# Research Papers — 2025-10 · bundle page 156 of 217

> Full-text markdown of papers #7751–7800 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/156
[← Previous bundle](https://aipod.dev/papers/archive/2025-10/page/155.md)
[Next bundle →](https://aipod.dev/papers/archive/2025-10/page/157.md)

### VersionRAG: Version-Aware Retrieval-Augmented Generation for Evolving Documents

Authors: Daniel Huwiler, Kurt Stockinger, Jonathan Fürst
Published: 2025-10-09
Categories: cs.IR, cs.AI, cs.CL
arXiv: https://arxiv.org/abs/2510.08109v1

Retrieval-Augmented Generation (RAG) systems fail when documents evolve through versioning-a ubiquitous characteristic of technical documentation. Existing approaches achieve only 58-64% accuracy on version-sensitive questions, retrieving semantically similar content without temporal validity checks. We present VersionRAG, a version-aware RAG framework that explicitly models document evolution through a hierarchical graph structure capturing version sequences, content boundaries, and changes between document states. During retrieval, VersionRAG routes queries through specialized paths based on intent classification, enabling precise version-aware filtering and change tracking. On our VersionQA benchmark-100 manually curated questions across 34 versioned technical documents-VersionRAG achieves 90% accuracy, outperforming naive RAG (58%) and GraphRAG (64%). VersionRAG reaches 60% accuracy on implicit change detection where baselines fail (0-10%), demonstrating its ability to track undocumented modifications. Additionally, VersionRAG requires 97% fewer tokens during indexing than GraphRAG, making it practical for large-scale deployment. Our work establishes versioned document QA as a distinct task and provides both a solution and benchmark for future research.

---

### Development of Mental Models in Human-AI Collaboration: A Conceptual Framework

Authors: Joshua Holstein, Gerhard Satzger
Published: 2025-10-09
Categories: cs.HC, cs.AI
arXiv: https://arxiv.org/abs/2510.08104v1

Artificial intelligence has become integral to organizational decision-making and while research has explored many facets of this human-AI collaboration, the focus has mainly been on designing the AI agent(s) and the way the collaboration is set up - generally assuming a human decision-maker to be "fixed". However, it has largely been neglected that decision-makers' mental models evolve through their continuous interaction with AI systems. This paper addresses this gap by conceptualizing how the design of human-AI collaboration influences the development of three complementary and interdependent mental models necessary for this collaboration. We develop an integrated socio-technical framework that identifies the mechanisms driving the mental model evolution: data contextualization, reasoning transparency, and performance feedback. Our work advances human-AI collaboration literature through three key contributions: introducing three distinct mental models (domain, information processing, complementarity-awareness); recognizing the dynamic nature of mental models; and establishing mechanisms that guide the purposeful design of effective human-AI collaboration.

---

### Lossless Vocabulary Reduction for Auto-Regressive Language Models

Authors: Daiki Chijiwa, Taku Hasegawa, Kyosuke Nishida, Shin'ya Yamaguchi, Tomoya Ohba, Tamao Sakao, Susumu Takeuchi
Published: 2025-10-09
Categories: cs.CL, cs.AI, cs.LG, stat.ML
arXiv: https://arxiv.org/abs/2510.08102v2

Tokenization -- the process of decomposing a given text into a sequence of subwords called tokens -- is one of the key components in the development of language models. Particularly, auto-regressive language models generate texts token by token, i.e., by predicting the next-token distribution given the previous ones, and thus tokenization directly affects their efficiency in text generation. Since each language model has their own vocabulary as a set of possible tokens, they struggle to cooperate with each other at the level of next-token distributions such as model ensemble. In this paper, we establish a theoretical framework of lossless vocabulary reduction, which efficiently converts a given auto-regressive language model into the one with an arbitrarily small vocabulary without any loss in accuracy. This framework allows language models with different tokenization to cooperate with each other efficiently by reduction to their maximal common vocabulary. Specifically, we empirically demonstrate its applicability to model ensemble with different tokenization.

---

### The Price of Thought: A Multilingual Analysis of Reasoning, Performance, and Cost of Negotiation in Large Language Models

Authors: Sherzod Hakimov, Roland Bernard, Tim Leiber, Karl Osswald, Kristina Richert, Ruilin Yang, Raffaella Bernardi, David Schlangen
Published: 2025-10-09
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2510.08098v2

Negotiation is a fundamental challenge for AI agents, as it requires an ability to reason strategically, model opponents, and balance cooperation with competition. We present the first comprehensive study that systematically evaluates how explicit reasoning training affects the negotiation abilities of both commercial and open-weight large language models, comparing these models to their vanilla counterparts across three languages. Using a self-play setup across three diverse dialogue games, we analyse trade-offs between performance and cost, the language consistency of reasoning processes, and the nature of strategic adaptation exhibited by models. Our findings show that enabling reasoning -- that is, scaling test time compute -- significantly improves negotiation outcomes by enhancing collaboration and helping models overcome task complexities, but comes at a substantial computational cost: reasoning improves GPT-5's performance by 31.4 % while increasing its cost by nearly 400 %. Most critically, we uncover a significant multilingual reasoning distinction: open-weight models consistently switch to English for their internal reasoning steps, even when negotiating in German or Italian (and thus possibly impacting potential explainability gains through the disclosure of reasoning traces), while a leading commercial model maintains language consistency between reasoning and final output.

---

### Efficient Label Refinement for Face Parsing Under Extreme Poses Using 3D Gaussian Splatting

Authors: Ankit Gahlawat, Anirban Mukherjee, Dinesh Babu Jayagopi
Published: 2025-10-09
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.08096v1

Accurate face parsing under extreme viewing angles remains a significant challenge due to limited labeled data in such poses. Manual annotation is costly and often impractical at scale. We propose a novel label refinement pipeline that leverages 3D Gaussian Splatting (3DGS) to generate accurate segmentation masks from noisy multiview predictions. By jointly fitting two 3DGS models, one to RGB images and one to their initial segmentation maps, our method enforces multiview consistency through shared geometry, enabling the synthesis of pose-diverse training data with only minimal post-processing. Fine-tuning a face parsing model on this refined dataset significantly improves accuracy on challenging head poses, while maintaining strong performance on standard views. Extensive experiments, including human evaluations, demonstrate that our approach achieves superior results compared to state-of-the-art methods, despite requiring no ground-truth 3D annotations and using only a small set of initial images. Our method offers a scalable and effective solution for improving face parsing robustness in real-world settings.

---

### Beyond Real Data: Synthetic Data through the Lens of Regularization

Authors: Amitis Shidani, Tyler Farghly, Yang Sun, Habib Ganjgahi, George Deligiannidis
Published: 2025-10-09
Categories: stat.ML, cs.LG
arXiv: https://arxiv.org/abs/2510.08095v2

Synthetic data can improve generalization when real data is scarce, but excessive reliance may introduce distributional mismatches that degrade performance. In this paper, we present a learning-theoretic framework to quantify the trade-off between synthetic and real data. Our approach leverages algorithmic stability to derive generalization error bounds, characterizing the optimal synthetic-to-real data ratio that minimizes expected test error as a function of the Wasserstein distance between the real and synthetic distributions. We motivate our framework in the setting of kernel ridge regression with mixed data, offering a detailed analysis that may be of independent interest. Our theory predicts the existence of an optimal ratio, leading to a U-shaped behavior of test error with respect to the proportion of synthetic data. Empirically, we validate this prediction on CIFAR-10 and a clinical brain MRI dataset. Our theory extends to the important scenario of domain adaptation, showing that carefully blending synthetic target data with limited source data can mitigate domain shift and enhance generalization. We conclude with practical guidance for applying our results to both in-domain and out-of-domain scenarios.

---

### DarkHash: A Data-Free Backdoor Attack Against Deep Hashing

Authors: Ziqi Zhou, Menghao Deng, Yufei Song, Hangtao Zhang, Wei Wan, Shengshan Hu, Minghui Li, Leo Yu Zhang, Dezhong Yao
Published: 2025-10-09
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.08094v1

Benefiting from its superior feature learning capabilities and efficiency, deep hashing has achieved remarkable success in large-scale image retrieval. Recent studies have demonstrated the vulnerability of deep hashing models to backdoor attacks. Although these studies have shown promising attack results, they rely on access to the training dataset to implant the backdoor. In the real world, obtaining such data (e.g., identity information) is often prohibited due to privacy protection and intellectual property concerns. Embedding backdoors into deep hashing models without access to the training data, while maintaining retrieval accuracy for the original task, presents a novel and challenging problem. In this paper, we propose DarkHash, the first data-free backdoor attack against deep hashing. Specifically, we design a novel shadow backdoor attack framework with dual-semantic guidance. It embeds backdoor functionality and maintains original retrieval accuracy by fine-tuning only specific layers of the victim model using a surrogate dataset. We consider leveraging the relationship between individual samples and their neighbors to enhance backdoor attacks during training. By designing a topological alignment loss, we optimize both individual and neighboring poisoned samples toward the target sample, further enhancing the attack capability. Experimental results on four image datasets, five model architectures, and two hashing methods demonstrate the high effectiveness of DarkHash, outperforming existing state-of-the-art backdoor attack methods. Defense experiments show that DarkHash can withstand existing mainstream backdoor defense methods.

---

### Computations and ML for surjective rational maps

Authors: Ilya Karzhemanov
Published: 2025-10-09
Categories: math.AG, cs.LG
arXiv: https://arxiv.org/abs/2510.08093v2

The present note studies \emph{surjective rational endomorphisms} $f: \mathbb{P}^2 \dashrightarrow \mathbb{P}^2$ with \emph{cubic} terms and the indeterminacy locus $I_f \ne \emptyset$. We develop an experimental approach, based on some Python programming and Machine Learning, towards the classification of such maps; a couple of new explicit $f$ is constructed in this way. We also prove (via pure projective geometry) that a general non-regular cubic endomorphism $f$ of $\mathbb{P}^2$ is surjective if and only if the set $I_f$ has cardinality at least $3$.

---

### Arguments that Alter Minds: LLM Rationales Sway Human (and LLM) Notions of Plausibility

Authors: Shramay Palta, Peter Rankel, Sarah Wiegreffe, Rachel Rudinger
Published: 2025-10-09
Categories: cs.CL, cs.AI, cs.HC
arXiv: https://arxiv.org/abs/2510.08091v3

We investigate the degree to which human (and LLM) plausibility judgments of multiple-choice commonsense benchmark answers are subject to influence by (im)plausibility arguments for or against an answer, in particular, using rationales generated by LLMs. We collect 3,000 plausibility judgments from humans and another 13,600 judgments from LLMs. Overall, we observe increases and decreases in mean human plausibility ratings in the presence of LLM-generated PRO and CON rationales, respectively, suggesting that, on the whole, human judges find these rationales convincing. Experiments with LLMs reveal similar patterns of influence. Our findings demonstrate a novel use of LLMs for studying aspects of human cognition, while also raising practical concerns that, even in domains where humans are ``experts'' (i.e., common sense), LLMs have the potential to exert considerable influence on people's beliefs.

---

### From Ethical Declarations to Provable Independence: An Ontology-Driven Optimal-Transport Framework for Certifiably Fair AI Systems

Authors: Sukriti Bhattacharya, Chitro Majumdar
Published: 2025-10-09
Categories: cs.AI
arXiv: https://arxiv.org/abs/2510.08086v1

This paper presents a framework for provably fair AI that overcomes the limits of current bias mitigation methods by systematically removing all sensitive information and its proxies. Using ontology engineering in OWL 2 QL, it formally defines sensitive attributes and infers their proxies through logical reasoning, constructing a sigma algebra G that captures the full structure of biased patterns. Fair representations are then obtained via Delbaen Majumdar optimal transport, which generates variables independent of G while minimizing L2 distance to preserve accuracy. This guarantees true independence rather than mere decorrelation. By modeling bias as dependence between sigma algebras, compiling ontological knowledge into measurable structures, and using optimal transport as the unique fair transformation, the approach ensures complete fairness in tasks like loan approval, where proxies such as ZIP code reveal race. The result is a certifiable and mathematically grounded method for trustworthy AI.

---

### A Novel Ensemble Learning Approach for Enhanced IoT Attack Detection: Redefining Security Paradigms in Connected Systems

Authors: Hikmat A. M. Abdeljaber, Md. Alamgir Hossain, Sultan Ahmad, Ahmed Alsanad, Md Alimul Haque, Sudan Jha, Jabeen Nazeer
Published: 2025-10-09
Categories: cs.CR, cs.AI, cs.LG
arXiv: https://arxiv.org/abs/2510.08084v1

The rapid expansion of Internet of Things (IoT) devices has transformed industries and daily life by enabling widespread connectivity and data exchange. However, this increased interconnection has introduced serious security vulnerabilities, making IoT systems more exposed to sophisticated cyber attacks. This study presents a novel ensemble learning architecture designed to improve IoT attack detection. The proposed approach applies advanced machine learning techniques, specifically the Extra Trees Classifier, along with thorough preprocessing and hyperparameter optimization. It is evaluated on several benchmark datasets including CICIoT2023, IoTID20, BotNeTIoT L01, ToN IoT, N BaIoT, and BoT IoT. The results show excellent performance, achieving high recall, accuracy, and precision with very low error rates. These outcomes demonstrate the model efficiency and superiority compared to existing approaches, providing an effective and scalable method for securing IoT environments. This research establishes a solid foundation for future progress in protecting connected devices from evolving cyber threats.

---

### AutoQual: An LLM Agent for Automated Discovery of Interpretable Features for Review Quality Assessment

Authors: Xiaochong Lan, Jie Feng, Yinxing Liu, Xinlei Shi, Yong Li
Published: 2025-10-09
Categories: cs.AI, cs.CL
arXiv: https://arxiv.org/abs/2510.08081v1

Ranking online reviews by their intrinsic quality is a critical task for e-commerce platforms and information services, impacting user experience and business outcomes. However, quality is a domain-dependent and dynamic concept, making its assessment a formidable challenge. Traditional methods relying on hand-crafted features are unscalable across domains and fail to adapt to evolving content patterns, while modern deep learning approaches often produce black-box models that lack interpretability and may prioritize semantics over quality. To address these challenges, we propose AutoQual, an LLM-based agent framework that automates the discovery of interpretable features. While demonstrated on review quality assessment, AutoQual is designed as a general framework for transforming tacit knowledge embedded in data into explicit, computable features. It mimics a human research process, iteratively generating feature hypotheses through reflection, operationalizing them via autonomous tool implementation, and accumulating experience in a persistent memory. We deploy our method on a large-scale online platform with a billion-level user base. Large-scale A/B testing confirms its effectiveness, increasing average reviews viewed per user by 0.79% and the conversion rate of review readers by 0.27%.

---

### Detecting and Mitigating Insertion Hallucination in Video-to-Audio Generation

Authors: Liyang Chen, Hongkai Chen, Yujun Cai, Sifan Li, Qingwen Ye, Yiwei Wang
Published: 2025-10-09
Categories: cs.SD, cs.LG
arXiv: https://arxiv.org/abs/2510.08078v5

Video-to-Audio generation has made remarkable strides in automatically synthesizing sound for video. However, existing evaluation metrics, which focus on semantic and temporal alignment, overlook a critical failure mode: models often generate acoustic events, particularly speech and music, that have no corresponding visual source. We term this phenomenon Insertion Hallucination and identify it as a systemic risk driven by dataset biases, such as the prevalence of off-screen sounds, that remains completely undetected by current metrics. To address this challenge, we first develop a systematic evaluation framework that employs a majority-voting ensemble of multiple audio event detectors. We also introduce two novel metrics to quantify the prevalence and severity of this issue: IH@vid (the fraction of videos with hallucinations) and IH@dur (the fraction of hallucinated duration). Building on this, we introduce HALCON to mitigate IH. HALCON follows a three-stage procedure: it first generates initial audio to expose hallucinated segments, then identifies and masks the corresponding unreliable video features, and finally regenerates the audio using the corrected conditioning. Experiments on several mainstream V2A benchmarks first reveal that state-of-the-art models suffer from severe IH. In contrast, our HALCON method reduces both the prevalence and duration of hallucinations by over 50\% on average, without degrading, and in some cases even improving, conventional metrics for audio quality and temporal synchronization. Our work is the first to formally define, systematically measure, and effectively mitigate Insertion Hallucination, paving the way for more reliable and faithful V2A models.

---

### Multi-Condition Conformal Selection

Authors: Qingyang Hao, Wenbo Liao, Bingyi Jing, Hongxin Wei
Published: 2025-10-09
Categories: cs.AI
arXiv: https://arxiv.org/abs/2510.08075v2

Selecting high-quality candidates from large-scale datasets is critically important in resource-constrained applications such as drug discovery, precision medicine, and the alignment of large language models. While conformal selection methods offer a rigorous solution with False Discovery Rate (FDR) control, their applicability is confined to single-threshold scenarios (i.e., y > c) and overlooks practical needs for multi-condition selection, such as conjunctive or disjunctive conditions. In this work, we propose the Multi-Condition Conformal Selection (MCCS) algorithm, which extends conformal selection to scenarios with multiple conditions. In particular, we introduce a novel nonconformity score with regional monotonicity for conjunctive conditions and a global Benjamini-Hochberg (BH) procedure for disjunctive conditions, thereby establishing finite-sample FDR control with theoretical guarantees. The integration of these components enables the proposed method to achieve rigorous FDR-controlled selection in various multi-condition environments. Extensive experiments validate the superiority of MCCS over baselines, its generalizability across diverse condition combinations, different real-world modalities, and multi-task scalability.

---

### Physics-Driven Spatiotemporal Modeling for AI-Generated Video Detection

Authors: Shuhai Zhang, ZiHao Lian, Jiahao Yang, Daiyuan Li, Guoxuan Pang, Feng Liu, Bo Han, Shutao Li, Mingkui Tan
Published: 2025-10-09
Categories: cs.CV, cs.LG
arXiv: https://arxiv.org/abs/2510.08073v2

AI-generated videos have achieved near-perfect visual realism (e.g., Sora), urgently necessitating reliable detection mechanisms. However, detecting such videos faces significant challenges in modeling high-dimensional spatiotemporal dynamics and identifying subtle anomalies that violate physical laws. In this paper, we propose the first physics-driven AI-generated video detection paradigm based on probability flow conservation principles. Specifically, we propose a statistic called Normalized Spatiotemporal Gradient (NSG), which quantifies the ratio of spatial probability gradients to temporal density changes, explicitly capturing deviations from natural video dynamics. Leveraging pre-trained diffusion models, we develop an NSG estimator through spatial gradients approximation and motion-aware temporal modeling without complex motion decomposition while preserving physical constraints. Building on this, we propose an NSG-based video detection method (NSG-VD) that computes the Maximum Mean Discrepancy (MMD) between NSG features of the test and real videos as a detection metric. Last, we derive an upper bound of NSG feature distances between real and generated videos, proving that generated videos exhibit amplified discrepancies due to distributional shifts. Extensive experiments confirm that NSG-VD outperforms state-of-the-art baselines by 16.00% in Recall and 10.75% in F1-Score, validating the superior performance of NSG-VD. The source code is available at https://github.com/ZSHsh98/NSG-VD.

---

### An Adaptive Multi Agent Bitcoin Trading System

Authors: Aadi Singhi
Published: 2025-10-09
Categories: q-fin.PM, cs.AI, q-fin.CP
arXiv: https://arxiv.org/abs/2510.08068v2

This paper presents a Multi Agent Bitcoin Trading system that utilizes Large Language Models (LLMs) for alpha generation and portfolio management in the cryptocurrencies market. Unlike equities, cryptocurrencies exhibit extreme volatility and are heavily influenced by rapidly shifting market sentiments and regulatory announcements, making them difficult to model using static regression models or neural networks trained solely on historical data. The proposed framework overcomes this by structuring LLMs into specialised agents for technical analysis, sentiment evaluation, decision-making, and performance reflection. The agents improve over time via a novel verbal feedback mechanism where a Reflect agent provides daily and weekly natural-language critiques of trading decisions. These textual evaluations are then injected into future prompts of the agents, allowing them to adjust allocation logic without weight updates or finetuning. Back-testing on Bitcoin price data from July 2024 to April 2025 shows consistent outperformance across market regimes: the Quantitative agent delivered over 30\% higher returns in bullish phases and 15\% overall gains versus buy-and-hold, while the sentiment-driven agent turned sideways markets from a small loss into a gain of over 100\%. Adding weekly feedback further improved total performance by 31\% and reduced bearish losses by 10\%. The results demonstrate that verbal feedback represents a new, scalable, and low-cost approach of tuning LLMs for financial goals.

---

### Towards Real-World Deepfake Detection: A Diverse In-the-wild Dataset of Forgery Faces

Authors: Junyu Shi, Minghui Li, Junguo Zuo, Zhifei Yu, Yipeng Lin, Shengshan Hu, Ziqi Zhou, Yechao Zhang, Wei Wan, Yinzhe Xu, Leo Yu Zhang
Published: 2025-10-09
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.08067v1

Deepfakes, leveraging advanced AIGC (Artificial Intelligence-Generated Content) techniques, create hyper-realistic synthetic images and videos of human faces, posing a significant threat to the authenticity of social media. While this real-world threat is increasingly prevalent, existing academic evaluations and benchmarks for detecting deepfake forgery often fall short to achieve effective application for their lack of specificity, limited deepfake diversity, restricted manipulation techniques.To address these limitations, we introduce RedFace (Real-world-oriented Deepfake Face), a specialized facial deepfake dataset, comprising over 60,000 forged images and 1,000 manipulated videos derived from authentic facial features, to bridge the gap between academic evaluations and real-world necessity. Unlike prior benchmarks, which typically rely on academic methods to generate deepfakes, RedFace utilizes 9 commercial online platforms to integrate the latest deepfake technologies found "in the wild", effectively simulating real-world black-box scenarios.Moreover, RedFace's deepfakes are synthesized using bespoke algorithms, allowing it to capture diverse and evolving methods used by real-world deepfake creators. Extensive experimental results on RedFace (including cross-domain, intra-domain, and real-world social network dissemination simulations) verify the limited practicality of existing deepfake detection schemes against real-world applications. We further perform a detailed analysis of the RedFace dataset, elucidating the reason of its impact on detection performance compared to conventional datasets. Our dataset is available at: https://github.com/kikyou-220/RedFace.

---

### Attribution-by-design: Ensuring Inference-Time Provenance in Generative Music Systems

Authors: Fabio Morreale, Wiebke Hutiri, Joan Serrà, Alice Xiang, Yuki Mitsufuji
Published: 2025-10-09
Categories: cs.SD, cs.AI, cs.HC
arXiv: https://arxiv.org/abs/2510.08062v1

The rise of AI-generated music is diluting royalty pools and revealing structural flaws in existing remuneration frameworks, challenging the well-established artist compensation systems in the music industry. Existing compensation solutions, such as piecemeal licensing agreements, lack scalability and technical rigour, while current data attribution mechanisms provide only uncertain estimates and are rarely implemented in practice. This paper introduces a framework for a generative music infrastructure centred on direct attribution, transparent royalty distribution, and granular control for artists and rights' holders. We distinguish ontologically between the training set and the inference set, which allows us to propose two complementary forms of attribution: training-time attribution and inference-time attribution. We here favour inference-time attribution, as it enables direct, verifiable compensation whenever an artist's catalogue is used to condition a generated output. Besides, users benefit from the ability to condition generations on specific songs and receive transparent information about attribution and permitted usage. Our approach offers an ethical and practical solution to the pressing need for robust compensation mechanisms in the era of AI-generated music, ensuring that provenance and fairness are embedded at the core of generative systems.

---

### A class-driven hierarchical ResNet for classification of multispectral remote sensing images

Authors: Giulio Weikmann, Gianmarco Perantoni, Lorenzo Bruzzone
Published: 2025-10-09
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.08060v1

This work presents a multitemporal class-driven hierarchical Residual Neural Network (ResNet) designed for modelling the classification of Time Series (TS) of multispectral images at different semantical class levels. The architecture consists of a modification of the ResNet where we introduce additional branches to perform the classification at the different hierarchy levels and leverage on hierarchy-penalty maps to discourage incoherent hierarchical transitions within the classification. In this way, we improve the discrimination capabilities of classes at different levels of semantic details and train a modular architecture that can be used as a backbone network for introducing new specific classes and additional tasks considering limited training samples available. We exploit the class-hierarchy labels to train efficiently the different layers of the architecture, allowing the first layers to train faster on the first levels of the hierarchy modeling general classes (i.e., the macro-classes) and the intermediate classes, while using the last ones to discriminate more specific classes (i.e., the micro-classes). In this way, the targets are constrained in following the hierarchy defined, improving the classification of classes at the most detailed level. The proposed modular network has intrinsic adaptation capability that can be obtained through fine tuning. The experimental results, obtained on two tiles of the Amazonian Forest on 12 monthly composites of Sentinel 2 images acquired during 2019, demonstrate the effectiveness of the hierarchical approach in both generalizing over different hierarchical levels and learning discriminant features for an accurate classification at the micro-class level on a new target area, with a better representation of the minoritarian classes.

---

### Mitigating Subject Dependency in EEG Decoding with Subject-Specific Low-Rank Adapters

Authors: Timon Klein, Piotr Minakowski, Sebastian Sager, Steffen Schotthöfer
Published: 2025-10-09
Categories: cs.LG
arXiv: https://arxiv.org/abs/2510.08059v2

Subject-specific distribution shifts represent a fundamental obstacle to developing foundation models for brain decoding. We propose the Subject-Specific Low-Rank Adapter (SuLoRA), a drop-in replacement for standard linear or convolutional layers that captures inter-subject variability by decomposing weights into a shared, subject-invariant component and a lightweight, low-rank correction unique to each subject. This explicit separation enables existing architectures to become robust to subject shifts without architectural redesign. We evaluate SuLoRA on MEG speech perception and EEG motor imagery tasks across CNN and transformer architectures. In the speech decoding task, SuLoRA exceeds the baseline performance with half of the parameters. On motor imagery dataset, SuLoRA outperforms both subject-agnostic models and independently trained subject-specific models. SuLoRA offers a practical path towards effective cross-subject foundation models for brain signal applications.

---

### FedDTRE: Federated Dialogue Generation Models Powered by Trustworthiness Evaluation

Authors: Shule Lu, Lingxiang Wang, Sijia Wen, Ziwei Wang, Hainan Zhang
Published: 2025-10-09
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2510.08058v1

With the rapid development of artificial intelligence, dialogue systems have become a prominent form of human-computer interaction. However, traditional centralized or fully local training approaches face challenges in balancing privacy preservation and personalization due to data privacy concerns and heterogeneous device capabilities. Federated learning, as a representative distributed paradigm, offers a promising solution. However, existing methods often suffer from overfitting under limited client data and tend to forget global information after multiple training rounds, leading to poor generalization. To address these issues, we propose FedDTRE, a Federated adaptive aggregation strategy for Dialogue generation based on Trustworthiness Evaluation. Instead of directly replacing local models with the global model, FedDTRE leverages trustworthiness scores of both global and local models on a fairness-oriented evaluation dataset to dynamically regulate the global model's contribution during local updates. Experimental results demonstrate that FedDTRE can improve dialogue model performance and enhance the quality of dialogue generation.

---

### From Tokens to Layers: Redefining Stall-Free Scheduling for MoE Serving with Layered Prefill

Authors: Gunjun Lee, Jiwon Kim, Jaiyoung Park, Younjoo Lee, Jung Ho Ahn
Published: 2025-10-09
Categories: cs.LG, cs.DC
arXiv: https://arxiv.org/abs/2510.08055v2

Large Language Model (LLM) inference in production must meet stringent service-level objectives for both time-to-first-token (TTFT) and time-between-token (TBT) while maximizing throughput under fixed compute, memory, and interconnect budgets. Modern serving systems adopt stall-free scheduling techniques such as chunked prefill, which splits the processing of long prompts along the token dimension and interleaves prefill with ongoing decode iterations. While effective at stabilizing TBT, chunked prefill incurs substantial overhead in Mixture-of-Experts (MoE) models: redundant expert weight loads increase memory traffic by up to 39% and inflate energy consumption. We propose layered prefill, a new scheduling paradigm that treats transformer layer groups as the primary scheduling unit, specifically targeting MoE serving. By vertically partitioning the model into contiguous layer groups and interleaving prefill and decode across the groups, layered prefill sustains stall-free decoding while eliminating chunk-induced MoE weight reloads. It reduces off-chip bandwidth demand, lowering TTFT by up to 70%, end-to-end latency by 41% and per-token energy by up to 22%. Evaluations show that layered prefill consistently improves the TTFT--TBT Pareto frontier over chunked prefill, reducing expert-load traffic and energy cost while maintaining stall-free decoding. Overall, shifting the scheduling axis from tokens to layers unlocks a new operating regime for high-efficiency, energy-aware MoE serving in co-located environments.

---

### RetouchLLM: Training-free Code-based Image Retouching with Vision Language Models

Authors: Moon Ye-Bin, Roy Miles, Tae-Hyun Oh, Ismail Elezi, Jiankang Deng
Published: 2025-10-09
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.08054v2

Image retouching not only enhances visual quality but also serves as a means of expressing personal preferences and emotions. However, existing learning-based approaches require large-scale paired data and operate as black boxes, making the retouching process opaque and limiting their adaptability to handle diverse, user- or image-specific adjustments. In this work, we propose RetouchLLM, a training-free white-box image retouching system, which requires no training data and performs interpretable, code-based retouching directly on high-resolution images. Our framework progressively enhances the image in a manner similar to how humans perform multi-step retouching, allowing exploration of diverse adjustment paths. It comprises of two main modules: a visual critic that identifies differences between the input and reference images, and a code generator that produces executable codes. Experiments demonstrate that our approach generalizes well across diverse retouching styles, while natural language-based user interaction enables interpretable and controllable adjustments tailored to user intent.

---

### RASALoRE: Region Aware Spatial Attention with Location-based Random Embeddings for Weakly Supervised Anomaly Detection in Brain MRI Scans

Authors: Bheeshm Sharma, Karthikeyan Jaganathan, Balamurugan Palaniappan
Published: 2025-10-09
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.08052v2

Weakly Supervised Anomaly detection (WSAD) in brain MRI scans is an important challenge useful to obtain quick and accurate detection of brain anomalies when precise pixel-level anomaly annotations are unavailable and only weak labels (e.g., slice-level) are available. In this work, we propose RASALoRE: Region Aware Spatial Attention with Location-based Random Embeddings, a novel two-stage WSAD framework. In the first stage, we introduce a Discriminative Dual Prompt Tuning (DDPT) mechanism that generates high-quality pseudo weak masks based on slice-level labels, serving as coarse localization cues. In the second stage, we propose a segmentation network with a region-aware spatial attention mechanism that relies on fixed location-based random embeddings. This design enables the model to effectively focus on anomalous regions. Our approach achieves state-of-the-art anomaly detection performance, significantly outperforming existing WSAD methods while utilizing less than 8 million parameters. Extensive evaluations on the BraTS20, BraTS21, BraTS23, and MSD datasets demonstrate a substantial performance improvement coupled with a significant reduction in computational complexity. Code is available at: https://github.com/BheeshmSharma/RASALoRE-BMVC-2025/.

---

### A Survey of Process Reward Models: From Outcome Signals to Process Supervisions for Large Language Models

Authors: Congmin Zheng, Jiachen Zhu, Zhuoying Ou, Yuxiang Chen, Kangning Zhang, Rong Shan, Zeyu Zheng, Mengyue Yang, Jianghao Lin, Yong Yu, Weinan Zhang
Published: 2025-10-09
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2510.08049v3

Although Large Language Models (LLMs) exhibit advanced reasoning ability, conventional alignment remains largely dominated by outcome reward models (ORMs) that judge only final answers. Process Reward Models(PRMs) address this gap by evaluating and guiding reasoning at the step or trajectory level. This survey provides a systematic overview of PRMs through the full loop: how to generate process data, build PRMs, and use PRMs for test-time scaling and reinforcement learning. We summarize applications across math, code, text, multimodal reasoning, robotics, and agents, and review emerging benchmarks. Our goal is to clarify design spaces, reveal open challenges, and guide future research toward fine-grained, robust reasoning alignment.

---

### TaoSR-AGRL: Adaptive Guided Reinforcement Learning Framework for E-commerce Search Relevance

Authors: Jianhui Yang, Yiming Jin, Pengkun Jiao, Chenhe Dong, Zerui Huang, Shaowei Yao, Xiaojiang Zhou, Dan Ou, Haihong Tang
Published: 2025-10-09
Categories: cs.IR, cs.AI, cs.CL
arXiv: https://arxiv.org/abs/2510.08048v4

Query-product relevance prediction is fundamental to e-commerce search and has become even more critical in the era of AI-powered shopping, where semantic understanding and complex reasoning directly shape the user experience and business conversion. Large Language Models (LLMs) enable generative, reasoning-based approaches, typically aligned via supervised fine-tuning (SFT) or preference optimization methods like Direct Preference Optimization (DPO). However, the increasing complexity of business rules and user queries exposes the inability of existing methods to endow models with robust reasoning capacity for long-tail and challenging cases. Efforts to address this via reinforcement learning strategies like Group Relative Policy Optimization (GRPO) often suffer from sparse terminal rewards, offering insufficient guidance for multi-step reasoning and slowing convergence. To address these challenges, we propose TaoSR-AGRL, an Adaptive Guided Reinforcement Learning framework for LLM-based relevance prediction in Taobao Search Relevance. TaoSR-AGRL introduces two key innovations: (1) Rule-aware Reward Shaping, which decomposes the final relevance judgment into dense, structured rewards aligned with domain-specific relevance criteria; and (2) Adaptive Guided Replay, which identifies low-accuracy rollouts during training and injects targeted ground-truth guidance to steer the policy away from stagnant, rule-violating reasoning patterns toward compliant trajectories. TaoSR-AGRL was evaluated on large-scale real-world datasets and through online side-by-side human evaluations on Taobao Search. It consistently outperforms DPO and standard GRPO baselines in offline experiments, improving relevance accuracy, rule adherence, and training stability. The model trained with TaoSR-AGRL has been successfully deployed in the main search scenario on Taobao, serving hundreds of millions of users.

---

### Pseudo2Real: Task Arithmetic for Pseudo-Label Correction in Automatic Speech Recognition

Authors: Yi-Cheng Lin, Yu-Hsuan Li Liang, Hsuan Su, Tzu-Quan Lin, Shang-Tse Chen, Yun-Nung Chen, Hung-yi Lee
Published: 2025-10-09
Categories: eess.AS, cs.CL
arXiv: https://arxiv.org/abs/2510.08047v2

Robust ASR under domain shift is crucial because real-world systems encounter unseen accents and domains with limited labeled data. Although pseudo-labeling offers a practical workaround, it often introduces systematic, accent-specific errors that filtering fails to fix. We ask: How can we correct these recurring biases without target ground truth? We propose a simple parameter-space correction: in a source domain containing both real and pseudo-labeled data, two ASR models are fine-tuned from the same initialization, one on ground-truth labels and the other on pseudo-labels, and their weight difference forms a correction vector that captures pseudo-label biases. When applied to a pseudo-labeled target model, this vector enhances recognition, achieving up to a 35% relative Word Error Rate (WER) reduction on AfriSpeech-200 across ten African accents with the Whisper tiny model.

---

### LinguaSim: Interactive Multi-Vehicle Testing Scenario Generation via Natural Language Instruction Based on Large Language Models

Authors: Qingyuan Shi, Qingwen Meng, Hao Cheng, Qing Xu, Jianqiang Wang
Published: 2025-10-09
Categories: cs.AI
arXiv: https://arxiv.org/abs/2510.08046v1

The generation of testing and training scenarios for autonomous vehicles has drawn significant attention. While Large Language Models (LLMs) have enabled new scenario generation methods, current methods struggle to balance command adherence accuracy with the realism of real-world driving environments. To reduce scenario description complexity, these methods often compromise realism by limiting scenarios to 2D, or open-loop simulations where background vehicles follow predefined, non-interactive behaviors. We propose LinguaSim, an LLM-based framework that converts natural language into realistic, interactive 3D scenarios, ensuring both dynamic vehicle interactions and faithful alignment between the input descriptions and the generated scenarios. A feedback calibration module further refines the generation precision, improving fidelity to user intent. By bridging the gap between natural language and closed-loop, interactive simulations, LinguaSim constrains adversarial vehicle behaviors using both the scenario description and the autonomous driving model guiding them. This framework facilitates the creation of high-fidelity scenarios that enhance safety testing and training. Experiments show LinguaSim can generate scenarios with varying criticality aligned with different natural language descriptions (ACT: 0.072 s for dangerous vs. 3.532 s for safe descriptions; comfortability: 0.654 vs. 0.764), and its refinement module effectively reduces excessive aggressiveness in LinguaSim's initial outputs, lowering the crash rate from 46.9% to 6.3% to better match user intentions.

---

### Verifying Quantized GNNs With Readout Is Decidable But Highly Intractable

Authors: Artem Chernobrovkin, Marco Sälzer, François Schwarzentruber, Nicolas Troquard
Published: 2025-10-09
Categories: cs.LO, cs.AI, cs.CC, cs.LG
arXiv: https://arxiv.org/abs/2510.08045v2

We introduce a logical language for reasoning about quantized aggregate-combine graph neural networks with global readout (ACR-GNNs). We provide a logical characterization and use it to prove that verification tasks for quantized GNNs with readout are (co)NEXPTIME-complete. This result implies that the verification of quantized GNNs is computationally intractable, prompting substantial research efforts toward ensuring the safety of GNN-based systems. We also experimentally demonstrate that quantized ACR-GNN models are lightweight while maintaining good accuracy and generalization capabilities with respect to non-quantized models.

---

### Towards Reliable LLM-based Robot Planning via Combined Uncertainty Estimation

Authors: Shiyuan Yin, Chenjia Bai, Zihao Zhang, Junwei Jin, Xinxin Zhang, Chi Zhang, Xuelong Li
Published: 2025-10-09
Categories: cs.RO, cs.AI
arXiv: https://arxiv.org/abs/2510.08044v1

Large language models (LLMs) demonstrate advanced reasoning abilities, enabling robots to understand natural language instructions and generate high-level plans with appropriate grounding. However, LLM hallucinations present a significant challenge, often leading to overconfident yet potentially misaligned or unsafe plans. While researchers have explored uncertainty estimation to improve the reliability of LLM-based planning, existing studies have not sufficiently differentiated between epistemic and intrinsic uncertainty, limiting the effectiveness of uncertainty estimation. In this paper, we present Combined Uncertainty estimation for Reliable Embodied planning (CURE), which decomposes the uncertainty into epistemic and intrinsic uncertainty, each estimated separately. Furthermore, epistemic uncertainty is subdivided into task clarity and task familiarity for more accurate evaluation. The overall uncertainty assessments are obtained using random network distillation and multi-layer perceptron regression heads driven by LLM features. We validated our approach in two distinct experimental settings: kitchen manipulation and tabletop rearrangement experiments. The results show that, compared to existing methods, our approach yields uncertainty estimates that are more closely aligned with the actual execution outcomes.

---

### Climate Knowledge in Large Language Models

Authors: Ivan Kuznetsov, Jacopo Grassi, Dmitrii Pantiukhin, Boris Shapkin, Thomas Jung, Nikolay Koldunov
Published: 2025-10-09
Categories: cs.CL, cs.LG, physics.ao-ph
arXiv: https://arxiv.org/abs/2510.08043v1

Large language models (LLMs) are increasingly deployed for climate-related applications, where understanding internal climatological knowledge is crucial for reliability and misinformation risk assessment. Despite growing adoption, the capacity of LLMs to recall climate normals from parametric knowledge remains largely uncharacterized. We investigate the capacity of contemporary LLMs to recall climate normals without external retrieval, focusing on a prototypical query: mean July 2-m air temperature 1991-2020 at specified locations. We construct a global grid of queries at 1° resolution land points, providing coordinates and location descriptors, and validate responses against ERA5 reanalysis. Results show that LLMs encode non-trivial climate structure, capturing latitudinal and topographic patterns, with root-mean-square errors of 3-6 °C and biases of $\pm$1 °C. However, spatially coherent errors remain, particularly in mountains and high latitudes. Performance degrades sharply above 1500 m, where RMSE reaches 5-13 °C compared to 2-4 °C at lower elevations. We find that including geographic context (country, city, region) reduces errors by 27% on average, with larger models being most sensitive to location descriptors. While models capture the global mean magnitude of observed warming between 1950-1974 and 2000-2024, they fail to reproduce spatial patterns of temperature change, which directly relate to assessing climate change. This limitation highlights that while LLMs may capture present-day climate distributions, they struggle to represent the regional and local expression of long-term shifts in temperature essential for understanding climate dynamics. Our evaluation framework provides a reproducible benchmark for quantifying parametric climate knowledge in LLMs and complements existing climate communication assessments.

---

### ChatGPT as a Translation Engine: A Case Study on Japanese-English

Authors: Vincent Michael Sutanto, Giovanni Gatti De Giacomo, Toshiaki Nakazawa, Masaru Yamada
Published: 2025-10-09
Categories: cs.CL
arXiv: https://arxiv.org/abs/2510.08042v1

This study investigates ChatGPT for Japanese-English translation, exploring simple and enhanced prompts and comparing against commercially available translation engines. Performing both automatic and MQM-based human evaluations, we found that document-level translation outperforms sentence-level translation for ChatGPT. On the other hand, we were not able to determine if enhanced prompts performed better than simple prompts in our experiments. We also discovered that ChatGPT-3.5 was preferred by automatic evaluation, but a tradeoff exists between accuracy (ChatGPT-3.5) and fluency (ChatGPT-4). Lastly, ChatGPT yields competitive results against two widely-known translation systems.

---

### MRI-derived quantification of hepatic vessel-to-volume ratios in chronic liver disease using a deep learning approach

Authors: Alexander Herold, Daniel Sobotka, Lucian Beer, Nina Bastati, Sarah Poetter-Lang, Michael Weber, Thomas Reiberger, Mattias Mandorfer, Georg Semmler, Benedikt Simbrunner, Barbara D. Wichtmann, Sami A. Ba-Ssalamah et al.
Published: 2025-10-09
Categories: physics.med-ph, cs.AI
arXiv: https://arxiv.org/abs/2510.08039v1

Background: We aimed to quantify hepatic vessel volumes across chronic liver disease stages and healthy controls using deep learning-based magnetic resonance imaging (MRI) analysis, and assess correlations with biomarkers for liver (dys)function and fibrosis/portal hypertension. Methods: We assessed retrospectively healthy controls, non-advanced and advanced chronic liver disease (ACLD) patients using a 3D U-Net model for hepatic vessel segmentation on portal venous phase gadoxetic acid-enhanced 3-T MRI. Total (TVVR), hepatic (HVVR), and intrahepatic portal vein-to-volume ratios (PVVR) were compared between groups and correlated with: albumin-bilirubin (ALBI) and model for end-stage liver disease-sodium (MELD-Na) score, and fibrosis/portal hypertension (Fibrosis-4 (FIB-4) score, liver stiffness measurement (LSM), hepatic venous pressure gradient (HVPG), platelet count (PLT), and spleen volume). Results: We included 197 subjects, aged 54.9 $\pm$ 13.8 years (mean $\pm$ standard deviation), 111 males (56.3\%): 35 healthy controls, 44 non-ACLD, and 118 ACLD patients. TVVR and HVVR were highest in controls (3.9; 2.1), intermediate in non-ACLD (2.8; 1.7), and lowest in ACLD patients (2.3; 1.0) ($p \leq 0.001$). PVVR was reduced in both non-ACLD and ACLD patients (both 1.2) compared to controls (1.7) ($p \leq 0.001$), but showed no difference between CLD groups ($p = 0.999$). HVVR significantly correlated indirectly with FIB-4, ALBI, MELD-Na, LSM, and spleen volume ($ρ$ ranging from -0.27 to -0.40), and directly with PLT ($ρ= 0.36$). TVVR and PVVR showed similar but weaker correlations. Conclusions: Deep learning-based hepatic vessel volumetry demonstrated differences between healthy liver and chronic liver disease stages and shows correlations with established markers of disease severity.

---

### AILoRA: Function-Aware Asymmetric Initialization for Low-Rank Adaptation of Large Language Models

Authors: Xiaoshuang Ji, Zhendong Zhao, Xiaoyan Gu, Xiaojun Chen, Xin Zhao, Zeyao Liu
Published: 2025-10-09
Categories: cs.AI
arXiv: https://arxiv.org/abs/2510.08034v1

Parameter-efficient finetuning (PEFT) aims to mitigate the substantial computational and memory overhead involved in adapting large-scale pretrained models to diverse downstream tasks. Among numerous PEFT strategies, Low-Rank Adaptation (LoRA) has emerged as one of the most widely adopted approaches due to its robust empirical performance and low implementation complexity. In practical deployment, LoRA is typically applied to the $W^Q$ and $W^V$ projection matrices of self-attention modules, enabling an effective trade-off between model performance and parameter efficiency. While LoRA has achieved considerable empirical success, it still encounters challenges such as suboptimal performance and slow convergence. To address these limitations, we introduce \textbf{AILoRA}, a novel parameter-efficient method that incorporates function-aware asymmetric low-rank priors. Our empirical analysis reveals that the projection matrices $W^Q$ and $W^V$ in the self-attention mechanism exhibit distinct parameter characteristics, stemming from their functional differences. Specifically, $W^Q$ captures task-specific semantic space knowledge essential for attention distributions computation, making its parameters highly sensitive to downstream task variations. In contrast, $W^V$ encodes token-level feature representations that tend to remain stable across tasks and layers. Leveraging these insights, AILoRA performs a function-aware initialization by injecting the principal components of $W^Q$ to retain task-adaptive capacity, and the minor components of $W^V$ to preserve generalizable feature representations. This asymmetric initialization strategy enables LoRA modules to better capture the specialized roles of attention parameters, thereby enhancing both finetuning performance and convergence efficiency.

---

### PEAR: Phase Entropy Aware Reward for Efficient Reasoning

Authors: Chen Huang, Wei Lu, Wenxuan Zhang
Published: 2025-10-09
Categories: cs.AI
arXiv: https://arxiv.org/abs/2510.08026v2

Large Reasoning Models (LRMs) have achieved impressive performance on complex reasoning tasks by generating detailed chain-of-thought (CoT) explanations. However, these responses are often excessively long, containing redundant reasoning steps that inflate inference cost and reduce usability. Controlling the length of generated reasoning without sacrificing accuracy remains an open challenge. Through a systematic empirical analysis, we reveal a consistent positive correlation between model entropy and response length at different reasoning stages across diverse LRMs: the thinking phase exhibits higher entropy, reflecting exploratory behavior of longer responses, while the final answer phase shows lower entropy, indicating a more deterministic solution. This observation suggests that entropy at different reasoning stages can serve as a control knob for balancing conciseness and performance. Based on this insight, this paper introduces Phase Entropy Aware Reward (PEAR), a reward mechanism that incorporating phase-dependent entropy into the reward design. Instead of treating all tokens uniformly, PEAR penalize excessive entropy during the thinking phase and allowing moderate exploration at the final answer phase, which encourages models to generate concise reasoning traces that retain sufficient flexibility to solve the task correctly. This enables adaptive control of response length without relying on explicit length targets or rigid truncation rules. Extensive experiments across four benchmarks demonstrate that PEAR consistently reduces response length while sustaining competitive accuracy across model scales. In addition, PEAR demonstrates strong out-of-distribution (OOD) robustness beyond the training distribution. Our code is available at: https://github.com/iNLP-Lab/PEAR.

---

### Do We Really Need Permutations? Impact of Model Width on Linear Mode Connectivity

Authors: Akira Ito, Masanori Yamada, Daiki Chijiwa, Atsutoshi Kumagai
Published: 2025-10-09
Categories: cs.LG
arXiv: https://arxiv.org/abs/2510.08023v3

Recently, Ainsworth et al. empirically demonstrated that, given two independently trained models, applying a parameter permutation that preserves the input-output behavior allows the two models to be connected by a low-loss linear path. When such a path exists, the models are said to achieve linear mode connectivity (LMC). Prior studies, including Ainsworth et al.(2023), have reported that achieving LMC requires not only an appropriate permutation search but also sufficiently wide models (e.g., a 32 $\times$ width multiplier for ResNet-20). This is broadly believed to be because increasing the model width ensures a large enough space of candidate permutations, increasing the chance of finding one that yields LMC. In this work, we empirically demonstrate that, even without any permutations, simply widening the models is sufficient for achieving LMC when using a suitable softmax temperature calibration. We further explain why this phenomenon arises by analyzing intermediate layer outputs. Specifically, we introduce layerwise exponentially weighted connectivity (LEWC), which states that the output of each layer of the merged model can be represented as an exponentially weighted sum of the outputs of the corresponding layers of the original models. Consequently the merged model's output matches that of an ensemble of the original models, facilitating LMC. To the best of our knowledge, this work is the first to show that widening the model not only facilitates nonlinear mode connectivity, as suggested in prior research, but also significantly increases the possibility of achieving linear mode connectivity.

---

### FastUMI-100K: Advancing Data-driven Robotic Manipulation with a Large-scale UMI-style Dataset

Authors: Kehui Liu, Zhongjie Jia, Yang Li, Zhaxizhuoma, Pengan Chen, Song Liu, Xin Liu, Pingrui Zhang, Haoming Song, Xinyi Ye, Nieqing Cao, Zhigang Wang et al.
Published: 2025-10-09
Categories: cs.RO, cs.AI
arXiv: https://arxiv.org/abs/2510.08022v1

Data-driven robotic manipulation learning depends on large-scale, high-quality expert demonstration datasets. However, existing datasets, which primarily rely on human teleoperated robot collection, are limited in terms of scalability, trajectory smoothness, and applicability across different robotic embodiments in real-world environments. In this paper, we present FastUMI-100K, a large-scale UMI-style multimodal demonstration dataset, designed to overcome these limitations and meet the growing complexity of real-world manipulation tasks. Collected by FastUMI, a novel robotic system featuring a modular, hardware-decoupled mechanical design and an integrated lightweight tracking system, FastUMI-100K offers a more scalable, flexible, and adaptable solution to fulfill the diverse requirements of real-world robot demonstration data. Specifically, FastUMI-100K contains over 100K+ demonstration trajectories collected across representative household environments, covering 54 tasks and hundreds of object types. Our dataset integrates multimodal streams, including end-effector states, multi-view wrist-mounted fisheye images and textual annotations. Each trajectory has a length ranging from 120 to 500 frames. Experimental results demonstrate that FastUMI-100K enables high policy success rates across various baseline algorithms, confirming its robustness, adaptability, and real-world applicability for solving complex, dynamic manipulation challenges. The source code and dataset will be released in this link https://github.com/MrKeee/FastUMI-100K.

---

### RayFusion: Ray Fusion Enhanced Collaborative Visual Perception

Authors: Shaohong Wang, Bin Lu, Xinyu Xiao, Hanzhi Zhong, Bowen Pang, Tong Wang, Zhiyu Xiang, Hangguan Shan, Eryun Liu
Published: 2025-10-09
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.08017v1

Collaborative visual perception methods have gained widespread attention in the autonomous driving community in recent years due to their ability to address sensor limitation problems. However, the absence of explicit depth information often makes it difficult for camera-based perception systems, e.g., 3D object detection, to generate accurate predictions. To alleviate the ambiguity in depth estimation, we propose RayFusion, a ray-based fusion method for collaborative visual perception. Using ray occupancy information from collaborators, RayFusion reduces redundancy and false positive predictions along camera rays, enhancing the detection performance of purely camera-based collaborative perception systems. Comprehensive experiments show that our method consistently outperforms existing state-of-the-art models, substantially advancing the performance of collaborative visual perception. The code is available at https://github.com/wangsh0111/RayFusion.

---

### Backdoor Vectors: a Task Arithmetic View on Backdoor Attacks and Defenses

Authors: Stanisław Pawlak, Jan Dubiński, Daniel Marczak, Bartłomiej Twardowski
Published: 2025-10-09
Categories: cs.LG, cs.AI, cs.CR
arXiv: https://arxiv.org/abs/2510.08016v1

Model merging (MM) recently emerged as an effective method for combining large deep learning models. However, it poses significant security risks. Recent research shows that it is highly susceptible to backdoor attacks, which introduce a hidden trigger into a single fine-tuned model instance that allows the adversary to control the output of the final merged model at inference time. In this work, we propose a simple framework for understanding backdoor attacks by treating the attack itself as a task vector. $Backdoor\ Vector\ (BV)$ is calculated as the difference between the weights of a fine-tuned backdoored model and fine-tuned clean model. BVs reveal new insights into attacks understanding and a more effective framework to measure their similarity and transferability. Furthermore, we propose a novel method that enhances backdoor resilience through merging dubbed $Sparse\ Backdoor\ Vector\ (SBV)$ that combines multiple attacks into a single one. We identify the core vulnerability behind backdoor threats in MM: $inherent\ triggers$ that exploit adversarial weaknesses in the base model. To counter this, we propose $Injection\ BV\ Subtraction\ (IBVS)$ - an assumption-free defense against backdoors in MM. Our results show that SBVs surpass prior attacks and is the first method to leverage merging to improve backdoor effectiveness. At the same time, IBVS provides a lightweight, general defense that remains effective even when the backdoor threat is entirely unknown.

---

### Unsupervised Radio Map Construction in Mixed LoS/NLoS Indoor Environments

Authors: Zheng Xing, Junting Chen
Published: 2025-10-09
Categories: cs.LG
arXiv: https://arxiv.org/abs/2510.08015v1

Radio maps are essential for enhancing wireless communications and localization. However, existing methods for constructing radio maps typically require costly calibration processes to collect location-labeled channel state information (CSI) datasets. This paper aims to recover the data collection trajectory directly from the channel propagation sequence, eliminating the need for location calibration. The key idea is to employ a hidden Markov model (HMM)-based framework to conditionally model the channel propagation matrix, while simultaneously modeling the location correlation in the trajectory. The primary challenges involve modeling the complex relationship between channel propagation in multiple-input multiple-output (MIMO) networks and geographical locations, and addressing both line-of-sight (LOS) and non-line-of-sight (NLOS) indoor conditions. In this paper, we propose an HMM-based framework that jointly characterizes the conditional propagation model and the evolution of the user trajectory. Specifically, the channel propagation in MIMO networks is modeled separately in terms of power, delay, and angle, with distinct models for LOS and NLOS conditions. The user trajectory is modeled using a Gaussian-Markov model. The parameters for channel propagation, the mobility model, and LOS/NLOS classification are optimized simultaneously. Experimental validation using simulated MIMO-Orthogonal Frequency-Division Multiplexing (OFDM) networks with a multi-antenna uniform linear arrays (ULA) configuration demonstrates that the proposed method achieves an average localization accuracy of 0.65 meters in an indoor environment, covering both LOS and NLOS regions. Moreover, the constructed radio map enables localization with a reduced error compared to conventional supervised methods, such as k-nearest neighbors (KNN), support vector machine (SVM), and deep neural network (DNN).

---

### Accelerated Evolving Set Processes for Local PageRank Computation

Authors: Binbin Huang, Luo Luo, Yanghua Xiao, Deqing Yang, Baojian Zhou
Published: 2025-10-09
Categories: cs.LG
arXiv: https://arxiv.org/abs/2510.08010v4

This work proposes a novel framework based on nested evolving set processes to accelerate Personalized PageRank (PPR) computation. At each stage of the process, we employ a localized inexact proximal point iteration to solve a simplified linear system. We show that the time complexity of such localized methods is upper bounded by $\min\{\tilde{\mathcal{O}}(R^2/ε^2), \tilde{\mathcal{O}}(m)\}$ to obtain an $ε$-approximation of the PPR vector, where $m$ denotes the number of edges in the graph and $R$ is a constant defined via nested evolving set processes. Furthermore, the algorithms induced by our framework require solving only $\tilde{\mathcal{O}}(1/\sqrtα)$ such linear systems, where $α$ is the damping factor. When $1/ε^2\ll m$, this implies the existence of an algorithm that computes an $\ epsilon $-approximation of the PPR vector with an overall time complexity of $\tilde{\mathcal{O}}\left(R^2 / (\sqrtαε^2)\right)$, independent of the underlying graph size. Our result resolves an open conjecture from existing literature. Experimental results on real-world graphs validate the efficiency of our methods, demonstrating significant convergence in the early stages.

---

### Language Models Do Not Embed Numbers Continuously

Authors: Alex O. Davies, Roussel Nzoyem, Nirav Ajmeri, Telmo M. Silva Filho
Published: 2025-10-09
Categories: cs.AI, cs.LG
arXiv: https://arxiv.org/abs/2510.08009v1

Recent research has extensively studied how large language models manipulate integers in specific arithmetic tasks, and on a more fundamental level, how they represent numeric values. These previous works have found that language model embeddings can be used to reconstruct the original values, however, they do not evaluate whether language models actually model continuous values as continuous. Using expected properties of the embedding space, including linear reconstruction and principal component analysis, we show that language models not only represent numeric spaces as non-continuous but also introduce significant noise. Using models from three major providers (OpenAI, Google Gemini and Voyage AI), we show that while reconstruction is possible with high fidelity ($R^2 \geq 0.95$), principal components only explain a minor share of variation within the embedding space. This indicates that many components within the embedding space are orthogonal to the simple numeric input space. Further, both linear reconstruction and explained variance suffer with increasing decimal precision, despite the ordinal nature of the input space being fundamentally unchanged. The findings of this work therefore have implications for the many areas where embedding models are used, in-particular where high numerical precision, large magnitudes or mixed-sign values are common.

---

### Beyond Sunk Costs: Boosting LLM Pre-training Efficiency via Orthogonal Growth of Mixture-of-Experts

Authors: Ruizhe Wang, Yucheng Ding, Xiao Liu, Yaoxiang Wang, Peng Cheng, Baining Guo, Zhengjun Zha, Yeyun Gong
Published: 2025-10-09
Categories: cs.LG
arXiv: https://arxiv.org/abs/2510.08008v2

As the computational demands for pre-training Large Language Models (LLMs) continue to surge, the need for efficient training paradigms becomes critical. Despite the vast resources already invested in existing pre-trained checkpoints, these assets often remain under-leveraged due to architectural limitations. We introduce an "orthogonal growth" strategy designed to "recycle" these checkpoints by strategically expanding their parameters prior to continued training. Our method focuses on optimizing converged Mixture-of-Experts (MoE) models through two dimensions: interpositional layer copying for increased depth and noisy expert duplication for expanded width. Through extensive scaling laws analysis, we demonstrate a strong positive correlation between the "sunk cost" (prior investment) and the final model accuracy. Empirical results on models up to 70B parameters and 1T tokens show that our recycling approach yields a 10.6% accuracy improvement compared to training from scratch under identical extra compute budgets. This work provides a cost-effective blueprint for sustainable large-scale LLM development.

---

### Past, Present, and Future of Bug Tracking in the Generative AI Era

Authors: Utku Boran Torun, Mehmet Taha Demircan, Mahmut Furkan Gön, Eray Tüzün
Published: 2025-10-09
Categories: cs.SE, cs.AI
arXiv: https://arxiv.org/abs/2510.08005v3

Traditional bug-tracking systems rely heavily on manual reporting, reproduction, classification, and resolution, involving multiple stakeholders such as end users, customer support, developers, and testers. This division of responsibilities requires substantial coordination and human effort, widens the communication gap between non-technical users and developers, and significantly slows the process from bug discovery to deployment. Moreover, current solutions are highly asynchronous, often leaving users waiting long periods before receiving any feedback. In this paper, we examine the evolution of bug-tracking practices, from early paper-based methods to today's web-based platforms, and present a forward-looking vision of an AI-powered bug tracking framework. The framework augments existing systems with large language model (LLM) and agent-driven automation, and we report early adaptations of its key components, providing initial empirical grounding for its feasibility. The proposed framework aims to reduce time to resolution and coordination overhead by enabling end users to report bugs in natural language while AI agents refine reports, attempt reproduction, classify bugs, validate reports, suggest no-code fixes, generate patches, and support continuous integration and deployment. We discuss the challenges and opportunities of integrating LLMs into bug tracking and show how intelligent automation can transform software maintenance into a more efficient, collaborative, and user-centric process.

---

### CIR-CoT: Towards Interpretable Composed Image Retrieval via End-to-End Chain-of-Thought Reasoning

Authors: Weihuang Lin, Yiwei Ma, Jiayi Ji, Xiaoshuai Sun, Rongrong Ji
Published: 2025-10-09
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.08003v1

Composed Image Retrieval (CIR), which aims to find a target image from a reference image and a modification text, presents the core challenge of performing unified reasoning across visual and semantic modalities. While current approaches based on Vision-Language Models (VLMs, e.g., CLIP) and more recent Multimodal Large Language Models (MLLMs, e.g., Qwen-VL) have shown progress, they predominantly function as ``black boxes." This inherent opacity not only prevents users from understanding the retrieval rationale but also restricts the models' ability to follow complex, fine-grained instructions. To overcome these limitations, we introduce CIR-CoT, the first end-to-end retrieval-oriented MLLM designed to integrate explicit Chain-of-Thought (CoT) reasoning. By compelling the model to first generate an interpretable reasoning chain, CIR-CoT enhances its ability to capture crucial cross-modal interactions, leading to more accurate retrieval while making its decision process transparent. Since existing datasets like FashionIQ and CIRR lack the necessary reasoning data, a key contribution of our work is the creation of structured CoT annotations using a three-stage process involving a caption, reasoning, and conclusion. Our model is then fine-tuned to produce this structured output before encoding its final retrieval intent into a dedicated embedding. Comprehensive experiments show that CIR-CoT achieves highly competitive performance on in-domain datasets (FashionIQ, CIRR) and demonstrates remarkable generalization on the out-of-domain CIRCO dataset, establishing a new path toward more effective and trustworthy retrieval systems.

---

### Learning on the Job: An Experience-Driven Self-Evolving Agent for Long-Horizon Tasks

Authors: Cheng Yang, Xuemeng Yang, Licheng Wen, Daocheng Fu, Jianbiao Mei, Rong Wu, Pinlong Cai, Yufan Shen, Nianchen Deng, Botian Shi, Yu Qiao, Haifeng Li
Published: 2025-10-09
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2510.08002v1

Large Language Models have demonstrated remarkable capabilities across diverse domains, yet significant challenges persist when deploying them as AI agents for real-world long-horizon tasks. Existing LLM agents suffer from a critical limitation: they are test-time static and cannot learn from experience, lacking the ability to accumulate knowledge and continuously improve on the job. To address this challenge, we propose MUSE, a novel agent framework that introduces an experience-driven, self-evolving system centered around a hierarchical Memory Module. MUSE organizes diverse levels of experience and leverages them to plan and execute long-horizon tasks across multiple applications. After each sub-task execution, the agent autonomously reflects on its trajectory, converting the raw trajectory into structured experience and integrating it back into the Memory Module. This mechanism enables the agent to evolve beyond its static pretrained parameters, fostering continuous learning and self-evolution. We evaluate MUSE on the long-horizon productivity benchmark TAC. It achieves new SOTA performance by a significant margin using only a lightweight Gemini-2.5 Flash model. Sufficient Experiments demonstrate that as the agent autonomously accumulates experience, it exhibits increasingly superior task completion capabilities, as well as robust continuous learning and self-evolution capabilities. Moreover, the accumulated experience from MUSE exhibits strong generalization properties, enabling zero-shot improvement on new tasks. MUSE establishes a new paradigm for AI agents capable of real-world productivity task automation.

---

### DemandCast: Global hourly electricity demand forecasting

Authors: Kevin Steijn, Vamsi Priya Goli, Enrico Antonini
Published: 2025-10-09
Categories: cs.LG, physics.soc-ph
arXiv: https://arxiv.org/abs/2510.08000v1

This paper presents a machine learning framework for electricity demand forecasting across diverse geographical regions using the gradient boosting algorithm XGBoost. The model integrates historical electricity demand and comprehensive weather and socioeconomic variables to predict normalized electricity demand profiles. To enable robust training and evaluation, we developed a large-scale dataset spanning multiple years and countries, applying a temporal data-splitting strategy that ensures benchmarking of out-of-sample performance. Our approach delivers accurate and scalable demand forecasts, providing valuable insights for energy system planners and policymakers as they navigate the challenges of the global energy transition.

---

### Leveraging Author-Specific Context for Scientific Figure Caption Generation: 3rd SciCap Challenge

Authors: Watcharapong Timklaypachara, Monrada Chiewhawan, Nopporn Lekuthai, Titipat Achakulvisut
Published: 2025-10-09
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2510.07993v1

Scientific figure captions require both accuracy and stylistic consistency to convey visual information. Here, we present a domain-specific caption generation system for the 3rd SciCap Challenge that integrates figure-related textual context with author-specific writing styles using the LaMP-Cap dataset. Our approach uses a two-stage pipeline: Stage 1 combines context filtering, category-specific prompt optimization via DSPy's MIPROv2 and SIMBA, and caption candidate selection; Stage 2 applies few-shot prompting with profile figures for stylistic refinement. Our experiments demonstrate that category-specific prompts outperform both zero-shot and general optimized approaches, improving ROUGE-1 recall by +8.3\% while limiting precision loss to -2.8\% and BLEU-4 reduction to -10.9\%. Profile-informed stylistic refinement yields 40--48\% gains in BLEU scores and 25--27\% in ROUGE. Overall, our system demonstrates that combining contextual understanding with author-specific stylistic adaptation can generate captions that are both scientifically accurate and stylistically faithful to the source paper.

---

### GraphEnet: Event-driven Human Pose Estimation with a Graph Neural Network

Authors: Gaurvi Goyal, Pham Cong Thuong, Arren Glover, Masayoshi Mizuno, Chiara Bartolozzi
Published: 2025-10-09
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.07990v1

Human Pose Estimation is a crucial module in human-machine interaction applications and, especially since the rise in deep learning technology, robust methods are available to consumers using RGB cameras and commercial GPUs. On the other hand, event-based cameras have gained popularity in the vision research community for their low latency and low energy advantages that make them ideal for applications where those resources are constrained like portable electronics and mobile robots. In this work we propose a Graph Neural Network, GraphEnet, that leverages the sparse nature of event camera output, with an intermediate line based event representation, to estimate 2D Human Pose of a single person at a high frequency. The architecture incorporates a novel offset vector learning paradigm with confidence based pooling to estimate the human pose. This is the first work that applies Graph Neural Networks to event data for Human Pose Estimation. The code is open-source at https://github.com/event-driven-robotics/GraphEnet-NeVi-ICCV2025.

---

### ReInAgent: A Context-Aware GUI Agent Enabling Human-in-the-Loop Mobile Task Navigation

Authors: Haitao Jia, Ming He, Zimo Yin, Likang Wu, Jianping Fan, Jitao Sang
Published: 2025-10-09
Categories: cs.AI
arXiv: https://arxiv.org/abs/2510.07988v1

Mobile GUI agents exhibit substantial potential to facilitate and automate the execution of user tasks on mobile phones. However, exist mobile GUI agents predominantly privilege autonomous operation and neglect the necessity of active user engagement during task execution. This omission undermines their adaptability to information dilemmas including ambiguous, dynamically evolving, and conflicting task scenarios, leading to execution outcomes that deviate from genuine user requirements and preferences. To address these shortcomings, we propose ReInAgent, a context-aware multi-agent framework that leverages dynamic information management to enable human-in-the-loop mobile task navigation. ReInAgent integrates three specialized agents around a shared memory module: an information-managing agent for slot-based information management and proactive interaction with the user, a decision-making agent for conflict-aware planning, and a reflecting agent for task reflection and information consistency validation. Through continuous contextual information analysis and sustained user-agent collaboration, ReInAgent overcomes the limitation of existing approaches that rely on clear and static task assumptions. Consequently, it enables more adaptive and reliable mobile task navigation in complex, real-world scenarios. Experimental results demonstrate that ReInAgent effectively resolves information dilemmas and produces outcomes that are more closely aligned with genuine user preferences. Notably, on complex tasks involving information dilemmas, ReInAgent achieves a 25% higher success rate than Mobile-Agent-v2.

---