# Research Papers — 2025-12 · bundle page 163 of 175

> Full-text markdown of papers #8101–8150 of 8703 published in 2025-12, newest first. AI-readable mirrors of the full 8703-paper 2025-12 archive (50 papers per file).

Canonical page: https://aipod.dev/papers/archive/2025-12/page/163
[← Previous bundle](https://aipod.dev/papers/archive/2025-12/page/162.md)
[Next bundle →](https://aipod.dev/papers/archive/2025-12/page/164.md)

### Defense That Attacks: How Robust Models Become Better Attackers

Authors: Mohamed Awad, Mahmoud Akrm, Walid Gomaa
Published: 2025-12-02
Categories: cs.CV, cs.AI
arXiv: https://arxiv.org/abs/2512.02830v3

Deep learning has achieved great success in computer vision, but remains vulnerable to adversarial attacks. Adversarial training is the leading defense designed to improve model robustness. However, its effect on the transferability of attacks is underexplored. In this work, we ask whether adversarial training unintentionally increases the transferability of adversarial examples. To answer this, we trained a diverse zoo of 36 models, including CNNs and ViTs, and conducted comprehensive transferability experiments. Our results reveal a clear paradox: adversarially trained (AT) models produce perturbations that transfer more effectively than those from standard models, which introduce a new ecosystem risk. To enable reproducibility and further study, we release all models, code, and experimental scripts. Furthermore, we argue that robustness evaluations should assess not only the resistance of a model to transferred attacks but also its propensity to produce transferable adversarial examples.

---

### From Navigation to Refinement: Revealing the Two-Stage Nature of Flow-based Diffusion Models through Oracle Velocity

Authors: Haoming Liu, Jinnuo Liu, Yanhao Li, Liuyang Bai, Yunkai Ji, Yuanhe Guo, Shenji Wan, Hongyi Wen
Published: 2025-12-02
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2512.02826v3

Flow-based diffusion models have emerged as a leading paradigm for training generative models across images and videos. However, their memorization-generalization behavior remains poorly understood. In this work, we revisit the flow matching (FM) objective and study its marginal velocity field, which admits a closed-form expression, allowing exact computation of the oracle FM target. Analyzing this oracle velocity field reveals that flow-based diffusion models inherently formulate a two-stage training target: an early stage guided by a mixture of data modes, and a later stage dominated by the nearest data sample. The two-stage objective leads to distinct learning behaviors: the early navigation stage generalizes across data modes to form global layouts, whereas the later refinement stage increasingly memorizes fine-grained details. Leveraging these insights, we explain the effectiveness of practical techniques such as timestep-shifted schedules, classifier-free guidance intervals, and latent space design choices. Our study deepens the understanding of diffusion model training dynamics and offers principles for guiding future architectural and algorithmic improvements. Our project page is available at: https://maps-research.github.io/from-navigation-to-refinement/.

---

### Tempering the Bayes Filter towards Improved Model-Based Estimation

Authors: Menno van Zutphen, Domagoj Herceg, Giannis Delimpaltadakis, Duarte J. Antunes
Published: 2025-12-02
Categories: eess.SY, math.OC, stat.ML
arXiv: https://arxiv.org/abs/2512.02823v1

Model-based filtering is often carried out while subject to an imperfect model, as learning partially-observable stochastic systems remains a challenge. Recent work on Bayesian inference found that tempering the likelihood or full posterior of an imperfect model can improve predictive accuracy, as measured by expected negative log likelihood. In this paper, we develop the tempered Bayes filter, improving estimation performance through both of the aforementioned, and one newly introduced, modalities. The result admits a recursive implementation with a computational complexity no higher than that of the original Bayes filter. Our analysis reveals that -- besides the well-known fact in the field of Bayesian inference that likelihood tempering affects the balance between prior and likelihood -- full-posterior tempering tunes the level of entropy in the final belief distribution. We further find that a region of the tempering space can be understood as interpolating between the Bayes- and MAP filters, recovering these as special cases. Analytical results further establish conditions under which a tempered Bayes filter achieves improved predictive performance. Specializing the results to the linear Gaussian case, we obtain the tempered Kalman filter. In this context, we interpret how the parameters affect the Kalman state estimate and covariance propagation. Empirical results confirm that our method consistently improves predictive accuracy over the Bayes filter baseline.

---

### BOOM: Beyond Only One Modality KIT's Multimodal Multilingual Lecture Companion

Authors: Sai Koneru, Fabian Retkowski, Christian Huber, Lukas Hilgert, Seymanur Akti, Enes Yavuz Ugan, Alexander Waibel, Jan Niehues
Published: 2025-12-02
Categories: cs.CL
arXiv: https://arxiv.org/abs/2512.02817v2

The globalization of education and rapid growth of online learning have made localizing educational content a critical challenge. Lecture materials are inherently multimodal, combining spoken audio with visual slides, which requires systems capable of processing multiple input modalities. To provide an accessible and complete learning experience, translations must preserve all modalities: text for reading, slides for visual understanding, and speech for auditory learning. We present \textbf{BOOM}, a multimodal multilingual lecture companion that jointly translates lecture audio and slides to produce synchronized outputs across three modalities: translated text, localized slides with preserved visual elements, and synthesized speech. This end-to-end approach enables students to access lectures in their native language while aiming to preserve the original content in its entirety. Our experiments demonstrate that slide-aware transcripts also yield cascading benefits for downstream tasks such as summarization and question answering. The demo video and code can be found at https://ai4lt.github.io/boom/ \footnote{All released code and models are licensed under the MIT License}.

---

### A benchmark dataset for evaluating Syndrome Differentiation and Treatment in large language models

Authors: Kunning Li, Jianbin Guo, Zhaoyang Shang, Yiqing Liu, Hongmin Du, Lingling Liu, Yuping Zhao, Lifeng Dong
Published: 2025-12-02
Categories: cs.CL
arXiv: https://arxiv.org/abs/2512.02816v1

The emergence of Large Language Models (LLMs) within the Traditional Chinese Medicine (TCM) domain presents an urgent need to assess their clinical application capabilities. However, such evaluations are challenged by the individualized, holistic, and diverse nature of TCM's "Syndrome Differentiation and Treatment" (SDT). Existing benchmarks are confined to knowledge-based question-answering or the accuracy of syndrome differentiation, often neglecting assessment of treatment decision-making. Here, we propose a comprehensive, clinical case-based benchmark spearheaded by TCM experts, and a specialized reward model employed to quantify prescription-syndrome congruence. Data annotation follows a rigorous pipeline. This benchmark, designated TCM-BEST4SDT, encompasses four tasks, including TCM Basic Knowledge, Medical Ethics, LLM Content Safety, and SDT. The evaluation framework integrates three mechanisms, namely selected-response evaluation, judge model evaluation, and reward model evaluation. The effectiveness of TCM-BEST4SDT was corroborated through experiments on 15 mainstream LLMs, spanning both general and TCM domains. To foster the development of intelligent TCM research, TCM-BEST4SDT is now publicly available.

---

### Radiologist Copilot: An Agentic Framework Orchestrating Specialized Tools for Reliable Radiology Reporting

Authors: Yongrui Yu, Zhongzhen Huang, Linjie Mu, Shaoting Zhang, Xiaofan Zhang
Published: 2025-12-02
Categories: cs.AI
arXiv: https://arxiv.org/abs/2512.02814v2

In clinical practice, radiology reporting is an essential yet complex, time-intensive, and error-prone task, particularly for 3D medical images. Existing automated approaches based on medical vision-language models primarily focus on isolated report generation. However, real-world radiology reporting extends far beyond report writing, which requires meticulous image observation and interpretation, appropriate template selection, and rigorous quality control to ensure adherence to clinical standards. This multi-stage, planning-intensive workflow fundamentally exceeds the capabilities of single-pass models. To bridge this gap, we propose Radiologist Copilot, an agentic system that autonomously orchestrates specialized tools to complete the entire radiology reporting workflow rather than isolated report writing. Radiologist Copilot enables region image localization and region analysis planning to support detailed visual reasoning, adopts strategic template selection for standardized report writing, and incorporates dedicated report quality control via quality assessment and feedback-driven iterative refinement. By integrating localization, interpretation, template selection, report composition, and quality control, Radiologist Copilot delivers a comprehensive and clinically aligned radiology reporting workflow. Experimental results demonstrate that it significantly outperforms state-of-the-art methods, supporting radiologists throughout the entire radiology reporting process. The code will be released upon acceptance.

---

### Enhancing Automated Paper Reproduction via Prompt-Free Collaborative Agents

Authors: Zijie Lin, Qilin Cai, Liang Shen, Mingjun Xiao
Published: 2025-12-02
Categories: cs.AI
arXiv: https://arxiv.org/abs/2512.02812v1

Automated paper reproduction has emerged as a promising approach to accelerate scientific research, employing multi-step workflow frameworks to systematically convert academic papers into executable code. However, existing frameworks often lack mechanisms to verify and refine the outputs at each generation step, or rely heavily on manually designed prompts for self-refinement, which limits their adaptability and scalability. To address these limitations, we propose a prompt-free collaborative agent framework that automatically enhances the quality of paper-to-code generation. Our approach employs two collaborative agents: a verification agent that examines whether the outputs at each step satisfy the requirements specified in the corresponding system prompt, and a refinement agent that revises the outputs based on the identified issues. Unlike previous methods that require human experts to craft specific refinement prompts for each step, our framework achieves automatic verification and improvement by leveraging only the original system prompts. We integrate our collaborative agents into the Paper2Code framework and conduct comprehensive experiments on PaperBench Code-Dev and Paper2CodeBench datasets. Experimental results demonstrate that our approach significantly improves the accuracy and completeness of reproduced code, achieving performance gains of approximately 15\% and 13\%, respectively, compared to the baseline without our agents. Furthermore, comparative experiments against Self-Refine validate the robustness and consistency of our prompt-free approach across different datasets.

---

### Phase-Adaptive LLM Framework with Multi-Stage Validation for Construction Robot Task Allocation: A Systematic Benchmark Against Traditional Optimization Algorithms

Authors: Shyam prasad reddy Kaitha, Hongrui Yu
Published: 2025-12-02
Categories: cs.RO, cs.AI, cs.LG
arXiv: https://arxiv.org/abs/2512.02810v1

Multi-robot task allocation in construction automation has traditionally relied on optimization methods such as Dynamic Programming and Reinforcement Learning. This research introduces the LangGraph-based Task Allocation Agent (LTAA), an LLM-driven framework that integrates phase-adaptive allocation strategies, multi-stage validation with hierarchical retries, and dynamic prompting for efficient robot coordination. Although recent LLM approaches show potential for construction robotics, they largely lack rigorous validation and benchmarking against established algorithms. This paper presents the first systematic comparison of LLM-based task allocation with traditional methods in construction scenarios.The study validates LLM feasibility through SMART-LLM replication and addresses implementation challenges using a Self-Corrective Agent Architecture. LTAA leverages natural-language reasoning combined with structured validation mechanisms, achieving major computational gains reducing token usage by 94.6% and allocation time by 86% through dynamic prompting. The framework adjusts its strategy across phases: emphasizing execution feasibility early and workload balance in later allocations.The authors evaluate LTAA against Dynamic Programming, Q-learning, and Deep Q-Network (DQN) baselines using construction operations from the TEACh human-robot collaboration dataset. In the Heavy Excels setting, where robots have strong task specializations, LTAA achieves 77% task completion with superior workload balance, outperforming all traditional methods. These findings show that LLM-based reasoning with structured validation can match established optimization algorithms while offering additional advantages such as interpretability, adaptability, and the ability to update task logic without retraining.

---

### SR-GRPO: Stable Rank as an Intrinsic Geometric Reward for Large Language Model Alignment

Authors: Yixuan Tang, Yi Yang
Published: 2025-12-02
Categories: cs.CL
arXiv: https://arxiv.org/abs/2512.02807v1

Aligning Large Language Models (LLMs) with human preferences typically relies on external supervision, which faces critical limitations: human annotations are scarce and subjective, reward models are vulnerable to reward hacking, and self-evaluation methods suffer from prompt sensitivity and biases. In this work, we propose stable rank, an intrinsic, annotation-free quality signal derived from model representations. Stable rank measures the effective dimensionality of hidden states by computing the ratio of total variance to dominant-direction variance, capturing quality through how information distributes across representation dimensions. Empirically, stable rank achieves 84.04% accuracy on RewardBench and improves task accuracy by an average of 11.3 percentage points over greedy decoding via Best-of-N sampling. Leveraging this insight, we introduce Stable Rank Group Relative Policy Optimization (SR-GRPO), which uses stable rank as a reward signal for reinforcement learning. Without external supervision, SR-GRPO improves Qwen2.5-1.5B-Instruct by 10% on STEM and 19% on mathematical reasoning, outperforming both learned reward models and self-evaluation baselines. Our findings demonstrate that quality signals can be extracted from internal model geometry, offering a path toward scalable alignment without external supervision.

---

### TriLex: A Framework for Multilingual Sentiment Analysis in Low-Resource South African Languages

Authors: Mike Nkongolo, Hilton Vorster, Josh Warren, Trevor Naick, Deandre Vanmali, Masana Mashapha, Luke Brand, Alyssa Fernandes, Janco Calitz, Sibusiso Makhoba
Published: 2025-12-02
Categories: cs.CL
arXiv: https://arxiv.org/abs/2512.02799v1

Low-resource African languages remain underrepresented in sentiment analysis, limiting both lexical coverage and the performance of multilingual Natural Language Processing (NLP) systems. This study proposes TriLex, a three-stage retrieval augmented framework that unifies corpus-based extraction, cross lingual mapping, and retrieval augmented generation (RAG) driven lexical refinement to systematically expand sentiment lexicons for low-resource languages. Using the enriched lexicon, the performance of two prominent African pretrained language models (AfroXLMR and AfriBERTa) is evaluated across multiple case studies. Results demonstrate that AfroXLMR delivers superior performance, achieving F1-scores above 80% for isiXhosa and isiZulu and exhibiting strong cross-lingual stability. Although AfriBERTa lacks pre-training on these target languages, it still achieves reliable F1-scores around 64%, validating its utility in computationally constrained settings. Both models outperform traditional machine learning baselines, and ensemble analyses further enhance precision and robustness. The findings establish TriLex as a scalable and effective framework for multilingual sentiment lexicon expansion and sentiment modeling in low-resource South African languages.

---

### HUD: Hierarchical Uncertainty-Aware Disambiguation Network for Composed Video Retrieval

Authors: Zhiwei Chen, Yupeng Hu, Zixu Li, Zhiheng Fu, Haokun Wen, Weili Guan
Published: 2025-12-02
Categories: cs.CV, cs.MM
arXiv: https://arxiv.org/abs/2512.02792v2

Composed Video Retrieval (CVR) is a challenging video retrieval task that utilizes multi-modal queries, consisting of a reference video and modification text, to retrieve the desired target video. The core of this task lies in understanding the multi-modal composed query and achieving accurate composed feature learning. Within multi-modal queries, the video modality typically carries richer semantic content compared to the textual modality. However, previous works have largely overlooked the disparity in information density between these two modalities. This limitation can lead to two critical issues: 1) modification subject referring ambiguity and 2) limited detailed semantic focus, both of which degrade the performance of CVR models. To address the aforementioned issues, we propose a novel CVR framework, namely the Hierarchical Uncertainty-aware Disambiguation network (HUD). HUD is the first framework that leverages the disparity in information density between video and text to enhance multi-modal query understanding. It comprises three key components: (a) Holistic Pronoun Disambiguation, (b) Atomistic Uncertainty Modeling, and (c) Holistic-to-Atomistic Alignment. By exploiting overlapping semantics through holistic cross-modal interaction and fine-grained semantic alignment via atomistic-level cross-modal interaction, HUD enables effective object disambiguation and enhances the focus on detailed semantics, thereby achieving precise composed feature learning. Moreover, our proposed HUD is also applicable to the Composed Image Retrieval (CIR) task and achieves state-of-the-art performance across three benchmark datasets for both CVR and CIR tasks. The codes are available on https://zivchen-ty.github.io/HUD.github.io/.

---

### Making Dialogue Grounding Data Rich: A Three-Tier Data Synthesis Framework for Generalized Referring Expression Comprehension

Authors: Juexi Shao, Siyou Li, Yujian Gan, Chris Madge, Vanja Karan, Massimo Poesio
Published: 2025-12-02
Categories: cs.CL
arXiv: https://arxiv.org/abs/2512.02791v2

Dialogue-Based Generalized Referring Expression Comprehension (GREC) requires models to ground the expression and unlimited targets in complex visual scenes while resolving coreference across a long dialogue context. However, existing systems struggle under distribution shift between training and evaluation domains, a gap exacerbated by the scarcity of annotated dialogue grounding data. We address this challenge with a three-tier data-synthesis method that balances realism and controllability to produce scalable supervision for dialogue-conditioned grounding. Fine-tuning on the synthesized data yields consistent, substantial improvements over prior approaches across standard evaluation metrics.

---

### TrackNetV5: Residual-Driven Spatio-Temporal Refinement and Motion Direction Decoupling for Fast Object Tracking

Authors: Haonan Tang, Yanjun Chen, Lezhi Jiang, Qianfei Li, Xinyu Guo
Published: 2025-12-02
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.02789v4

The TrackNet series has established a strong baseline for fast-moving small object tracking in sports. However, existing iterations face significant limitations: V1-V3 struggle with occlusions due to a reliance on purely visual cues, while TrackNetV4, despite introducing motion inputs, suffers from directional ambiguity as its absolute difference method discards motion polarity. To overcome these bottlenecks, we propose TrackNetV5, a robust architecture integrating two novel mechanisms. First, to recover lost directional priors, we introduce the Motion Direction Decoupling (MDD) module. Unlike V4, MDD decomposes temporal dynamics into signed polarity fields, explicitly encoding both movement occurrence and trajectory direction. Second, we propose the Residual-Driven Spatio-Temporal Refinement (R-STR) head. Operating on a coarse-to-fine paradigm, this Transformer-based module leverages factorized spatio-temporal contexts to estimate a corrective residual, effectively recovering occluded targets. Extensive experiments on the TrackNetV2 dataset demonstrate that TrackNetV5 achieves a new state-of-the-art F1-score of 0.9859 and an accuracy of 0.9733, significantly outperforming previous versions. Notably, this performance leap is achieved with a marginal 3.7% increase in FLOPs compared to V4, maintaining real-time inference capabilities while delivering superior tracking precision.

---

### Diagnose, Correct, and Learn from Manipulation Failures via Visual Symbols

Authors: Xianchao Zeng, Xinyu Zhou, Youcheng Li, Jiayou Shi, Tianle Li, Liangming Chen, Lei Ren, Yong-Lu Li
Published: 2025-12-02
Categories: cs.RO, cs.CV
arXiv: https://arxiv.org/abs/2512.02787v3

Vision-Language-Action (VLA) models have recently achieved remarkable progress in robotic manipulation, yet they remain limited in failure diagnosis and learning from failures. Additionally, existing failure datasets are mostly generated programmatically in simulation, which limits their generalization to the real world. In light of these, we introduce ViFailback, a framework designed to diagnose robotic manipulation failures and provide both textual and visual correction guidance. Our framework utilizes explicit visual symbols to enhance annotation efficiency. We further release the ViFailback dataset, a large-scale collection of 58,126 Visual Question Answering (VQA) pairs along with their corresponding 5,202 real-world manipulation trajectories. Based on the dataset, we establish ViFailback-Bench, a benchmark of 11 fine-grained VQA tasks designed to assess the failure diagnosis and correction abilities of Vision-Language Models (VLMs), featuring ViFailback-Bench Lite for closed-ended and ViFailback-Bench Hard for open-ended evaluation. To demonstrate the effectiveness of our framework, we built the ViFailback-8B VLM, which not only achieves significant overall performance improvement on ViFailback-Bench but also generates visual symbols for corrective action guidance. Finally, by integrating ViFailback-8B with a VLA model, we conduct real-world robotic experiments demonstrating its ability to assist the VLA model in recovering from failures. Project Website: https://x1nyuzhou.github.io/vifailback.github.io/

---

### FiMMIA: scaling semantic perturbation-based membership inference across modalities

Authors: Anton Emelyanov, Sergei Kudriashov, Alena Fenogenova
Published: 2025-12-02
Categories: cs.LG
arXiv: https://arxiv.org/abs/2512.02786v1

Membership Inference Attacks (MIAs) aim to determine whether a specific data point was included in the training set of a target model. Although there are have been numerous methods developed for detecting data contamination in large language models (LLMs), their performance on multimodal LLMs (MLLMs) falls short due to the instabilities introduced through multimodal component adaptation and possible distribution shifts across multiple inputs. In this work, we investigate multimodal membership inference and address two issues: first, by identifying distribution shifts in the existing datasets, and second, by releasing an extended baseline pipeline to detect them. We also generalize the perturbation-based membership inference methods to MLLMs and release \textbf{FiMMIA} -- a modular \textbf{F}ramework for \textbf{M}ultimodal \textbf{MIA}.\footnote{The source code and framework have been made publicly available under the MIT license via \href{https://github.com/ai-forever/data_leakage_detect}{link}.The video demonstration is available on \href{https://youtu.be/a9L4-H80aSg}{YouTube}.} Our approach trains a neural network to analyze the target model's behavior on perturbed inputs, capturing distributional differences between members and non-members. Comprehensive evaluations on various fine-tuned multimodal models demonstrate the effectiveness of our perturbation-based membership inference attacks in multimodal domains.

---

### Perception of AI-Generated Music -- The Role of Composer Identity, Personality Traits, Music Preferences, and Perceived Humanness

Authors: David Stammer, Hannah Strauss, Peter Knees
Published: 2025-12-02
Categories: cs.HC, cs.AI, cs.CY
arXiv: https://arxiv.org/abs/2512.02785v1

The rapid rise of AI-generated art has sparked debate about potential biases in how audiences perceive and evaluate such works. This study investigates how composer information and listener characteristics shape the perception of AI-generated music, adopting a mixed-method approach. Using a diverse set of stimuli across various genres from two AI music models, we examine effects of perceived authorship on liking and emotional responses, and explore how attitudes toward AI, personality traits, and music-related variables influence evaluations. We further assess the influence of perceived humanness and analyze open-ended responses to uncover listener criteria for judging AI-generated music. Attitudes toward AI proved to be the best predictor of both liking and emotional intensity of AI-generated music. This quantitative finding was complemented by qualitative themes from our thematic analysis, which identified ethical, cultural, and contextual considerations as important criteria in listeners' evaluations of AI-generated music. Our results offer a nuanced view of how people experience music created by AI tools and point to key factors and methodological considerations for future research on music perception in human-AI interaction.

---

### Exploring Definitions of Quality and Diversity in Sonic Measurement Spaces

Authors: Björn Þór Jónsson, Çağrı Erdem, Stefano Fasciani, Kyrre Glette
Published: 2025-12-02
Categories: cs.SD, cs.NE
arXiv: https://arxiv.org/abs/2512.02783v1

Digital sound synthesis presents the opportunity to explore vast parameter spaces containing millions of configurations. Quality diversity (QD) evolutionary algorithms offer a promising approach to harness this potential, yet their success hinges on appropriate sonic feature representations. Existing QD methods predominantly employ handcrafted descriptors or supervised classifiers, potentially introducing unintended exploration biases and constraining discovery to familiar sonic regions. This work investigates unsupervised dimensionality reduction methods for automatically defining and dynamically reconfiguring sonic behaviour spaces during QD search. We apply Principal Component Analysis (PCA) and autoencoders to project high-dimensional audio features onto structured grids for MAP-Elites, implementing dynamic reconfiguration through model retraining at regular intervals. Comparison across two experimental scenarios shows that automatic approaches achieve significantly greater diversity than handcrafted behaviour spaces while avoiding expert-imposed biases. Dynamic behaviour-space reconfiguration maintains evolutionary pressure and prevents stagnation, with PCA proving most effective among the dimensionality reduction techniques. These results contribute to automated sonic discovery systems capable of exploring vast parameter spaces without manual intervention or supervised training constraints.

---

### LumiX: Structured and Coherent Text-to-Intrinsic Generation

Authors: Xu Han, Biao Zhang, Xiangjun Tang, Xianzhi Li, Peter Wonka
Published: 2025-12-02
Categories: cs.CV, cs.GR, cs.LG
arXiv: https://arxiv.org/abs/2512.02781v1

We present LumiX, a structured diffusion framework for coherent text-to-intrinsic generation. Conditioned on text prompts, LumiX jointly generates a comprehensive set of intrinsic maps (e.g., albedo, irradiance, normal, depth, and final color), providing a structured and physically consistent description of an underlying scene. This is enabled by two key contributions: 1) Query-Broadcast Attention, a mechanism that ensures structural consistency by sharing queries across all maps in each self-attention block. 2) Tensor LoRA, a tensor-based adaptation that parameter-efficiently models cross-map relations for efficient joint training. Together, these designs enable stable joint diffusion training and unified generation of multiple intrinsic properties. Experiments show that LumiX produces coherent and physically meaningful results, achieving 23% higher alignment and a better preference score (0.19 vs. -0.41) compared to the state of the art, and it can also perform image-conditioned intrinsic decomposition within the same framework.

---

### Rethinking Surgical Smoke: A Smoke-Type-Aware Laparoscopic Video Desmoking Method and Dataset

Authors: Qifan Liang, Junlin Li, Zhen Han, Xihao Wang, Zhongyuan Wang, Bin Mei
Published: 2025-12-02
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.02780v1

Electrocautery or lasers will inevitably generate surgical smoke, which hinders the visual guidance of laparoscopic videos for surgical procedures. The surgical smoke can be classified into different types based on its motion patterns, leading to distinctive spatio-temporal characteristics across smoky laparoscopic videos. However, existing desmoking methods fail to account for such smoke-type-specific distinctions. Therefore, we propose the first Smoke-Type-Aware Laparoscopic Video Desmoking Network (STANet) by introducing two smoke types: Diffusion Smoke and Ambient Smoke. Specifically, a smoke mask segmentation sub-network is designed to jointly conduct smoke mask and smoke type predictions based on the attention-weighted mask aggregation, while a smokeless video reconstruction sub-network is proposed to perform specially desmoking on smoky features guided by two types of smoke mask. To address the entanglement challenges of two smoke types, we further embed a coarse-to-fine disentanglement module into the mask segmentation sub-network, which yields more accurate disentangled masks through the smoke-type-aware cross attention between non-entangled and entangled regions. In addition, we also construct the first large-scale synthetic video desmoking dataset with smoke type annotations. Extensive experiments demonstrate that our method not only outperforms state-of-the-art approaches in quality evaluations, but also exhibits superior generalization across multiple downstream surgical tasks.

---

### AI-Driven Document Redaction in UK Public Authorities: Implementation Gaps, Regulatory Challenges, and the Human Oversight Imperative

Authors: Yijun Chen
Published: 2025-12-02
Categories: cs.CY, cs.AI
arXiv: https://arxiv.org/abs/2512.02774v1

Document redaction in public authorities faces critical challenges as traditional manual approaches struggle to balance growing transparency demands with increasingly stringent data protection requirements. This study investigates the implementation of AI-driven document redaction within UK public authorities through Freedom of Information (FOI) requests. While AI technologies offer potential solutions to redaction challenges, their actual implementation within public sector organizations remains underexplored. Based on responses from 44 public authorities across healthcare, government, and higher education sectors, this study reveals significant gaps between technological possibilities and organizational realities. Findings show highly limited AI adoption (only one authority reported using AI tools), widespread absence of formal redaction policies (50 percent reported "information not held"), and deficiencies in staff training. The study identifies three key barriers to effective AI implementation: poor record-keeping practices, lack of standardized redaction guidelines, and insufficient specialized training for human oversight. These findings highlight the need for a socio-technical approach that balances technological automation with meaningful human expertise. This research provides the first empirical assessment of AI redaction practices in UK public authorities and contributes evidence to support policymakers navigating the complex interplay between transparency obligations, data protection requirements, and emerging AI technologies in public administration.

---

### Towards Unification of Hallucination Detection and Fact Verification for Large Language Models

Authors: Weihang Su, Jianming Long, Changyue Wang, Shiyu Lin, Jingyan Xu, Ziyi Ye, Qingyao Ai, Yiqun Liu
Published: 2025-12-02
Categories: cs.CL, cs.IR
arXiv: https://arxiv.org/abs/2512.02772v1

Large Language Models (LLMs) frequently exhibit hallucinations, generating content that appears fluent and coherent but is factually incorrect. Such errors undermine trust and hinder their adoption in real-world applications. To address this challenge, two distinct research paradigms have emerged: model-centric Hallucination Detection (HD) and text-centric Fact Verification (FV). Despite sharing the same goal, these paradigms have evolved in isolation, using distinct assumptions, datasets, and evaluation protocols. This separation has created a research schism that hinders their collective progress. In this work, we take a decisive step toward bridging this divide. We introduce UniFact, a unified evaluation framework that enables direct, instance-level comparison between FV and HD by dynamically generating model outputs and corresponding factuality labels. Through large-scale experiments across multiple LLM families and detection methods, we reveal three key findings: (1) No paradigm is universally superior; (2) HD and FV capture complementary facets of factual errors; and (3) hybrid approaches that integrate both methods consistently achieve state-of-the-art performance. Beyond benchmarking, we provide the first in-depth analysis of why FV and HD diverged, as well as empirical evidence supporting the need for their unification. The comprehensive experimental results call for a new, integrated research agenda toward unifying Hallucination Detection and Fact Verification in LLMs. We have open-sourced all the code, data, and baseline implementation at: https://github.com/oneal2000/UniFact/

---

### PEFT-Factory: Unified Parameter-Efficient Fine-Tuning of Autoregressive Large Language Models

Authors: Robert Belanec, Ivan Srba, Maria Bielikova
Published: 2025-12-02
Categories: cs.CL
arXiv: https://arxiv.org/abs/2512.02764v3

Parameter-Efficient Fine-Tuning (PEFT) methods address the increasing size of Large Language Models (LLMs). Currently, many newly introduced PEFT methods are challenging to replicate, deploy, or compare with one another. To address this, we introduce PEFT-Factory, a unified framework for efficient fine-tuning LLMs using both off-the-shelf and custom PEFT methods. While its modular design supports extensibility, it natively provides a representative set of 19 PEFT methods, 27 classification and text generation datasets addressing 12 tasks, and both standard and PEFT-specific evaluation metrics. As a result, PEFT-Factory provides a ready-to-use, controlled, and stable environment, improving replicability and benchmarking of PEFT methods. PEFT-Factory is a downstream framework that originates from the popular LLaMA-Factory, and is publicly available at https://github.com/kinit-sk/PEFT-Factory.

---

### SurveyEval: Towards Comprehensive Evaluation of LLM-Generated Academic Surveys

Authors: Jiahao Zhao, Shuaixing Zhang, Nan Xu, Lei Wang
Published: 2025-12-02
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2512.02763v1

LLM-based automatic survey systems are transforming how users acquire information from the web by integrating retrieval, organization, and content synthesis into end-to-end generation pipelines. While recent works focus on developing new generation pipelines, how to evaluate such complex systems remains a significant challenge. To this end, we introduce SurveyEval, a comprehensive benchmark that evaluates automatically generated surveys across three dimensions: overall quality, outline coherence, and reference accuracy. We extend the evaluation across 7 subjects and augment the LLM-as-a-Judge framework with human references to strengthen evaluation-human alignment. Evaluation results show that while general long-text or paper-writing systems tend to produce lower-quality surveys, specialized survey-generation systems are able to deliver substantially higher-quality results. We envision SurveyEval as a scalable testbed to understand and improve automatic survey systems across diverse subjects and evaluation criteria.

---

### AttMetNet: Attention-Enhanced Deep Neural Network for Methane Plume Detection in Sentinel-2 Satellite Imagery

Authors: Rakib Ahsan, MD Sadik Hossain Shanto, Md Sultanul Arifin, Tanzima Hashem
Published: 2025-12-02
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.02751v1

Methane is a powerful greenhouse gas that contributes significantly to global warming. Accurate detection of methane emissions is the key to taking timely action and minimizing their impact on climate change. We present AttMetNet, a novel attention-enhanced deep learning framework for methane plume detection with Sentinel-2 satellite imagery. The major challenge in developing a methane detection model is to accurately identify methane plumes from Sentinel-2's B11 and B12 bands while suppressing false positives caused by background variability and diverse land cover types. Traditional detection methods typically depend on the differences or ratios between these bands when comparing the scenes with and without plumes. However, these methods often require verification by a domain expert because they generate numerous false positives. Recent deep learning methods make some improvements using CNN-based architectures, but lack mechanisms to prioritize methane-specific features. AttMetNet introduces a methane-aware architecture that fuses the Normalized Difference Methane Index (NDMI) with an attention-enhanced U-Net. By jointly exploiting NDMI's plume-sensitive cues and attention-driven feature selection, AttMetNet selectively amplifies methane absorption features while suppressing background noise. This integration establishes a first-of-its-kind architecture tailored for robust methane plume detection in real satellite imagery. Additionally, we employ focal loss to address the severe class imbalance arising from both limited positive plume samples and sparse plume pixels within imagery. Furthermore, AttMetNet is trained on the real methane plume dataset, making it more robust to practical scenarios. Extensive experiments show that AttMetNet surpasses recent methods in methane plume detection with a lower false positive rate, better precision recall balance, and higher IoU.

---

### Reasoning-Aware Multimodal Fusion for Hateful Video Detection

Authors: Shuonan Yang, Tailin Chen, Jiangbei Yue, Guangliang Cheng, Jianbo Jiao, Zeyu Fu
Published: 2025-12-02
Categories: cs.CV, cs.AI
arXiv: https://arxiv.org/abs/2512.02743v2

Hate speech in online videos is posing an increasingly serious threat to digital platforms, especially as video content becomes increasingly multimodal and context-dependent. Existing methods often struggle to effectively fuse the complex semantic relationships between modalities and lack the ability to understand nuanced hateful content. To address these issues, we propose an innovative Reasoning-Aware Multimodal Fusion (RAMF) framework. To tackle the first challenge, we design Local-Global Context Fusion (LGCF) to capture both local salient cues and global temporal structures, and propose Semantic Cross Attention (SCA) to enable fine-grained multimodal semantic interaction. To tackle the second challenge, we introduce adversarial reasoning-a structured three-stage process where a vision-language model generates (i) objective descriptions, (ii) hate-assumed inferences, and (iii) non-hate-assumed inferences-providing complementary semantic perspectives that enrich the model's contextual understanding of nuanced hateful intent. Evaluations on two real-world hateful video datasets demonstrate that our method achieves robust generalisation performance, improving upon state-of-the-art methods by 3% and 7% in Macro-F1 and hate class recall, respectively. The source codes and data required to reproduce our results are available at https://github.com/Multimodal-Intelligence-Lab-MIL/RAMF.

---

### Adversarial Jamming for Autoencoder Distribution Matching

Authors: Waleed El-Geresy, Deniz Gündüz
Published: 2025-12-02
Categories: cs.LG, cs.IT
arXiv: https://arxiv.org/abs/2512.02740v1

We propose the use of adversarial wireless jamming to regularise the latent space of an autoencoder to match a diagonal Gaussian distribution. We consider the minimisation of a mean squared error distortion, where a jammer attempts to disrupt the recovery of a Gaussian source encoded and transmitted over the adversarial channel. A straightforward consequence of existing theoretical results is the fact that the saddle point of a minimax game - involving such an encoder, its corresponding decoder, and an adversarial jammer - consists of diagonal Gaussian noise output by the jammer. We use this result as inspiration for a novel approach to distribution matching in the latent space, utilising jamming as an auxiliary objective to encourage the aggregated latent posterior to match a diagonal Gaussian distribution. Using this new technique, we achieve distribution matching comparable to standard variational autoencoders and to Wasserstein autoencoders. This approach can also be generalised to other latent distributions.

---

### Probabilistic energy profiler for statically typed JVM-based programming languages

Authors: Joel Nyholm, Wojciech Mostowski, Christoph Reichenbach
Published: 2025-12-02
Categories: cs.PL, cs.CL
arXiv: https://arxiv.org/abs/2512.02738v1

Energy consumption is a growing concern in several fields, from mobile devices to large data centers. Developers need detailed data on the energy consumption of their software to mitigate consumption issues. Previous approaches have a broader focus, such as on specific functions or programs, rather than source code statements. They primarily focus on estimating the CPU's energy consumption using point estimates, thereby disregarding other hardware effects and limiting their use for statistical reasoning and explainability. We developed a novel methodology to address the limitations of measuring only the CPU's consumption and using point estimates, focusing on predicting the energy usage of statically typed JVM-based programming languages, such as Java and Scala. We measure the energy consumption of Bytecode patterns, the translation from the programming language's source code statement to their Java Bytecode representation. With the energy measurements, we construct a statistical model using Bayesian statistics, which allows us to predict the energy consumption through statistical distributions and analyze individual factors. The model includes three factors we obtain statically from the code: data size, data type, operation, and one factor about the hardware platform the code executes on: device. To validate our methodology, we implemented it for Java and evaluated its energy predictions on unseen programs. We observe that all four factors are influential, notably that two devices of the same model may differ in energy consumption and that the operations and data types cause consumption differences. The experiments also show that the energy prediction of programs closely follows the program's real energy consumption, validating our approach. Our work presents a methodology for constructing an energy model that future work, such as verification tools, can use for their energy estimates.

---

### Beyond Paired Data: Self-Supervised UAV Geo-Localization from Reference Imagery Alone

Authors: Tristan Amadei, Enric Meinhardt-Llopis, Benedicte Bascle, Corentin Abgrall, Gabriele Facciolo
Published: 2025-12-02
Categories: cs.CV, cs.LG
arXiv: https://arxiv.org/abs/2512.02737v1

Image-based localization in GNSS-denied environments is critical for UAV autonomy. Existing state-of-the-art approaches rely on matching UAV images to geo-referenced satellite images; however, they typically require large-scale, paired UAV-satellite datasets for training. Such data are costly to acquire and often unavailable, limiting their applicability. To address this challenge, we adopt a training paradigm that removes the need for UAV imagery during training by learning directly from satellite-view reference images. This is achieved through a dedicated augmentation strategy that simulates the visual domain shift between satellite and real-world UAV views. We introduce CAEVL, an efficient model designed to exploit this paradigm, and validate it on ViLD, a new and challenging dataset of real-world UAV images that we release to the community. Our method achieves competitive performance compared to approaches trained with paired data, demonstrating its effectiveness and strong generalization capabilities.

---

### A Framework for Causal Concept-based Model Explanations

Authors: Anna Rodum Bjøru, Jacob Lysnæs-Larsen, Oskar Jørgensen, Inga Strümke, Helge Langseth
Published: 2025-12-02
Categories: cs.AI
arXiv: https://arxiv.org/abs/2512.02735v1

This work presents a conceptual framework for causal concept-based post-hoc Explainable Artificial Intelligence (XAI), based on the requirements that explanations for non-interpretable models should be understandable as well as faithful to the model being explained. Local and global explanations are generated by calculating the probability of sufficiency of concept interventions. Example explanations are presented, generated with a proof-of-concept model made to explain classifiers trained on the CelebA dataset. Understandability is demonstrated through a clear concept-based vocabulary, subject to an implicit causal interpretation. Fidelity is addressed by highlighting important framework assumptions, stressing that the context of explanation interpretation must align with the context of explanation generation.

---

### Self-Improving AI Agents through Self-Play

Authors: Przemyslaw Chojecki
Published: 2025-12-02
Categories: cs.AI, cs.LG
arXiv: https://arxiv.org/abs/2512.02731v1

We extend the moduli-theoretic framework of psychometric batteries to the domain of dynamical systems. While previous work established the AAI capability score as a static functional on the space of agent representations, this paper formalizes the agent as a flow $ν_r$ parameterized by computational resource $r$, governed by a recursive Generator-Verifier-Updater (GVU) operator. We prove that this operator generates a vector field on the parameter manifold $Θ$, and we identify the coefficient of self-improvement $κ$ as the Lie derivative of the capability functional along this flow. The central contribution of this work is the derivation of the Variance Inequality, a spectral condition that is sufficient (under mild regularity) for the stability of self-improvement. We show that a sufficient condition for $κ> 0$ is that, up to curvature and step-size effects, the combined noise of generation and verification must be small enough. We then apply this formalism to unify the recent literature on Language Self-Play (LSP), Self-Correction, and Synthetic Data bootstrapping. We demonstrate that architectures such as STaR, SPIN, Reflexion, GANs and AlphaZero are specific topological realizations of the GVU operator that satisfy the Variance Inequality through filtration, adversarial discrimination, or grounding in formal systems.

---

### DF-Mamba: Deformable State Space Modeling for 3D Hand Pose Estimation in Interactions

Authors: Yifan Zhou, Takehiko Ohkawa, Guwenxiao Zhou, Kanoko Goto, Takumi Hirose, Yusuke Sekikawa, Nakamasa Inoue
Published: 2025-12-02
Categories: cs.CV, cs.AI, cs.LG
arXiv: https://arxiv.org/abs/2512.02727v1

Modeling daily hand interactions often struggles with severe occlusions, such as when two hands overlap, which highlights the need for robust feature learning in 3D hand pose estimation (HPE). To handle such occluded hand images, it is vital to effectively learn the relationship between local image features (e.g., for occluded joints) and global context (e.g., cues from inter-joints, inter-hands, or the scene). However, most current 3D HPE methods still rely on ResNet for feature extraction, and such CNN's inductive bias may not be optimal for 3D HPE due to its limited capability to model the global context. To address this limitation, we propose an effective and efficient framework for visual feature extraction in 3D HPE using recent state space modeling (i.e., Mamba), dubbed Deformable Mamba (DF-Mamba). DF-Mamba is designed to capture global context cues beyond standard convolution through Mamba's selective state modeling and the proposed deformable state scanning. Specifically, for local features after convolution, our deformable scanning aggregates these features within an image while selectively preserving useful cues that represent the global context. This approach significantly improves the accuracy of structured 3D HPE, with comparable inference speed to ResNet-50. Our experiments involve extensive evaluations on five divergent datasets including single-hand and two-hand scenarios, hand-only and hand-object interactions, as well as RGB and depth-based estimation. DF-Mamba outperforms the latest image backbones, including VMamba and Spatial-Mamba, on all datasets and achieves state-of-the-art performance.

---

### AuditCopilot: Leveraging LLMs for Fraud Detection in Double-Entry Bookkeeping

Authors: Md Abdul Kadir, Sai Suresh Macharla Vasu, Sidharth S. Nair, Daniel Sonntag
Published: 2025-12-02
Categories: cs.AI
arXiv: https://arxiv.org/abs/2512.02726v1

Auditors rely on Journal Entry Tests (JETs) to detect anomalies in tax-related ledger records, but rule-based methods generate overwhelming false positives and struggle with subtle irregularities. We investigate whether large language models (LLMs) can serve as anomaly detectors in double-entry bookkeeping. Benchmarking SoTA LLMs such as LLaMA and Gemma on both synthetic and real-world anonymized ledgers, we compare them against JETs and machine learning baselines. Our results show that LLMs consistently outperform traditional rule-based JETs and classical ML baselines, while also providing natural-language explanations that enhance interpretability. These results highlight the potential of \textbf{AI-augmented auditing}, where human auditors collaborate with foundation models to strengthen financial integrity.

---

### Credal Graph Neural Networks

Authors: Matteo Tolloso, Davide Bacciu
Published: 2025-12-02
Categories: cs.LG
arXiv: https://arxiv.org/abs/2512.02722v1

Uncertainty quantification is essential for deploying reliable Graph Neural Networks (GNNs), where existing approaches primarily rely on Bayesian inference or ensembles. In this paper, we introduce the first credal graph neural networks (CGNNs), which extend credal learning to the graph domain by training GNNs to output set-valued predictions in the form of credal sets. To account for the distinctive nature of message passing in GNNs, we develop a complementary approach to credal learning that leverages different aspects of layer-wise information propagation. We assess our approach on uncertainty quantification in node classification under out-of-distribution conditions. Our analysis highlights the critical role of the graph homophily assumption in shaping the effectiveness of uncertainty estimates. Extensive experiments demonstrate that CGNNs deliver more reliable representations of epistemic uncertainty and achieve state-of-the-art performance under distributional shift on heterophilic graphs.

---

### Generative modeling using evolved quantum Boltzmann machines

Authors: Mark M. Wilde
Published: 2025-12-02
Categories: quant-ph, cs.LG, math.OC
arXiv: https://arxiv.org/abs/2512.02721v1

Born-rule generative modeling, a central task in quantum machine learning, seeks to learn probability distributions that can be efficiently sampled by measuring complex quantum states. One hope is for quantum models to efficiently capture probability distributions that are difficult to learn and simulate by classical means alone. Quantum Boltzmann machines were proposed about one decade ago for this purpose, yet efficient training methods have remained elusive. In this paper, I overcome this obstacle by proposing a practical solution that trains quantum Boltzmann machines for Born-rule generative modeling. Two key ingredients in the proposal are the Donsker-Varadhan variational representation of the classical relative entropy and the quantum Boltzmann gradient estimator of (Patel et al., arXiv:2410.12935). I present the main result for a more general ansatz known as an evolved quantum Boltzmann machine (Minervini et al., arXiv:2501.03367), which combines parameterized real- and imaginary-time evolution. I also show how to extend the findings to other distinguishability measures beyond relative entropy. Finally, I present four different hybrid quantum-classical algorithms for the minimax optimization underlying training, and I discuss their theoretical convergence guarantees.

---

### StockMem: An Event-Reflection Memory Framework for Stock Forecasting

Authors: He Wang, Wenyilin Xiao, Songqiao Han, Hailiang Huang
Published: 2025-12-02
Categories: cs.AI, cs.LG
arXiv: https://arxiv.org/abs/2512.02720v1

Stock price prediction is challenging due to market volatility and its sensitivity to real-time events. While large language models (LLMs) offer new avenues for text-based forecasting, their application in finance is hindered by noisy news data and the lack of explicit answers in text. General-purpose memory architectures struggle to identify the key drivers of price movements. To address this, we propose StockMem, an event-reflection dual-layer memory framework. It structures news into events and mines them along two dimensions: horizontal consolidation integrates daily events, while longitudinal tracking captures event evolution to extract incremental information reflecting market expectation discrepancies. This builds a temporal event knowledge base. By analyzing event-price dynamics, the framework further forms a reflection knowledge base of causal experiences. For prediction, it retrieves analogous historical scenarios and reasons with current events, incremental data, and past experiences. Experiments show StockMem outperforms existing memory architectures and provides superior, explainable reasoning by tracing the information chain affecting prices, enhancing decision transparency in financial forecasting.

---

### Emergent Bayesian Behaviour and Optimal Cue Combination in LLMs

Authors: Julian Ma, Jun Wang, Zafeirios Fountas
Published: 2025-12-02
Categories: cs.CL, cs.AI, cs.CV, cs.LG, q-bio.NC
arXiv: https://arxiv.org/abs/2512.02719v1

Large language models (LLMs) excel at explicit reasoning, but their implicit computational strategies remain underexplored. Decades of psychophysics research show that humans intuitively process and integrate noisy signals using near-optimal Bayesian strategies in perceptual tasks. We ask whether LLMs exhibit similar behaviour and perform optimal multimodal integration without explicit training or instruction. Adopting the psychophysics paradigm, we infer computational principles of LLMs from systematic behavioural studies. We introduce a behavioural benchmark - BayesBench: four magnitude estimation tasks (length, location, distance, and duration) over text and image, inspired by classic psychophysics, and evaluate a diverse set of nine LLMs alongside human judgments for calibration. Through controlled ablations of noise, context, and instruction prompts, we measure performance, behaviour and efficiency in multimodal cue-combination. Beyond accuracy and efficiency metrics, we introduce a Bayesian Consistency Score that detects Bayes-consistent behavioural shifts even when accuracy saturates. Our results show that while capable models often adapt in Bayes-consistent ways, accuracy does not guarantee robustness. Notably, GPT-5 Mini achieves perfect text accuracy but fails to integrate visual cues efficiently. This reveals a critical dissociation between capability and strategy, suggesting accuracy-centric benchmarks may over-index on performance while missing brittle uncertainty handling. These findings reveal emergent principled handling of uncertainty and highlight the correlation between accuracy and Bayesian tendencies. We release our psychophysics benchmark and consistency metric (https://bayes-bench.github.io) as evaluation tools and to inform future multimodal architecture designs.

---

### Menta: A Small Language Model for On-Device Mental Health Prediction

Authors: Tianyi Zhang, Xiangyuan Xue, Lingyan Ruan, Shiya Fu, Feng Xia, Simon D'Alfonso, Vassilis Kostakos, Ting Dang, Hong Jia
Published: 2025-12-02
Categories: cs.AI
arXiv: https://arxiv.org/abs/2512.02716v3

Mental health conditions affect hundreds of millions globally, yet early detection remains limited. While large language models (LLMs) have shown promise in mental health applications, their size and computational demands hinder practical deployment. Small language models (SLMs) offer a lightweight alternative, but their use for social media--based mental health prediction remains largely underexplored. In this study, we introduce Menta, the first optimized SLM fine-tuned specifically for multi-task mental health prediction from social media data. Menta is jointly trained across six classification tasks using a LoRA-based framework, a cross-dataset strategy, and a balanced accuracy--oriented loss. Evaluated against nine state-of-the-art SLM baselines, Menta achieves an average improvement of 15.2\% across tasks covering depression, stress, and suicidality compared with the best-performing non--fine-tuned SLMs. It also achieves higher accuracy on depression and stress classification tasks compared to 13B-parameter LLMs, while being approximately 3.25x smaller. Moreover, we demonstrate real-time, on-device deployment of Menta on an iPhone 15 Pro Max, requiring only approximately 3GB RAM. Supported by a comprehensive benchmark against existing SLMs and LLMs, Menta highlights the potential for scalable, privacy-preserving mental health monitoring. Code is available at: https://hong-labs.github.io/menta-project/

---

### GeoViS: Geospatially Rewarded Visual Search for Remote Sensing Visual Grounding

Authors: Peirong Zhang, Yidan Zhang, Luxiao Xu, Jinliang Lin, Zonghao Guo, Fengxiang Wang, Xue Yang, Kaiwen Wei, Lei Wang
Published: 2025-12-02
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.02715v1

Recent advances in multimodal large language models(MLLMs) have led to remarkable progress in visual grounding, enabling fine-grained cross-modal alignment between textual queries and image regions. However, transferring such capabilities to remote sensing imagery remains challenging, as targets are often extremely small within kilometer-scale scenes, and queries typically involve intricate geospatial relations such as relative positions, spatial hierarchies, or contextual dependencies across distant objects. To address these challenges, we propose GeoViS, a Geospatially Rewarded Visual Search framework that reformulates remote sensing visual grounding as a progressive search-and-reasoning process. Rather than directly predicting the target location in a single step, GeoViS actively explores the global image through a tree-structured sequence of visual cues, integrating multimodal perception, spatial reasoning, and reward-guided exploration to refine geospatial hypotheses iteratively. This design enables the model to detect subtle small-scale targets while maintaining holistic scene awareness. Extensive experiments on five remote sensing grounding benchmarks demonstrate that GeoViS achieves precise geospatial understanding and consistently surpasses existing methods across key visual grounding metrics, highlighting its strong cross-domain generalization and interpretability.

---

### Training Data Attribution for Image Generation using Ontology-Aligned Knowledge Graphs

Authors: Theodoros Aivalis, Iraklis A. Klampanos, Antonis Troumpoukis, Joemon M. Jose
Published: 2025-12-02
Categories: cs.AI
arXiv: https://arxiv.org/abs/2512.02713v1

As generative models become powerful, concerns around transparency, accountability, and copyright violations have intensified. Understanding how specific training data contributes to a model's output is critical. We introduce a framework for interpreting generative outputs through the automatic construction of ontologyaligned knowledge graphs (KGs). While automatic KG construction from natural text has advanced, extracting structured and ontology-consistent representations from visual content remains challenging -- due to the richness and multi-object nature of images. Leveraging multimodal large language models (LLMs), our method extracts structured triples from images, aligned with a domain-specific ontology. By comparing the KGs of generated and training images, we can trace potential influences, enabling copyright analysis, dataset transparency, and interpretable AI. We validate our method through experiments on locally trained models via unlearning, and on large-scale models through a style-specific experiment. Our framework supports the development of AI systems that foster human collaboration, creativity and stimulate curiosity.

---

### CREST: Universal Safety Guardrails Through Cluster-Guided Cross-Lingual Transfer

Authors: Lavish Bansal, Naman Mishra
Published: 2025-12-02
Categories: cs.CL, cs.LG
arXiv: https://arxiv.org/abs/2512.02711v2

Ensuring content safety in large language models (LLMs) is essential for their deployment in real-world applications. However, existing safety guardrails are predominantly tailored for high-resource languages, leaving a significant portion of the world's population underrepresented who communicate in low-resource languages. To address this, we introduce CREST (CRoss-lingual Efficient Safety Transfer), a parameter-efficient multilingual safety classification model that supports 100 languages with only 0.5B parameters. By training on a strategically chosen subset of only 13 high-resource languages, our model utilizes cluster-based cross-lingual transfer from a few to 100 languages, enabling effective generalization to both unseen high-resource and low-resource languages. This approach addresses the challenge of limited training data in low-resource settings. We conduct comprehensive evaluations across six safety benchmarks to demonstrate that CREST outperforms existing state-of-the-art guardrails of comparable scale and achieves competitive results against models with significantly larger parameter counts (2.5B parameters and above). Our findings highlight the limitations of language-specific guardrails and underscore the importance of developing universal, language-agnostic safety systems that can scale effectively to serve global populations.

---

### Empirical Assessment of the Perception of Software Product Line Engineering by an SME before Migrating its Code Base

Authors: Thomas Georges, Marianne Huchard, Mélanie König, Clémentine Nebut, Chouki Tibermacine
Published: 2025-12-02
Categories: cs.SE, cs.AI
arXiv: https://arxiv.org/abs/2512.02707v1

Migrating a set of software variants into a software product line (SPL) is an expensive and potentially challenging endeavor. Indeed, SPL engineering can significantly impact a company's development process and often requires changes to established developer practices. The work presented in this paper stems from a collaboration with a Small and Medium-sized Enterprise (SME) that decided to migrate its existing code base into an SPL. In this study, we conducted an in-depth evaluation of the company's current development processes and practices, as well as the anticipated benefits and risks associated with the migration. Key stakeholders involved in software development participated in this evaluation to provide insight into their perceptions of the migration and their potential resistance to change. This paper describes the design of the interviews conducted with these stakeholders and presents an analysis of the results. Among the qualitative findings, we observed that all participants, regardless of their role in the development process, identified benefits of the migration relevant to their own activities. Furthermore, our results suggest that an effective risk mitigation strategy involves keeping stakeholders informed and engaged throughout the process, preserving as many good practices as possible, and actively involving them in the migration to ensure a smooth transition and minimize potential challenges.

---

### FGC-Comp: Adaptive Neighbor-Grouped Attribute Completion for Graph-based Anomaly Detection

Authors: Junpeng Wu, Pinheng Zong
Published: 2025-12-02
Categories: cs.LG, cs.IR
arXiv: https://arxiv.org/abs/2512.02705v1

Graph-based Anomaly Detection models have gained widespread adoption in recent years, identifying suspicious nodes by aggregating neighborhood information. However, most existing studies overlook the pervasive issues of missing and adversarially obscured node attributes, which can undermine aggregation stability and prediction reliability. To mitigate this, we propose FGC-Comp, a lightweight, classifier-agnostic, and deployment-friendly attribute completion module-designed to enhance neighborhood aggregation under incomplete attributes. We partition each node's neighbors into three label-based groups, apply group-specific transforms to the labeled groups while a node-conditioned gate handles unknowns, fuse messages via residual connections, and train end-to-end with a binary classification objective to improve aggregation stability and prediction reliability under missing attributes. Experiments on two real-world fraud datasets validate the effectiveness of the approach with negligible computational overhead.

---

### Conformal Correction for Efficiency May be at Odds with Entropy

Authors: Senrong Xu, Tianyu Wang, Zenan Li, Yuan Yao, Taolue Chen, Feng Xu, Xiaoxing Ma
Published: 2025-12-02
Categories: cs.LG
arXiv: https://arxiv.org/abs/2512.02704v1

Conformal prediction (CP) provides a comprehensive framework to produce statistically rigorous uncertainty sets for black-box machine learning models. To further improve the efficiency of CP, conformal correction is proposed to fine-tune or wrap the base model with an extra module using a conformal-aware inefficiency loss. In this work, we empirically and theoretically identify a trade-off between the CP efficiency and the entropy of model prediction. We then propose an entropy-constrained conformal correction method, exploring a better Pareto optimum between efficiency and entropy. Extensive experimental results on both computer vision and graph datasets demonstrate the efficacy of the proposed method. For instance, it can significantly improve the efficiency of state-of-the-art CP methods by up to 34.4%, given an entropy threshold.

---

### A method for tissue-mask supported whole-body image registration in the UK Biobank

Authors: Yasemin Utkueri, Elin Lundström, Håkan Ahlström, Johan Öfverstedt, Joel Kullberg
Published: 2025-12-02
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.02702v3

The UK Biobank is a large-scale study collecting imaging and non-imaging health data. Robust and accurate inter-subject image registration of these whole-body MR images would enable their spatial standardization, and region/voxel-wise correlation analysis of non-imaging data with image-derived parameters. We propose a sex-stratified inter-subject whole-body MR image registration approach that uses subcutaneous adipose tissue- and muscle-masks from the state-of-the-art VIBESegmentator method to augment intensity-based graph-cut registration. The proposed mask-supported method was evaluated on 4000 subjects by comparisons to an intensity-only method as well as two other methods, uniGradICON and MIRTK. The evaluation included Jacobian Determinant (JD) folding frequency, Dice scores, and voxel-wise label error frequency. MRSegmentator and TotalSegmentator were also used for independent Dice score evaluations. Additionally, voxel-wise correlation between age and both fat content and tissue volume was studied to exemplify the usefulness for medical research. The proposed method showed lower frequency of JD folding when compared to the intensity-based method. The mask-supported method exhibited a mean Dice score of 0.77 / 0.74 when evaluated on the VIBESegmentator masks for males / females, respectively. In comparison to the intensity-only method, the mean values were 6 pp higher for both sexes, and the label error frequency decreased in most regions. These differences were 9pp / 8pp against uniGradICON and 12pp / 13pp against MIRTK. The proposed method showed higher mean Dice scores in comparison to the other methods when evaluated on MRSegmentator and TotalSegmentator. The age-correlation maps were less noisy and showed higher anatomical alignment. In conclusion, the image registration method using two tissue masks improves whole-body registration of UK Biobank images.

---

### VLM-Pruner: Buffering for Spatial Sparsity in an Efficient VLM Centrifugal Token Pruning Paradigm

Authors: Zhenkai Wu, Xiaowen Ma, Zhenliang Ni, Dengming Zhang, Han Shu, Xin Jiang, Xinghao Chen
Published: 2025-12-02
Categories: cs.CV, cs.LG
arXiv: https://arxiv.org/abs/2512.02700v4

Vision-language models (VLMs) excel at image understanding tasks, but the large number of visual tokens imposes significant computational costs, hindering deployment on mobile devices. Many pruning methods rely solely on token importance and thus overlook inter-token redundancy, retaining numerous duplicated tokens and wasting capacity. Although some redundancy-aware approaches have been proposed, they often ignore the spatial relationships among visual tokens. This can lead to overly sparse selections of retained tokens that fail to adequately cover the regions of target objects. To address these limitations, we propose VLM-Pruner, a training-free token pruning algorithm that explicitly balances redundancy and spatial sparsity. We introduce a centrifugal token pruning paradigm that enables near-to-far selection while prioritizing the preservation of fine-grained object details. Moreover, we design a Buffering for Spatial Sparsity (BSS) criterion that defers the selection of spatially distant tokens. We further adopt a parallel greedy strategy to conduct token selection efficiently. To mitigate information loss from pruning, we selectively fuse salient information from the discarded tokens into the retained ones. Comprehensive comparisons demonstrate that VLM-Pruner consistently outperforms strong baselines across five VLMs with an 88.9\% pruning rate, while delivering an end-to-end inference speedup. The code is available at https://github.com/Casey-bit/VLMPruner.

---

### Learning What to Attend First: Modality-Importance-Guided Reasoning for Reliable Multimodal Emotion Understanding

Authors: Hyeongseop Rha, Jeong Hun Yeo, Junil Won, Se Jin Park, Yong Man Ro
Published: 2025-12-02
Categories: cs.AI
arXiv: https://arxiv.org/abs/2512.02699v1

In this paper, we present Modality-Importance-Guided Reasoning (MIGR), a framework designed to improve the reliability of reasoning-based multimodal emotion understanding in multimodal large language models. Although existing methods have advanced emotion understanding, they often suffer from reasoning drift: models gradually rely on their own generated text instead of multimodal evidence, and their explanations are overly shaped by visually initiated reasoning paths. To address these issues, we introduce Modality Importance (MI), a simple yet effective mechanism for identifying the emotion-dominant modality. Using MI, MIGR reorganizes reasoning sequences so that explanations begin from the modality most critical to the target emotion, preventing early reasoning from being misled by less informative cues. Our two-stage framework-comprising modality-aligned supervised fine-tuning and modality-aware reward optimization-encourages models to generate emotionally grounded, causally relevant, and coherence-preserving explanations. Experimental results on the DFEW benchmark show that MIGR substantially improves reasoning reliability, decreasing instances of correct predictions accompanied by emotionally inconsistent explanations from 18.10% to 7.37%. These results confirm the benefit of initiating reasoning from the emotion-dominant modality.

---

### GeoBridge++: Fact-Guided Geo-Semantic Bridging for Unified Cross-View Geo-Localization

Authors: Zixuan Song, Jing Zhang, Di Wang, Zhiming Luo, Wenbin Liu, Haonan Guo, En Wang, Bo Du, Liangpei Zhang
Published: 2025-12-02
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.02697v5

Cross-view geo-localization infers a location by retrieving geo-tagged reference images matching a query image. However, the traditional satellite-centric paradigm limits robustness when high-resolution or up-to-date satellite imagery is unavailable and underexploits complementary cues across views and modalities. To address these challenges, we propose GeoBridge, a novel model that performs bidirectional matching across views and supports language-to-image retrieval. GeoBridge builds on a novel semantic-anchor mechanism that bridges multi-view features through textual descriptions for robust, flexible localization. We further extend GeoBridge to propose GeoBridge++, a fact-guided geo-semantic bridging framework incorporating real-world geographic knowledge to reduce the ambiguity and instability in appearance-dominated supervision. It integrates structured geographic attributes with visual observations to construct factual descriptions and applies targeted guidance based on modality-specific observable content, thereby enhancing geographic discriminability. GeoBridge++ exploits explicit spatial structures encoded by static maps to build a geo-semantic bridge that adaptively aggregates complementary multi-view information and promotes cross-view consistency. In support of this task, we further construct GeoLoc-MM, a million-scale, multi-view, and multi-scale dataset with aligned drone, satellite, street-view, and static-map imagery at six spatial extents per location, enabling systematic evaluation of arbitrary cross-view retrieval, scale robustness, and cross-view generalization. Extensive experiments show that GeoBridge supports robust cross-view and cross-modal geo-localization, while GeoBridge++ achieves consistent improvements across multiple benchmarks. Code and dataset will be released at https://github.com/MiliLab/GeoBridge.

---

### ALDI-ray: Adapting the ALDI Framework for Security X-ray Object Detection

Authors: Omid Reza Heidari, Yang Wang, Xinxin Zuo
Published: 2025-12-02
Categories: cs.CV, cs.LG
arXiv: https://arxiv.org/abs/2512.02696v1

Domain adaptation in object detection is critical for real-world applications where distribution shifts degrade model performance. Security X-ray imaging presents a unique challenge due to variations in scanning devices and environmental conditions, leading to significant domain discrepancies. To address this, we apply ALDI++, a domain adaptation framework that integrates self-distillation, feature alignment, and enhanced training strategies to mitigate domain shift effectively in this area. We conduct extensive experiments on the EDS dataset, demonstrating that ALDI++ surpasses the state-of-the-art (SOTA) domain adaptation methods across multiple adaptation scenarios. In particular, ALDI++ with a Vision Transformer for Detection (ViTDet) backbone achieves the highest mean average precision (mAP), confirming the effectiveness of transformer-based architectures for cross-domain object detection. Additionally, our category-wise analysis highlights consistent improvements in detection accuracy, reinforcing the robustness of the model across diverse object classes. Our findings establish ALDI++ as an efficient solution for domain-adaptive object detection, setting a new benchmark for performance stability and cross-domain generalization in security X-ray imagery.

---

### Embedding networks with the random walk first return time distribution

Authors: Vedanta Thapar, Renaud Lambiotte, George T. Cantwell
Published: 2025-12-02
Categories: cs.SI, cs.LG
arXiv: https://arxiv.org/abs/2512.02694v3

We propose the first return time distribution (FRTD) of a random walk as an interpretable and mathematically grounded node embedding. The FRTD assigns a probability mass function to each node, allowing us to define a distance between any pair of nodes using standard metrics for discrete distributions. We present several arguments to motivate the FRTD embedding. First, we show that FRTDs are strictly more informative than eigenvalue spectra, yet insufficient for complete graph identification, thus placing FRTD equivalence between cospectrality and isomorphism. Second, we argue that FRTD equivalence between nodes captures structural similarity. Third, we empirically demonstrate that the FRTD embedding outperforms manually designed graph metrics in network alignment tasks. Finally, we show that random networks that approximately match the FRTD of a desired target also preserve other salient features. Together these results demonstrate the FRTD as a simple and mathematically principled embedding for complex networks.

---

### An Empirical Survey of Model Merging Algorithms for Social Bias Mitigation

Authors: Daiki Shirafuji, Tatsuhiko Saito, Yasutomo Kimura
Published: 2025-12-02
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2512.02689v1

Large language models (LLMs) are known to inherit and even amplify societal biases present in their pre-training corpora, threatening fairness and social trust. To address this issue, recent work has explored ``editing'' LLM parameters to mitigate social bias with model merging approaches; however, there is no empirical comparison. In this work, we empirically survey seven algorithms: Linear, Karcher Mean, SLERP, NuSLERP, TIES, DELLA, and Nearswap, applying 13 open weight models in the GPT, LLaMA, and Qwen families. We perform a comprehensive evaluation using three bias datasets (BBQ, BOLD, and HONEST) and measure the impact of these techniques on LLM performance in downstream tasks of the SuperGLUE benchmark. We find a trade-off between bias reduction and downstream performance: methods achieving greater bias mitigation degrade accuracy, particularly on tasks requiring reading comprehension and commonsense and causal reasoning. Among the merging algorithms, Linear, SLERP, and Nearswap consistently reduce bias while maintaining overall performance, with SLERP at moderate interpolation weights emerging as the most balanced choice. These results highlight the potential of model merging algorithms for bias mitigation, while indicating that excessive debiasing or inappropriate merging methods may lead to the degradation of important linguistic abilities.

---