in Austria, which occurred between January 2022 and September 2024. These hailstones have maximum diameters ranging from 2 to 11cm. We estimate the hail diameters and compare four different models utilizing one-stage and two-stage prompting strategies. The latter utilizes additional size cues from reference objects, such as human hands, within the image. Our results show that pretrained models already have the potential to measure hailstone diameters from images with an average mean absolute error of 1.12cm for the best model. In comparison to a single-stage prompt, two-stage prompting improves the reliability of most models. Our study suggests that these off-the-shelf models, even without fine-tuning, can complement traditional hail sensors by extracting meaningful and spatially dense information from social media imagery, enabling faster and more detailed assessments of severe weather events. The automated real-time image harvesting from social media and other sources remains an open task, but it will make our approach directly applicable to future hail events.

---

### Uncertainty in Machine Learning

Authors: Hans Weytjens, Wouter Verbeke
Published: 2025-10-07
Categories: cs.LG
arXiv: https://arxiv.org/abs/2510.06007v1

This book chapter introduces the principles and practical applications of uncertainty quantification in machine learning. It explains how to identify and distinguish between different types of uncertainty and presents methods for quantifying uncertainty in predictive models, including linear regression, random forests, and neural networks. The chapter also covers conformal prediction as a framework for generating predictions with predefined confidence intervals. Finally, it explores how uncertainty estimation can be leveraged to improve business decision-making, enhance model reliability, and support risk-aware strategies.

---