# Research Papers — 2025-10 · bundle page 61 of 217 > Full-text markdown of papers #3001–3050 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/61 [← Previous bundle](https://aipod.dev/papers/archive/2025-10/page/60.md) [Next bundle →](https://aipod.dev/papers/archive/2025-10/page/62.md) ### TRUST: A Decentralized Framework for Auditing Large Language Model Reasoning Authors: Morris Yu-Chao Huang, Zhen Tan, Mohan Zhang, Pingzhi Li, Zhuo Zhang, Tianlong Chen Published: 2025-10-23 Categories: cs.AI arXiv: https://arxiv.org/abs/2510.20188v1 Large Language Models generate complex reasoning chains that reveal their decision-making, yet verifying the faithfulness and harmlessness of these intermediate steps remains a critical unsolved problem. Existing auditing methods are centralized, opaque, and hard to scale, creating significant risks for deploying proprietary models in high-stakes domains. We identify four core challenges: (1) Robustness: Centralized auditors are single points of failure, prone to bias or attacks. (2) Scalability: Reasoning traces are too long for manual verification. (3) Opacity: Closed auditing undermines public trust. (4) Privacy: Exposing full reasoning risks model theft or distillation. We propose TRUST, a transparent, decentralized auditing framework that overcomes these limitations via: (1) A consensus mechanism among diverse auditors, guaranteeing correctness under up to $30\%$ malicious participants. (2) A hierarchical DAG decomposition of reasoning traces, enabling scalable, parallel auditing. (3) A blockchain ledger that records all verification decisions for public accountability. (4) Privacy-preserving segmentation, sharing only partial reasoning steps to protect proprietary logic. We provide theoretical guarantees for the security and economic incentives of the TRUST framework. Experiments across multiple LLMs (GPT-OSS, DeepSeek-r1, Qwen) and reasoning tasks (math, medical, science, humanities) show TRUST effectively detects reasoning flaws and remains robust against adversarial auditors. Our work pioneers decentralized AI auditing, offering a practical path toward safe and trustworthy LLM deployment. --- ### Every Question Has Its Own Value: Reinforcement Learning with Explicit Human Values Authors: Dian Yu, Yulai Zhao, Kishan Panaganti, Linfeng Song, Haitao Mi, Dong Yu Published: 2025-10-23 Categories: cs.LG, cs.CL arXiv: https://arxiv.org/abs/2510.20187v1 We propose Reinforcement Learning with Explicit Human Values (RLEV), a method that aligns Large Language Model (LLM) optimization directly with quantifiable human value signals. While Reinforcement Learning with Verifiable Rewards (RLVR) effectively trains models in objective domains using binary correctness rewards, it overlooks that not all tasks are equally significant. RLEV extends this framework by incorporating human-defined value signals directly into the reward function. Using exam-style data with explicit ground-truth value labels, RLEV consistently outperforms correctness-only baselines across multiple RL algorithms and model scales. Crucially, RLEV policies not only improve value-weighted accuracy but also learn a value-sensitive termination policy: concise for low-value prompts, thorough for high-value ones. We demonstrate this behavior stems from value-weighted gradient amplification on end-of-sequence tokens. Ablation studies confirm the gain is causally linked to value alignment. RLEV remains robust under noisy value signals, such as difficulty-based labels, demonstrating that optimizing for an explicit utility function offers a practical path to aligning LLMs with human priorities. --- ### PEDRA: Evaluating the Realism of Pedestrian Dynamics in Video Generation Authors: Aaron Appelle, Jerome P. Lynch Published: 2025-10-23 Categories: cs.CV arXiv: https://arxiv.org/abs/2510.20182v2 Pedestrian simulation traditionally relies on expert-tuned, hand-crafted models that limit scalability and generalization. Meanwhile, large-scale video generation models have achieved high visual realism across diverse settings, motivating exploration of their potential as general-purpose world simulators. Existing benchmarks primarily assess single-subject realism rather than scenes with multiple interacting people, leaving the plausibility of multi-agent dynamics in generated videos untested. We propose a rigorous evaluation protocol to benchmark text-to-video (T2V) and image-to-video (I2V) models as implicit simulators of pedestrian dynamics. For I2V, we leverage start frames from established datasets to enable direct comparison with ground truth videos, while for T2V we design a prompt suite covering varied crowd densities and interaction types. A key component is a method to reconstruct 2D bird's-eye view trajectories from pixel-space without known camera parameters. Our analysis shows that leading models exhibit effective priors for plausible multi-agent behavior, though issues such as merging and disappearing pedestrians reveal limits to their physical consistency. --- ### PPMStereo: Pick-and-Play Memory Construction for Consistent Dynamic Stereo Matching Authors: Yun Wang, Junjie Hu, Qiaole Dong, Yongjian Zhang, Yanwei Fu, Tin Lun Lam, Dapeng Wu Published: 2025-10-23 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2510.20178v1 Temporally consistent depth estimation from stereo video is critical for real-world applications such as augmented reality, where inconsistent depth estimation disrupts the immersion of users. Despite its importance, this task remains challenging due to the difficulty in modeling long-term temporal consistency in a computationally efficient manner. Previous methods attempt to address this by aggregating spatio-temporal information but face a fundamental trade-off: limited temporal modeling provides only modest gains, whereas capturing long-range dependencies significantly increases computational cost. To address this limitation, we introduce a memory buffer for modeling long-range spatio-temporal consistency while achieving efficient dynamic stereo matching. Inspired by the two-stage decision-making process in humans, we propose a \textbf{P}ick-and-\textbf{P}lay \textbf{M}emory (PPM) construction module for dynamic \textbf{Stereo} matching, dubbed as \textbf{PPMStereo}. PPM consists of a `pick' process that identifies the most relevant frames and a `play' process that weights the selected frames adaptively for spatio-temporal aggregation. This two-stage collaborative process maintains a compact yet highly informative memory buffer while achieving temporally consistent information aggregation. Extensive experiments validate the effectiveness of PPMStereo, demonstrating state-of-the-art performance in both accuracy and temporal consistency. % Notably, PPMStereo achieves 0.62/1.11 TEPE on the Sintel clean/final (17.3\% \& 9.02\% improvements over BiDAStereo) with fewer computational costs. Codes are available at \textcolor{blue}{https://github.com/cocowy1/PPMStereo}. --- ### Mixture-of-Minds: Multi-Agent Reinforcement Learning for Table Understanding Authors: Yuhang Zhou, Mingrui Zhang, Ke Li, Mingyi Wang, Qiao Liu, Qifei Wang, Jiayi Liu, Fei Liu, Serena Li, Weiwei Li, Mingze Gao, Abhishek Kumar et al. Published: 2025-10-23 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2510.20176v2 Understanding and reasoning over tables is a critical capability for many real-world applications. Large language models (LLMs) have shown promise on this task, but current approaches remain limited. Fine-tuning based methods strengthen language reasoning; yet they are prone to arithmetic errors and hallucination. In contrast, tool-based methods enable precise table manipulation but rely on rigid schemas and lack semantic understanding. These complementary drawbacks highlight the need for approaches that integrate robust reasoning with reliable table processing. In this work, we propose Mixture-of-Minds, a multi-agent framework that decomposes table reasoning into three specialized roles: planning, coding, and answering. This design enables each agent to focus on a specific aspect of the task while leveraging code execution for precise table manipulation. Building on this workflow, we introduce a self-improvement training framework that employs Monte Carlo Tree Search (MCTS) rollouts to generate pseudo-gold trajectories and optimize agents with reinforcement learning (RL). Extensive experiments show that Mixture-of-Minds delivers substantial gains, reaching 62.13% on TableBench and surpassing OpenAI-o4-mini-high. These results demonstrate the promise of combining structured multi-agent workflows with RL to advance table understanding. --- ### Collective Communication for 100k+ GPUs Authors: Min Si, Pavan Balaji, Yongzhou Chen, Ching-Hsiang Chu, Adi Gangidi, Saif Hasan, Subodh Iyengar, Dan Johnson, Bingzhe Liu, Regina Ren, Deep Shah, Ashmitha Jeevaraj Shetty et al. Published: 2025-10-23 Categories: cs.DC, cs.AI, cs.NI arXiv: https://arxiv.org/abs/2510.20171v4 The increasing scale of large language models (LLMs) necessitates highly efficient collective communication frameworks, particularly as training workloads extend to hundreds of thousands of GPUs. Traditional communication methods face significant throughput and latency limitations at this scale, hindering both the development and deployment of state-of-the-art models. This paper presents the NCCLX collective communication framework, developed at Meta, engineered to optimize performance across the full LLM lifecycle, from the synchronous demands of large-scale training to the low-latency requirements of inference. The framework is designed to support complex workloads on clusters exceeding 100,000 GPUs, ensuring reliable, high-throughput, and low-latency data exchange. Empirical evaluation on the Llama4 model demonstrates substantial improvements in communication efficiency. This research contributes a robust solution for enabling the next generation of LLMs to operate at unprecedented scales. --- ### Empowering Targeted Neighborhood Search via Hyper Tour for Large-Scale TSP Authors: Tongkai Lu, Shuai Ma, Chongyang Tao Published: 2025-10-23 Categories: cs.LG arXiv: https://arxiv.org/abs/2510.20169v3 Traveling Salesman Problem (TSP) is a classic NP-hard problem that has garnered significant attention from both academia and industry. While neural-based methods have shown promise for solving TSPs, they still face challenges in scaling to larger instances, particularly in memory constraints associated with global heatmaps, edge weights, or access matrices, as well as in generating high-quality initial solutions and insufficient global guidance for efficiently navigating vast search spaces. To address these challenges, we propose a Hyper Tour Guided Neighborhood Search (HyperNS) method for large-scale TSP instances. Inspired by the ``clustering first, route second" strategy, our approach initially divides the TSP instance into clusters using a sparse heatmap graph and abstracts them as supernodes, followed by the generation of a hyper tour to guide both the initialization and optimization processes. This method reduces the search space by focusing on edges relevant to the hyper tour, leading to more efficient and effective optimization. Experimental results on both synthetic and real-world datasets demonstrate that our approach outperforms existing neural-based methods, particularly in handling larger-scale instances, offering a significant reduction in the gap to the optimal solution. --- ### DeepWideSearch: Benchmarking Depth and Width in Agentic Information Seeking Authors: Tian Lan, Bin Zhu, Qianghuai Jia, Junyang Ren, Haijun Li, Longyue Wang, Zhao Xu, Weihua Luo, Kaifu Zhang Published: 2025-10-23 Categories: cs.CL arXiv: https://arxiv.org/abs/2510.20168v1 Current search agents fundamentally lack the ability to simultaneously perform \textit{deep} reasoning over multi-hop retrieval and \textit{wide}-scale information collection-a critical deficiency for real-world applications like comprehensive market analysis and business development. To bridge this gap, we introduce DeepWideSearch, the first benchmark explicitly designed to evaluate agents to integrate depth and width in information seeking. In DeepWideSearch, agents must process a large volume of data, each requiring deep reasoning over multi-hop retrieval paths. Specifically, we propose two methods to converse established datasets, resulting in a curated collection of 220 questions spanning 15 diverse domains. Extensive experiments demonstrate that even state-of-the-art agents achieve only 2.39% average success rate on DeepWideSearch, highlighting the substantial challenge of integrating depth and width search in information-seeking tasks. Furthermore, our error analysis reveals four failure modes: lack of reflection, overreliance on internal knowledge, insufficient retrieval, and context overflow-exposing key limitations in current agent architectures. We publicly release DeepWideSearch to catalyze future research on more capable and robust information-seeking agents. --- ### IB-GAN: Disentangled Representation Learning with Information Bottleneck Generative Adversarial Networks Authors: Insu Jeon, Wonkwang Lee, Myeongjang Pyeon, Gunhee Kim Published: 2025-10-23 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2510.20165v1 We propose a new GAN-based unsupervised model for disentangled representation learning. The new model is discovered in an attempt to utilize the Information Bottleneck (IB) framework to the optimization of GAN, thereby named IB-GAN. The architecture of IB-GAN is partially similar to that of InfoGAN but has a critical difference; an intermediate layer of the generator is leveraged to constrain the mutual information between the input and the generated output. The intermediate stochastic layer can serve as a learnable latent distribution that is trained with the generator jointly in an end-to-end fashion. As a result, the generator of IB-GAN can harness the latent space in a disentangled and interpretable manner. With the experiments on dSprites and Color-dSprites dataset, we demonstrate that IB-GAN achieves competitive disentanglement scores to those of state-of-the-art \b{eta}-VAEs and outperforms InfoGAN. Moreover, the visual quality and the diversity of samples generated by IB-GAN are often better than those by \b{eta}-VAEs and Info-GAN in terms of FID score on CelebA and 3D Chairs dataset. --- ### TOMCAT: Test-time Comprehensive Knowledge Accumulation for Compositional Zero-Shot Learning Authors: Xudong Yan, Songhe Feng Published: 2025-10-23 Categories: cs.CV arXiv: https://arxiv.org/abs/2510.20162v2 Compositional Zero-Shot Learning (CZSL) aims to recognize novel attribute-object compositions based on the knowledge learned from seen ones. Existing methods suffer from performance degradation caused by the distribution shift of label space at test time, which stems from the inclusion of unseen compositions recombined from attributes and objects. To overcome the challenge, we propose a novel approach that accumulates comprehensive knowledge in both textual and visual modalities from unsupervised data to update multimodal prototypes at test time. Building on this, we further design an adaptive update weight to control the degree of prototype adjustment, enabling the model to flexibly adapt to distribution shift during testing. Moreover, a dynamic priority queue is introduced that stores high-confidence images to acquire visual knowledge from historical images for inference. Considering the semantic consistency of multimodal knowledge, we align textual and visual prototypes by multimodal collaborative representation learning. Extensive experiments indicate that our approach achieves state-of-the-art performance on four benchmark datasets under both closed-world and open-world settings. Code will be available at https://github.com/xud-yan/TOMCAT . --- ### Monocular Visual 8D Pose Estimation for Articulated Bicycles and Cyclists Authors: Eduardo R. Corral-Soto, Yang Liu, Yuan Ren, Bai Dongfeng, Liu Bingbing Published: 2025-10-23 Categories: cs.CV arXiv: https://arxiv.org/abs/2510.20158v1 In Autonomous Driving, cyclists belong to the safety-critical class of Vulnerable Road Users (VRU), and accurate estimation of their pose is critical for cyclist crossing intention classification, behavior prediction, and collision avoidance. Unlike rigid objects, articulated bicycles are composed of movable rigid parts linked by joints and constrained by a kinematic structure. 6D pose methods can estimate the 3D rotation and translation of rigid bicycles, but 6D becomes insufficient when the steering/pedals angles of the bicycle vary. That is because: 1) varying the articulated pose of the bicycle causes its 3D bounding box to vary as well, and 2) the 3D box orientation is not necessarily aligned to the orientation of the steering which determines the actual intended travel direction. In this work, we introduce a method for category-level 8D pose estimation for articulated bicycles and cyclists from a single RGB image. Besides being able to estimate the 3D translation and rotation of a bicycle from a single image, our method also estimates the rotations of its steering handles and pedals with respect to the bicycle body frame. These two new parameters enable the estimation of a more fine-grained bicycle pose state and travel direction. Our proposed model jointly estimates the 8D pose and the 3D Keypoints of articulated bicycles, and trains with a mix of synthetic and real image data to generalize on real images. We include an evaluation section where we evaluate the accuracy of our estimated 8D pose parameters, and our method shows promising results by achieving competitive scores when compared against state-of-the-art category-level 6D pose estimators that use rigid canonical object templates for matching. --- ### ADP-VRSGP: Decentralized Learning with Adaptive Differential Privacy via Variance-Reduced Stochastic Gradient Push Authors: Xiaoming Wu, Teng Liu, Xin Wang, Ming Yang, Jiguo Yu Published: 2025-10-23 Categories: cs.LG, cs.DC arXiv: https://arxiv.org/abs/2510.20157v1 Differential privacy is widely employed in decentralized learning to safeguard sensitive data by introducing noise into model updates. However, existing approaches that use fixed-variance noise often degrade model performance and reduce training efficiency. To address these limitations, we propose a novel approach called decentralized learning with adaptive differential privacy via variance-reduced stochastic gradient push (ADP-VRSGP). This method dynamically adjusts both the noise variance and the learning rate using a stepwise-decaying schedule, which accelerates training and enhances final model performance while providing node-level personalized privacy guarantees. To counteract the slowed convergence caused by large-variance noise in early iterations, we introduce a progressive gradient fusion strategy that leverages historical gradients. Furthermore, ADP-VRSGP incorporates decentralized push-sum and aggregation techniques, making it particularly suitable for time-varying communication topologies. Through rigorous theoretical analysis, we demonstrate that ADP-VRSGP achieves robust convergence with an appropriate learning rate, significantly improving training stability and speed. Experimental results validate that our method outperforms existing baselines across multiple scenarios, highlighting its efficacy in addressing the challenges of privacy-preserving decentralized learning. --- ### PartNeXt: A Next-Generation Dataset for Fine-Grained and Hierarchical 3D Part Understanding Authors: Penghao Wang, Yiyang He, Xin Lv, Yukai Zhou, Lan Xu, Jingyi Yu, Jiayuan Gu Published: 2025-10-23 Categories: cs.CV arXiv: https://arxiv.org/abs/2510.20155v3 Understanding objects at the level of their constituent parts is fundamental to advancing computer vision, graphics, and robotics. While datasets like PartNet have driven progress in 3D part understanding, their reliance on untextured geometries and expert-dependent annotation limits scalability and usability. We introduce PartNeXt, a next-generation dataset addressing these gaps with over 23,000 high-quality, textured 3D models annotated with fine-grained, hierarchical part labels across 50 categories. We benchmark PartNeXt on two tasks: (1) class-agnostic part segmentation, where state-of-the-art methods (e.g., PartField, SAMPart3D) struggle with fine-grained and leaf-level parts, and (2) 3D part-centric question answering, a new benchmark for 3D-LLMs that reveals significant gaps in open-vocabulary part grounding. Additionally, training Point-SAM on PartNeXt yields substantial gains over PartNet, underscoring the dataset's superior quality and diversity. By combining scalable annotation, texture-aware labels, and multi-task evaluation, PartNeXt opens new avenues for research in structured 3D understanding. --- ### Are Stereotypes Leading LLMs' Zero-Shot Stance Detection ? Authors: Anthony Dubreuil, Antoine Gourru, Christine Largeron, Amine Trabelsi Published: 2025-10-23 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2510.20154v1 Large Language Models inherit stereotypes from their pretraining data, leading to biased behavior toward certain social groups in many Natural Language Processing tasks, such as hateful speech detection or sentiment analysis. Surprisingly, the evaluation of this kind of bias in stance detection methods has been largely overlooked by the community. Stance Detection involves labeling a statement as being against, in favor, or neutral towards a specific target and is among the most sensitive NLP tasks, as it often relates to political leanings. In this paper, we focus on the bias of Large Language Models when performing stance detection in a zero-shot setting. We automatically annotate posts in pre-existing stance detection datasets with two attributes: dialect or vernacular of a specific group and text complexity/readability, to investigate whether these attributes influence the model's stance detection decisions. Our results show that LLMs exhibit significant stereotypes in stance detection tasks, such as incorrectly associating pro-marijuana views with low text complexity and African American dialect with opposition to Donald Trump. --- ### BoundRL: Efficient Structured Text Segmentation through Reinforced Boundary Generation Authors: Haoyuan Li, Zhengyuan Shen, Sullam Jeoung, Yueyan Chen, Jiayu Li, Qi Zhu, Shuai Wang, Vassilis Ioannidis, Huzefa Rangwala Published: 2025-10-23 Categories: cs.CL arXiv: https://arxiv.org/abs/2510.20151v2 Structured texts refer to texts containing structured elements beyond plain texts, such as code snippets and placeholders. Such structured texts increasingly require segmentation into semantically meaningful components, which cannot be effectively handled by conventional sentence-level segmentation methods. To address this, we propose BoundRL, a novel approach that jointly performs efficient token-level text segmentation and label prediction for long structured texts. Instead of generating full texts for each segment, it generates only starting tokens and reconstructs the complete texts by locating these tokens within the original texts, thereby reducing output tokens by 90% and minimizing hallucination. To train the models for the boundary generation, BoundRL~performs reinforcement learning with verifiable rewards (RLVR) that jointly optimizes document reconstruction fidelity and semantic alignment. It further mitigates entropy collapse by constructing intermediate candidates by perturbing segment boundaries and labels to create stepping stones toward higher-quality solutions. Experiments show that BoundRL enables small language models (1.7B parameters) to outperform few-shot prompting with much larger models as well as SFT and standard RLVR baselines on complex prompts used for LLM applications. --- ### Understanding Mechanistic Role of Structural and Functional Connectivity in Tau Propagation Through Multi-Layer Modeling Authors: Tingting Dan, Xinwei Huang, Jiaqi Ding, Yinggang Zheng, Guorong Wu Published: 2025-10-23 Categories: cs.LG, math.DS, physics.med-ph arXiv: https://arxiv.org/abs/2510.20148v1 Emerging neuroimaging evidence shows that pathological tau proteins build up along specific brain networks, suggesting that large-scale network architecture plays a key role in the progression of Alzheimer's disease (AD). However, how structural connectivity (SC) and functional connectivity (FC) interact to influence tau propagation remains unclear. Leveraging an unprecedented volume of longitudinal neuroimaging data, we examine SC-FC interactions through a multi-layer graph diffusion model. Beyond showing that connectome architecture constrains tau spread, our model reveals a regionally asymmetric contribution of SC and FC. Specifically, FC predominantly drives tau spread in subcortical areas, the insula, frontal and temporal cortices, whereas SC plays a larger role in occipital, parietal, and limbic regions. The relative dominance of SC versus FC shifts over the course of disease, with FC generally prevailing in early AD and SC becoming primary in later stages. Spatial patterns of SC- and FC-dominant regions strongly align with the regional expression of AD-associated genes involved in inflammation, apoptosis, and lysosomal function, including CHUK (IKK-alpha), TMEM106B, MCL1, NOTCH1, and TH. In parallel, other non-modifiable risk factors (e.g., APOE genotype, sex) and biological mechanisms (e.g., amyloid deposition) selectively reshape tau propagation by shifting dominant routes between anatomical and functional pathways in a region-specific manner. Findings are validated in an independent AD cohort. --- ### Compositional Generation for Long-Horizon Coupled PDEs Authors: Somayajulu L. N. Dhulipala, Deep Ray, Nicholas Forman Published: 2025-10-23 Categories: stat.ML, cs.LG arXiv: https://arxiv.org/abs/2510.20141v1 Simulating coupled PDE systems is computationally intensive, and prior efforts have largely focused on training surrogates on the joint (coupled) data, which requires a large amount of data. In the paper, we study compositional diffusion approaches where diffusion models are only trained on the decoupled PDE data and are composed at inference time to recover the coupled field. Specifically, we investigate whether the compositional strategy can be feasible under long time horizons involving a large number of time steps. In addition, we compare a baseline diffusion model with that trained using the v-parameterization strategy. We also introduce a symmetric compositional scheme for the coupled fields based on the Euler scheme. We evaluate on Reaction-Diffusion and modified Burgers with longer time grids, and benchmark against a Fourier Neural Operator trained on coupled data. Despite seeing only decoupled training data, the compositional diffusion models recover coupled trajectories with low error. v-parameterization can improve accuracy over a baseline diffusion model, while the neural operator surrogate remains strongest given that it is trained on the coupled data. These results show that compositional diffusion is a viable strategy towards efficient, long-horizon modeling of coupled PDEs. --- ### Revisiting Logit Distributions for Reliable Out-of-Distribution Detection Authors: Jiachen Liang, Ruibing Hou, Minyang Hu, Hong Chang, Shiguang Shan, Xilin Chen Published: 2025-10-23 Categories: cs.CV arXiv: https://arxiv.org/abs/2510.20134v1 Out-of-distribution (OOD) detection is critical for ensuring the reliability of deep learning models in open-world applications. While post-hoc methods are favored for their efficiency and ease of deployment, existing approaches often underexploit the rich information embedded in the model's logits space. In this paper, we propose LogitGap, a novel post-hoc OOD detection method that explicitly exploits the relationship between the maximum logit and the remaining logits to enhance the separability between in-distribution (ID) and OOD samples. To further improve its effectiveness, we refine LogitGap by focusing on a more compact and informative subset of the logit space. Specifically, we introduce a training-free strategy that automatically identifies the most informative logits for scoring. We provide both theoretical analysis and empirical evidence to validate the effectiveness of our approach. Extensive experiments on both vision-language and vision-only models demonstrate that LogitGap consistently achieves state-of-the-art performance across diverse OOD detection scenarios and benchmarks. Code is available at https://github.com/GIT-LJc/LogitGap. --- ### Inverse Image-Based Rendering for Light Field Generation from Single Images Authors: Hyunjun Jung, Hae-Gon Jeon Published: 2025-10-23 Categories: cs.CV arXiv: https://arxiv.org/abs/2510.20132v1 A concept of light-fields computed from multiple view images on regular grids has proven its benefit for scene representations, and supported realistic renderings of novel views and photographic effects such as refocusing and shallow depth of field. In spite of its effectiveness of light flow computations, obtaining light fields requires either computational costs or specialized devices like a bulky camera setup and a specialized microlens array. In an effort to broaden its benefit and applicability, in this paper, we propose a novel view synthesis method for light field generation from only single images, named inverse image-based rendering. Unlike previous attempts to implicitly rebuild 3D geometry or to explicitly represent objective scenes, our method reconstructs light flows in a space from image pixels, which behaves in the opposite way to image-based rendering. To accomplish this, we design a neural rendering pipeline to render a target ray in an arbitrary viewpoint. Our neural renderer first stores the light flow of source rays from the input image, then computes the relationships among them through cross-attention, and finally predicts the color of the target ray based on these relationships. After the rendering pipeline generates the first novel view from a single input image, the generated out-of-view contents are updated to the set of source rays. This procedure is iteratively performed while ensuring the consistent generation of occluded contents. We demonstrate that our inverse image-based rendering works well with various challenging datasets without any retraining or finetuning after once trained on synthetic dataset, and outperforms relevant state-of-the-art novel view synthesis methods. --- ### SAID: Safety-Aware Intent Defense via Prefix Probing for Large Language Models Authors: Yulong Chen, Qi Zhang, Jiawen Zhang, Yadong Liu, Mu Li, Jie Wen, Yong Xu Published: 2025-10-23 Categories: cs.CR, cs.AI arXiv: https://arxiv.org/abs/2510.20129v2 Large Language Models (LLMs) remain vulnerable to jailbreak attacks, where adversarially crafted prompts induce policy-violating responses despite safety alignment. Existing defenses typically improve safety through external filtering, auxiliary guardrails, or decoding-time control. However, these interventions often reduce practical deployability because they may require additional model access, introduce extra inference cost, or affect benign-task utility. In this paper, we propose Safety-Aware Intent Defense (SAID), a training-free jailbreak defense framework based on intent-level safety probing. SAID first distills potentially obfuscated user inputs into concise core intents using the target model itself. It then applies a validated safety prefix to probe each distilled intent and elicit the model's safety-aware response. Finally, a conservative aggregation rule rejects the original request if any distilled intent is identified as unsafe. This design enables black-box-compatible defense without updating model parameters or modifying the decoding process. Experiments on four open-source LLMs under six representative jailbreak attacks show that SAID achieves state-of-the-art defense performance in reducing harmful responses while maintaining competitive utility on benign tasks. Further analyses on prefix variants, hierarchical distillation, and inference efficiency demonstrate that SAID provides a practical safety-utility trade-off for securing LLMs against jailbreak threats. --- ### Physics-Guided Fusion for Robust 3D Tracking of Fast Moving Small Objects Authors: Prithvi Raj Singh, Raju Gottumukkala, Anthony S. Maida, Alan B. Barhorst, Vijaya Gopu Published: 2025-10-23 Categories: cs.CV arXiv: https://arxiv.org/abs/2510.20126v1 While computer vision has advanced considerably for general object detection and tracking, the specific problem of fast-moving tiny objects remains underexplored. This paper addresses the significant challenge of detecting and tracking rapidly moving small objects using an RGB-D camera. Our novel system combines deep learning-based detection with physics-based tracking to overcome the limitations of existing approaches. Our contributions include: (1) a comprehensive system design for object detection and tracking of fast-moving small objects in 3D space, (2) an innovative physics-based tracking algorithm that integrates kinematics motion equations to handle outliers and missed detections, and (3) an outlier detection and correction module that significantly improves tracking performance in challenging scenarios such as occlusions and rapid direction changes. We evaluated our proposed system on a custom racquetball dataset. Our evaluation shows our system surpassing kalman filter based trackers with up to 70\% less Average Displacement Error. Our system has significant applications for improving robot perception on autonomous platforms and demonstrates the effectiveness of combining physics-based models with deep learning approaches for real-time 3D detection and tracking of challenging small objects. --- ### There is No "apple" in Timeseries: Rethinking TSFM through the Lens of Invariance Authors: Arian Prabowo, Flora D. Salim Published: 2025-10-23 Categories: cs.LG arXiv: https://arxiv.org/abs/2510.20119v1 Timeseries foundation models (TSFMs) have multiplied, yet lightweight supervised baselines and even classical models often match them. We argue this gap stems from the naive importation of NLP or CV pipelines. In language and vision, large web-scale corpora densely capture human concepts i.e. there are countless images and text of apples. In contrast, timeseries data is built to complement the image and text modalities. There are no timeseries dataset that contains the concept apple. As a result, the scrape-everything-online paradigm fails for TS. We posit that progress demands a shift from opportunistic aggregation to principled design: constructing datasets that systematically span the space of invariance that preserve temporal semantics. To this end, we suggest that the ontology of timeseries invariances should be built based on first principles. Only by ensuring representational completeness through invariance coverage can TSFMs achieve the aligned structure necessary for generalisation, reasoning, and truly emergent behaviour. --- ### AsyncHZP: Hierarchical ZeRO Parallelism with Asynchronous Scheduling for Scalable LLM Training Authors: Huawei Bai, Yifan Huang, Wenqi Shi, Ansheng You, Feifan Shao, Tengfei Han, Minghui Yu Published: 2025-10-23 Categories: cs.DC, cs.LG arXiv: https://arxiv.org/abs/2510.20111v1 The training efficiency and scalability of language models on massive clusters currently remain a critical bottleneck. Mainstream approaches like ND parallelism are often cumbersome and complex, while flexible alternatives such as the Zero Redundancy Optimizer (ZeRO) are frequently hampered by communication overhead. In this paper, we propose Asynchronous Hierarchical Zero Parallelism (AsyncHZP), a novel asynchronous variant of ZeRO designed to achieve superior performance while maintaining simplicity and memory efficiency. Unlike traditional ZeRO, which employs over-fine-grained sharding that can lead to inefficient communication, AsyncHZP adaptively reshards parameters, gradients, and optimizer states across different replica groups. This strategy optimizes device memory utilization and significantly reduces communication overhead. In addition, we also design a multi-stream asynchronous scheduling method that executes parameter all-gather and gradient reduce-scatter operations in dedicated background threads, effectively overlapping communication with computation while incurring negligible memory fragmentation. Empirical evaluations on both Dense and Mixture-of-Experts (MoE) models confirm that AsyncHZP maintains robust stability at scale. It consistently outperforms classic ND parallelism, achieving state-of-the-art performance without complex strategic tuning, thereby simplifying the path to efficient large-scale training. --- ### The Verification-Value Paradox: A Normative Critique of Gen AI in Legal Practice Authors: Joshua Yuvaraj Published: 2025-10-23 Categories: cs.AI arXiv: https://arxiv.org/abs/2510.20109v1 It is often claimed that machine learning-based generative AI products will drastically streamline and reduce the cost of legal practice. This enthusiasm assumes lawyers can effectively manage AI's risks. Cases in Australia and elsewhere in which lawyers have been reprimanded for submitting inaccurate AI-generated content to courts suggest this paradigm must be revisited. This paper argues that a new paradigm is needed to evaluate AI use in practice, given (a) AI's disconnection from reality and its lack of transparency, and (b) lawyers' paramount duties like honesty, integrity, and not to mislead the court. It presents an alternative model of AI use in practice that more holistically reflects these features (the verification-value paradox). That paradox suggests increases in efficiency from AI use in legal practice will be met by a correspondingly greater imperative to manually verify any outputs of that use, rendering the net value of AI use often negligible to lawyers. The paper then sets out the paradox's implications for legal practice and legal education, including for AI use but also the values that the paradox suggests should undergird legal practice: fidelity to the truth and civic responsibility. --- ### Why Prototypes Collapse: Diagnosing and Preventing Partial Collapse in Prototypical Self-Supervised Learning Authors: Gabriel Y. Arteaga, Marius Aasan, Rwiddhi Chakraborty, Martine Hjelkrem-Tan, Thalles Silva, Michael Kampffmeyer, Adín Ramírez Rivera Published: 2025-10-23 Categories: cs.LG, cs.CV arXiv: https://arxiv.org/abs/2510.20108v2 Prototypical self-supervised learning methods consistently suffer from partial prototype collapse, where multiple prototypes converge to nearly identical representations. This undermines their central purpose -- providing diverse and informative targets to guide encoders toward rich representations -- and has led practitioners to over-parameterize prototype sets or add ad-hoc regularizers, which mitigate symptoms rather than address the root cause. We empirically trace the collapse to the joint optimization of encoders and prototypes, which encourages a type of shortcut learning: early in training prototypes drift toward redundant representations that minimize loss without necessarily enhancing representation diversity. To break the joint optimization, we introduce a fully decoupled training strategy that learns prototypes and encoders under separate objectives. Concretely, we model prototypes as a Gaussian mixture updated with an online EM-style procedure, independent of the encoder's loss. This simple yet principled decoupling eliminates prototype collapse without explicit regularization and yields consistently diverse prototypes and stronger downstream performance. --- ### On pattern classification with weighted dimensions Authors: Ayatullah Faruk Mollah Published: 2025-10-23 Categories: cs.LG arXiv: https://arxiv.org/abs/2510.20107v1 Studies on various facets of pattern classification is often imperative while working with multi-dimensional samples pertaining to diverse application scenarios. In this notion, weighted dimension-based distance measure has been one of the vital considerations in pattern analysis as it reflects the degree of similarity between samples. Though it is often presumed to be settled with the pervasive use of Euclidean distance, plethora of issues often surface. In this paper, we present (a) a detail analysis on the impact of distance measure norms and weights of dimensions along with visualization, (b) a novel weighting scheme for each dimension, (c) incorporation of this dimensional weighting schema into a KNN classifier, and (d) pattern classification on a variety of synthetic as well as realistic datasets with the developed model. It has performed well across diverse experiments in comparison to the traditional KNN under the same experimental setups. Specifically, for gene expression datasets, it yields significant and consistent gain in classification accuracy (around 10%) in all cross-validation experiments with different values of k. As such datasets contain limited number of samples of high dimensions, meaningful selection of nearest neighbours is desirable, and this requirement is reasonably met by regulating the shape and size of the region enclosing the k number of reference samples with the developed weighting schema and appropriate norm. It, therefore, stands as an important generalization of KNN classifier powered by weighted Minkowski distance with the present weighting schema. --- ### Competition is the key: A Game Theoretic Causal Discovery Approach Authors: Amartya Roy, Souvik Chakraborty Published: 2025-10-23 Categories: cs.LG arXiv: https://arxiv.org/abs/2510.20106v1 Causal discovery remains a central challenge in machine learning, yet existing methods face a fundamental gap: algorithms like GES and GraN-DAG achieve strong empirical performance but lack finite-sample guarantees, while theoretically principled approaches fail to scale. We close this gap by introducing a game-theoretic reinforcement learning framework for causal discovery, where a DDQN agent directly competes against a strong baseline (GES or GraN-DAG), always warm-starting from the opponent's solution. This design yields three provable guarantees: the learned graph is never worse than the opponent, warm-starting strictly accelerates convergence, and most importantly, with high probability the algorithm selects the true best candidate graph. To the best of our knowledge, our result makes a first-of-its-kind progress in explaining such finite-sample guarantees in causal discovery: on synthetic SEMs (30 nodes), the observed error probability decays with n, tightly matching theory. On real-world benchmarks including Sachs, Asia, Alarm, Child, Hepar2, Dream, and Andes, our method consistently improves upon GES and GraN-DAG while remaining theoretically safe. Remarkably, it scales to large graphs such as Hepar2 (70 nodes), Dream (100 nodes), and Andes (220 nodes). Together, these results establish a new class of RL-based causal discovery algorithms that are simultaneously provably consistent, sample-efficient, and practically scalable, marking a decisive step toward unifying empirical performance with rigorous finite-sample theory. --- ### Extending machine learning model for implicit solvation to free energy calculations Authors: Rishabh Dey, Michael Brocidiacono, Kushal Koirala, Alexander Tropsha, Konstantin I. Popov Published: 2025-10-23 Categories: physics.chem-ph, cs.LG arXiv: https://arxiv.org/abs/2510.20103v2 The implicit solvent approach offers a computationally efficient framework to model solvation effects in molecular simulations. However, its accuracy often falls short compared to explicit solvent models, limiting its use in precise thermodynamic calculations. Recent advancements in machine learning (ML) present an opportunity to overcome these limitations by leveraging neural networks to develop more precise implicit solvent potentials for diverse applications. A major drawback of current ML-based methods is their reliance on force-matching alone, which can lead to energy predictions that differ by an arbitrary constant and are therefore unsuitable for absolute free energy comparisons. Here, we introduce a novel methodology with a graph neural network (GNN)-based implicit solvent model, dubbed Lambda Solvation Neural Network (LSNN). In addition to force-matching, this network was trained to match the derivatives of alchemical variables, ensuring that solvation free energies can be meaningfully compared across chemical species. Trained on a dataset of approximately 300,000 small molecules, LSNN achieves free energy predictions with accuracy comparable to explicit-solvent alchemical simulations, while offering a computational speedup and establishing a foundational framework for future applications in drug discovery. --- ### Human-Centered LLM-Agent System for Detecting Anomalous Digital Asset Transactions Authors: Gyuyeon Na, Minjung Park, Hyeonjeong Cha, Sangmi Chai Published: 2025-10-23 Categories: cs.AI arXiv: https://arxiv.org/abs/2510.20102v3 We present HCLA, a human-centered multi-agent system for anomaly detection in digital-asset transactions. The system integrates three cognitively aligned roles: Rule Abstraction, Evidence Scoring, and Expert-Style Justification. These roles operate in a conversational workflow that enables non-experts to express analytical intent in natural language, inspect structured risk evidence, and obtain traceable, context-aware reasoning. Implemented with an open-source, web-based interface, HCLA translates user intent into explicit analytical rules, applies classical anomaly detectors to quantify evidential risk, and reconstructs expert-style justifications grounded in observable transactional signals. Experiments on a cryptocurrency anomaly dataset show that, while the underlying detector achieves strong predictive accuracy, HCLA substantially improves interpretability, interaction, and decision transparency. Importantly, HCLA is not designed to explain a black-box model in the conventional XAI sense. Instead, we reconstruct a traceable expert reasoning process that aligns algorithmic evidence with regulatory and investigative judgment. By explicitly separating evidence scoring from expert-style justification, the framework emphasizes accountability beyond explainability and addresses practical requirements for regulatory, audit, and compliance-driven financial forensics. We describe the system architecture, closed-loop interaction design, datasets, evaluation protocol, and limitations. We argue that a human-in-the-loop reasoning reconstruction paradigm is essential for achieving transparency, accountability, and trust in high-stakes financial environments. Keywords: Human-Centered AI; LLM-Agent System; Multi-Agent Architecture; Anomaly Detection; Digital Asset Transactions; Cryptocurrency Forensics; Blockchain Analytics; Human-in-the-Loop; Explainable AI (XAI); Interpretability --- ### AI PB: A Grounded Generative Agent for Personalized Investment Insights Authors: Daewoo Park, Suho Park, Inseok Hong, Hanwool Lee, Junkyu Park, Sangjun Lee, Jeongman An, Hyunbin Loh Published: 2025-10-23 Categories: cs.AI, cs.CE, cs.CL arXiv: https://arxiv.org/abs/2510.20099v1 We present AI PB, a production-scale generative agent deployed in real retail finance. Unlike reactive chatbots that answer queries passively, AI PB proactively generates grounded, compliant, and user-specific investment insights. It integrates (i) a component-based orchestration layer that deterministically routes between internal and external LLMs based on data sensitivity, (ii) a hybrid retrieval pipeline using OpenSearch and the finance-domain embedding model, and (iii) a multi-stage recommendation mechanism combining rule heuristics, sequential behavioral modeling, and contextual bandits. Operating fully on-premises under Korean financial regulations, the system employs Docker Swarm and vLLM across 24 X NVIDIA H100 GPUs. Through human QA and system metrics, we demonstrate that grounded generation with explicit routing and layered safety can deliver trustworthy AI insights in high-stakes finance. --- ### Leveraging the Power of Large Language Models in Entity Linking via Adaptive Routing and Targeted Reasoning Authors: Yajie Li, Albert Galimov, Mitra Datta Ganapaneni, Pujitha Thejaswi, De Meng, Priyanshu Kumar, Saloni Potdar Published: 2025-10-23 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2510.20098v2 Entity Linking (EL) has traditionally relied on large annotated datasets and extensive model fine-tuning. While recent few-shot methods leverage large language models (LLMs) through prompting to reduce training requirements, they often suffer from inefficiencies due to expensive LLM-based reasoning. ARTER (Adaptive Routing and Targeted Entity Reasoning) presents a structured pipeline that achieves high performance without deep fine-tuning by strategically combining candidate generation, context-based scoring, adaptive routing, and selective reasoning. ARTER computes a small set of complementary signals(both embedding and LLM-based) over the retrieved candidates to categorize contextual mentions into easy and hard cases. The cases are then handled by a low-computational entity linker (e.g. ReFinED) and more expensive targeted LLM-based reasoning respectively. On standard benchmarks, ARTER outperforms ReFinED by up to +4.47%, with an average gain of +2.53% on 5 out of 6 datasets, and performs comparably to pipelines using LLM-based reasoning for all mentions, while being as twice as efficient in terms of the number of LLM tokens. --- ### BioCAP: Exploiting Synthetic Captions Beyond Labels in Biological Foundation Models Authors: Ziheng Zhang, Xinyue Ma, Arpita Chowdhury, Elizabeth G. Campolongo, Matthew J. Thompson, Net Zhang, Samuel Stevens, Hilmar Lapp, Tanya Berger-Wolf, Yu Su, Wei-Lun Chao, Jianyang Gu Published: 2025-10-23 Categories: cs.CV, cs.CL, cs.LG arXiv: https://arxiv.org/abs/2510.20095v3 This work investigates descriptive captions as an additional source of supervision for biological multimodal foundation models. Images and captions can be viewed as complementary samples from the latent morphospace of a species, each capturing certain biological traits. Incorporating captions during training encourages alignment with this shared latent structure, emphasizing potentially diagnostic characters while suppressing spurious correlations. The main challenge, however, lies in obtaining faithful, instance-specific captions at scale. This requirement has limited the utilization of natural language supervision in organismal biology compared with many other scientific domains. We complement this gap by generating synthetic captions with multimodal large language models (MLLMs), guided by Wikipedia-derived visual information and taxon-tailored format examples. These domain-specific contexts help reduce hallucination and yield accurate, instance-based descriptive captions. Using these captions, we train BioCAP (i.e., BioCLIP with Captions), a biological foundation model that captures rich semantics and achieves strong performance in species classification and text-image retrieval. These results demonstrate the value of descriptive captions beyond labels in bridging biological images with multimodal foundation models. --- ### On the Structure of Stationary Solutions to McKean-Vlasov Equations with Applications to Noisy Transformers Authors: Krishnakumar Balasubramanian, Sayan Banerjee, Philippe Rigollet Published: 2025-10-23 Categories: math.PR, cs.AI, cs.LG, math.AP, stat.ML arXiv: https://arxiv.org/abs/2510.20094v2 We study stationary solutions of McKean-Vlasov equations on the circle. Our main contributions stem from observing an exact equivalence between solutions of the stationary McKean-Vlasov equation and an infinite-dimensional quadratic system of equations over Fourier coefficients, which allows explicit characterization of the stationary states in a sequence space rather than a function space. This framework provides a transparent description of local bifurcations, characterizing their periodicity, and resonance structures, while accommodating singular potentials. We derive analytic expressions that characterize the emergence, form and shape (supercritical, critical, subcritical or transcritical) of bifurcations involving possibly multiple Fourier modes and connect them with discontinuous phase transitions. We also characterize, under suitable assumptions, the detailed structure of the stationary bifurcating solutions that are accurate upto an arbitrary number of Fourier modes. At the global level, we establish regularity and concavity properties of the free energy landscape, proving existence, compactness, and coexistence of globally minimizing stationary measures, further identifying discontinuous phase transitions with points of non-differentiability of the minimum free energy map. As an application, we specialize the theory to the Noisy Mean-Field Transformer model, where we show how changing the inverse temperature parameter $β$ affects the geometry of the infinitely many bifurcations from the uniform measure. We also explain how increasing $β$ can lead to a rich class of approximate multi-mode stationary solutions which can be seen as `metastable states'. Further, a sharp transition from continuous to discontinuous (first-order) phase behavior is observed as $β$ increases. --- ### StableSketcher: Enhancing Diffusion Model for Pixel-based Sketch Generation via Visual Question Answering Feedback Authors: Jiho Park, Sieun Choi, Jaeyoon Seo, Jihie Kim Published: 2025-10-23 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2510.20093v2 Although recent advancements in diffusion models have significantly enriched the quality of generated images, challenges remain in synthesizing pixel-based human-drawn sketches, a representative example of abstract expression. To combat these challenges, we propose StableSketcher, a novel framework that empowers diffusion models to generate hand-drawn sketches with high prompt fidelity. Within this framework, we fine-tune the variational autoencoder to optimize latent decoding, enabling it to better capture the characteristics of sketches. In parallel, we integrate a new reward function for reinforcement learning based on visual question answering, which improves text-image alignment and semantic consistency. Extensive experiments demonstrate that StableSketcher generates sketches with improved stylistic fidelity, achieving better alignment with prompts compared to the Stable Diffusion baseline. Additionally, we introduce SketchDUO, to the best of our knowledge, the first dataset comprising instance-level sketches paired with captions and question-answer pairs, thereby addressing the limitations of existing datasets that rely on image-label pairs. Our code and dataset will be made publicly available upon acceptance. Project page: https://zihos.github.io/StableSketcher --- ### Attentive Convolution: Unifying the Expressivity of Self-Attention with Convolutional Efficiency Authors: Hao Yu, Haoyu Chen, Yan Jiang, Wei Peng, Zhaodong Sun, Samuel Kaski, Guoying Zhao Published: 2025-10-23 Categories: cs.CV arXiv: https://arxiv.org/abs/2510.20092v1 Self-attention (SA) has become the cornerstone of modern vision backbones for its powerful expressivity over traditional Convolutions (Conv). However, its quadratic complexity remains a critical bottleneck for practical applications. Given that Conv offers linear complexity and strong visual priors, continuing efforts have been made to promote the renaissance of Conv. However, a persistent performance chasm remains, highlighting that these modernizations have not yet captured the intrinsic expressivity that defines SA. In this paper, we re-examine the design of the CNNs, directed by a key question: what principles give SA its edge over Conv? As a result, we reveal two fundamental insights that challenge the long-standing design intuitions in prior research (e.g., Receptive field). The two findings are: (1) \textit{Adaptive routing}: SA dynamically regulates positional information flow according to semantic content, whereas Conv employs static kernels uniformly across all positions. (2) \textit{Lateral inhibition}: SA induces score competition among token weighting, effectively suppressing redundancy and sharpening representations, whereas Conv filters lack such inhibitory dynamics and exhibit considerable redundancy. Based on this, we propose \textit{Attentive Convolution} (ATConv), a principled reformulation of the convolutional operator that intrinsically injects these principles. Interestingly, with only $3\times3$ kernels, ATConv consistently outperforms various SA mechanisms in fundamental vision tasks. Building on ATConv, we introduce AttNet, a CNN family that can attain \textbf{84.4\%} ImageNet-1K Top-1 accuracy with only 27M parameters. In diffusion-based image generation, replacing all SA with the proposed $3\times 3$ ATConv in SiT-XL/2 reduces ImageNet FID by 0.15 in 400k steps with faster sampling. Code is available at: github.com/price112/Attentive-Convolution. --- ### CreativityPrism: A Cross-Domain Evaluation Framework for Large Language Model Creativity Authors: Zhaoyi Joey Hou, Bowei Alvin Zhang, Yining Lu, Bhiman Kumar Baghel, Anneliese Brei, Ximing Lu, Meng Jiang, Faeze Brahman, Snigdha Chaturvedi, Haw-Shiuan Chang, Daniel Khashabi, Xiang Lorraine Li Published: 2025-10-23 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2510.20091v3 Creativity is often seen as a hallmark of human intelligence. While large language models(LLMs) are increasingly perceived as generating creative text, there is still no cross-domain and scalable framework to evaluate their creativity across diverse scenarios. Existing methods of LLM creativity evaluation either heavily rely on humans, limiting speed and scalability, or are fragmented across different domains and different definitions of creativity. To address this gap, we propose CreativityPrism, an evaluation and analysis framework that consolidates eight tasks from three domains: divergent thinking, creative writing, and logical reasoning, into a taxonomy of creativity that emphasizes three dimensions: quality, novelty, and diversity of LLM generations. The framework is designed to be scalable with reliable automatic evaluation judges that have been validated against human annotations. We evaluate 17 state-of-the-art (SoTA) LLMs on CreativityPrism and find that while frontier-scale LLMs dominate creative writing and logical reasoning tasks by a .10 (or 15%) lead over locally-deployable open models, they offer no significant advantage in divergent thinking, a domain much less explored in existing post-training regimes. Our analysis also shows that high performance in one creative dimension or domain rarely generalizes to others; specifically, novelty metrics often show weak or negative correlations with other metrics. This fragmentation confirms that a cross-domain, multi-dimensional framework like CreativityPrism is essential for any meaningful assessment of LLM creativity. --- ### Endoshare: A Publicly Available, Surgeons-Friendly Solution to De-Identify and Manage Surgical Videos Authors: Lorenzo Arboit, Dennis N. Schneider, Britty Baby, Vinkle Srivastav, Pietro Mascagni, Nicolas Padoy Published: 2025-10-23 Categories: cs.CV arXiv: https://arxiv.org/abs/2510.20087v2 Video-based assessment and surgical data science can advance surgical training, research, and quality improvement, yet adoption remains limited by heterogeneous recording formats and privacy concerns linked to video sharing. This work develops, evaluates, and publicly releases Endoshare, a surgeon-friendly application that merges, standardizes, and de-identifies endoscopic videos. Development followed an iterative, user-centered software life cycle. In the analysis phase, an internal survey of four clinicians and four computer scientists, based on 10 usability heuristics, identified early requirements and guided a cross-platform, privacy-by-design architecture. Prototype testing reported high usability for clinicians (4.68 +/- 0.40 out of 5) and for computer scientists (4.03 +/- 0.51 out of 5), with the lowest score (4.00 +/- 0.93 out of 5) relating to label clarity, prompting interface refinement to streamline case selection, video merging, automated out-of-body removal, and filename pseudonymization. In the testing phase, ten surgeons completed an external survey combining the same heuristics with Technology Acceptance Model constructs, reporting high perceived usefulness (5.07 +/- 1.75 out of 7), ease of use (5.15 +/- 1.71 out of 7), heuristic usability (4.38 +/- 0.48 out of 5), and strong recommendation likelihood (9.20 +/- 0.79 out of 10). A performance assessment across different hardware and configurations showed that processing time increased proportionally with video duration and was consistently lower in fast mode. Endoshare is a publicly available solution to manage surgical videos, with potential to support training, research, and quality improvement. Compliance certification and broader interoperability validation are needed to establish it as a reliable tool for surgical video management. The software is available at https://camma-public.github.io/Endoshare --- ### Hierarchical Dual-Head Model for Suicide Risk Assessment via MentalRoBERTa Authors: Chang Yang, Ziyi Wang, Wangfeng Tan, Zhiting Tan, Changrui Ji, Zhiming Zhou Published: 2025-10-23 Categories: cs.LG, cs.CY arXiv: https://arxiv.org/abs/2510.20085v1 Social media platforms have become important sources for identifying suicide risk, but automated detection systems face multiple challenges including severe class imbalance, temporal complexity in posting patterns, and the dual nature of risk levels as both ordinal and categorical. This paper proposes a hierarchical dual-head neural network based on MentalRoBERTa for suicide risk classification into four levels: indicator, ideation, behavior, and attempt. The model employs two complementary prediction heads operating on a shared sequence representation: a CORAL (Consistent Rank Logits) head that preserves ordinal relationships between risk levels, and a standard classification head that enables flexible categorical distinctions. A 3-layer Transformer encoder with 8-head multi-head attention models temporal dependencies across post sequences, while explicit time interval embeddings capture posting behavior dynamics. The model is trained with a combined loss function (0.5 CORAL + 0.3 Cross-Entropy + 0.2 Focal Loss) that simultaneously addresses ordinal structure preservation, overconfidence reduction, and class imbalance. To improve computational efficiency, we freeze the first 6 layers (50%) of MentalRoBERTa and employ mixed-precision training. The model is evaluated using 5-fold stratified cross-validation with macro F1 score as the primary metric. --- ### ShapeX: Shapelet-Driven Post Hoc Explanations for Time Series Classification Models Authors: Bosong Huang, Ming Jin, Yuxuan Liang, Johan Barthelemy, Debo Cheng, Qingsong Wen, Chenghao Liu, Shirui Pan Published: 2025-10-23 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2510.20084v2 Explaining time series classification models is crucial, particularly in high-stakes applications such as healthcare and finance, where transparency and trust play a critical role. Although numerous time series classification methods have identified key subsequences, known as shapelets, as core features for achieving state-of-the-art performance and validating their pivotal role in classification outcomes, existing post-hoc time series explanation (PHTSE) methods primarily focus on timestep-level feature attribution. These explanation methods overlook the fundamental prior that classification outcomes are predominantly driven by key shapelets. To bridge this gap, we present ShapeX, an innovative framework that segments time series into meaningful shapelet-driven segments and employs Shapley values to assess their saliency. At the core of ShapeX lies the Shapelet Describe-and-Detect (SDD) framework, which effectively learns a diverse set of shapelets essential for classification. We further demonstrate that ShapeX produces explanations which reveal causal relationships instead of just correlations, owing to the atomicity properties of shapelets. Experimental results on both synthetic and real-world datasets demonstrate that ShapeX outperforms existing methods in identifying the most relevant subsequences, enhancing both the precision and causal fidelity of time series explanations. --- ### DyPBP: Dynamic Peer Beneficialness Prediction for Cryptocurrency P2P Networking Authors: Nazmus Sakib, Simeon Wuthier, Amanul Islam, Xiaobo Zhou, Jinoh Kim, Ikkyun Kim, Sang-Yoon Chang Published: 2025-10-22 Categories: cs.NI, cs.LG arXiv: https://arxiv.org/abs/2511.17523v1 Distributed peer-to-peer (P2P) networking delivers the new blocks and transactions and is critical for the cryptocurrency blockchain system operations. Having poor P2P connectivity reduces the financial rewards from the mining consensus protocol. Previous research defines beneficalness of each Bitcoin peer connection and estimates the beneficialness based on the observations of the blocks and transactions delivery, which are after they are delivered. However, due to the infrequent block arrivals and the sporadic and unstable peer connections, the peers do not stay connected long enough to have the beneficialness score to converge to its expected beneficialness. We design and build Dynamic Peer Beneficialness Prediction (DyPBP) which predicts a peer's beneficialness by using networking behavior observations beyond just the block and transaction arrivals. DyPBP advances the previous research by estimating the beneficialness of a peer connection before it delivers new blocks and transactions. To achieve such goal, DyPBP introduces a new feature for remembrance to address the dynamic connectivity issue, as Bitcoin's peers using distributed networking often disconnect and re-connect. We implement DyPBP on an active Bitcoin node connected to the Mainnet and use machine learning for the beneficialness prediction. Our experimental results validate and evaluate the effectiveness of DyPBP; for example, the error performance improves by 2 to 13 orders of magnitude depending on the machine-learning model selection. DyPBP's use of the remembrance feature also informs our model selection. DyPBP enables the P2P connection's beneficialness estimation from the connection start before a new block arrives. --- ### Affordable EEG, Actionable Insights: An Open Dataset and Evaluation Framework for Epilepsy Patient Stratification Authors: HM Shadman Tabib, Md. Hasnaen Adil, Ayesha Rahman, Ahmmad Nur Swapnil, Maoyejatun Hasana, Ahmed Hossain Chowdhury, A. B. M. Alim Al Islam Published: 2025-10-22 Categories: eess.SP, cs.LG arXiv: https://arxiv.org/abs/2511.01879v1 Access to clinical multi-channel EEG remains limited in many regions worldwide. We present NEUROSKY-EPI, the first open dataset of single-channel, consumer-grade EEG for epilepsy, collected in a South Asian clinical setting along with rich contextual metadata. To explore its utility, we introduce EmbedCluster, a patient-stratification pipeline that transfers representations from EEGNet models trained on clinical data and enriches them with contextual autoencoder embeddings, followed by unsupervised clustering of patients based on EEG patterns. Results show that low-cost, single-channel data can support meaningful stratification. Beyond algorithmic performance, we emphasize human-centered concerns such as deployability in resource-constrained environments, interpretability for non-specialists, and safeguards for privacy, inclusivity, and bias. By releasing the dataset and code, we aim to catalyze interdisciplinary research across health technology, human-computer interaction, and machine learning, advancing the goal of affordable and actionable EEG-based epilepsy care. --- ### A Machine Learning-Based Framework to Shorten the Questionnaire for Assessing Autism Intervention Authors: Audrey Dong, Claire Xu, Samuel R. Guo, Kevin Yang, Xue-Jun Kong Published: 2025-10-22 Categories: stat.AP, cs.LG arXiv: https://arxiv.org/abs/2510.26808v1 Caregivers of individuals with autism spectrum disorder (ASD) often find the 77-item Autism Treatment Evaluation Checklist (ATEC) burdensome, limiting its use for routine monitoring. This study introduces a generalizable machine learning framework that seeks to shorten assessments while maintaining evaluative accuracy. Using longitudinal ATEC data from 60 autistic children receiving therapy, we applied feature selection and cross-validation techniques to identify the most predictive items across two assessment goals: longitudinal therapy tracking and point-in-time severity estimation. For progress monitoring, the framework identified 16 items (21% of the original questionnaire) that retained strong correlation with total score change and full subdomain coverage. We also generated smaller subsets (1-7 items) for efficient approximations. For point-in-time severity assessment, our model achieved over 80% classification accuracy using just 13 items (17% of the original set). While demonstrated on ATEC, the methodology-based on subset optimization, model interpretability, and statistical rigor-is broadly applicable to other high-dimensional psychometric tools. The resulting framework could potentially enable more accessible, frequent, and scalable assessments and offer a data-driven approach for AI-supported interventions across neurodevelopmental and psychiatric contexts. --- ### Dingtalk DeepResearch: A Unified Multi Agent Framework for Adaptive Intelligence in Enterprise Environments Authors: Mengyuan Chen, Chengjun Dai, Xinyang Dong, Chengzhe Feng, Kewei Fu, Jianshe Li, Zhihan Peng, Yongqi Tong, Junshao Zhang, Hong Zhu Published: 2025-10-22 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2510.24760v1 We present Dingtalk DeepResearch, a unified multi agent intelligence framework for real world enterprise environments, delivering deep research, heterogeneous table reasoning, and multimodal report generation. --- ### Genotype-Phenotype Integration through Machine Learning and Personalized Gene Regulatory Networks for Cancer Metastasis Prediction Authors: Jiwei Fu, Chunyu Yang Published: 2025-10-22 Categories: q-bio.OT, cs.AI arXiv: https://arxiv.org/abs/2510.23620v2 Metastasis is the leading cause of cancer-related mortality, yet most predictive models rely on shallow architectures and neglect patient-specific regulatory mechanisms. Here, we integrate classical machine learning and deep learning to predict metastatic potential across multiple cancer types. Gene expression profiles from the Cancer Cell Line Encyclopedia were combined with a transcription factor-target prior from DoRothEA, focusing on nine metastasis-associated regulators. After selecting differential genes using the Kruskal-Wallis test, ElasticNet, Random Forest, and XGBoost models were trained for benchmarking. Personalized gene regulatory networks were then constructed using PANDA and LIONESS and analyzed through a graph attention neural network (GATv2) to learn topological and expression-based representations. While XGBoost achieved the highest AUROC (0.7051), the GNN captured non-linear regulatory dependencies at the patient level. These results demonstrate that combining traditional machine learning with graph-based deep learning enables a scalable and interpretable framework for metastasis risk prediction in precision oncology. --- ### SIGN: Schema-Induced Games for Naming Authors: Ryan Zhang, Herbert Woisetschläger Published: 2025-10-22 Categories: cs.AI, cs.CL, cs.LG, cs.MA arXiv: https://arxiv.org/abs/2510.21855v1 Real-world AI systems are tackling increasingly complex problems, often through interactions among large language model (LLM) agents. When these agents develop inconsistent conventions, coordination can break down. Applications such as collaborative coding and distributed planning therefore require reliable, consistent communication, and scalability is a central concern as systems grow. We introduce Schema-Induced Games for Naming (SIGN), a naming game that examines how lightweight structure can steer convention formation. We compare schema-induced communication to unconstrained natural language and find faster convergence with up to 5.8x higher agreement. These results suggest that minimal structure can act as a simple control knob for efficient multi-agent coordination, pointing toward broader applications beyond the naming game. --- ### Policy Optimization Prefers The Path of Least Resistance Authors: Debdeep Sanyal, Aakash Sen Sharma, Dhruv Kumar, Saurabh Deshpande, Murari Mandal Published: 2025-10-22 Categories: cs.CL arXiv: https://arxiv.org/abs/2510.21853v1 Policy optimization (PO) algorithms are used to refine Large Language Models for complex, multi-step reasoning. Current state-of-the-art pipelines enforce a strict think-then-answer format to elicit chain-of-thought (CoT); however, the behavior of PO when these rigid constraints are relaxed into an open-ended CoT structure remains an under-studied question. We investigate this gap with an extensive suite of controlled experiments and identify a consistent principle: \textit{policy optimization consistently follows the path of least resistance}. When afforded the flexibility to interleave reasoning and response, policy optimization consistently learns to discard explicit reasoning, causing the policy to degenerate to a direct \texttt{<answer>}-only format. This outcome holds true across various models and algorithms. We find that this collapse in format is persistent even when the complex \texttt{<think><answer>} format is assigned up to 4x larger reward weights. We formalize this principle through a series of controlled reward decomposition experiments, demonstrating a clear hierarchy: PO systematically optimizes for the simplest reward component first, a preference that holds even when faced with mutually exclusive choices or strong incentives for more complex behaviors. Finally, we show that successful convergence on the high-reward shortcut is not a low-effort drift but is driven by the optimization process that requires the KL-regularized policy to have sufficient freedom to make a significant shift from its initial prior. Our findings reveal that granting policies the freedom to diverge is a double-edged sword: while necessary for discovering high-reward shortcuts, it also creates a powerful incentive to game the simplest aspects of the reward function, posing a critical challenge for reward hacking under alignment. --- ### Interpretable Diagnostics and Adaptive Data Assimilation for Neural ODEs via Discrete Empirical Interpolation Authors: Hojin Kim, Romit Maulik Published: 2025-10-22 Categories: cs.LG, physics.flu-dyn arXiv: https://arxiv.org/abs/2510.21852v2 We present a framework that leverages the Discrete Empirical Interpolation Method (DEIM) for interpretable deep learning and dynamical system analysis. Although DEIM efficiently approximates nonlinear terms in projection-based reduced-order models (POD-ROM), its fixed interpolation points are repurposed for identifying dynamically representative spatial structures in learned models. We apply DEIM as an interpretability tool to examine the learned dynamics of a pre-trained Neural Ordinary Differential Equation (NODE) for two-dimensional vortex-merging and backward-facing step flows. DEIM trajectories reveal physically meaningful structures in NODE predictions and expose failure modes when extrapolating to unseen flow configurations. Building on this diagnostic capability, we further introduce a DEIM-guided data assimilation strategy that injects sparse, dynamically representative corrections into the NODE rollout. By allocating a limited nudging budget to DEIM-identified sampling locations, the framework significantly improves long-term stability and predictive accuracy in out-of-distribution scenarios for the two-dimensional vortex-merging flow. Additional experiments for a flow over a backward-facing step reveal regime-dependent gains, with alternative sampling strategies performing competitively as well. These results demonstrate that DEIM can serve as an interpretable diagnostic and control framework for understanding and enhancing neural differential equation models. --- ### Data-Driven Approach to Capitation Reform in Rwanda Authors: Babaniyi Olaniyi, Ina Kalisa, Ana Fernández del Río, Jean Marie Vianney Hakizayezu, Enric Jané, Eniola Olaleye, Juan Francisco Garamendi, Ivan Nazarov, Aditya Rastogi, Mateo Diaz-Quiroz, África Periáñez, Regis Hitimana Published: 2025-10-22 Categories: cs.CY, cs.LG, stat.AP arXiv: https://arxiv.org/abs/2510.21851v2 As part of Rwanda's transition toward universal health coverage, the national Community-Based Health Insurance (CBHI) scheme is moving from retrospective fee-for-service reimbursements to prospective capitation payments for public primary healthcare providers. This work outlines a data-driven approach to designing, calibrating, and monitoring the capitation model using individual-level claims data from the Intelligent Health Benefits System (IHBS). We introduce a transparent, interpretable formula for allocating payments to Health Centers and their affiliated Health Posts. The formula is based on catchment population, service utilization patterns, and patient inflows, with parameters estimated via regression models calibrated on national claims data. Repeated validation exercises show the payment scheme closely aligns with historical spending while promoting fairness and adaptability across diverse facilities. In addition to payment design, the same dataset enables actionable behavioral insights. We highlight the use case of monitoring antibiotic prescribing patterns, particularly in pediatric care, to flag potential overuse and guideline deviations. Together, these capabilities lay the groundwork for a learning health financing system: one that connects digital infrastructure, resource allocation, and service quality to support continuous improvement and evidence-informed policy reform. --- ### SCoPE VLM: Selective Context Processing for Efficient Document Navigation in Vision-Language Models Authors: Gyubeum Lim, Yemo Koo, Vijay Krishna Madisetti Published: 2025-10-22 Categories: cs.CV, cs.CL arXiv: https://arxiv.org/abs/2510.21850v2 Understanding long-context visual information remains a fundamental challenge for vision-language models, particularly in agentic tasks such as GUI control and web navigation. While web pages and GUI environments are inherently structured documents, current VLMs typically neglect decision-oriented document understanding in their training objectives. Existing approaches primarily extend visual embeddings to process long, high-resolution inputs, but these methods are memory-intensive and impractical for locally deployable solutions. To address these issues, we propose SCoPE VLM, a document navigation expert that leverages a novel Chain of Scroll mechanism to selectively and recursively navigate documents, focusing exclusively on relevant segments. We introduce a dedicated data generation pipeline to construct informative Chain of Scroll trajectories and Episodic Group Relative Policy Optimization, a tailored reinforcement learning method to bridge the gap between training and inference. Our method substantially reduces memory usage and effectively models human-like reading behaviors. To the best of our knowledge, SCoPE VLM is the first framework to explicitly model agentic reading patterns in multi-page document question answering, advancing the capabilities of multimodal agents. --- ### TowerVision: Understanding and Improving Multilinguality in Vision-Language Models Authors: André G. Viveiros, Patrick Fernandes, Saul Santos, Sonal Sannigrahi, Emmanouil Zaranis, Nuno M. Guerreiro, Amin Farajian, Pierre Colombo, Graham Neubig, André F. T. Martins Published: 2025-10-22 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2510.21849v3 Despite significant advances in vision-language models (VLMs), most existing work follows an English-centric design process, limiting their effectiveness in multilingual settings. In this work, we provide a comprehensive empirical study analyzing the impact of several multilingual design choices, such as training data composition, encoder selection, and text backbones. The result is TowerVision, a family of open multilingual VLMs for both image-text and video-text tasks, built upon the multilingual text-only model Tower+. TowerVision achieves competitive performance on multiple multimodal multilingual benchmarks and shows particular strength in culturally grounded tasks and multimodal translation. By incorporating visual and cultural context during fine-tuning, our models surpass existing approaches trained on substantially larger datasets, as demonstrated on ALM-Bench and Multi30K (image tasks) and ViMUL-Bench (video tasks). Alongside the models, we release VisionBlocks, a high-quality, curated vision-language dataset. Our findings highlight that multilingual vision-language training data substantially improves cross-lingual generalization -- both from high-resource to underrepresented languages and vice versa -- and that instruction-tuned LLMs are not always the optimal initialization point. To support further research, we publicly release all models, data, and training recipes. ---