# Research Papers — 2025-10 · bundle page 75 of 217

> Full-text markdown of papers #3701–3750 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/75
[← Previous bundle](https://aipod.dev/papers/archive/2025-10/page/74.md)
[Next bundle →](https://aipod.dev/papers/archive/2025-10/page/76.md)

### Towards Fair ASR For Second Language Speakers Using Fairness Prompted Finetuning

Authors: Monorama Swain, Bubai Maji, Jagabandhu Mishra, Markus Schedl, Anders Søgaard, Jesper Rindom Jensen
Published: 2025-10-21
Categories: cs.CL
arXiv: https://arxiv.org/abs/2510.18374v2

In this work, we address the challenge of building fair English ASR systems for second-language speakers. Our analysis of widely used ASR models, Whisper and Seamless-M4T, reveals large fluctuations in word error rate (WER) across 26 accent groups, indicating significant fairness gaps. To mitigate this, we propose fairness-prompted finetuning with lightweight adapters, incorporating Spectral Decoupling (SD), Group Distributionally Robust Optimization (Group-DRO), and Invariant Risk Minimization (IRM). Our proposed fusion of traditional empirical risk minimization (ERM) with cross-entropy and fairness-driven objectives (SD, Group DRO, and IRM) enhances fairness across accent groups while maintaining overall recognition accuracy. In terms of macro-averaged word error rate, our approach achieves a relative improvement of 58.7% and 58.5% over the large pretrained Whisper and SeamlessM4T, and 9.7% and 7.8% over them, finetuning with standard empirical risk minimization with cross-entropy loss.

---

### Training Diverse Graph Experts for Ensembles: A Systematic Empirical Study

Authors: Gangda Deng, Yuxin Yang, Ömer Faruk Akgül, Hanqing Zeng, Yinglong Xia, Rajgopal Kannan, Viktor Prasanna
Published: 2025-10-21
Categories: cs.LG
arXiv: https://arxiv.org/abs/2510.18370v1

Graph Neural Networks (GNNs) have become essential tools for learning on relational data, yet the performance of a single GNN is often limited by the heterogeneity present in real-world graphs. Recent advances in Mixture-of-Experts (MoE) frameworks demonstrate that assembling multiple, explicitly diverse GNNs with distinct generalization patterns can significantly improve performance. In this work, we present the first systematic empirical study of expert-level diversification techniques for GNN ensembles. Evaluating 20 diversification strategies -- including random re-initialization, hyperparameter tuning, architectural variation, directionality modeling, and training data partitioning -- across 14 node classification benchmarks, we construct and analyze over 200 ensemble variants. Our comprehensive evaluation examines each technique in terms of expert diversity, complementarity, and ensemble performance. We also uncovers mechanistic insights into training maximally diverse experts. These findings provide actionable guidance for expert training and the design of effective MoE frameworks on graph data. Our code is available at https://github.com/Hydrapse/bench-gnn-diversification.

---

### KoSimpleQA: A Korean Factuality Benchmark with an Analysis of Reasoning LLMs

Authors: Donghyeon Ko, Kyubyung Chae, Yeguk Jin, Byungwook Lee, Chansong Jo, Sookyo In, Jaehong Lee, Taesup Kim, Donghyun Kwak
Published: 2025-10-21
Categories: cs.CL
arXiv: https://arxiv.org/abs/2510.18368v2

We present $\textbf{Korean SimpleQA (KoSimpleQA)}$, a benchmark for evaluating factuality in large language models (LLMs) with a focus on Korean cultural knowledge. KoSimpleQA is designed to be challenging yet easy to grade, consisting of 938 short, fact-seeking questions with unambiguous answers. We conduct a comprehensive evaluation across a diverse set of open-source LLMs of varying sizes that support Korean, and find that even the strongest model generates correct answer only 31.6% of the time, underscoring the challenging nature of KoSimpleQA. Notably, performance rankings on KoSimpleQA differ substantially from those on the English SimpleQA, highlighting the unique value of our dataset. Furthermore, we observe that reasoning helps mitigate the cross-lingual knowledge gap in LLMs, which refers to disparities in their ability to manifest knowledge across languages. KoSimpleQA can be found at https://github.com/naver-ai/KoSimpleQA.

---

### Towards Unsupervised Open-Set Graph Domain Adaptation via Dual Reprogramming

Authors: Zhen Zhang, Bingsheng He
Published: 2025-10-21
Categories: cs.LG
arXiv: https://arxiv.org/abs/2510.18363v1

Unsupervised Graph Domain Adaptation has become a promising paradigm for transferring knowledge from a fully labeled source graph to an unlabeled target graph. Existing graph domain adaptation models primarily focus on the closed-set setting, where the source and target domains share the same label spaces. However, this assumption might not be practical in the real-world scenarios, as the target domain might include classes that are not present in the source domain. In this paper, we investigate the problem of unsupervised open-set graph domain adaptation, where the goal is to not only correctly classify target nodes into the known classes, but also recognize previously unseen node types into the unknown class. Towards this end, we propose a novel framework called GraphRTA, which conducts reprogramming on both the graph and model sides. Specifically, we reprogram the graph by modifying target graph structure and node features, which facilitates better separation of known and unknown classes. Meanwhile, we also perform model reprogramming by pruning domain-specific parameters to reduce bias towards the source graph while preserving parameters that capture transferable patterns across graphs. Additionally, we extend the classifier with an extra dimension for the unknown class, thus eliminating the need of manually specified threshold in open-set recognition. Comprehensive experiments on several public datasets demonstrate that our proposed model can achieve satisfied performance compared with recent state-of-the-art baselines. Our source codes and datasets are publicly available at https://github.com/cszhangzhen/GraphRTA.

---

### FeatureFool: Zero-Query Fooling of Video Models via Feature Map

Authors: Duoxun Tang, Xi Xiao, Guangwu Hu, Kangkang Sun, Xiao Yang, Dongyang Chen, Qing Li, Yongjie Yin, Jiyao Wang
Published: 2025-10-21
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.18362v2

The vulnerability of deep neural networks (DNNs) has been preliminarily verified. Existing black-box adversarial attacks usually require multi-round interaction with the model and consume numerous queries, which is impractical in the real-world and hard to scale to recently emerged Video-LLMs. Moreover, no attack in the video domain directly leverages feature maps to shift the clean-video feature space. We therefore propose FeatureFool, a stealthy, video-domain, zero-query black-box attack that utilizes information extracted from a DNN to alter the feature space of clean videos. Unlike query-based methods that rely on iterative interaction, FeatureFool performs a zero-query attack by directly exploiting DNN-extracted information. This efficient approach is unprecedented in the video domain. Experiments show that FeatureFool achieves an attack success rate above 70\% against traditional video classifiers without any queries. Benefiting from the transferability of the feature map, it can also craft harmful content and bypass Video-LLM recognition. Additionally, adversarial videos generated by FeatureFool exhibit high quality in terms of SSIM, PSNR, and Temporal-Inconsistency, making the attack barely perceptible. This paper may contain violent or explicit content.

---

### Learning to Flow from Generative Pretext Tasks for Neural Architecture Encoding

Authors: Sunwoo Kim, Hyunjin Hwang, Kijung Shin
Published: 2025-10-21
Categories: cs.LG
arXiv: https://arxiv.org/abs/2510.18360v1

The performance of a deep learning model on a specific task and dataset depends heavily on its neural architecture, motivating considerable efforts to rapidly and accurately identify architectures suited to the target task and dataset. To achieve this, researchers use machine learning models-typically neural architecture encoders-to predict the performance of a neural architecture. Many state-of-the-art encoders aim to capture information flow within a neural architecture, which reflects how information moves through the forward pass and backpropagation, via a specialized model structure. However, due to their complicated structures, these flow-based encoders are significantly slower to process neural architectures compared to simpler encoders, presenting a notable practical challenge. To address this, we propose FGP, a novel pre-training method for neural architecture encoding that trains an encoder to capture the information flow without requiring specialized model structures. FGP trains an encoder to reconstruct a flow surrogate, our proposed representation of the neural architecture's information flow. Our experiments show that FGP boosts encoder performance by up to 106% in Precision-1%, compared to the same encoder trained solely with supervised learning.

---

### Ensembling Pruned Attention Heads For Uncertainty-Aware Efficient Transformers

Authors: Firas Gabetni, Giuseppe Curci, Andrea Pilzer, Subhankar Roy, Elisa Ricci, Gianni Franchi
Published: 2025-10-21
Categories: cs.LG, cs.CV
arXiv: https://arxiv.org/abs/2510.18358v2

Uncertainty quantification (UQ) is essential for deploying deep neural networks in safety-critical settings. Although methods like Deep Ensembles achieve strong UQ performance, their high computational and memory costs hinder scalability to large models. We introduce Hydra Ensembles, an efficient transformer-based ensemble that prunes attention heads to create diverse members and merges them via a new multi-head attention with grouped fully-connected layers. This yields a compact model with inference speed close to a single network, matching or surpassing Deep Ensembles in UQ performance without retraining from scratch. We also provide an in-depth analysis of pruning, showing that naive approaches can harm calibration, whereas Hydra Ensembles preserves robust uncertainty. Experiments on image and text classification tasks, with various architectures, show consistent gains over Deep Ensembles. Remarkably, in zero-shot classification on ImageNet-1k, our approach surpasses state of the art methods, even without requiring additional training.

---

### Learning Human-Object Interaction as Groups

Authors: Jiajun Hong, Jianan Wei, Wenguan Wang
Published: 2025-10-21
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.18357v1

Human-Object Interaction Detection (HOI-DET) aims to localize human-object pairs and identify their interactive relationships. To aggregate contextual cues, existing methods typically propagate information across all detected entities via self-attention mechanisms, or establish message passing between humans and objects with bipartite graphs. However, they primarily focus on pairwise relationships, overlooking that interactions in real-world scenarios often emerge from collective behaviors (multiple humans and objects engaging in joint activities). In light of this, we revisit relation modeling from a group view and propose GroupHOI, a framework that propagates contextual information in terms of geometric proximity and semantic similarity. To exploit the geometric proximity, humans and objects are grouped into distinct clusters using a learnable proximity estimator based on spatial features derived from bounding boxes. In each group, a soft correspondence is computed via self-attention to aggregate and dispatch contextual cues. To incorporate the semantic similarity, we enhance the vanilla transformer-based interaction decoder with local contextual cues from HO-pair features. Extensive experiments on HICO-DET and V-COCO benchmarks demonstrate the superiority of GroupHOI over the state-of-the-art methods. It also exhibits leading performance on the more challenging Nonverbal Interaction Detection (NVI-DET) task, which involves varied forms of higher-order interactions within groups.

---

### KrishokBondhu: A Retrieval-Augmented Voice-Based Agricultural Advisory Call Center for Bengali Farmers

Authors: Mohd Ruhul Ameen, Akif Islam, Farjana Aktar, M. Saifuzzaman Rafat
Published: 2025-10-21
Categories: cs.CL, cs.HC, cs.IR
arXiv: https://arxiv.org/abs/2510.18355v2

In Bangladesh, many farmers still struggle to access timely, expert-level agricultural guidance. This paper presents KrishokBondhu, a voice-enabled, call-centre-integrated advisory platform built on a Retrieval-Augmented Generation (RAG) framework for Bengali-speaking farmers. The system combines agricultural handbooks, extension manuals, and NGO publications, processes them through an OCR-based pipeline, and indexes the curated content in a vector database for semantic retrieval. Through a phone-based interface, farmers can receive real-time, context-aware advice: speech-to-text converts the Bengali query, the RAG module retrieves relevant information, a large language model (Gemma 3-4B) generates a grounded response, and text-to-speech delivers the answer in spoken Bengali. In a pilot evaluation, KrishokBondhu produced high-quality responses for 72.7% of diverse agricultural queries. Compared to the KisanQRS benchmark, it achieved a composite score of 4.53 versus 3.13 on a 5-point scale, with a 44.7% improvement and especially large gains in contextual richness and completeness, while maintaining comparable relevance and technical specificity. Semantic-similarity analysis further showed a strong correlation between retrieved context and answer quality. KrishokBondhu demonstrates the feasibility of combining call-centre accessibility, multilingual voice interaction, and modern RAG techniques to deliver expert-level agricultural guidance to remote Bangladeshi farmers.

---

### Ranking-based Preference Optimization for Diffusion Models from Implicit User Feedback

Authors: Yi-Lun Wu, Bo-Kai Ruan, Chiang Tseng, Hong-Han Shuai
Published: 2025-10-21
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.18353v1

Direct preference optimization (DPO) methods have shown strong potential in aligning text-to-image diffusion models with human preferences by training on paired comparisons. These methods improve training stability by avoiding the REINFORCE algorithm but still struggle with challenges such as accurately estimating image probabilities due to the non-linear nature of the sigmoid function and the limited diversity of offline datasets. In this paper, we introduce Diffusion Denoising Ranking Optimization (Diffusion-DRO), a new preference learning framework grounded in inverse reinforcement learning. Diffusion-DRO removes the dependency on a reward model by casting preference learning as a ranking problem, thereby simplifying the training objective into a denoising formulation and overcoming the non-linear estimation issues found in prior methods. Moreover, Diffusion-DRO uniquely integrates offline expert demonstrations with online policy-generated negative samples, enabling it to effectively capture human preferences while addressing the limitations of offline data. Comprehensive experiments show that Diffusion-DRO delivers improved generation quality across a range of challenging and unseen prompts, outperforming state-of-the-art baselines in both both quantitative metrics and user studies. Our source code and pre-trained models are available at https://github.com/basiclab/DiffusionDRO.

---

### Computable universal online learning

Authors: Dariusz Kalociński, Tomasz Steifer
Published: 2025-10-21
Categories: cs.LG, cs.LO
arXiv: https://arxiv.org/abs/2510.18352v1

Understanding when learning is possible is a fundamental task in the theory of machine learning. However, many characterizations known from the literature deal with abstract learning as a mathematical object and ignore the crucial question: when can learning be implemented as a computer program? We address this question for universal online learning, a generalist theoretical model of online binary classification, recently characterized by Bousquet et al. (STOC'21). In this model, there is no hypothesis fixed in advance; instead, Adversary -- playing the role of Nature -- can change their mind as long as local consistency with the given class of hypotheses is maintained. We require Learner to achieve a finite number of mistakes while using a strategy that can be implemented as a computer program. We show that universal online learning does not imply computable universal online learning, even if the class of hypotheses is relatively easy from a computability-theoretic perspective. We then study the agnostic variant of computable universal online learning and provide an exact characterization of classes that are learnable in this sense. We also consider a variant of proper universal online learning and show exactly when it is possible. Together, our results give a more realistic perspective on the existing theory of online binary classification and the related problem of inductive inference.

---

### PGTT: Phase-Guided Terrain Traversal for Perceptive Legged Locomotion

Authors: Alexandros Ntagkas, Chairi Kiourt, Konstantinos Chatzilygeroudis
Published: 2025-10-21
Categories: cs.RO, cs.AI, cs.LG
arXiv: https://arxiv.org/abs/2510.18348v2

State-of-the-art perceptive Reinforcement Learning controllers for legged robots typically either (i) impose oscillator-or IK-based gait priors that constrain the action space, bias policy optimization, and limit adaptability across robot morphologies, or (ii) operate "blind," making them unable to anticipate hind-leg terrain and brittle to observation noise. We propose Phase-Guided Terrain Traversal (PGTT), a perception-aware deep-RL approach that enforces gait structure through reward shaping, thereby reducing inductive bias compared to oscillator- or IK-conditioned action priors. PGTT encodes per-leg phase as a cubic Hermite spline, adapts swing height to local heightmap statistics, and adds a swing-phase contact penalty, while the policy acts directly in joint space for morphology-agnostic deployment. Trained in MuJoCo (MJX) on procedurally generated stair-like terrains with curriculum learning and domain randomization, PGTT achieves the highest success rate among the evaluated baselines under push disturbances (median +7.5% over the next-best baseline) and on discrete obstacles (+9%), while maintaining comparable velocity tracking. We validate PGTT on a Unitree Go2 using a real-time LiDAR elevation-to-heightmap pipeline and report preliminary results on ANYmal-C using the same hyperparameters. These results provide early evidence that terrain-adaptive, phase-guided reward shaping can transfer across platforms without platform-specific policy priors or extensive re-tuning.

---

### AV-Master: Dual-Path Comprehensive Perception Makes Better Audio-Visual Question Answering

Authors: Jiayu Zhang, Shuo Ye, Qilang Ye, Xun Lin, Zihan Song, Zitong Yu
Published: 2025-10-21
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.18346v3

Audio-Visual Question Answering (AVQA) requires models to effectively utilize both visual and auditory modalities to answer complex and diverse questions about audio-visual scenes. However, existing methods lack sufficient flexibility and dynamic adaptability in temporal sampling and modality preference awareness, making it difficult to focus on key information based on the question. This limits their reasoning capability in complex scenarios. To address these challenges, we propose a novel framework named AV-Master. It enhances the model's ability to extract key information from complex audio-visual scenes with substantial redundant content by dynamically modeling both temporal and modality dimensions. In the temporal dimension, we introduce a dynamic adaptive focus sampling mechanism that progressively focuses on audio-visual segments most relevant to the question, effectively mitigating redundancy and segment fragmentation in traditional sampling methods. In the modality dimension, we propose a preference-aware strategy that models each modality's contribution independently, enabling selective activation of critical features. Furthermore, we introduce a dual-path contrastive loss to reinforce consistency and complementarity across temporal and modality dimensions, guiding the model to learn question-specific cross-modal collaborative representations. Experiments on four large-scale benchmarks show that AV-Master significantly outperforms existing methods, especially in complex reasoning tasks.

---

### GPTFace: Generative Pre-training of Facial-Linguistic Transformer by Span Masking and Weakly Correlated Text-image Data

Authors: Yudong Li, Hao Li, Xianxu Hou, Linlin Shen
Published: 2025-10-21
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.18345v1

Compared to the prosperity of pre-training models in natural image understanding, the research on large-scale pre-training models for facial knowledge learning is still limited. Current approaches mainly rely on manually assembled and annotated face datasets for training, but labeling such datasets is labor-intensive and the trained models have limited scalability beyond the training data. To address these limitations, we present a generative pre-training model for facial knowledge learning that leverages large-scale web-built data for training. We use texts and images containing human faces crawled from the internet and conduct pre-training on self-supervised tasks, including masked image/language modeling (MILM) and image-text matching (ITM). During the generation stage, we further utilize the image-text matching loss to pull the generation distribution towards the control signal for controllable image/text generation. Experimental results demonstrate that our model achieves comparable performance to state-of-the-art pre-training models for various facial downstream tasks, such as attribution classification and expression recognition. Furthermore, our approach is also applicable to a wide range of face editing tasks, including face attribute editing, expression manipulation, mask removal, and photo inpainting.

---

### Combining Distantly Supervised Models with In Context Learning for Monolingual and Cross-Lingual Relation Extraction

Authors: Vipul Rathore, Malik Hammad Faisal, Parag Singla, Mausam
Published: 2025-10-21
Categories: cs.CL
arXiv: https://arxiv.org/abs/2510.18344v1

Distantly Supervised Relation Extraction (DSRE) remains a long-standing challenge in NLP, where models must learn from noisy bag-level annotations while making sentence-level predictions. While existing state-of-the-art (SoTA) DSRE models rely on task-specific training, their integration with in-context learning (ICL) using large language models (LLMs) remains underexplored. A key challenge is that the LLM may not learn relation semantics correctly, due to noisy annotation. In response, we propose HYDRE -- HYbrid Distantly Supervised Relation Extraction framework. It first uses a trained DSRE model to identify the top-k candidate relations for a given test sentence, then uses a novel dynamic exemplar retrieval strategy that extracts reliable, sentence-level exemplars from training data, which are then provided in LLM prompt for outputting the final relation(s). We further extend HYDRE to cross-lingual settings for RE in low-resource languages. Using available English DSRE training data, we evaluate all methods on English as well as a newly curated benchmark covering four diverse low-resource Indic languages -- Oriya, Santali, Manipuri, and Tulu. HYDRE achieves up to 20 F1 point gains in English and, on average, 17 F1 points on Indic languages over prior SoTA DSRE models. Detailed ablations exhibit HYDRE's efficacy compared to other prompting strategies.

---

### ShortcutBreaker: Low-Rank Noisy Bottleneck and Frequency Filtering Block for Multi-Class Unsupervised Anomaly Detection

Authors: Peng Tang, Xiaobin Hu, Tingcheng Li, Yang Nan, Tobias Lasser, Hongwei Bran Li
Published: 2025-10-21
Categories: cs.AI
arXiv: https://arxiv.org/abs/2510.18342v2

Multi-class unsupervised anomaly detection (MUAD) has garnered growing research interest, as it seeks to develop a unified model for anomaly detection across multiple classes, i.e., eliminating the need to train separate models for distinct objects and thereby saving substantial computational resources. Under the MUAD setting, while advanced Transformer-based architectures have brought significant performance improvements, identity shortcuts persist: they directly copy inputs to outputs, narrowing the gap in reconstruction errors between normal and abnormal cases, and thereby making the two harder to distinguish. Therefore, we propose ShortcutBreaker, a novel unified feature-reconstruction framework for MUAD tasks, featuring two key innovations to address the issue of shortcuts. First, drawing on matrix rank inequality, we design a low-rank noisy bottleneck (LRNB) to project highdimensional features into a low-rank latent space, and theoretically demonstrate its capacity to prevent trivial identity reproduction. Second, leveraging ViTs global modeling capability instead of merely focusing on local features, we incorporate a global perturbation attention to prevent information shortcuts in the decoders. Extensive experiments are performed on four widely used anomaly detection benchmarks, including three industrial datasets (MVTec-AD, ViSA, and Real-IAD) and one medical dataset (Universal Medical). The proposed method achieves a remarkable image-level AUROC of 99.8%, 98.9%, 90.6%, and 87.8% on these four datasets, respectively, consistently outperforming previous MUAD methods across different scenarios Our code will be released..

---

### ViSE: A Systematic Approach to Vision-Only Street-View Extrapolation

Authors: Kaiyuan Tan, Yingying Shen, Haiyang Sun, Bing Wang, Guang Chen, Hangjun Ye
Published: 2025-10-21
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.18341v1

Realistic view extrapolation is critical for closed-loop simulation in autonomous driving, yet it remains a significant challenge for current Novel View Synthesis (NVS) methods, which often produce distorted and inconsistent images beyond the original trajectory. This report presents our winning solution which ctook first place in the RealADSim Workshop NVS track at ICCV 2025. To address the core challenges of street view extrapolation, we introduce a comprehensive four-stage pipeline. First, we employ a data-driven initialization strategy to generate a robust pseudo-LiDAR point cloud, avoiding local minima. Second, we inject strong geometric priors by modeling the road surface with a novel dimension-reduced SDF termed 2D-SDF. Third, we leverage a generative prior to create pseudo ground truth for extrapolated viewpoints, providing auxilary supervision. Finally, a data-driven adaptation network removes time-specific artifacts. On the RealADSim-NVS benchmark, our method achieves a final score of 0.441, ranking first among all participants.

---

### Why Policy Gradient Algorithms Work for Undiscounted Total-Reward MDPs

Authors: Jongmin Lee, Ernest K. Ryu
Published: 2025-10-21
Categories: cs.LG
arXiv: https://arxiv.org/abs/2510.18340v2

The classical policy gradient method is the theoretical and conceptual foundation of modern policy-based reinforcement learning (RL) algorithms. Most rigorous analyses of such methods, particularly those establishing convergence guarantees, assume a discount factor $γ< 1$. In contrast, however, a recent line of work on policy-based RL for large language models uses the undiscounted total-reward setting with $γ= 1$, rendering much of the existing theory inapplicable. In this paper, we provide analyses of the policy gradient method for undiscounted expected total-reward infinite-horizon MDPs based on two key insights: (i) the classification of the MDP states into recurrent and transient states is invariant over the set of policies that assign strictly positive probability to every action (as is typical in deep RL models employing a softmax output layer) and (ii) the classical state visitation measure (which may be ill-defined when $γ= 1$) can be replaced with a new object that we call the transient visitation measure.

---

### ECG-LLM -- training and evaluation of domain-specific large language models for electrocardiography

Authors: Lara Ahrens, Wilhelm Haverkamp, Nils Strodthoff
Published: 2025-10-21
Categories: cs.CL, cs.LG
arXiv: https://arxiv.org/abs/2510.18339v1

Domain-adapted open-weight large language models (LLMs) offer promising healthcare applications, from queryable knowledge bases to multimodal assistants, with the crucial advantage of local deployment for privacy preservation. However, optimal adaptation strategies, evaluation methodologies, and performance relative to general-purpose LLMs remain poorly characterized. We investigated these questions in electrocardiography, an important area of cardiovascular medicine, by finetuning open-weight models on domain-specific literature and implementing a multi-layered evaluation framework comparing finetuned models, retrieval-augmented generation (RAG), and Claude Sonnet 3.7 as a representative general-purpose model. Finetuned Llama 3.1 70B achieved superior performance on multiple-choice evaluations and automatic text metrics, ranking second to Claude 3.7 in LLM-as-a-judge assessments. Human expert evaluation favored Claude 3.7 and RAG approaches for complex queries. Finetuned models significantly outperformed their base counterparts across nearly all evaluation modes. Our findings reveal substantial performance heterogeneity across evaluation methodologies, underscoring assessment complexity. Nevertheless, domain-specific adaptation through finetuning and RAG achieves competitive performance with proprietary models, supporting the viability of privacy-preserving, locally deployable clinical solutions.

---

### Position: LLM Watermarking Should Align Stakeholders' Incentives for Practical Adoption

Authors: Yepeng Liu, Xuandong Zhao, Dawn Song, Gregory W. Wornell, Yuheng Bu
Published: 2025-10-21
Categories: cs.CR, cs.CL
arXiv: https://arxiv.org/abs/2510.18333v2

Despite progress in watermarking algorithms for large language models (LLMs), real-world deployment remains limited. We argue that this gap stems from misaligned incentives among LLM providers, platforms, and end users, which manifest as three key barriers: competitive risk, detection-tool governance, and attribution issues. We revisit three classes of watermarking through this lens. \emph{Model watermarking} naturally aligns with LLM provider interests, yet faces new challenges in open-source ecosystems. \emph{LLM text watermarking} offers modest provider benefit when framed solely as an anti-misuse tool, but can gain traction in narrowly scoped settings such as dataset de-contamination or user-controlled provenance. \emph{In-context watermarking} (ICW) is tailored for trusted parties, such as conference organizers or educators, who embed hidden watermarking instructions into documents. If a dishonest reviewer or student submits this text to an LLM, the output carries a detectable watermark indicating misuse. This setup aligns incentives: users experience no quality loss, trusted parties gain a detection tool, and LLM providers remain neutral by simply following watermark instructions. We advocate for a broader exploration of incentive-aligned methods, with ICW as an example, in domains where trusted parties need reliable tools to detect misuse. More broadly, we distill design principles for incentive-aligned, domain-specific watermarking and outline future research directions. Our position is that the practical adoption of LLM watermarking requires aligning stakeholder incentives in targeted application domains and fostering active community engagement.

---

### Parametrising the Inhomogeneity Inducing Capacity of a Training Set, and its Impact on Supervised Learning

Authors: Gargi Roy, Dalia Chakrabarty
Published: 2025-10-21
Categories: stat.ML, cs.LG
arXiv: https://arxiv.org/abs/2510.18332v1

We introduce parametrisation of that property of the available training dataset, that necessitates an inhomogeneous correlation structure for the function that is learnt as a model of the relationship between the pair of variables, observations of which comprise the considered training data. We refer to a parametrisation of this property of a given training set, as its ``inhomogeneity parameter''. It is easy to compute this parameter for small-to-large datasets, and we demonstrate such computation on multiple publicly-available datasets, while also demonstrating that conventional ``non-stationarity'' of data does not imply a non-zero inhomogeneity parameter of the dataset. We prove that - within the probabilistic Gaussian Process-based learning approach - a training set with a non-zero inhomogeneity parameter renders it imperative, that the process that is invoked to model the sought function, be non-stationary. Following the learning of a real-world multivariate function with such a Process, quality and reliability of predictions at test inputs, are demonstrated to be affected by the inhomogeneity parameter of the training data.

---

### Scalable, Explainable and Provably Robust Anomaly Detection with One-Step Flow Matching

Authors: Zhong Li, Qi Huang, Yuxuan Zhu, Lincen Yang, Mohammad Mohammadi Amiri, Niki van Stein, Matthijs van Leeuwen
Published: 2025-10-21
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2510.18328v1

We introduce Time-Conditioned Contraction Matching (TCCM), a novel method for semi-supervised anomaly detection in tabular data. TCCM is inspired by flow matching, a recent generative modeling framework that learns velocity fields between probability distributions and has shown strong performance compared to diffusion models and generative adversarial networks. Instead of directly applying flow matching as originally formulated, TCCM builds on its core idea -- learning velocity fields between distributions -- but simplifies the framework by predicting a time-conditioned contraction vector toward a fixed target (the origin) at each sampled time step. This design offers three key advantages: (1) a lightweight and scalable training objective that removes the need for solving ordinary differential equations during training and inference; (2) an efficient scoring strategy called one time-step deviation, which quantifies deviation from expected contraction behavior in a single forward pass, addressing the inference bottleneck of existing continuous-time models such as DTE (a diffusion-based model with leading anomaly detection accuracy but heavy inference cost); and (3) explainability and provable robustness, as the learned velocity field operates directly in input space, making the anomaly score inherently feature-wise attributable; moreover, the score function is Lipschitz-continuous with respect to the input, providing theoretical guarantees under small perturbations. Extensive experiments on the ADBench benchmark show that TCCM strikes a favorable balance between detection accuracy and inference cost, outperforming state-of-the-art methods -- especially on high-dimensional and large-scale datasets. The source code is available at our GitHub repository.

---

### Enhancing Few-Shot Classification of Benchmark and Disaster Imagery with ABHFA-Net

Authors: Gao Yu Lee, Tanmoy Dam, Md Meftahul Ferdaus, Daniel Puiu Poenar, Vu Duong
Published: 2025-10-21
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.18326v4

The rising incidence of natural and human-induced disasters necessitates robust visual recognition systems capable of operating under limited labeled data conditions. However, disaster-related image classification remains challenging due to data scarcity, high intra-class variability, and domain-specific complexities in remote sensing imagery. To address these challenges, we propose the Attention Bhattacharyya Distance-based Feature Aggregation Network (ABHFA-Net), a novel few-shot learning (FSL) framework that models class prototypes as probability distributions and performs classification via Bhattacharyya distance-based comparison. Our approach integrates a spatial channel attention mechanism to enhance discrimiantive feature learning in the few-shot context and introduces a Bhattacharyya-based contrastive softmax loss for improved class separability. Extensive experiments on both benchmark datasets (CIFAR-FS, FC-100, miniImageNet, tieredImageNet) and real-world disaster datasets (AIDER, CDD, MEDIC) demonstrate the effectiveness of the proposed method. In particular, ABHFA-Net achieves 80.7% and 92.3% accuracy on CIFAR-FS under 5-way 1-shot and 5-shot settings, respectively, outperforming existing state-of-the-art methods. On disaster datasets, the model consistently improves classification performance, achieving up to 68.2% (1-shot) and 78.3% (5-shot) accuracy on AIDER, highlighting its robustness in real-world scenarios. These results establish ABHFA-Net as a strong and practical solution for few-shot disaster image classification, particularly in data-scarce and time-critical remote sensing applications. The code repository for our work is available at https://github.com/GreedYLearner1146/ABHFA-Net.

---

### Uncertainty Estimation by Flexible Evidential Deep Learning

Authors: Taeseong Yoon, Heeyoung Kim
Published: 2025-10-21
Categories: cs.LG, stat.ML
arXiv: https://arxiv.org/abs/2510.18322v2

Uncertainty quantification (UQ) is crucial for deploying machine learning models in high-stakes applications, where overconfident predictions can lead to serious consequences. An effective UQ method must balance computational efficiency with the ability to generalize across diverse scenarios. Evidential deep learning (EDL) achieves efficiency by modeling uncertainty through the prediction of a Dirichlet distribution over class probabilities. However, the restrictive assumption of Dirichlet-distributed class probabilities limits EDL's robustness, particularly in complex or unforeseen situations. To address this, we propose \textit{flexible evidential deep learning} ($\mathcal{F}$-EDL), which extends EDL by predicting a flexible Dirichlet distribution -- a generalization of the Dirichlet distribution -- over class probabilities. This approach provides a more expressive and adaptive representation of uncertainty, significantly enhancing UQ generalization and reliability under challenging scenarios. We theoretically establish several advantages of $\mathcal{F}$-EDL and empirically demonstrate its state-of-the-art UQ performance across diverse evaluation settings, including classical, long-tailed, and noisy in-distribution scenarios.

---

### Beyond Single Models: Mitigating Multimodal Hallucinations via Adaptive Token Ensemble Decoding

Authors: Jinlin Li, Yuran Wang, Yifei Yuan, Xiao Zhou, Yingying Zhang, Xixian Yong, Yefeng Zheng, Xian Wu
Published: 2025-10-21
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.18321v1

Large Vision-Language Models (LVLMs) have recently achieved impressive results in multimodal tasks such as image captioning and visual question answering. However, they remain prone to object hallucination -- generating descriptions of nonexistent or misidentified objects. Prior work has partially mitigated this via auxiliary training objectives or external modules, but challenges remain in terms of scalability, adaptability, and model independence. To address these limitations, we propose Adaptive Token Ensemble Decoding (ATED), a training-free, token-level ensemble framework that mitigates hallucination by aggregating predictions from multiple LVLMs during inference. ATED dynamically computes uncertainty-based weights for each model, reflecting their reliability at each decoding step. It also integrates diverse decoding paths to improve contextual grounding and semantic consistency. Experiments on standard hallucination detection benchmarks demonstrate that ATED significantly outperforms state-of-the-art methods, reducing hallucination without compromising fluency or relevance. Our findings highlight the benefits of adaptive ensembling and point to a promising direction for improving LVLM robustness in high-stakes applications. The code is available at https://github.com/jinlin2021/ATED.

---

### Earth AI: Unlocking Geospatial Insights with Foundation Models and Cross-Modal Reasoning

Authors: Aaron Bell, Amit Aides, Amr Helmy, Arbaaz Muslim, Aviad Barzilai, Aviv Slobodkin, Bolous Jaber, David Schottlander, George Leifman, Joydeep Paul, Mimi Sun, Nadav Sherman et al.
Published: 2025-10-21
Categories: cs.AI
arXiv: https://arxiv.org/abs/2510.18318v4

Geospatial data offers immense potential for understanding our planet. However, the sheer volume and diversity of this data along with its varied resolutions, timescales, and sparsity pose significant challenges for thorough analysis and interpretation. This paper introduces Earth AI, a family of geospatial AI models and agentic reasoning that enables significant advances in our ability to unlock novel and profound insights into our planet. This approach is built upon foundation models across three key domains--Planet-scale Imagery, Population, and Environment--and an intelligent Gemini-powered reasoning engine. We present rigorous benchmarks showcasing the power and novel capabilities of our foundation models and validate that when used together, they provide complementary value for geospatial inference and their synergies unlock superior predictive capabilities. To handle complex, multi-step queries, we developed a Gemini-powered agent that jointly reasons over our multiple foundation models along with large geospatial data sources and tools. On a new benchmark of real-world crisis scenarios, our agent demonstrates the ability to deliver critical and timely insights, effectively bridging the gap between raw geospatial data and actionable understanding.

---

### MoMaGen: Generating Demonstrations under Soft and Hard Constraints for Multi-Step Bimanual Mobile Manipulation

Authors: Chengshu Li, Mengdi Xu, Arpit Bahety, Hang Yin, Yunfan Jiang, Huang Huang, Josiah Wong, Sujay Garlanka, Cem Gokmen, Ruohan Zhang, Weiyu Liu, Jiajun Wu et al.
Published: 2025-10-21
Categories: cs.RO, cs.AI, cs.LG
arXiv: https://arxiv.org/abs/2510.18316v4

Imitation learning from large-scale, diverse human demonstrations has been shown to be effective for training robots, but collecting such data is costly and time-consuming. This challenge intensifies for multi-step bimanual mobile manipulation, where humans must teleoperate both the mobile base and two high-DoF arms. Prior X-Gen works have developed automated data generation frameworks for static (bimanual) manipulation tasks, augmenting a few human demos in simulation with novel scene configurations to synthesize large-scale datasets. However, prior works fall short for bimanual mobile manipulation tasks for two major reasons: 1) a mobile base introduces the problem of how to place the robot base to enable downstream manipulation (reachability) and 2) an active camera introduces the problem of how to position the camera to generate data for a visuomotor policy (visibility). To address these challenges, MoMaGen formulates data generation as a constrained optimization problem that satisfies hard constraints (e.g., reachability) while balancing soft constraints (e.g., visibility while navigation). This formulation generalizes across most existing automated data generation approaches and offers a principled foundation for developing future methods. We evaluate on four multi-step bimanual mobile manipulation tasks and find that MoMaGen enables the generation of much more diverse datasets than previous methods. As a result of the dataset diversity, we also show that the data generated by MoMaGen can be used to train successful imitation learning policies using a single source demo. Furthermore, the trained policy can be fine-tuned with a very small amount of real-world data (40 demos) to be succesfully deployed on real robotic hardware. More details are on our project page: momagen.github.io.

---

### Higher Embedding Dimension Creates a Stronger World Model for a Simple Sorting Task

Authors: Brady Bhalla, Honglu Fan, Nancy Chen, Tony Yue YU
Published: 2025-10-21
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2510.18315v2

We investigate how embedding dimension affects the emergence of an internal "world model" in a transformer trained with reinforcement learning to perform bubble-sort-style adjacent swaps. Models achieve high accuracy even with very small embedding dimensions, but larger dimensions yield more faithful, consistent, and robust internal representations. In particular, higher embedding dimensions strengthen the formation of structured internal representation and lead to better interpretability. After hundreds of experiments, we observe two consistent mechanisms: (1) the last row of the attention weight matrix monotonically encodes the global ordering of tokens; and (2) the selected transposition aligns with the largest adjacent difference of these encoded values. Our results provide quantitative evidence that transformers build structured internal world models and that model size improves representation quality in addition to end performance. We release our metrics and analyses, which can be used to probe similar algorithmic tasks.

---

### Genesis: Evolving Attack Strategies for LLM Web Agent Red-Teaming

Authors: Zheng Zhang, Jiarui He, Yuchen Cai, Deheng Ye, Peilin Zhao, Ruili Feng, Hao Wang
Published: 2025-10-21
Categories: cs.AI
arXiv: https://arxiv.org/abs/2510.18314v2

As large language model (LLM) agents increasingly automate complex web tasks, they boost productivity while simultaneously introducing new security risks. However, relevant studies on web agent attacks remain limited. Existing red-teaming approaches mainly rely on manually crafted attack strategies or static models trained offline. Such methods fail to capture the underlying behavioral patterns of web agents, making it difficult to generalize across diverse environments. In web agent attacks, success requires the continuous discovery and evolution of attack strategies. To this end, we propose Genesis, a novel agentic framework composed of three modules: Attacker, Scorer, and Strategist. The Attacker generates adversarial injections by integrating the genetic algorithm with a hybrid strategy representation. The Scorer evaluates the target web agent's responses to provide feedback. The Strategist dynamically uncovers effective strategies from interaction logs and compiles them into a continuously growing strategy library, which is then re-deployed to enhance the Attacker's effectiveness. Extensive experiments across various web tasks show that our framework discovers novel strategies and consistently outperforms existing attack baselines. Our code is available at https://github.com/CjangCjengh/web_agent_attack.

---

### OmniNWM: Omniscient Driving Navigation World Models

Authors: Bohan Li, Zhuang Ma, Dalong Du, Baorui Peng, Zhujin Liang, Zhenqiang Liu, Xianda Guo, Zheng Zhu, Chao Ma, Yueming Jin, Xin Jin, Hao Zhao et al.
Published: 2025-10-21
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.18313v6

Autonomous driving world models are expected to work effectively across three core dimensions: state, action, and reward. However, existing methods are typically restricted to fragmented modality modeling, short-horizon drift, and imprecise action control, while lacking intrinsic mechanisms for policy evaluation. In this paper, we introduce OmniNWM, an Omniscient panoramic Navigation World Model that addresses all three dimensions within a consistent probabilistic framework. For State, OmniNWM generates panoramic videos of RGB, semantics, metric depth, and 3D occupancy, ensuring pixel-level alignment across modalities with joint distribution modeling. To mitigate autoregressive exposure bias, we propose a structured panoramic forcing strategy to stabilize long-horizon generation via stochastic manifold thickening. For Action, we introduce canonical geometric action encoding with normalized panoramic Plücker ray-maps. This representation decouples motion dynamics from sensor intrinsics, enabling precise, zero-shot trajectory control across heterogeneous datasets and camera configurations. For Reward, we derive intrinsic occupancy-grounded dense rewards directly from generated 3D volumes, establishing a reliable closed-loop simulation cycle for evaluating diverse planning agents. Extensive experiments demonstrate that OmniNWM achieves SOTA performance in generation fidelity and control precision, with remarkable zero-shot robustness to novel scenes on NuPlan and in-house datasets with distinct camera rigs. Project page is available at https://arlo0o.github.io/OmniNWM/.

---

### Towards Identifiability of Hierarchical Temporal Causal Representation Learning

Authors: Zijian Li, Minghao Fu, Junxian Huang, Yifan Shen, Ruichu Cai, Yuewen Sun, Guangyi Chen, Kun Zhang
Published: 2025-10-21
Categories: cs.LG, stat.ME
arXiv: https://arxiv.org/abs/2510.18310v1

Modeling hierarchical latent dynamics behind time series data is critical for capturing temporal dependencies across multiple levels of abstraction in real-world tasks. However, existing temporal causal representation learning methods fail to capture such dynamics, as they fail to recover the joint distribution of hierarchical latent variables from \textit{single-timestep observed variables}. Interestingly, we find that the joint distribution of hierarchical latent variables can be uniquely determined using three conditionally independent observations. Building on this insight, we propose a Causally Hierarchical Latent Dynamic (CHiLD) identification framework. Our approach first employs temporal contextual observed variables to identify the joint distribution of multi-layer latent variables. Sequentially, we exploit the natural sparsity of the hierarchical structure among latent variables to identify latent variables within each layer. Guided by the theoretical results, we develop a time series generative model grounded in variational inference. This model incorporates a contextual encoder to reconstruct multi-layer latent variables and normalize flow-based hierarchical prior networks to impose the independent noise condition of hierarchical latent dynamics. Empirical evaluations on both synthetic and real-world datasets validate our theoretical claims and demonstrate the effectiveness of CHiLD in modeling hierarchical latent dynamics.

---

### The Impact of Image Resolution on Biomedical Multimodal Large Language Models

Authors: Liangyu Chen, James Burgess, Jeffrey J Nirschl, Orr Zohar, Serena Yeung-Levy
Published: 2025-10-21
Categories: cs.CV, cs.CL
arXiv: https://arxiv.org/abs/2510.18304v1

Imaging technologies are fundamental to biomedical research and modern medicine, requiring analysis of high-resolution images across various modalities. While multimodal large language models (MLLMs) show promise for biomedical image analysis, most are designed for low-resolution images from general-purpose datasets, risking critical information loss. We investigate how image resolution affects MLLM performance in biomedical applications and demonstrate that: (1) native-resolution training and inference significantly improve performance across multiple tasks, (2) misalignment between training and inference resolutions severely degrades performance, and (3) mixed-resolution training effectively mitigates misalignment and balances computational constraints with performance requirements. Based on these findings, we recommend prioritizing native-resolution inference and mixed-resolution datasets to optimize biomedical MLLMs for transformative impact in scientific research and clinical applications.

---

### Proactive Reasoning-with-Retrieval Framework for Medical Multimodal Large Language Models

Authors: Lehan Wang, Yi Qin, Honglong Yang, Xiaomeng Li
Published: 2025-10-21
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.18303v1

Incentivizing the reasoning ability of Multimodal Large Language Models (MLLMs) is essential for medical applications to transparently analyze medical scans and provide reliable diagnosis. However, existing medical MLLMs rely solely on internal knowledge during reasoning, leading to hallucinated reasoning and factual inaccuracies when encountering cases beyond their training scope. Although recent Agentic Retrieval-Augmented Generation (RAG) methods elicit the medical model's proactive retrieval ability during reasoning, they are confined to unimodal LLMs, neglecting the crucial visual information during reasoning and retrieval. Consequently, we propose the first Multimodal Medical Reasoning-with-Retrieval framework, Med-RwR, which actively retrieves external knowledge by querying observed symptoms or domain-specific medical concepts during reasoning. Specifically, we design a two-stage reinforcement learning strategy with tailored rewards that stimulate the model to leverage both visual diagnostic findings and textual clinical information for effective retrieval. Building on this foundation, we further propose a Confidence-Driven Image Re-retrieval (CDIR) method for test-time scaling when low prediction confidence is detected. Evaluation on various public medical benchmarks demonstrates Med-RwR's significant improvements over baseline models, proving the effectiveness of enhancing reasoning capabilities with external knowledge integration. Furthermore, Med-RwR demonstrates remarkable generalizability to unfamiliar domains, evidenced by 8.8% performance gain on our proposed EchoCardiography Benchmark (ECBench), despite the scarcity of echocardiography data in the training corpus. Our data, model, and codes will be made publicly available at https://github.com/xmed-lab/Med-RwR.

---

### A Distributed Framework for Causal Modeling of Performance Variability in GPU Traces

Authors: Ankur Lahiry, Ayush Pokharel, Banooqa Banday, Seth Ockerman, Amal Gueroudji, Mohammad Zaeed, Tanzima Z. Islam, Line Pouchard
Published: 2025-10-21
Categories: cs.DC, cs.LG
arXiv: https://arxiv.org/abs/2510.18300v1

Large-scale GPU traces play a critical role in identifying performance bottlenecks within heterogeneous High-Performance Computing (HPC) architectures. However, the sheer volume and complexity of a single trace of data make performance analysis both computationally expensive and time-consuming. To address this challenge, we present an end-to-end parallel performance analysis framework designed to handle multiple large-scale GPU traces efficiently. Our proposed framework partitions and processes trace data concurrently and employs causal graph methods and parallel coordinating chart to expose performance variability and dependencies across execution flows. Experimental results demonstrate a 67% improvement in terms of scalability, highlighting the effectiveness of our pipeline for analyzing multiple traces independently.

---

### Physics-Informed Parametric Bandits for Beam Alignment in mmWave Communications

Authors: Hao Qin, Thang Duong, Ming F. Li, Chicheng Zhang
Published: 2025-10-21
Categories: cs.LG
arXiv: https://arxiv.org/abs/2510.18299v2

In millimeter wave (mmWave) communications, beam alignment and tracking are crucial to combat the significant path loss. As scanning the entire directional space is inefficient, designing an efficient and robust method to identify the optimal beam directions is essential. Since traditional bandit algorithms require a long time horizon to converge under large beam spaces, many existing works propose efficient bandit algorithms for beam alignment by relying on unimodality or multimodality assumptions on the reward function's structure. However, such assumptions often do not hold (or cannot be strictly satisfied) in practice, which causes such algorithms to converge to choosing suboptimal beams. In this work, we propose two physics-informed bandit algorithms \textit{pretc} and \textit{prgreedy} that exploit the sparse multipath property of mmWave channels - a generic but realistic assumption - which is connected to the Phase Retrieval Bandit problem. Our algorithms treat the parameters of each path as black boxes and maintain optimal estimates of them based on sampled historical rewards. \textit{pretc} starts with a random exploration phase and then commits to the optimal beam under the estimated reward function. \textit{prgreedy} performs such estimation in an online manner and chooses the best beam under current estimates. Our algorithms can also be easily adapted to beam tracking in the mobile setting. Through experiments using both the synthetic DeepMIMO dataset and the real-world DeepSense6G dataset, we demonstrate that both algorithms outperform existing approaches in a wide range of scenarios across diverse channel environments, showing their generalizability and robustness.

---

### From Retrieval to Generation: Unifying External and Parametric Knowledge for Medical Question Answering

Authors: Lei Li, Xiao Zhou, Yingying Zhang, Xian Wu
Published: 2025-10-21
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2510.18297v1

Medical question answering (QA) requires extensive access to domain-specific knowledge. A promising direction is to enhance large language models (LLMs) with external knowledge retrieved from medical corpora or parametric knowledge stored in model parameters. Existing approaches typically fall into two categories: Retrieval-Augmented Generation (RAG), which grounds model reasoning on externally retrieved evidence, and Generation-Augmented Generation (GAG), which depends solely on the models internal knowledge to generate contextual documents. However, RAG often suffers from noisy or incomplete retrieval, while GAG is vulnerable to hallucinated or inaccurate information due to unconstrained generation. Both issues can mislead reasoning and undermine answer reliability. To address these challenges, we propose MedRGAG, a unified retrieval-generation augmented framework that seamlessly integrates external and parametric knowledge for medical QA. MedRGAG comprises two key modules: Knowledge-Guided Context Completion (KGCC), which directs the generator to produce background documents that complement the missing knowledge revealed by retrieval; and Knowledge-Aware Document Selection (KADS), which adaptively selects an optimal combination of retrieved and generated documents to form concise yet comprehensive evidence for answer generation. Extensive experiments on five medical QA benchmarks demonstrate that MedRGAG achieves a 12.5% improvement over MedRAG and a 4.5% gain over MedGENIE, highlighting the effectiveness of unifying retrieval and generation for knowledge-intensive reasoning. Our code and data are publicly available at https://anonymous.4open.science/r/MedRGAG

---

### GeoDiff: Geometry-Guided Diffusion for Metric Depth Estimation

Authors: Tuan Pham, Thanh-Tung Le, Xiaohui Xie, Stephan Mandt
Published: 2025-10-21
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.18291v1

We introduce a novel framework for metric depth estimation that enhances pretrained diffusion-based monocular depth estimation (DB-MDE) models with stereo vision guidance. While existing DB-MDE methods excel at predicting relative depth, estimating absolute metric depth remains challenging due to scale ambiguities in single-image scenarios. To address this, we reframe depth estimation as an inverse problem, leveraging pretrained latent diffusion models (LDMs) conditioned on RGB images, combined with stereo-based geometric constraints, to learn scale and shift for accurate depth recovery. Our training-free solution seamlessly integrates into existing DB-MDE frameworks and generalizes across indoor, outdoor, and complex environments. Extensive experiments demonstrate that our approach matches or surpasses state-of-the-art methods, particularly in challenging scenarios involving translucent and specular surfaces, all without requiring retraining.

---

### Food4All: An Agentic Framework and Benchmark for Food Resource Navigation with Adaptive User Understanding

Authors: Yiyang Li, Weixiang Sun, Tianyi Ma, Kaiwen Shi, Zheyuan Zhang, Yanfang Ye
Published: 2025-10-21
Categories: cs.CL, cs.CY, cs.MA
arXiv: https://arxiv.org/abs/2510.18289v3

Food assistance referral requires conversational agents to translate underspecified, often noisy help-seeking dialogues into locally valid resource recommendations. We present Food4All, an agentic food-resource referral framework and benchmark grounded in 686 structured Indiana food resources. Food4All couples a food-specific search tool with 300 multi-turn evaluation tasks spanning single food needs, composite cases with access or document constraints, and five non-ideal user interaction traits: unreasonable demands, rambling responses, impatience, incomplete answers, and inconsistent information. We evaluate six Large Language Models (LLMs) on requirement grounding, resource retrieval, final referral correctness, and interaction efficiency. Although the strongest model achieves 96.33% referral accuracy, our diagnostics reveal persistent failures in grounding schedule, eligibility, intake, and document constraints, as well as failures to preserve valid retrieved resources in the final recommendation. Trait-level analysis further shows that different non-ideal behaviors stress different parts of the referral pipeline. Food4All provides a controlled testbed for studying tool-calling agents in constraint-sensitive food assistance referral under realistic user interaction challenges.

---

### BrailleLLM: Braille Instruction Tuning with Large Language Models for Braille Domain Tasks

Authors: Tianyuan Huang, Zepeng Zhu, Hangdi Xing, Zirui Shao, Zhi Yu, Chaoxiong Yang, Jiaxian He, Xiaozhong Liu, Jiajun Bu
Published: 2025-10-21
Categories: cs.CL
arXiv: https://arxiv.org/abs/2510.18288v1

Braille plays a vital role in education and information accessibility for visually impaired individuals. However, Braille information processing faces challenges such as data scarcity and ambiguities in mixed-text contexts. We construct English and Chinese Braille Mixed Datasets (EBMD/CBMD) with mathematical formulas to support diverse Braille domain research, and propose a syntax tree-based augmentation method tailored for Braille data. To address the underperformance of traditional fine-tuning methods in Braille-related tasks, we investigate Braille Knowledge-Based Fine-Tuning (BKFT), which reduces the learning difficulty of Braille contextual features. BrailleLLM employs BKFT via instruction tuning to achieve unified Braille translation, formula-to-Braille conversion, and mixed-text translation. Experiments demonstrate that BKFT achieves significant performance improvements over conventional fine-tuning in Braille translation scenarios. Our open-sourced datasets and methodologies establish a foundation for low-resource multilingual Braille research.

---

### Efficient Few-shot Identity Preserving Attribute Editing for 3D-aware Deep Generative Models

Authors: Vishal Vinod
Published: 2025-10-21
Categories: cs.CV, cs.LG
arXiv: https://arxiv.org/abs/2510.18287v1

Identity preserving editing of faces is a generative task that enables modifying the illumination, adding/removing eyeglasses, face aging, editing hairstyles, modifying expression etc., while preserving the identity of the face. Recent progress in 2D generative models have enabled photorealistic editing of faces using simple techniques leveraging the compositionality in GANs. However, identity preserving editing for 3D faces with a given set of attributes is a challenging task as the generative model must reason about view consistency from multiple poses and render a realistic 3D face. Further, 3D portrait editing requires large-scale attribute labelled datasets and presents a trade-off between editability in low-resolution and inflexibility to editing in high resolution. In this work, we aim to alleviate some of the constraints in editing 3D faces by identifying latent space directions that correspond to photorealistic edits. To address this, we present a method that builds on recent advancements in 3D-aware deep generative models and 2D portrait editing techniques to perform efficient few-shot identity preserving attribute editing for 3D-aware generative models. We aim to show from experimental results that using just ten or fewer labelled images of an attribute is sufficient to estimate edit directions in the latent space that correspond to 3D-aware attribute editing. In this work, we leverage an existing face dataset with masks to obtain the synthetic images for few attribute examples required for estimating the edit directions. Further, to demonstrate the linearity of edits, we investigate one-shot stylization by performing sequential editing and use the (2D) Attribute Style Manipulation (ASM) technique to investigate a continuous style manifold for 3D consistent identity preserving face aging. Code and results are available at: https://vishal-vinod.github.io/gmpi-edit/

---

### Online Time Series Forecasting with Theoretical Guarantees

Authors: Zijian Li, Changze Zhou, Minghao Fu, Sanjay Manjunath, Fan Feng, Guangyi Chen, Yingyao Hu, Ruichu Cai, Kun Zhang
Published: 2025-10-21
Categories: cs.LG
arXiv: https://arxiv.org/abs/2510.18281v1

This paper is concerned with online time series forecasting, where unknown distribution shifts occur over time, i.e., latent variables influence the mapping from historical to future observations. To develop an automated way of online time series forecasting, we propose a Theoretical framework for Online Time-series forecasting (TOT in short) with theoretical guarantees. Specifically, we prove that supplying a forecaster with latent variables tightens the Bayes risk, the benefit endures under estimation uncertainty of latent variables and grows as the latent variables achieve a more precise identifiability. To better introduce latent variables into online forecasting algorithms, we further propose to identify latent variables with minimal adjacent observations. Based on these results, we devise a model-agnostic blueprint by employing a temporal decoder to match the distribution of observed variables and two independent noise estimators to model the causal inference of latent variables and mixing procedures of observed variables, respectively. Experiment results on synthetic data support our theoretical claims. Moreover, plug-in implementations built on several baselines yield general improvement across multiple benchmarks, highlighting the effectiveness in real-world applications.

---

### Text or Pixels? It Takes Half: On the Token Efficiency of Visual Text Inputs in Multimodal LLMs

Authors: Yanhong Li, Zixuan Lan, Jiawei Zhou
Published: 2025-10-21
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2510.18279v2

Large language models (LLMs) and their multimodal variants can now process visual inputs, including images of text. This raises an intriguing question: can we compress textual inputs by feeding them as images to reduce token usage while preserving performance? In this paper, we show that visual text representations are a practical and surprisingly effective form of input compression for decoder LLMs. We exploit the idea of rendering long text inputs as a single image and provide it directly to the model. This leads to dramatically reduced number of decoder tokens required, offering a new form of input compression. Through experiments on two distinct benchmarks RULER (long-context retrieval) and CNN/DailyMail (document summarization) we demonstrate that this text-as-image method yields substantial token savings (often nearly half) without degrading task performance.

---

### StreamingTOM: Streaming Token Compression for Efficient Video Understanding

Authors: Xueyi Chen, Keda Tao, Kele Shao, Huan Wang
Published: 2025-10-21
Categories: cs.CV, cs.AI
arXiv: https://arxiv.org/abs/2510.18269v2

Unlike offline processing, streaming video vision-language models face two fundamental constraints: causality and accumulation. Causality prevents access to future frames that offline methods exploit, while accumulation causes tokens to grow unbounded, creating efficiency bottlenecks. However, existing approaches only regulate post-LLM kv-cache, leaving costly pre-LLM prefill unchanged. We introduce StreamingTOM, a training-free, plug-and-play two-stage framework that addresses both pre-LLM and post-LLM bottlenecks. Causal Temporal Reduction imposes a fixed per-frame budget and selects tokens based on adjacent-frame changes and token saliency, drastically reducing per-frame prefill cost by processing only a compact subset of visual tokens, ensuring predictable latency. Online Quantized Memory stores tokens in 4-bit format, retrieves relevant groups on demand, and dequantizes them, keeping the active kv-cache bounded regardless of stream length. Experiments demonstrate our method achieves $15.7\times$ kv-cache compression ratio; compared to prior SOTA (LiveVLM), it delivers $1.2\times$ lower peak memory and $2\times$ faster TTFT. StreamingTOM achieves state-of-the-art accuracy among training-free methods with an average of $63.8\%$ on offline benchmarks and $55.8\%$ accuracy and $3.7$ score on RVS. These results demonstrate that real-time streaming video understanding with bounded active memory is achievable without model retraining.

---

### TreeFedDG: Alleviating Global Drift in Federated Domain Generalization for Medical Image Segmentation

Authors: Yucheng Song, Chenxi Li, Haokang Ding, Zhining Liao, Zhifang Liao
Published: 2025-10-21
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.18268v1

In medical image segmentation tasks, Domain Generalization (DG) under the Federated Learning (FL) framework is crucial for addressing challenges related to privacy protection and data heterogeneity. However, traditional federated learning methods fail to account for the imbalance in information aggregation across clients in cross-domain scenarios, leading to the Global Drift (GD) problem and a consequent decline in model generalization performance. This motivates us to delve deeper and define a new critical issue: global drift in federated domain generalization for medical imaging (FedDG-GD). In this paper, we propose a novel tree topology framework called TreeFedDG. First, starting from the distributed characteristics of medical images, we design a hierarchical parameter aggregation method based on a tree-structured topology to suppress deviations in the global model direction. Second, we introduce a parameter difference-based style mixing method (FedStyle), which enforces mixing among clients with maximum parameter differences to enhance robustness against drift. Third, we develop a a progressive personalized fusion strategy during model distribution, ensuring a balance between knowledge transfer and personalized features. Finally, during the inference phase, we use feature similarity to guide the retrieval of the most relevant model chain from the tree structure for ensemble decision-making, thereby fully leveraging the advantages of hierarchical knowledge. We conducted extensive experiments on two publicly available datasets. The results demonstrate that our method outperforms other state-of-the-art domain generalization approaches in these challenging tasks and achieves better balance in cross-domain performance.

---

### Latent-Info and Low-Dimensional Learning for Human Mesh Recovery and Parallel Optimization

Authors: Xiang Zhang, Suping Wu, Sheng Yang
Published: 2025-10-21
Categories: cs.CV, cs.AI
arXiv: https://arxiv.org/abs/2510.18267v1

Existing 3D human mesh recovery methods often fail to fully exploit the latent information (e.g., human motion, shape alignment), leading to issues with limb misalignment and insufficient local details in the reconstructed human mesh (especially in complex scenes). Furthermore, the performance improvement gained by modelling mesh vertices and pose node interactions using attention mechanisms comes at a high computational cost. To address these issues, we propose a two-stage network for human mesh recovery based on latent information and low dimensional learning. Specifically, the first stage of the network fully excavates global (e.g., the overall shape alignment) and local (e.g., textures, detail) information from the low and high-frequency components of image features and aggregates this information into a hybrid latent frequency domain feature. This strategy effectively extracts latent information. Subsequently, utilizing extracted hybrid latent frequency domain features collaborates to enhance 2D poses to 3D learning. In the second stage, with the assistance of hybrid latent features, we model the interaction learning between the rough 3D human mesh template and the 3D pose, optimizing the pose and shape of the human mesh. Unlike existing mesh pose interaction methods, we design a low-dimensional mesh pose interaction method through dimensionality reduction and parallel optimization that significantly reduces computational costs without sacrificing reconstruction accuracy. Extensive experimental results on large publicly available datasets indicate superiority compared to the most state-of-the-art.

---

### SPIKE: Stable Physics-Informed Kernel Evolution Method for Solving Hyperbolic Conservation Laws

Authors: Hua Su, Lei Zhang, Jin Zhao
Published: 2025-10-21
Categories: math.NA, cs.AI, cs.LG, math.AP
arXiv: https://arxiv.org/abs/2510.18266v1

We introduce the Stable Physics-Informed Kernel Evolution (SPIKE) method for numerical computation of inviscid hyperbolic conservation laws. SPIKE resolves a fundamental paradox: how strong-form residual minimization can capture weak solutions containing discontinuities. SPIKE employs reproducing kernel representations with regularized parameter evolution, where Tikhonov regularization provides a smooth transition mechanism through shock formation, allowing the dynamics to traverse shock singularities. This approach automatically maintains conservation, tracks characteristics, and captures shocks satisfying Rankine-Hugoniot conditions within a unified framework requiring no explicit shock detection or artificial viscosity. Numerical validation across scalar and vector-valued conservation laws confirms the method's effectiveness.

---

### From Competition to Synergy: Unlocking Reinforcement Learning for Subject-Driven Image Generation

Authors: Ziwei Huang, Ying Shu, Hao Fang, Quanyu Long, Wenya Wang, Qiushi Guo, Tiezheng Ge, Leilei Gan
Published: 2025-10-21
Categories: cs.LG, cs.CV, cs.GR
arXiv: https://arxiv.org/abs/2510.18263v2

Subject-driven image generation models face a fundamental trade-off between identity preservation (fidelity) and prompt adherence (editability). While online reinforcement learning (RL), specifically GPRO, offers a promising solution, we find that a naive application of GRPO leads to competitive degradation, as the simple linear aggregation of rewards with static weights causes conflicting gradient signals and a misalignment with the temporal dynamics of the diffusion process. To overcome these limitations, we propose Customized-GRPO, a novel framework featuring two key innovations: (i) Synergy-Aware Reward Shaping (SARS), a non-linear mechanism that explicitly penalizes conflicted reward signals and amplifies synergistic ones, providing a sharper and more decisive gradient. (ii) Time-Aware Dynamic Weighting (TDW), which aligns the optimization pressure with the model's temporal dynamics by prioritizing prompt-following in the early, identity preservation in the later. Extensive experiments demonstrate that our method significantly outperforms naive GRPO baselines, successfully mitigating competitive degradation. Our model achieves a superior balance, generating images that both preserve key identity features and accurately adhere to complex textual prompts.

---

### UWBench: A Comprehensive Vision-Language Benchmark for Underwater Understanding

Authors: Da Zhang, Chenggang Rong, Bingyu Li, Feiyu Wang, Zhiyuan Zhao, Junyu Gao, Xuelong Li
Published: 2025-10-21
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.18262v1

Large vision-language models (VLMs) have achieved remarkable success in natural scene understanding, yet their application to underwater environments remains largely unexplored. Underwater imagery presents unique challenges including severe light attenuation, color distortion, and suspended particle scattering, while requiring specialized knowledge of marine ecosystems and organism taxonomy. To bridge this gap, we introduce UWBench, a comprehensive benchmark specifically designed for underwater vision-language understanding. UWBench comprises 15,003 high-resolution underwater images captured across diverse aquatic environments, encompassing oceans, coral reefs, and deep-sea habitats. Each image is enriched with human-verified annotations including 15,281 object referring expressions that precisely describe marine organisms and underwater structures, and 124,983 question-answer pairs covering diverse reasoning capabilities from object recognition to ecological relationship understanding. The dataset captures rich variations in visibility, lighting conditions, and water turbidity, providing a realistic testbed for model evaluation. Based on UWBench, we establish three comprehensive benchmarks: detailed image captioning for generating ecologically informed scene descriptions, visual grounding for precise localization of marine organisms, and visual question answering for multimodal reasoning about underwater environments. Extensive experiments on state-of-the-art VLMs demonstrate that underwater understanding remains challenging, with substantial room for improvement. Our benchmark provides essential resources for advancing vision-language research in underwater contexts and supporting applications in marine science, ecological monitoring, and autonomous underwater exploration. Our code and benchmark will be available.

---

### Learning under Quantization for High-Dimensional Linear Regression

Authors: Dechen Zhang, Junwei Su, Difan Zou
Published: 2025-10-21
Categories: stat.ML, cs.AI, cs.LG
arXiv: https://arxiv.org/abs/2510.18259v3

The use of low-bit quantization has emerged as an indispensable technique for enabling the efficient training of large-scale models. Despite its widespread empirical success, a rigorous theoretical understanding of its impact on learning performance remains notably absent, even in the simplest linear regression setting. We present the first systematic theoretical study of this fundamental question, analyzing finite-step stochastic gradient descent (SGD) for high-dimensional linear regression under a comprehensive range of quantization targets: data, label, parameter, activation, and gradient. Our novel analytical framework establishes precise algorithm-dependent and data-dependent excess risk bounds that characterize how different quantization affects learning: parameter, activation, and gradient quantization amplify noise during training; data quantization distorts the data spectrum and introduces additional approximation error. Crucially, we distinguish the effects of two quantization schemes: we prove that for additive quantization (with constant quantization steps), the noise amplification benefits from a suppression effect scaled by the batch size, while multiplicative quantization (with input-dependent quantization steps) largely preserves the spectral structure, thereby reducing the spectral distortion. Furthermore, under common polynomial-decay data spectra, we quantitatively compare the risks of multiplicative and additive quantization, drawing a parallel to the comparison between FP and integer quantization methods. Our theory provides a powerful lens to characterize how quantization shapes the learning dynamics of optimization algorithms, paving the way to further explore learning theory under practical hardware constraints.

---

### NTKMTL: Mitigating Task Imbalance in Multi-Task Learning from Neural Tangent Kernel Perspective

Authors: Xiaohan Qin, Xiaoxing Wang, Ning Liao, Junchi Yan
Published: 2025-10-21
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2510.18258v1

Multi-Task Learning (MTL) enables a single model to learn multiple tasks simultaneously, leveraging knowledge transfer among tasks for enhanced generalization, and has been widely applied across various domains. However, task imbalance remains a major challenge in MTL. Although balancing the convergence speeds of different tasks is an effective approach to address this issue, it is highly challenging to accurately characterize the training dynamics and convergence speeds of multiple tasks within the complex MTL system. To this end, we attempt to analyze the training dynamics in MTL by leveraging Neural Tangent Kernel (NTK) theory and propose a new MTL method, NTKMTL. Specifically, we introduce an extended NTK matrix for MTL and adopt spectral analysis to balance the convergence speeds of multiple tasks, thereby mitigating task imbalance. Based on the approximation via shared representation, we further propose NTKMTL-SR, achieving training efficiency while maintaining competitive performance. Extensive experiments demonstrate that our methods achieve state-of-the-art performance across a wide range of benchmarks, including both multi-task supervised learning and multi-task reinforcement learning. Source code is available at https://github.com/jianke0604/NTKMTL.

---