# Research Papers — 2025-10 · bundle page 148 of 217

> Full-text markdown of papers #7351–7400 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/148
[← Previous bundle](https://aipod.dev/papers/archive/2025-10/page/147.md)
[Next bundle →](https://aipod.dev/papers/archive/2025-10/page/149.md)

### ReFIne: A Framework for Trustworthy Large Reasoning Models with Reliability, Faithfulness, and Interpretability

Authors: Chung-En Sun, Ge Yan, Akshay Kulkarni, Tsui-Wei Weng
Published: 2025-10-10
Categories: cs.CL
arXiv: https://arxiv.org/abs/2510.09062v2

Recent advances in long chain-of-thought (CoT) reasoning have largely prioritized answer accuracy and token efficiency, while overlooking aspects critical to trustworthiness. We argue that usable reasoning systems must be trustworthy, characterized by three properties: interpretability, faithfulness, and reliability. To this end, we propose ReFIne, a new training framework that integrates supervised fine-tuning with GRPO to encourage models to: (i) improve interpretability by producing structured, tag-based traces with high-level planning that are easier for humans to follow; (ii) enhance faithfulness by explicitly disclosing the decisive information guiding each solution, with consistent cross-section references; and (iii) promote reliability by providing self-assessments of both the derivation's soundness and the confidence of the final answer. We apply ReFIne to the Qwen3 models at multiple scales (1.7B/4B/8B) and evaluate across mathematical benchmarks of varying difficulty. Our experimental results show that ReFIne models generate clearer and better-structured reasoning traces (interpretability +44.0%), more faithfully expose their underlying decision process (faithfulness +18.8%), and offer informative confidence estimates (reliability +42.4%). These findings highlight an overlooked but important direction: reasoning models should be optimized not only for accuracy, but also for broader dimensions of trustworthiness. Our code is available at: https://github.com/Trustworthy-ML-Lab/Training_Trustworthy_LRM_with_Refine

---

### Letting Trajectories Spread: Quality-Preserving Control for Diverse Flow Matching

Authors: Jingxuan Wu, Zhenglin Wan, Xingrui Yu, Yuzhe Yang, Bo An, Ivor Tsang, Yang You
Published: 2025-10-10
Categories: cs.AI, cs.CV
arXiv: https://arxiv.org/abs/2510.09060v2

Flow-based text-to-image models follow deterministic trajectories, making it costly to explore diverse modes under limited sampling budgets. Existing approaches to improving diversity often rely on retraining or degrade image fidelity. To address this limitation, we present a training-free, inference-time control mechanism that makes the flow itself diversity-aware. Our core insight is to encourage diversity through guidance that is geometrically decoupled from the mode's quality-seeking direction. Our method simultaneously encourages lateral spread among trajectories via a feature-space objective and reintroduces uncertainty through a time-scheduled stochastic perturbation. Crucially, this perturbation is projected to be orthogonal to the generation flow, a geometric constraint that allows it to boost variation without degrading image details or prompt fidelity. Theoretically, we show that this design monotonically increases a volume surrogate while approximately preserving the marginal distribution, providing a principled explanation for the robustness of generation quality. Empirically, across multiple text-to-image settings under fixed sampling budgets, our method consistently improves diversity metrics such as the Vendi Score and Brisque over strong baselines, while upholding image quality and alignment.

---

### Lesion-Aware Post-Training of Latent Diffusion Models for Synthesizing Diffusion MRI from CT Perfusion

Authors: Junhyeok Lee, Hyunwoong Kim, Hyungjin Chung, Heeseong Eom, Joon Jang, Chul-Ho Sohn, Kyu Sung Choi
Published: 2025-10-10
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.09056v1

Image-to-Image translation models can help mitigate various challenges inherent to medical image acquisition. Latent diffusion models (LDMs) leverage efficient learning in compressed latent space and constitute the core of state-of-the-art generative image models. However, this efficiency comes with a trade-off, potentially compromising crucial pixel-level detail essential for high-fidelity medical images. This limitation becomes particularly critical when generating clinically significant structures, such as lesions, which often occupy only a small portion of the image. Failure to accurately reconstruct these regions can severely impact diagnostic reliability and clinical decision-making. To overcome this limitation, we propose a novel post-training framework for LDMs in medical image-to-image translation by incorporating lesion-aware medical pixel space objectives. This approach is essential, as it not only enhances overall image quality but also improves the precision of lesion delineation. We evaluate our framework on brain CT-to-MRI translation in acute ischemic stroke patients, where early and accurate diagnosis is critical for optimal treatment selection and improved patient outcomes. While diffusion MRI is the gold standard for stroke diagnosis, its clinical utility is often constrained by high costs and low accessibility. Using a dataset of 817 patients, we demonstrate that our framework improves overall image quality and enhances lesion delineation when synthesizing DWI and ADC images from CT perfusion scans, outperforming existing image-to-image translation models. Furthermore, our post-training strategy is easily adaptable to pre-trained LDMs and exhibits substantial potential for broader applications across diverse medical image translation tasks.

---

### Alif: Advancing Urdu Large Language Models via Multilingual Synthetic Data Distillation

Authors: Muhammad Ali Shafique, Kanwal Mehreen, Muhammad Arham, Maaz Amjad, Sabur Butt, Hamza Farooq
Published: 2025-10-10
Categories: cs.CL, cs.AI, cs.LG
arXiv: https://arxiv.org/abs/2510.09051v1

Developing a high-performing large language models (LLMs) for low-resource languages such as Urdu, present several challenges. These challenges include the scarcity of high-quality datasets, multilingual inconsistencies, and safety concerns. Existing multilingual LLMs often address these issues by translating large volumes of available data. However, such translations often lack quality and cultural nuance while also incurring significant costs for data curation and training. To address these issues, we propose Alif-1.0-8B-Instruct, a multilingual Urdu-English model, that tackles these challenges with a unique approach. We train the model on a high-quality, multilingual synthetic dataset (Urdu-Instruct), developed using a modified self-instruct technique. By using unique prompts and seed values for each task along with a global task pool, this dataset incorporates Urdu-native chain-of-thought based reasoning, bilingual translation, cultural relevance, and ethical safety alignments. This technique significantly enhances the comprehension of Alif-1.0-8B-Instruct model for Urdu-specific tasks. As a result, Alif-1.0-8B-Instruct, built upon the pretrained Llama-3.1-8B, demonstrates superior performance compared to Llama-3.1-8B-Instruct for Urdu specific-tasks. It also outperformed leading multilingual LLMs, including Mistral-7B-Instruct-v0.3, Qwen-2.5-7B-Instruct, and Cohere-Aya-Expanse-8B, all within a training budget of under $100. Our results demonstrate that high-performance and low-resource language LLMs can be developed efficiently and culturally aligned using our modified self-instruct approach. All datasets, models, and code are publicly available at: https://github.com/traversaal-ai/alif-urdu-llm.

---

### MEC$^3$O: Multi-Expert Consensus for Code Time Complexity Prediction

Authors: Joonghyuk Hahn, Soohan Lim, Yo-Sub Han
Published: 2025-10-10
Categories: cs.AI, cs.SE
arXiv: https://arxiv.org/abs/2510.09049v1

Predicting the complexity of source code is essential for software development and algorithm analysis. Recently, Baik et al. (2025) introduced CodeComplex for code time complexity prediction. The paper shows that LLMs without fine-tuning struggle with certain complexity classes. This suggests that no single LLM excels at every class, but rather each model shows advantages in certain classes. We propose MEC$^3$O, a multi-expert consensus system, which extends the multi-agent debate frameworks. MEC$^3$O assigns LLMs to complexity classes based on their performance and provides them with class-specialized instructions, turning them into experts. These experts engage in structured debates, and their predictions are integrated through a weighted consensus mechanism. Our expertise assignments to LLMs effectively handle Degeneration-of-Thought, reducing reliance on a separate judge model, and preventing convergence to incorrect majority opinions. Experiments on CodeComplex show that MEC$^3$O outperforms the open-source baselines, achieving at least 10% higher accuracy and macro-F1 scores. It also surpasses GPT-4o-mini in macro-F1 scores on average and demonstrates competitive on-par F1 scores to GPT-4o and GPT-o4-mini on average. This demonstrates the effectiveness of multi-expert debates and weight consensus strategy to generate the final predictions. Our code and data is available at https://github.com/suhanmen/MECO.

---

### Spatio-Temporal Graph Convolutional Networks for EV Charging Demand Forecasting Using Real-World Multi-Modal Data Integration

Authors: Jose Tupayachi, Mustafa C. Camur, Kevin Heaslip, Xueping Li
Published: 2025-10-10
Categories: cs.LG
arXiv: https://arxiv.org/abs/2510.09048v3

Transportation remains a major contributor to greenhouse gas emissions, highlighting the urgency of transitioning toward sustainable alternatives such as electric vehicles (EVs). Yet, uneven spatial distribution and irregular utilization of charging infrastructure create challenges for both power grid stability and investment planning. This study introduces TW-GCN, a spatio-temporal forecasting framework that combines Graph Convolutional Networks with temporal architectures to predict EV charging demand in Tennessee, United States (U.S.). We utilize real-world traffic flows, weather conditions, and proprietary data provided by one of the largest EV infrastructure company in the U.S. to capture both spatial dependencies and temporal dynamics. Extensive experiments across varying lag horizons, clustering strategies, and sequence lengths reveal that mid-horizon (3-hour) forecasts achieve the best balance between responsiveness and stability, with 1DCNN consistently outperforming other temporal models. Regional analysis shows disparities in predictive accuracy across East, Middle, and West Tennessee, reflecting how station density, population, and local demand variability shape model performance. The proposed TW-GCN framework advances the integration of data-driven intelligence into EV infrastructure planning, supporting both sustainable mobility transitions and resilient grid management.

---

### LLM Based Long Code Translation using Identifier Replacement

Authors: Manojit Chakraborty, Madhusudan Ghosh, Rishabh Gupta
Published: 2025-10-10
Categories: cs.SE, cs.AI, cs.IR, cs.LG
arXiv: https://arxiv.org/abs/2510.09045v2

In the domain of software development, LLMs have been utilized to automate tasks such as code translation, where source code from one programming language is translated to another while preserving its functionality. However, LLMs often struggle with long source codes that don't fit into the context window, which produces inaccurate translations. To address this, we propose a novel zero-shot code translation method that incorporates identifier replacement. By substituting user-given long identifiers with generalized placeholders during translation, our method allows the LLM to focus on the logical structure of the code, by reducing token count and memory usage, which improves the efficiency and cost-effectiveness of long code translation. Our empirical results demonstrate that our approach preserves syntactical and hierarchical information and produces translation results with reduced tokens.

---

### Humanoid Artificial Consciousness Designed with Large Language Model Based on Psychoanalysis and Personality Theory

Authors: Sang Hun Kim, Jongmin Lee, Dongkyu Park, So Young Lee, Yosep Chong
Published: 2025-10-10
Categories: cs.AI
arXiv: https://arxiv.org/abs/2510.09043v2

Human consciousness is still a concept hard to define with current scientific understanding. Although Large Language Models (LLMs) have recently demonstrated significant advancements across various domains including translation and summarization, human consciousness is not something to imitate with current upfront technology owing to so-called hallucination. This study, therefore, proposes a novel approach to address these challenges by integrating psychoanalysis and the Myers-Briggs Type Indicator (MBTI) into constructing consciousness and personality modules. We developed three artificial consciousnesses (self-awareness, unconsciousness, and preconsciousness) based on the principles of psychoanalysis. Additionally, we designed 16 characters with different personalities representing the sixteen MBTI types, with several attributes such as needs, status, and memories. To determine if our model's artificial consciousness exhibits human-like cognition, we created ten distinct situations considering seven attributes such as emotional understanding and logical thinking. The decision-making process of artificial consciousness and the final action were evaluated in three ways: survey evaluation, three-tier classification via ChatGPT, and qualitative review. Both quantitative and qualitative analyses indicated a high likelihood of well-simulated consciousness, although the difference in response between different characters and consciousnesses was not very significant. This implies that the developed models incorporating elements of psychoanalysis and personality theory can lead to building a more intuitive and adaptable AI system with humanoid consciousness. Therefore, this study contributes to opening up new avenues for improving AI interactions in complex cognitive contexts.

---

### MAKO: Meta-Adaptive Koopman Operators for Learning-based Model Predictive Control of Parametrically Uncertain Nonlinear Systems

Authors: Minghao Han, Kiwan Wong, Adrian Wing-Keung Law, Xunyuan Yin
Published: 2025-10-10
Categories: eess.SY, cs.LG
arXiv: https://arxiv.org/abs/2510.09042v2

In this work, we propose a meta-learning-based Koopman modeling and predictive control approach for nonlinear systems with parametric uncertainties. An adaptive deep meta-learning-based modeling approach, called Meta Adaptive Koopman Operator (MAKO), is proposed. Without knowledge of the parametric uncertainty, the proposed MAKO approach can learn a meta-model from a multi-modal dataset and efficiently adapt to new systems with previously unseen parameter settings by using online data. Based on the learned meta Koopman model, a predictive control scheme is developed, and the stability of the closed-loop system is ensured even in the presence of previously unseen parameter settings. Through extensive simulations, our proposed approach demonstrates superior performance in both modeling accuracy and control efficacy as compared to competitive baselines.

---

### Robust Driving Control for Autonomous Vehicles: An Intelligent General-sum Constrained Adversarial Reinforcement Learning Approach

Authors: Junchao Fan, Qi Wei, Ruichen Zhang, Yang Lu, Jianhua Wang, Xiaolin Chang, Bo Ai
Published: 2025-10-10
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2510.09041v3

Deep reinforcement learning (DRL) has demonstrated remarkable success in developing autonomous driving policies. However, its vulnerability to adversarial attacks remains a critical barrier to real-world deployment. Although existing robust methods have achieved success, they still suffer from three key issues: (i) these methods are trained against myopic adversarial attacks, limiting their abilities to respond to more strategic threats, (ii) they have trouble causing truly safety-critical events (e.g., collisions), but instead often result in minor consequences, and (iii) these methods can introduce learning instability and policy drift during training due to the lack of robust constraints. To address these issues, we propose Intelligent General-sum Constrained Adversarial Reinforcement Learning (IGCARL), a novel robust autonomous driving approach that consists of a strategic targeted adversary and a robust driving agent. The strategic targeted adversary is designed to leverage the temporal decision-making capabilities of DRL to execute strategically coordinated multi-step attacks. In addition, it explicitly focuses on inducing safety-critical events by adopting a general-sum objective. The robust driving agent learns by interacting with the adversary to develop a robust autonomous driving policy against adversarial attacks. To ensure stable learning in adversarial environments and to mitigate policy drift caused by attacks, the agent is optimized under a constrained formulation. Extensive experiments show that IGCARL improves the success rate by at least 27.9% over state-of-the-art methods, demonstrating superior robustness to adversarial attacks and enhancing the safety and reliability of DRL-based autonomous driving.

---

### Auto-scaling Continuous Memory for GUI Agent

Authors: Wenyi Wu, Kun Zhou, Ruoxin Yuan, Vivian Yu, Stephen Wang, Zhiting Hu, Biwei Huang
Published: 2025-10-10
Categories: cs.AI, cs.CL, cs.CV, cs.CY, cs.LG
arXiv: https://arxiv.org/abs/2510.09038v1

We study how to endow GUI agents with scalable memory that help generalize across unfamiliar interfaces and long-horizon tasks. Prior GUI agents compress past trajectories into text tokens, which balloons context length and misses decisive visual cues (e.g., exact widget size and position). We propose a continuous memory that encodes each GUI trajectory into a fixed-length sequence of continuous embeddings using the VLM itself as an encoder; these embeddings are plugged directly into the backbone's input layer, sharply reducing context cost while preserving fine-grained visual information. As memory size and retrieval depth increase, performance improves monotonically, unlike text memories that degrade with long prompts. To grow memory at low cost, we introduce an auto-scaling data flywheel that (i) discovers new environments via search, (ii) synthesizes tasks with an open-source VLM, (iii) rolls out trajectories with the agent, and (iv) verifies success with the same VLM. Using this pipeline, we collect 100k+ trajectories for about \$4000 and fine-tune only the memory encoder (LoRA on a Q-Former, 1.2\% parameters) with 1,500 samples. On real-world GUI benchmarks, our memory-augmented agent consistently improves success rates under long horizons and distribution shifts. Notably, Qwen-2.5-VL-7B + continuous memory achieves performance comparable to state-of-the-art closed-source models (e.g., GPT-4o, Claude-4).

---

### Repairing Regex Vulnerabilities via Localization-Guided Instructions

Authors: Sicheol Sung, Joonghyuk Hahn, Yo-Sub Han
Published: 2025-10-10
Categories: cs.AI, cs.PL
arXiv: https://arxiv.org/abs/2510.09037v1

Regular expressions (regexes) are foundational to modern computing for critical tasks like input validation and data parsing, yet their ubiquity exposes systems to regular expression denial of service (ReDoS), a vulnerability requiring automated repair methods. Current approaches, however, are hampered by a trade-off. Symbolic, rule-based system are precise but fails to repair unseen or complex vulnerability patterns. Conversely, large language models (LLMs) possess the necessary generalizability but are unreliable for tasks demanding strict syntactic and semantic correctness. We resolve this impasse by introducing a hybrid framework, localized regex repair (LRR), designed to harness LLM generalization while enforcing reliability. Our core insight is to decouple problem identification from the repair process. First, a deterministic, symbolic module localizes the precise vulnerable subpattern, creating a constrained and tractable problem space. Then, the LLM invoked to generate a semantically equivalent fix for this isolated segment. This combined architecture successfully resolves complex repair cases intractable for rule-based repair while avoiding the semantic errors of LLM-only approaches. Our work provides a validated methodology for solving such problems in automated repair, improving the repair rate by 15.4%p over the state-of-the-art. Our code is available at https://github.com/cdltlehf/LRR.

---

### Exploring Single Domain Generalization of LiDAR-based Semantic Segmentation under Imperfect Labels

Authors: Weitong Kong, Zichao Zeng, Di Wen, Jiale Wei, Kunyu Peng, June Moh Goo, Jan Boehm, Rainer Stiefelhagen
Published: 2025-10-10
Categories: cs.CV, cs.LG, cs.RO
arXiv: https://arxiv.org/abs/2510.09035v2

Accurate perception is critical for vehicle safety, with LiDAR as a key enabler in autonomous driving. To ensure robust performance across environments, sensor types, and weather conditions without costly re-annotation, domain generalization in LiDAR-based 3D semantic segmentation is essential. However, LiDAR annotations are often noisy due to sensor imperfections, occlusions, and human errors. Such noise degrades segmentation accuracy and is further amplified under domain shifts, threatening system reliability. While noisy-label learning is well-studied in images, its extension to 3D LiDAR segmentation under domain generalization remains largely unexplored, as the sparse and irregular structure of point clouds limits direct use of 2D methods. To address this gap, we introduce the novel task Domain Generalization for LiDAR Semantic Segmentation under Noisy Labels (DGLSS-NL) and establish the first benchmark by adapting three representative noisy-label learning strategies from image classification to 3D segmentation. However, we find that existing noisy-label learning approaches adapt poorly to LiDAR data. We therefore propose DuNe, a dual-view framework with strong and weak branches that enforce feature-level consistency and apply cross-entropy loss based on confidence-aware filtering of predictions. Our approach shows state-of-the-art performance by achieving 56.86% mIoU on SemanticKITTI, 42.28% on nuScenes, and 52.58% on SemanticPOSS under 10% symmetric label noise, with an overall Arithmetic Mean (AM) of 49.57% and Harmonic Mean (HM) of 48.50%, thereby demonstrating robust domain generalization in DGLSS-NL tasks. The code is available on our project page.

---

### Convergence of optimizers implies eigenvalues filtering at equilibrium

Authors: Jerome Bolte, Quoc-Tung Le, Edouard Pauwels
Published: 2025-10-10
Categories: cs.LG, math.DS, math.OC
arXiv: https://arxiv.org/abs/2510.09034v1

Ample empirical evidence in deep neural network training suggests that a variety of optimizers tend to find nearly global optima. In this article, we adopt the reversed perspective that convergence to an arbitrary point is assumed rather than proven, focusing on the consequences of this assumption. From this viewpoint, in line with recent advances on the edge-of-stability phenomenon, we argue that different optimizers effectively act as eigenvalue filters determined by their hyperparameters. Specifically, the standard gradient descent method inherently avoids the sharpest minima, whereas Sharpness-Aware Minimization (SAM) algorithms go even further by actively favoring wider basins. Inspired by these insights, we propose two novel algorithms that exhibit enhanced eigenvalue filtering, effectively promoting wider minima. Our theoretical analysis leverages a generalized Hadamard--Perron stable manifold theorem and applies to general semialgebraic $C^2$ functions, without requiring additional non-degeneracy conditions or global Lipschitz bound assumptions. We support our conclusions with numerical experiments on feed-forward neural networks.

---

### Do LLMs Really Know What They Don't Know? Internal States Mainly Reflect Knowledge Recall Rather Than Truthfulness

Authors: Chi Seng Cheang, Hou Pong Chan, Wenxuan Zhang, Yang Deng
Published: 2025-10-10
Categories: cs.CL
arXiv: https://arxiv.org/abs/2510.09033v3

Recent work suggests that LLMs "know what they don't know", positing that hallucinated and factually correct outputs arise from distinct internal processes and can therefore be distinguished using internal signals. However, hallucinations have multifaceted causes: beyond simple knowledge gaps, they can emerge from training incentives that encourage models to exploit statistical shortcuts or spurious associations learned during pretraining. In this paper, we argue that when LLMs rely on such learned associations to produce hallucinations, their internal processes are mechanistically similar to those of factual recall, as both stem from strong statistical correlations encoded in the model's parameters. To verify this, we propose a novel taxonomy categorizing hallucinations into Unassociated Hallucinations (UHs), where outputs lack parametric grounding, and Associated Hallucinations (AHs), which are driven by spurious associations. Through mechanistic analysis, we compare their computational processes and hidden-state geometries with factually correct outputs. Our results show that hidden states primarily reflect whether the model is recalling parametric knowledge rather than the truthfulness of the output itself. Consequently, AHs exhibit hidden-state geometries that largely overlap with factual outputs, rendering standard detection methods ineffective. In contrast, UHs exhibit distinctive, clustered representations that facilitate reliable detection.

---

### Exploring Cross-Lingual Knowledge Transfer via Transliteration-Based MLM Fine-Tuning for Critically Low-resource Chakma Language

Authors: Adity Khisa, Nusrat Jahan Lia, Tasnim Mahfuz Nafis, Zarif Masud, Tanzir Pial, Shebuti Rayana, Ahmedul Kabir
Published: 2025-10-10
Categories: cs.CL
arXiv: https://arxiv.org/abs/2510.09032v2

As an Indo-Aryan language with limited available data, Chakma remains largely underrepresented in language models. In this work, we introduce a novel corpus of contextually coherent Bangla-transliterated Chakma, curated from Chakma literature, and validated by native speakers. Using this dataset, we fine-tune six encoder-based transformer models, including multilingual (mBERT, XLM-RoBERTa, DistilBERT), regional (BanglaBERT, IndicBERT), and monolingual English (DeBERTaV3) variants on masked language modeling (MLM) tasks. Our experiments show that fine-tuned multilingual models outperform their pre-trained counterparts when adapted to Bangla-transliterated Chakma, achieving up to 73.54% token accuracy and a perplexity as low as 2.90. Our analysis further highlights the impact of data quality on model performance and shows the limitations of OCR pipelines for morphologically rich Indic scripts. Our research demonstrates that Bangla-transliterated Chakma can be very effective for transfer learning for Chakma language, and we release our dataset to encourage further research on multilingual language modeling for low-resource languages.

---

### Automated Refinement of Essay Scoring Rubrics for Language Models via Reflect-and-Revise

Authors: Keno Harada, Lui Yoshida, Takeshi Kojima, Yusuke Iwasawa, Yutaka Matsuo
Published: 2025-10-10
Categories: cs.CL
arXiv: https://arxiv.org/abs/2510.09030v1

The performance of Large Language Models (LLMs) is highly sensitive to the prompts they are given. Drawing inspiration from the field of prompt optimization, this study investigates the potential for enhancing Automated Essay Scoring (AES) by refining the scoring rubrics used by LLMs. Specifically, our approach prompts models to iteratively refine rubrics by reflecting on models' own scoring rationales and observed discrepancies with human scores on sample essays. Experiments on the TOEFL11 and ASAP datasets using GPT-4.1, Gemini-2.5-Pro, and Qwen-3-Next-80B-A3B-Instruct show Quadratic Weighted Kappa (QWK) improvements of up to 0.19 and 0.47, respectively. Notably, even with a simple initial rubric, our approach achieves comparable or better QWK than using detailed human-authored rubrics. Our findings highlight the importance of iterative rubric refinement in LLM-based AES to enhance alignment with human evaluations.

---

### Déréverbération non-supervisée de la parole par modèle hybride

Authors: Louis Bahrman, Mathieu Fontaine, Gaël Richard
Published: 2025-10-10
Categories: cs.SD, cs.AI, eess.AS
arXiv: https://arxiv.org/abs/2510.09025v1

This paper introduces a new training strategy to improve speech dereverberation systems in an unsupervised manner using only reverberant speech. Most existing algorithms rely on paired dry/reverberant data, which is difficult to obtain. Our approach uses limited acoustic information, like the reverberation time (RT60), to train a dereverberation system. Experimental results demonstrate that our method achieves more consistent performance across various objective metrics than the state-of-the-art.

---

### The Attacker Moves Second: Stronger Adaptive Attacks Bypass Defenses Against Llm Jailbreaks and Prompt Injections

Authors: Milad Nasr, Nicholas Carlini, Chawin Sitawarin, Sander V. Schulhoff, Jamie Hayes, Michael Ilie, Juliette Pluto, Shuang Song, Harsh Chaudhari, Ilia Shumailov, Abhradeep Thakurta, Kai Yuanqing Xiao et al.
Published: 2025-10-10
Categories: cs.LG, cs.CR
arXiv: https://arxiv.org/abs/2510.09023v1

How should we evaluate the robustness of language model defenses? Current defenses against jailbreaks and prompt injections (which aim to prevent an attacker from eliciting harmful knowledge or remotely triggering malicious actions, respectively) are typically evaluated either against a static set of harmful attack strings, or against computationally weak optimization methods that were not designed with the defense in mind. We argue that this evaluation process is flawed. Instead, we should evaluate defenses against adaptive attackers who explicitly modify their attack strategy to counter a defense's design while spending considerable resources to optimize their objective. By systematically tuning and scaling general optimization techniques-gradient descent, reinforcement learning, random search, and human-guided exploration-we bypass 12 recent defenses (based on a diverse set of techniques) with attack success rate above 90% for most; importantly, the majority of defenses originally reported near-zero attack success rates. We believe that future defense work must consider stronger attacks, such as the ones we describe, in order to make reliable and convincing claims of robustness.

---

### The Environmental Impacts of Language Model Training Keep Rising Now is the Time to Catch Impacts on the Rebound

Authors: Clément Morand, Anne-Laure Ligozat, Aurélie Névéol
Published: 2025-10-10
Categories: cs.LG
arXiv: https://arxiv.org/abs/2510.09022v2

Recent Machine Learning (ML) approaches have shown increased performance on benchmarks at the cost of escalating compute demands. Hardware, algorithmic and carbon optimizations have been proposed to curb energy use and environmental impacts. We estimate the environmental impacts associated with training models documented in the Epoch AI database over the last decade, with a particular focus on impacts associated with Large Language Models and the hardware used to train them. We find that energy use and environmental impacts associated with training ML models have increased exponentially, even when considering impact reduction strategies such as using less carbon intensive electricity mixes or more efficient hardware. Optimization strategies do not mitigate the impacts induced by model training, suggesting rebound effect. We show that the impacts of hardware must be considered over the entire life cycle rather than the sole use phase in order to avoid impact shifting. Our study demonstrates that increasing efficiency alone does not ensure sustainability. There is an urgent need to systematically integrate environmental impacts in NLP evaluation practices to better inform the community and support the use of impact as a feature in research planning and decision making.

---

### RefGrader: Automated Grading of Mathematical Competition Proofs using Agentic Workflows

Authors: Hamed Mahdavi, Pouria Mahdavinia, Samira Malek, Pegah Mohammadipour, Alireza Hashemi, Majid Daliri, Alireza Farhadi, Amir Khasahmadi, Niloofar Mireshghallah, Vasant Honavar
Published: 2025-10-10
Categories: cs.AI, cs.LG
arXiv: https://arxiv.org/abs/2510.09021v1

State-of-the-art (SOTA) LLMs have progressed from struggling on proof-based Olympiad problems to solving most of the IMO 2025 problems, with leading systems reportedly handling 5 of 6 problems. Given this progress, we assess how well these models can grade proofs: detecting errors, judging their severity, and assigning fair scores beyond binary correctness. We study proof-analysis capabilities using a corpus of 90 Gemini 2.5 Pro-generated solutions that we grade on a 1-4 scale with detailed error annotations, and on MathArena solution sets for IMO/USAMO 2025 scored on a 0-7 scale. Our analysis shows that models can reliably flag incorrect (including subtly incorrect) solutions but exhibit calibration gaps in how partial credit is assigned. To address this, we introduce agentic workflows that extract and analyze reference solutions and automatically derive problem-specific rubrics for a multi-step grading process. We instantiate and compare different design choices for the grading workflows, and evaluate their trade-offs. Across our annotated corpus and MathArena, our proposed workflows achieve higher agreement with human grades and more consistent handling of partial credit across metrics. We release all code, data, and prompts/logs to facilitate future research.

---

### MagicDock: Toward Docking-oriented De Novo Ligand Design via Gradient Inversion

Authors: Zekai Chen, Xunkai Li, Sirui Zhang, Henan Sun, Jia Li, Zhenjun Li, Bing Zhou, Rong-Hua Li, Guoren Wang
Published: 2025-10-10
Categories: cs.LG
arXiv: https://arxiv.org/abs/2510.09020v1

De novo ligand design is a fundamental task that seeks to generate protein or molecule candidates that can effectively dock with protein receptors and achieve strong binding affinity entirely from scratch. It holds paramount significance for a wide spectrum of biomedical applications. However, most existing studies are constrained by the \textbf{Pseudo De Novo}, \textbf{Limited Docking Modeling}, and \textbf{Inflexible Ligand Type}. To address these issues, we propose MagicDock, a forward-looking framework grounded in the progressive pipeline and differentiable surface modeling. (1) We adopt a well-designed gradient inversion framework. To begin with, general docking knowledge of receptors and ligands is incorporated into the backbone model. Subsequently, the docking knowledge is instantiated as reverse gradient flows by binding prediction, which iteratively guide the de novo generation of ligands. (2) We emphasize differentiable surface modeling in the docking process, leveraging learnable 3D point-cloud representations to precisely capture binding details, thereby ensuring that the generated ligands preserve docking validity through direct and interpretable spatial fingerprints. (3) We introduce customized designs for different ligand types and integrate them into a unified gradient inversion framework with flexible triggers, thereby ensuring broad applicability. Moreover, we provide rigorous theoretical guarantees for each component of MagicDock. Extensive experiments across 9 scenarios demonstrate that MagicDock achieves average improvements of 27.1\% and 11.7\% over SOTA baselines specialized for protein or molecule ligand design, respectively.

---

### Slim Scheduler: A Runtime-Aware RL and Scheduler System for Efficient CNN Inference

Authors: Ian Harshbarger, Calvin Chidambaram
Published: 2025-10-10
Categories: cs.LG
arXiv: https://arxiv.org/abs/2510.09018v1

Most neural network scheduling research focuses on optimizing static, end-to-end models of fixed width, overlooking dynamic approaches that adapt to heterogeneous hardware and fluctuating runtime conditions. We present Slim Scheduler, a hybrid scheduling framework that integrates a Proximal Policy Optimization (PPO) reinforcement learning policy with algorithmic, greedy schedulers to coordinate distributed inference for slimmable models. Each server runs a local greedy scheduler that batches compatible requests and manages instance scaling based on VRAM and utilization constraints, while the PPO router learns global routing policies for device selection, width ratio, and batch configuration. This hierarchical design reduces search space complexity, mitigates overfitting to specific hardware, and balances efficiency and throughput. Compared to a purely randomized task distribution baseline, Slim Scheduler can achieve various accuracy and latency trade-offs such as: A 96.45% reduction in mean latency and a 97.31% reduction in energy usage dropping accuracy to the slimmest model available (70.3%). It can then accomplish an overall reduction in average latency plus energy consumption with an increase in accuracy at the cost of higher standard deviations of said latency and energy, effecting overall task throughput.

---

### Value-State Gated Attention for Mitigating Extreme-Token Phenomena in Transformers

Authors: Rui Bu, Haofeng Zhong, Wenzheng Chen, Yangyan Li
Published: 2025-10-10
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2510.09017v3

Large models based on the Transformer architecture are susceptible to extreme-token phenomena, such as attention sinks and value-state drains. These issues, which degrade model performance, quantization fidelity, and interpretability, arise from a problematic mutual reinforcement mechanism where the model learns an inefficient 'no-op' behavior by focusing attention on tokens with near-zero value states. In this paper, we propose Value-State Gated Attention (VGA), a simple, dedicated, and stable architectural mechanism for performing 'no-op' attention efficiently by directly breaking this cycle. VGA introduces a learnable, data-dependent gate, computed directly from the value vectors (V), to modulate the output. Through a theoretical analysis of the underlying gradients, we show that gating the value-state with a function of itself is more effective at decoupling value and attention score updates than prior methods that gate on input embeddings. This creates a direct regulatory pathway that allows the model to suppress a token's contribution based on its emergent value representation. Our experiments demonstrate that VGA significantly mitigates the formation of attention sinks and stabilizes value-state norms, leading to improved performance, robust quantization fidelity, and enhanced model interpretability.

---

### DiTSinger: Scaling Singing Voice Synthesis with Diffusion Transformer and Implicit Alignment

Authors: Zongcai Du, Guilin Deng, Xiaofeng Guo, Xin Gao, Linke Li, Kaichang Cheng, Fubo Han, Siyu Yang, Peng Liu, Pan Zhong, Qiang Fu
Published: 2025-10-10
Categories: cs.SD, cs.AI, eess.AS
arXiv: https://arxiv.org/abs/2510.09016v2

Recent progress in diffusion-based Singing Voice Synthesis (SVS) demonstrates strong expressiveness but remains limited by data scarcity and model scalability. We introduce a two-stage pipeline: a compact seed set of human-sung recordings is constructed by pairing fixed melodies with diverse LLM-generated lyrics, and melody-specific models are trained to synthesize over 500 hours of high-quality Chinese singing data. Building on this corpus, we propose DiTSinger, a Diffusion Transformer with RoPE and qk-norm, systematically scaled in depth, width, and resolution for enhanced fidelity. Furthermore, we design an implicit alignment mechanism that obviates phoneme-level duration labels by constraining phoneme-to-acoustic attention within character-level spans, thereby improving robustness under noisy or uncertain alignments. Extensive experiments validate that our approach enables scalable, alignment-free, and high-fidelity SVS.

---

### LitE-SQL: A Lightweight and Efficient Text-to-SQL Framework with Vector-based Schema Linking and Execution-Guided Self-Correction

Authors: Shengmin Piao, Jieun Lee, Sanghyun Park
Published: 2025-10-10
Categories: cs.CL
arXiv: https://arxiv.org/abs/2510.09014v2

The Text-to-SQL task translates natural language questions into SQL queries, enabling intuitive database interaction for non-experts. While recent methods leveraging Large Language Models (LLMs) achieve strong performance, their reliance on proprietary models raise concerns about deployment feasibility and data privacy. In this work, we introduce LitE-SQL, a Lightweight and Efficient framework with two components: (i) a Schema Retriever that performs efficient schema linking using a vector database of pre-computed schema embeddings, optimized with a hard-negative supervised contrastive objective to distinguish semantically similar but functionally irrelevant columns, and (ii) a SQL Generator fine-tuned in two stages-supervised fine-tuning followed by execution-guided reinforcement-enabling execution-guided self-correction without multi-candidate sampling, which is commonly required by prior LLM-based approaches. On BIRD, LitE-SQL achieves 72.10% execution accuracy, and on Spider 1.0 it reaches 88.45%, demonstrating comparable or superior performance to LLM-based methods despite using 2x to 30x fewer parameters. Our findings demonstrate that high-quality Text-to-SQL generation is feasible with lightweight models, offering a practical solution for privacy-sensitive and resource-constrained settings.

---

### Towards Better & Faster Autoregressive Image Generation: From the Perspective of Entropy

Authors: Xiaoxiao Ma, Feng Zhao, Pengyang Ling, Haibo Qiu, Zhixiang Wei, Hu Yu, Jie Huang, Zhixiong Zeng, Lin Ma
Published: 2025-10-10
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.09012v2

In this work, we first revisit the sampling issues in current autoregressive (AR) image generation models and identify that image tokens, unlike text tokens, exhibit lower information density and non-uniform spatial distribution. Accordingly, we present an entropy-informed decoding strategy that facilitates higher autoregressive generation quality with faster synthesis speed. Specifically, the proposed method introduces two main innovations: 1) dynamic temperature control guided by spatial entropy of token distributions, enhancing the balance between content diversity, alignment accuracy, and structural coherence in both mask-based and scale-wise models, without extra computational overhead, and 2) entropy-aware acceptance rules in speculative decoding, achieving near-lossless generation at about 85\% of the inference cost of conventional acceleration methods. Extensive experiments across multiple benchmarks using diverse AR image generation models demonstrate the effectiveness and generalizability of our approach in enhancing both generation quality and sampling speed.

---

### TripScore: Aligning LLMs for Real-World Travel Planning via Expert-Calibrated Reward

Authors: Yincen Qu, Huan Xiao, Feng Li, Gregory Li, Hui Zhou, Xiangying Dai, Xiaoru Dai, Xuan Huang
Published: 2025-10-10
Categories: cs.AI, cs.CL
arXiv: https://arxiv.org/abs/2510.09011v4

In our deployed travel-planning service, most users give minimal inputs or free-form requests rather than the structured constraint checklists assumed by existing benchmarks. We therefore present TripScore, a behavior-grounded benchmark and evaluation framework built from real user logs and calibrated against 1,468 pairwise judgments by 203 travel experts. TripScore couples a hierarchical feasibility gate (format and commonsense) with a unified, point-wise reward that aggregates soft quality and preference fulfillment. Using TripScore as both evaluator and reward signal, we benchmark direct prompting, test-time compute, neuro-symbolic solvers, code agents, and fine-tuning. We find that reinforcement learning fine-tuning (e.g., GRPO) provides consistent gains over other approaches under the same base model and practical latency.

---

### On Epistemic Uncertainty of Visual Tokens for Object Hallucinations in Large Vision-Language Models

Authors: Hoigi Seo, Dong Un Kang, Hyunjin Cho, Joohoon Lee, Se Young Chun
Published: 2025-10-10
Categories: cs.CV, cs.AI, cs.CL
arXiv: https://arxiv.org/abs/2510.09008v1

Large vision-language models (LVLMs), which integrate a vision encoder (VE) with a large language model, have achieved remarkable success across various tasks. However, there are still crucial challenges in LVLMs such as object hallucination, generating descriptions of objects that are not in the input image. Here, we argue that uncertain visual tokens within the VE is a key factor that contributes to object hallucination. Our statistical analysis found that there are positive correlations between visual tokens with high epistemic uncertainty and the occurrence of hallucinations. Furthermore, we show theoretically and empirically that visual tokens in early VE layers that exhibit large representation deviations under small adversarial perturbations indicate high epistemic uncertainty. Based on these findings, we propose a simple yet effective strategy to mitigate object hallucination by modifying the VE only. Our method comprises a proxy method with adversarial perturbations for identifying uncertain visual tokens efficiently and a method to mask these uncertain visual tokens during the self-attention process in the middle layers of the VE, suppressing their influence on visual encoding and thus alleviating hallucinations. Extensive experiments show that our method significantly reduces object hallucinations in LVLMs and can synergistically work with other prior arts.

---

### LLM Unlearning on Noisy Forget Sets: A Study of Incomplete, Rewritten, and Watermarked Data

Authors: Changsheng Wang, Yihua Zhang, Dennis Wei, Jinghan Jia, Pin-Yu Chen, Sijia Liu
Published: 2025-10-10
Categories: cs.LG
arXiv: https://arxiv.org/abs/2510.09007v1

Large language models (LLMs) exhibit remarkable generative capabilities but raise ethical and security concerns by memorizing sensitive data, reinforcing biases, and producing harmful content. These risks have spurred interest in LLM unlearning, the task of removing knowledge associated with undesirable data from pre-trained models. However, most existing methods assume access to clean, well-defined forget data samples, whereas real-world forget data could often be low-quality, synthetically rewritten, or watermarked, casting doubt on the reliability of unlearning. This work presents the first study of unlearning under perturbed or low-fidelity forget data, referred to as noisy forget sets. By systematically benchmarking state-of-the-art LLM unlearning methods, RMU and NPO, on such noisy forget sets, we find that unlearning remains surprisingly robust to perturbations, provided that core semantic signals are preserved. To explain this robustness, we propose a saliency-based interpretation: key semantic components that drive forgetting remain consistently influential despite substantial variation in surface form. This suggests that unlearning algorithms are primarily guided by deep semantic cues rather than shallow lexical patterns.

---

### Decoupling Safety into Orthogonal Subspace: Cost-Efficient and Performance-Preserving Alignment for Large Language Models

Authors: Yutao Mou, Xiaoling Zhou, Yuxiao Luo, Shikun Zhang, Wei Ye
Published: 2025-10-10
Categories: cs.CL
arXiv: https://arxiv.org/abs/2510.09004v1

Safety alignment is essential for building trustworthy artificial intelligence, yet it remains challenging to enhance model safety without degrading general performance. Current approaches require computationally expensive searches for the optimal proportion of safety-critical and general-purpose data to balance safety and general performance, incurring high costs with limited gains. In this work, we show that LoRA-based Refusal-training enables performance-preserving safety alignment even when trained solely on safety data, demonstrating that LoRA serves as cost-efficient, performance-preserving, and plug-and-play safety patches. Beyond empirical findings, we provide both theoretical and experimental evidence that LoRA effectively decouples safety into a low-rank subspace largely orthogonal to the model's intrinsic transformation space, ensuring that safety enhancements do not interfere with inherent capabilities.

---

### DARO: Difficulty-Aware Reweighting Policy Optimization

Authors: Jingyu Zhou, Lu Ma, Hao Liang, Chengyu Shen, Bin Cui, Wentao Zhang
Published: 2025-10-10
Categories: cs.CL
arXiv: https://arxiv.org/abs/2510.09001v1

Recent advances in large language models (LLMs) have shown that reasoning ability can be significantly enhanced through Reinforcement Learning with Verifiable Rewards (RLVR). Group Relative Policy Optimization (GRPO) has emerged as the de facto approach for RLVR, inspiring numerous variants. However, our mathematical analysis reveals that these methods are fundamentally weighted variations of GRPO. We provide a unified view, demonstrating that their reliance on static or overly simplistic weighting schemes tied to sample difficulty prevents adaptation to a model's evolving capabilities. This creates a significant loss scale issue, where training disproportionately focuses on certain difficulty levels at the expense of others, hindering overall performance. To address these limitations, we introduce \textbf{Difficulty-Aware Reweighting Policy Optimization (DARO)}, a method that dynamically adjusts the loss contribution of each difficulty group based on the model's learning state. Extensive experiments on Qwen2.5-Math-1.5B, Qwen2.5-Math-7B, and Llama3.1-8B show that DARO outperforms four leading baselines across six math benchmarks, achieving significantly faster convergence and superior final performance.

---

### SQS: Bayesian DNN Compression through Sparse Quantized Sub-distributions

Authors: Ziyi Wang, Nan Jiang, Guang Lin, Qifan Song
Published: 2025-10-10
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2510.08999v2

Compressing large-scale neural networks is essential for deploying models on resource-constrained devices. Most existing methods adopt weight pruning or low-bit quantization individually, often resulting in suboptimal compression rates to preserve acceptable performance drops. We introduce a unified framework for simultaneous pruning and low-bit quantization via Bayesian variational learning (\method), which achieves higher compression rates than prior baselines while maintaining comparable performance. The key idea is to employ a spike-and-slab prior to induce sparsity and model quantized weights using Gaussian Mixture Models (GMMs) to enable low-bit precision. Due to the intractability of the objective involving spike-and-slab priors with GMMs, we derive an efficient approximation that facilitates effective compression with minimal accuracy loss. In theory, we provide a consistent result for our proposed variational approach to a sparse and quantized deep neural network. Extensive experiments on compressing ResNet, BERT-base, Llama3.2, and Qwen2.5 models show that our method achieves higher compression rates than a line of existing methods with comparable performance drops. Project page: https://comeusr.github.io/SQS_Webpage.

---

### Saving SWE-Bench: A Benchmark Mutation Approach for Realistic Agent Evaluation

Authors: Spandan Garg, Benjamin Steenhoek, Yufan Huang
Published: 2025-10-10
Categories: cs.SE, cs.AI
arXiv: https://arxiv.org/abs/2510.08996v4

Current benchmarks for evaluating software engineering agents, such as SWE-Bench Verified, are predominantly derived from GitHub issues and fail to accurately reflect how developers interact with chat-based coding assistants in integrated development environments (IDEs). We posit that this mismatch leads to a systematic overestimation of agent's capabilities in real-world scenarios, especially bug fixing. We introduce a novel benchmarking framework that transforms existing formal benchmarks into realistic user queries through systematic analysis of developer interaction patterns with chat-based agents. Our methodology is flexible and can be easily extended to existing benchmarks. In this paper, we apply our testing framework to SWE-Bench Verified, the TypeScript subset of Multi-SWE-Bench and a private benchmark, SWE-Bench C# and transform formal GitHub issue descriptions into realistic user-style queries based on telemetry analysis of a popular chat-based agent interactions. Our findings reveal that existing benchmarks significantly overestimate agent capabilities for some models by >50% over baseline performance for public benchmarks and ~10-16% for our internal benchmark. This work establishes a new paradigm for evaluating interactive chat-based software engineering agents through benchmark mutation techniques.

---

### Speculative Jacobi-Denoising Decoding for Accelerating Autoregressive Text-to-image Generation

Authors: Yao Teng, Fuyun Wang, Xian Liu, Zhekai Chen, Han Shi, Yu Wang, Zhenguo Li, Weiyang Liu, Difan Zou, Xihui Liu
Published: 2025-10-10
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.08994v1

As a new paradigm of visual content generation, autoregressive text-to-image models suffer from slow inference due to their sequential token-by-token decoding process, often requiring thousands of model forward passes to generate a single image. To address this inefficiency, we propose Speculative Jacobi-Denoising Decoding (SJD2), a framework that incorporates the denoising process into Jacobi iterations to enable parallel token generation in autoregressive models. Our method introduces a next-clean-token prediction paradigm that enables the pre-trained autoregressive models to accept noise-perturbed token embeddings and predict the next clean tokens through low-cost fine-tuning. This denoising paradigm guides the model towards more stable Jacobi trajectories. During inference, our method initializes token sequences with Gaussian noise and performs iterative next-clean-token-prediction in the embedding space. We employ a probabilistic criterion to verify and accept multiple tokens in parallel, and refine the unaccepted tokens for the next iteration with the denoising trajectory. Experiments show that our method can accelerate generation by reducing model forward passes while maintaining the visual quality of generated images.

---

### PlatformX: An End-to-End Transferable Platform for Energy-Efficient Neural Architecture Search

Authors: Xiaolong Tu, Dawei Chen, Kyungtae Han, Onur Altintas, Haoxin Wang
Published: 2025-10-10
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2510.08993v1

Hardware-Aware Neural Architecture Search (HW-NAS) has emerged as a powerful tool for designing efficient deep neural networks (DNNs) tailored to edge devices. However, existing methods remain largely impractical for real-world deployment due to their high time cost, extensive manual profiling, and poor scalability across diverse hardware platforms with complex, device-specific energy behavior. In this paper, we present PlatformX, a fully automated and transferable HW-NAS framework designed to overcome these limitations. PlatformX integrates four key components: (i) an energy-driven search space that expands conventional NAS design by incorporating energy-critical configurations, enabling exploration of high-efficiency architectures; (ii) a transferable kernel-level energy predictor across devices and incrementally refined with minimal on-device samples; (iii) a Pareto-based multi-objective search algorithm that balances energy and accuracy to identify optimal trade-offs; and (iv) a high-resolution runtime energy profiling system that automates on-device power measurement using external monitors without human intervention. We evaluate PlatformX across multiple mobile platforms, showing that it significantly reduces search overhead while preserving accuracy and energy fidelity. It identifies models with up to 0.94 accuracy or as little as 0.16 mJ per inference, both outperforming MobileNet-V2 in accuracy and efficiency. Code and tutorials are available at github.com/amai-gsu/PlatformX.

---

### Constraints-of-Thought: A Framework for Constrained Reasoning in Language-Model-Guided Search

Authors: Kamel Alrashedy, Vriksha Srihari, Zulfiqar Zaidi, Ridam Srivastava, Pradyumna Tambwekar, Matthew Gombolay
Published: 2025-10-10
Categories: cs.LG
arXiv: https://arxiv.org/abs/2510.08992v3

While researchers have made significant progress in enabling large language models (LLMs) to perform multi-step planning, LLMs struggle to ensure that those plans align with high-level user intent and satisfy symbolic constraints, especially in complex, multi-step domains. Existing reasoning approaches such as Chain-of-Thought (CoT), Tree-of-Thought (ToT), and verifier-augmented methods, expand the search space but often yield infeasible actions or hallucinated steps. To overcome these limitations, we propose Constraints-of-Thought (Const-o-T), a framework that provides a structured prior that enables Monte Carlo Tree Search (MCTS) focus search on semantically meaningful paths. Each reasoning step is represented as an (intent, constraint) pair, which serves both to compress the search space and enforce validity. Unlike prior methods that merely generate reasoning traces or validate outputs post hoc, Const-o-T uses (intent, constraint)pairs to actively focus the search toward feasible and meaningful plans. We integrate Const-o-T into MCTS using a structured representation of intent-constraint pairs constraints prune infeasible branches and guide exploration toward semantically valid actions, improving planning efficiency and verifiable decision-making. We demonstrate across three domains Risk game, CAD code generation, and arithmetic reasoning that our approach outperforms baselines, yielding higher accuracy and stronger structural alignment. Our contribution is to demonstrate that Const-of-T offers a generalizable foundation for constraint-guided reasoning, enabling more efficient, constraint-aligned, and domain-adaptable planning with LLMs.

---

### MASA: LLM-Driven Multi-Agent Systems for Autoformalization

Authors: Lan Zhang, Marco Valentino, André Freitas
Published: 2025-10-10
Categories: cs.CL, cs.FL
arXiv: https://arxiv.org/abs/2510.08988v1

Autoformalization serves a crucial role in connecting natural language and formal reasoning. This paper presents MASA, a novel framework for building multi-agent systems for autoformalization driven by Large Language Models (LLMs). MASA leverages collaborative agents to convert natural language statements into their formal representations. The architecture of MASA is designed with a strong emphasis on modularity, flexibility, and extensibility, allowing seamless integration of new agents and tools to adapt to a fast-evolving field. We showcase the effectiveness of MASA through use cases on real-world mathematical definitions and experiments on formal mathematics datasets. This work highlights the potential of multi-agent systems powered by the interaction of LLMs and theorem provers in enhancing the efficiency and reliability of autoformalization, providing valuable insights and support for researchers and practitioners in the field.

---

### Towards Efficient Multimodal Unified Reasoning Model via Model Merging

Authors: Qixiang Yin, Huanjin Yao, Jianghao Chen, Jiaxing Huang, Zhicheng Zhao, Fei Su
Published: 2025-10-10
Categories: cs.AI
arXiv: https://arxiv.org/abs/2510.08987v2

Although Multimodal Large Language Models (MLLMs) have demonstrated remarkable capabilities across diverse tasks, they encounter challenges in terms of reasoning efficiency, large model size and overthinking. However, existing lightweight MLLMs lack the capability to balance high efficiency and performance at a small scale. To this end, we propose Tiny-R1V, a novel lightweight 3B model that achieves faster inference and higher accuracy via a two-stage optimization, while unifying multimodal reasoning across multiple tasks with fewer inference tokens. In the first stage, Tiny-R1V introduces Length-Informed Relative Policy Optimization (LIPO), a new reinforcement learning method, to train each reasoning model, including mathematical reasoning, chart reasoning, and OCR capability. The LIPO dynamically adjusts the advantages of responses within groups by prioritizing concise yet high-quality responses to encourage the generation of shorter and more accurate responses. In the second stage, we propose Adaptive Model Merging (AMM), a training-free model merging method that merges multiple specialist models into a unified architecture. Specifically, AMM adaptively adjusts the weights of task vectors via a novel gradient projection regularization loss function, thus mitigating redundant conflicts between them. Extensive evaluations on ten widely-used reasoning benchmarks covering mathematics, structured data (charts, tables, documents), OCR, and general capabilities showcase the superior performance of Tiny-R1V, enabling lightweight models to excel in diverse multimodal reasoning tasks. Code will be available at \href{https://github.com/buptyqx/Tiny-R1V}{https://github.com/buptyqx/Tiny-R1V}

---

### CAPC-CG: A Large-Scale, Expert-Directed LLM-Annotated Corpus of Adaptive Policy Communication in China

Authors: Bolun Sun, Charles Chang, Yuen Yuen Ang, Ruotong Mu, Yuchen Xu, Zhengxin Zhang, Pingxu Hao
Published: 2025-10-10
Categories: cs.CL, cs.CE, cs.CY
arXiv: https://arxiv.org/abs/2510.08986v3

We introduce CAPC-CG, the Chinese Adaptive Policy Communication (Central Government) Corpus, the first open dataset of Chinese policy directives annotated with a five-color taxonomy of clear and ambiguous language categories, building on Ang's theory of adaptive policy communication. Spanning 1949-2023, this corpus includes national laws, administrative regulations, and ministerial rules issued by China's top authorities. Each document is segmented into paragraphs, producing a total of 3.3 million units. Alongside the corpus, we release comprehensive metadata, a two-round labeling framework, and a gold-standard annotation set developed by expert and trained coders. Inter-annotator agreement achieves a Fleiss's kappa of K = 0.86 on directive labels, indicating high reliability for supervised modeling. We provide baseline classification results with several large language models (LLMs), together with our annotation codebook, and describe patterns from the dataset. This release aims to support downstream tasks and multilingual NLP research in policy communication.

---

### FedL2T: Personalized Federated Learning with Two-Teacher Distillation for Seizure Prediction

Authors: Jionghao Lou, Jian Zhang, Zhongmei Li, Lanlan Chen, Enbo Feng
Published: 2025-10-10
Categories: cs.LG, cs.NE
arXiv: https://arxiv.org/abs/2510.08984v1

The training of deep learning models in seizure prediction requires large amounts of Electroencephalogram (EEG) data. However, acquiring sufficient labeled EEG data is difficult due to annotation costs and privacy constraints. Federated Learning (FL) enables privacy-preserving collaborative training by sharing model updates instead of raw data. However, due to the inherent inter-patient variability in real-world scenarios, existing FL-based seizure prediction methods struggle to achieve robust performance under heterogeneous client settings. To address this challenge, we propose FedL2T, a personalized federated learning framework that leverages a novel two-teacher knowledge distillation strategy to generate superior personalized models for each client. Specifically, each client simultaneously learns from a globally aggregated model and a dynamically assigned peer model, promoting more direct and enriched knowledge exchange. To ensure reliable knowledge transfer, FedL2T employs an adaptive multi-level distillation strategy that aligns both prediction outputs and intermediate feature representations based on task confidence. In addition, a proximal regularization term is introduced to constrain personalized model updates, thereby enhancing training stability. Extensive experiments on two EEG datasets demonstrate that FedL2T consistently outperforms state-of-the-art FL methods, particularly under low-label conditions. Moreover, FedL2T exhibits rapid and stable convergence toward optimal performance, thereby reducing the number of communication rounds and associated overhead. These results underscore the potential of FedL2T as a reliable and personalized solution for seizure prediction in privacy-sensitive healthcare scenarios.

---

### SEER: Sustainability Enhanced Engineering of Software Requirements

Authors: Mandira Roy, Novarun Deb, Nabendu Chaki, Agostino Cortesi
Published: 2025-10-10
Categories: cs.SE, cs.AI
arXiv: https://arxiv.org/abs/2510.08981v1

The rapid expansion of software development has significant environmental, technical, social, and economic impacts. Achieving the United Nations Sustainable Development Goals by 2030 compels developers to adopt sustainable practices. Existing methods mostly offer high-level guidelines, which are time-consuming to implement and rely on team adaptability. Moreover, they focus on design or implementation, while sustainability assessment should start at the requirements engineering phase. In this paper, we introduce SEER, a framework which addresses sustainability concerns in the early software development phase. The framework operates in three stages: (i) it identifies sustainability requirements (SRs) relevant to a specific software product from a general taxonomy; (ii) it evaluates how sustainable system requirements are based on the identified SRs; and (iii) it optimizes system requirements that fail to satisfy any SR. The framework is implemented using the reasoning capabilities of large language models and the agentic RAG (Retrieval Augmented Generation) approach. SEER has been experimented on four software projects from different domains. Results generated using Gemini 2.5 reasoning model demonstrate the effectiveness of the proposed approach in accurately identifying a broad range of sustainability concerns across diverse domains.

---

### Uncolorable Examples: Preventing Unauthorized AI Colorization via Perception-Aware Chroma-Restrictive Perturbation

Authors: Yuki Nii, Futa Waseda, Ching-Chun Chang, Isao Echizen
Published: 2025-10-10
Categories: cs.CV, cs.LG
arXiv: https://arxiv.org/abs/2510.08979v2

AI-based colorization has shown remarkable capability in generating realistic color images from grayscale inputs. However, it poses risks of copyright infringement -- for example, the unauthorized colorization and resale of monochrome manga and films. Despite these concerns, no effective method currently exists to prevent such misuse. To address this, we introduce the first defensive paradigm, Uncolorable Examples, which embed imperceptible perturbations into grayscale images to invalidate unauthorized colorization. To ensure real-world applicability, we establish four criteria: effectiveness, imperceptibility, transferability, and robustness. Our method, Perception-Aware Chroma-Restrictive Perturbation (PAChroma), generates Uncolorable Examples that meet these four criteria by optimizing imperceptible perturbations with a Laplacian filter to preserve perceptual quality, and applying diverse input transformations during optimization to enhance transferability across models and robustness against common post-processing (e.g., compression). Experiments on ImageNet and Danbooru datasets demonstrate that PAChroma effectively degrades colorization quality while maintaining the visual appearance. This work marks the first step toward protecting visual content from illegitimate AI colorization, paving the way for copyright-aware defenses in generative media.

---

### HandEval: Taking the First Step Towards Hand Quality Evaluation in Generated Images

Authors: Zichuan Wang, Bo Peng, Songlin Yang, Zhenchen Tang, Jing Dong
Published: 2025-10-10
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.08978v1

Although recent text-to-image (T2I) models have significantly improved the overall visual quality of generated images, they still struggle in the generation of accurate details in complex local regions, especially human hands. Generated hands often exhibit structural distortions and unrealistic textures, which can be very noticeable even when the rest of the body is well-generated. However, the quality assessment of hand regions remains largely neglected, limiting downstream task performance like human-centric generation quality optimization and AIGC detection. To address this, we propose the first quality assessment task targeting generated hand regions and showcase its abundant downstream applications. We first introduce the HandPair dataset for training hand quality assessment models. It consists of 48k images formed by high- and low-quality hand pairs, enabling low-cost, efficient supervision without manual annotation. Based on it, we develop HandEval, a carefully designed hand-specific quality assessment model. It leverages the powerful visual understanding capability of Multimodal Large Language Model (MLLM) and incorporates prior knowledge of hand keypoints, gaining strong perception of hand quality. We further construct a human-annotated test set with hand images from various state-of-the-art (SOTA) T2I models to validate its quality evaluation capability. Results show that HandEval aligns better with human judgments than existing SOTA methods. Furthermore, we integrate HandEval into image generation and AIGC detection pipelines, prominently enhancing generated hand realism and detection accuracy, respectively, confirming its universal effectiveness in downstream applications. Code and dataset will be available.

---

### Breaking the Self-Confirming Loop: Diagnosing and Mitigating Systemic Reward Bias in Self-Rewarding RL

Authors: Chuyi Tan, Peiwen Yuan, Xinglin Wang, Yiwei Li, Shaoxiong Feng, Yueqi Zhang, Jiayi Shi, Ji Zhang, Boyuan Pan, Yao Hu, Kan Li
Published: 2025-10-10
Categories: cs.LG, cs.CL
arXiv: https://arxiv.org/abs/2510.08977v2

Reinforcement learning with verifiable rewards (RLVR) efficiently scales the reasoning ability of large language models (LLMs) but is bottlenecked by scarce labeled data. Reinforcement learning with intrinsic rewards (RLIR) offers a scalable alternative via self-rewarding, yet often suffers from instability and inferior performance. We trace this gap to a systemic bias in confidence-coupled self-rewarding: the model tends to over-reward high-confidence mistakes, forming a self-confirming loop. We quantify this feedback-loop bias with three metrics: reward noise magnitude (rho_noise), policy-reward coupling (rho_selfbias), and over-/under-reward skew (rho_symbias). Our analyses show a compounding effect where strong coupling amplifies confidence-conditioned errors and drives a drift toward over-reward, leading to instability and a lower performance ceiling. To mitigate this, we propose reinforcement learning with ensembled rewards (RLER), which aggregates diverse models with adaptive reward interpolation and disagreement-aware rollout selection to reduce coupling and suppress over-reward drift. Extensive experiments show that RLER improves by 6.2% over the best RLIR baseline and is within 3.6% of RLVR, while exhibiting stable scaling on unlabeled samples.

---

### MIRAGE: Runtime Scheduling for Multi-Vector Image Retrieval with Hierarchical Decomposition

Authors: Maoliang Li, Ke Li, Yaoyang Liu, Jiayu Chen, Zihao Zheng, Yinjun Wu, Chenchen Liu, Xiang Chen
Published: 2025-10-10
Categories: cs.CV, cs.DC, cs.IR
arXiv: https://arxiv.org/abs/2510.08976v3

To effectively leverage user-specific data, retrieval augmented generation (RAG) is employed in multimodal large language model (MLLM) applications. However, conventional retrieval approaches often suffer from limited retrieval accuracy. Recent advances in multi-vector retrieval (MVR) improve accuracy by decomposing queries and matching against segmented images. They still suffer from sub-optimal accuracy and efficiency, overlooking alignment between the query and varying image objects and redundant fine-grained image segments. In this work, we present an efficient scheduling framework for image retrieval - MIRAGE. First, we introduce a novel hierarchical paradigm, employing multiple intermediate granularities for varying image objects to enhance alignment. Second, we minimize redundancy in retrieval by leveraging cross-hierarchy similarity consistency and hierarchy sparsity to minimize unnecessary matching computation. Furthermore, we configure parameters for each dataset automatically for practicality across diverse scenarios. Our empirical study shows that, MIRAGE not only achieves substantial accuracy improvements but also reduces computation by up to 3.5 times over the existing MVR system.

---

### mmJoints: Expanding Joint Representations Beyond (x,y,z) in mmWave-Based 3D Pose Estimation

Authors: Zhenyu Wang, Mahathir Monjur, Shahriar Nirjon
Published: 2025-10-10
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.08970v1

In mmWave-based pose estimation, sparse signals and weak reflections often cause models to infer body joints from statistical priors rather than sensor data. While prior knowledge helps in learning meaningful representations, over-reliance on it degrades performance in downstream tasks like gesture and activity recognition. In this paper, we introduce mmJoints, a framework that augments a pre-trained, black-box mmWave-based 3D pose estimator's output with additional joint descriptors. Rather than mitigating bias, mmJoints makes it explicit by estimating the likelihood of a joint being sensed and the reliability of its predicted location. These descriptors enhance interpretability and improve downstream task accuracy. Through extensive evaluations using over 115,000 signal frames across 13 pose estimation settings, we show that mmJoints estimates descriptors with an error rate below 4.2%. mmJoints also improves joint position accuracy by up to 12.5% and boosts activity recognition by up to 16% over state-of-the-art methods.

---

### Learning Regularizers: Learning Optimizers that can Regularize

Authors: Suraj Kumar Sahoo, Narayanan C Krishnan
Published: 2025-10-10
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2510.08968v1

Learned Optimizers (LOs), a type of Meta-learning, have gained traction due to their ability to be parameterized and trained for efficient optimization. Traditional gradient-based methods incorporate explicit regularization techniques such as Sharpness-Aware Minimization (SAM), Gradient-norm Aware Minimization (GAM), and Gap-guided Sharpness-Aware Minimization (GSAM) to enhance generalization and convergence. In this work, we explore a fundamental question: \textbf{Can regularizers be learned?} We empirically demonstrate that LOs can be trained to learn and internalize the effects of traditional regularization techniques without explicitly applying them to the objective function. We validate this through extensive experiments on standard benchmarks (including MNIST, FMNIST, CIFAR and Neural Networks such as MLP, MLP-Relu and CNN), comparing LOs trained with and without access to explicit regularizers. Regularized LOs consistently outperform their unregularized counterparts in terms of test accuracy and generalization. Furthermore, we show that LOs retain and transfer these regularization effects to new optimization tasks by inherently seeking minima similar to those targeted by these regularizers. Our results suggest that LOs can inherently learn regularization properties, \textit{challenging the conventional necessity of explicit optimizee loss regularization.

---

### SAM2-3dMed: Empowering SAM2 for 3D Medical Image Segmentation

Authors: Yeqing Yang, Le Xu, Lixia Tian
Published: 2025-10-10
Categories: eess.IV, cs.CV
arXiv: https://arxiv.org/abs/2510.08967v1

Accurate segmentation of 3D medical images is critical for clinical applications like disease assessment and treatment planning. While the Segment Anything Model 2 (SAM2) has shown remarkable success in video object segmentation by leveraging temporal cues, its direct application to 3D medical images faces two fundamental domain gaps: 1) the bidirectional anatomical continuity between slices contrasts sharply with the unidirectional temporal flow in videos, and 2) precise boundary delineation, crucial for morphological analysis, is often underexplored in video tasks. To bridge these gaps, we propose SAM2-3dMed, an adaptation of SAM2 for 3D medical imaging. Our framework introduces two key innovations: 1) a Slice Relative Position Prediction (SRPP) module explicitly models bidirectional inter-slice dependencies by guiding SAM2 to predict the relative positions of different slices in a self-supervised manner; 2) a Boundary Detection (BD) module enhances segmentation accuracy along critical organ and tissue boundaries. Extensive experiments on three diverse medical datasets (the Lung, Spleen, and Pancreas in the Medical Segmentation Decathlon (MSD) dataset) demonstrate that SAM2-3dMed significantly outperforms state-of-the-art methods, achieving superior performance in segmentation overlap and boundary precision. Our approach not only advances 3D medical image segmentation performance but also offers a general paradigm for adapting video-centric foundation models to spatial volumetric data.

---

### Beyond Prefixes: Graph-as-Memory Cross-Attention for Knowledge Graph Completion with Large Language Models

Authors: Ruitong Liu, Boxu Lin, Peize Li, Siyuan Li, Yunjia Wu, Te Sun, Chaohan Wu
Published: 2025-10-10
Categories: cs.AI, cs.CL
arXiv: https://arxiv.org/abs/2510.08966v2

Fusing Knowledge Graphs with Large Language Models (LLMs) is crucial for knowledge-intensive tasks like knowledge graph completion. Existing LLM-based approaches typically inject graph information via prefix concatenation, resulting in shallow interactions that fail to support fine-grained evidence retrieval during generation. Beyond prefixes, we propose Graph-as-Memory Tuning (GMT), a new paradigm that represents local graph structure as explicit graph memory and injects it into LLMs via deep, token-wise cross-attention. Specifically, GMT first employs a Semantic Graph Module to encode context-aware semantics from local neighborhoods guided by knowledge-enhanced relations, and compresses them into a fixed number of graph memory tokens. A Graph-as-Memory Cross-Attention Fusion Module then integrates these tokens into multiple Transformer layers, allowing LLM hidden state to dynamically retrieve relevant graph evidence. To enable efficient adaptation, GMT applies LoRA only to the memory cross-attention while keeping the base LLM frozen. Extensive experiments show that GMT significantly outperforms prefix-tuning and other strong baselines, providing more potent signals for robust reasoning. The code is published at https://github.com/tongruiliu/GMT.

---