# Research Papers — 2025-10 · bundle page 178 of 217 > Full-text markdown of papers #8851–8900 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/178 [← Previous bundle](https://aipod.dev/papers/archive/2025-10/page/177.md) [Next bundle →](https://aipod.dev/papers/archive/2025-10/page/179.md) ### Physics-informed Attention-enhanced Fourier Neural Operator for Solar Magnetic Field Extrapolations Authors: Jinghao Cao, Qin Li, Mengnan Du, Haimin Wang, Bo Shen Published: 2025-10-06 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2510.05351v1 We propose Physics-informed Attention-enhanced Fourier Neural Operator (PIANO) to solve the Nonlinear Force-Free Field (NLFFF) problem in solar physics. Unlike conventional approaches that rely on iterative numerical methods, our proposed PIANO directly learns the 3D magnetic field structure from 2D boundary conditions. Specifically, PIANO integrates Efficient Channel Attention (ECA) mechanisms with Dilated Convolutions (DC), which enhances the model's ability to capture multimodal input by prioritizing critical channels relevant to the magnetic field's variations. Furthermore, we apply physics-informed loss by enforcing the force-free and divergence-free conditions in the training process so that our prediction is consistent with underlying physics with high accuracy. Experimental results on the ISEE NLFFF dataset show that our PIANO not only outperforms state-of-the-art neural operators in terms of accuracy but also shows strong consistency with the physical characteristics of NLFFF data across magnetic fields reconstructed from various solar active regions. The GitHub of this project is available https://github.com/Autumnstar-cjh/PIANO --- ### Margin Adaptive DPO: Leveraging Reward Model for Granular Control in Preference Optimization Authors: Hyung Gyu Rho Published: 2025-10-06 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2510.05342v2 Direct Preference Optimization (DPO) has emerged as a simple and effective method for aligning large language models. However, its reliance on a fixed temperature parameter leads to suboptimal training on diverse preference data, causing overfitting on easy examples and under-learning from informative ones. Recent methods have emerged to counter this. While IPO addresses general overfitting, its uniform regularization can be overly conservative. The more targeted approach of $β$-DPO suffers from its own limitations: its batch-level adaptation applies a single, compromised temperature to mixed-margin pairs, its linear update rule can produce unstable negative $β$ values, and its filtering mechanism discards potentially useful training signals. In this work, we introduce Margin-Adaptive Direct Preference Optimization (MADPO), a method that provides a stable, data-preserving, and instance-level solution. MADPO employs a practical two-step approach: it first trains a reward model to estimate preference margins and then uses these margins to apply a continuous, adaptive weight to the DPO loss for each individual training sample. This re-weighting scheme creates an effective target margin that is amplified for hard pairs and dampened for easy pairs, allowing for granular control over the learning signal. We provide a comprehensive theoretical analysis, proving that MADPO has a well-behaved optimization landscape and is robust to reward model estimation errors. We validate our theory with experiments on a summarization task using human preference data. MADPO consistently outperforms strong baselines across a comprehensive sweep of decoding temperatures. --- ### Integrating Bayesian methods with neural network--based model predictive control: a review Authors: Asli Karacelik Published: 2025-10-06 Categories: cs.AI, stat.ML arXiv: https://arxiv.org/abs/2510.05338v1 In this review, we assess the use of Bayesian methods in model predictive control (MPC), focusing on neural-network-based modeling, control design, and uncertainty quantification. We systematically analyze individual studies and how they are implemented in practice. While Bayesian approaches are increasingly adopted to capture and propagate uncertainty in MPC, reported gains in performance and robustness remain fragmented, with inconsistent baselines and limited reliability analyses. We therefore argue for standardized benchmarks, ablation studies, and transparent reporting to rigorously determine the effectiveness of Bayesian techniques for MPC. --- ### WeatherArchive-Bench: Benchmarking Retrieval-Augmented Reasoning for Historical Weather Archives Authors: Yongan Yu, Xianda Du, Qingchen Hu, Jiahao Liang, Jingwei Ni, Dan Qiang, Kaiyu Huang, Grant McKenzie, Renee Sieber, Fengran Mo Published: 2025-10-06 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2510.05336v2 Historical archives on weather events are collections of enduring primary source records that offer rich, untapped narratives of how societies have experienced and responded to extreme weather events. These qualitative accounts provide insights into societal vulnerability and resilience that are largely absent from meteorological records, making them valuable for climate scientists to understand societal responses. However, their vast scale, noisy digitized quality, and archaic language make it difficult to transform them into structured knowledge for climate research. To address this challenge, we introduce WeatherArchive-Bench, the first benchmark for evaluating retrieval-augmented generation (RAG) systems on historical weather archives. WeatherArchive-Bench comprises two tasks: WeatherArchive-Retrieval, which measures a system's ability to locate historically relevant passages from over one million archival news segments, and WeatherArchive-Assessment, which evaluates whether Large Language Models (LLMs) can classify societal vulnerability and resilience indicators from extreme weather narratives. Extensive experiments across sparse, dense, and re-ranking retrievers, as well as a diverse set of LLMs, reveal that dense retrievers often fail on historical terminology, while LLMs frequently misinterpret vulnerability and resilience concepts. These findings highlight key limitations in reasoning about complex societal indicators and provide insights for designing more robust climate-focused RAG systems from archival contexts. The constructed dataset and evaluation framework are publicly available at https://anonymous.4open.science/r/WeatherArchive-Bench/. --- ### Biomedical reasoning in action: Multi-agent System for Auditable Biomedical Evidence Synthesis Authors: Oskar Wysocki, Magdalena Wysocka, Mauricio Jacobo, Harriet Unsworth, André Freitas Published: 2025-10-06 Categories: cs.AI arXiv: https://arxiv.org/abs/2510.05335v1 We present M-Reason, a demonstration system for transparent, agent-based reasoning and evidence integration in the biomedical domain, with a focus on cancer research. M-Reason leverages recent advances in large language models (LLMs) and modular agent orchestration to automate evidence retrieval, appraisal, and synthesis across diverse biomedical data sources. Each agent specializes in a specific evidence stream, enabling parallel processing and fine-grained analysis. The system emphasizes explainability, structured reporting, and user auditability, providing complete traceability from source evidence to final conclusions. We discuss critical tradeoffs between agent specialization, system complexity, and resource usage, as well as the integration of deterministic code for validation. An open, interactive user interface allows researchers to directly observe, explore and evaluate the multi-agent workflow. Our evaluation demonstrates substantial gains in efficiency and output consistency, highlighting M-Reason's potential as both a practical tool for evidence synthesis and a testbed for robust multi-agent LLM systems in scientific research, available at https://m-reason.digitalecmt.com. --- ### Tensor-on-tensor Regression Neural Networks for Process Modeling with High-dimensional Data Authors: Qian Wang, Mohammad N. Bisheh, Kamran Paynabar Published: 2025-10-06 Categories: cs.LG, stat.ML arXiv: https://arxiv.org/abs/2510.05329v1 Modern sensing and metrology systems now stream terabytes of heterogeneous, high-dimensional (HD) data profiles, images, and dense point clouds, whose natural representation is multi-way tensors. Understanding such data requires regression models that preserve tensor geometry, yet remain expressive enough to capture the pronounced nonlinear interactions that dominate many industrial and mechanical processes. Existing tensor-based regressors meet the first requirement but remain essentially linear. Conversely, conventional neural networks offer nonlinearity only after flattening, thereby discarding spatial structure and incurring prohibitive parameter counts. This paper introduces a Tensor-on-Tensor Regression Neural Network (TRNN) that unifies these two paradigms. --- ### DeepV: A Model-Agnostic Retrieval-Augmented Framework for Verilog Code Generation with a High-Quality Knowledge Base Authors: Zahin Ibnat, Paul E. Calzada, Rasin Mohammed Ihtemam, Sujan Kumar Saha, Jingbo Zhou, Farimah Farahmandi, Mark Tehranipoor Published: 2025-10-06 Categories: cs.AR, cs.AI arXiv: https://arxiv.org/abs/2510.05327v1 As large language models (LLMs) continue to be integrated into modern technology, there has been an increased push towards code generation applications, which also naturally extends to hardware design automation. LLM-based solutions for register transfer level (RTL) code generation for intellectual property (IP) designs have grown, especially with fine-tuned LLMs, prompt engineering, and agentic approaches becoming popular in literature. However, a gap has been exposed in these techniques, as they fail to integrate novel IPs into the model's knowledge base, subsequently resulting in poorly generated code. Additionally, as general-purpose LLMs continue to improve, fine-tuned methods on older models will not be able to compete to produce more accurate and efficient designs. Although some retrieval augmented generation (RAG) techniques exist to mitigate challenges presented in fine-tuning approaches, works tend to leverage low-quality codebases, incorporate computationally expensive fine-tuning in the frameworks, or do not use RAG directly in the RTL generation step. In this work, we introduce DeepV: a model-agnostic RAG framework to generate RTL designs by enhancing context through a large, high-quality dataset without any RTL-specific training. Our framework benefits the latest commercial LLM, OpenAI's GPT-5, with a near 17% increase in performance on the VerilogEval benchmark. We host DeepV for use by the community in a Hugging Face (HF) Space: https://huggingface.co/spaces/FICS-LLM/DeepV. --- ### Fine-Tuned CNN-Based Approach for Multi-Class Mango Leaf Disease Detection Authors: Jalal Ahmmed, Faruk Ahmed, Rashedul Hasan Shohan, Md. Mahabub Rana, Mahdi Hasan Published: 2025-10-06 Categories: cs.CV arXiv: https://arxiv.org/abs/2510.05326v1 Mango is an important fruit crop in South Asia, but its cultivation is frequently hampered by leaf diseases that greatly impact yield and quality. This research examines the performance of five pre-trained convolutional neural networks, DenseNet201, InceptionV3, ResNet152V2, SeResNet152, and Xception, for multi-class identification of mango leaf diseases across eight classes using a transfer learning strategy with fine-tuning. The models were assessed through standard evaluation metrics, such as accuracy, precision, recall, F1-score, and confusion matrices. Among the architectures tested, DenseNet201 delivered the best results, achieving 99.33% accuracy with consistently strong metrics for individual classes, particularly excelling in identifying Cutting Weevil and Bacterial Canker. Moreover, ResNet152V2 and SeResNet152 provided strong outcomes, whereas InceptionV3 and Xception exhibited lower performance in visually similar categories like Sooty Mould and Powdery Mildew. The training and validation plots demonstrated stable convergence for the highest-performing models. The capability of fine-tuned transfer learning models, for precise and dependable multi-class mango leaf disease detection in intelligent agricultural applications. --- ### Dynamic Functional Connectivity Features for Brain State Classification: Insights from the Human Connectome Project Authors: Valeriya Kirova, Dzerassa Kadieva, Daniil Vlasenko, Isak B. Blank, Fedor Ratnikov Published: 2025-10-06 Categories: q-bio.NC, cs.AI arXiv: https://arxiv.org/abs/2510.05325v1 We analyze functional magnetic resonance imaging (fMRI) data from the Human Connectome Project (HCP) to match brain activities during a range of cognitive tasks. Our findings demonstrate that even basic linear machine learning models can effectively classify brain states and achieve state-of-the-art accuracy, particularly for tasks related to motor functions and language processing. Feature importance ranking allows to identify distinct sets of brain regions whose activation patterns are uniquely associated with specific cognitive functions. These discriminative features provide strong support for the hypothesis of functional specialization across cortical and subcortical areas of the human brain. Additionally, we investigate the temporal dynamics of the identified brain regions, demonstrating that the time-dependent structure of fMRI signals are essential for shaping functional connectivity between regions: uncorrelated areas are least important for classification. This temporal perspective provides deeper insights into the formation and modulation of brain neural networks involved in cognitive processing. --- ### BIRD-INTERACT: Re-imagining Text-to-SQL Evaluation for Large Language Models via Lens of Dynamic Interactions Authors: Nan Huo, Xiaohan Xu, Jinyang Li, Per Jacobsson, Shipei Lin, Bowen Qin, Binyuan Hui, Xiaolong Li, Ge Qu, Shuzheng Si, Linheng Han, Edward Alexander et al. Published: 2025-10-06 Categories: cs.AI arXiv: https://arxiv.org/abs/2510.05318v3 Large language models (LLMs) have demonstrated remarkable performance on single-turn text-to-SQL tasks, but real-world database applications predominantly require multi-turn interactions to handle ambiguous queries, execution errors, and evolving user requirements. Existing multi-turn benchmarks fall short by treating conversation histories as static context or limiting evaluation to read-only operations, failing to reflect production-grade database assistant challenges. We introduce BIRD-INTERACT, a benchmark that restores this realism through: (1) a comprehensive interaction environment coupling each database with a hierarchical knowledge base, metadata files, and a function-driven user simulator, enabling models to solicit clarifications, retrieve knowledge, and recover from errors without human supervision; (2) two evaluation settings consisting of a pre-defined conversational protocol (c-Interact) and an open-ended agentic setting (a-Interact) where models autonomously decide when to query the user simulator or explore the environment; (3) a challenging task suite covering the full CRUD spectrum for business-intelligence and operational use cases, guarded by executable test cases. Each task features ambiguous and follow-up sub-tasks requiring dynamic interaction. The suite comprises BIRD-INTERACT-FULL (600 tasks, up to 11,796 interactions) for comprehensive performance assessment, and BIRD-INTERACT-LITE (300 tasks with simplified databases) for detailed behavioral analysis and rapid method development. Our empirical results highlight BIRD-INTERACT's difficulty: GPT-5 completes only 8.67% of tasks in c-Interact and 17.00% in a-Interact. Analysis via memory grafting and Interaction Test-time Scaling validates the importance of effective interaction for complex, dynamic text-to-SQL tasks. --- ### RegMix: Adversarial Mutual and Generalization Regularization for Enhancing DNN Robustness Authors: Zhenyu Liu, Varun Ojha Published: 2025-10-06 Categories: cs.LG, cs.CV arXiv: https://arxiv.org/abs/2510.05317v1 Adversarial training is the most effective defense against adversarial attacks. The effectiveness of the adversarial attacks has been on the design of its loss function and regularization term. The most widely used loss function in adversarial training is cross-entropy and mean squared error (MSE) as its regularization objective. However, MSE enforces overly uniform optimization between two output distributions during training, which limits its robustness in adversarial training scenarios. To address this issue, we revisit the idea of mutual learning (originally designed for knowledge distillation) and propose two novel regularization strategies tailored for adversarial training: (i) weighted adversarial mutual regularization and (ii) adversarial generalization regularization. In the former, we formulate a decomposed adversarial mutual Kullback-Leibler divergence (KL-divergence) loss, which allows flexible control over the optimization process by assigning unequal weights to the main and auxiliary objectives. In the latter, we introduce an additional clean target distribution into the adversarial training objective, improving generalization and enhancing model robustness. Extensive experiments demonstrate that our proposed methods significantly improve adversarial robustness compared to existing regularization-based approaches. --- ### DeepAf: One-Shot Spatiospectral Auto-Focus Model for Digital Pathology Authors: Yousef Yeganeh, Maximilian Frantzen, Michael Lee, Kun-Hsing Yu, Nassir Navab, Azade Farshad Published: 2025-10-06 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2510.05315v1 While Whole Slide Imaging (WSI) scanners remain the gold standard for digitizing pathology samples, their high cost limits accessibility in many healthcare settings. Other low-cost solutions also face critical limitations: automated microscopes struggle with consistent focus across varying tissue morphology, traditional auto-focus methods require time-consuming focal stacks, and existing deep-learning approaches either need multiple input images or lack generalization capability across tissue types and staining protocols. We introduce a novel automated microscopic system powered by DeepAf, a novel auto-focus framework that uniquely combines spatial and spectral features through a hybrid architecture for single-shot focus prediction. The proposed network automatically regresses the distance to the optimal focal point using the extracted spatiospectral features and adjusts the control parameters for optimal image outcomes. Our system transforms conventional microscopes into efficient slide scanners, reducing focusing time by 80% compared to stack-based methods while achieving focus accuracy of 0.18 μm on the same-lab samples, matching the performance of dual-image methods (0.19 μm) with half the input requirements. DeepAf demonstrates robust cross-lab generalization with only 0.72% false focus predictions and 90% of predictions within the depth of field. Through an extensive clinical study of 536 brain tissue samples, our system achieves 0.90 AUC in cancer classification at 4x magnification, a significant achievement at lower magnification than typical 20x WSI scans. This results in a comprehensive hardware-software design enabling accessible, real-time digital pathology in resource-constrained settings while maintaining diagnostic accuracy. --- ### RAG Makes Guardrails Unsafe? Investigating Robustness of Guardrails under RAG-style Contexts Authors: Yining She, Daniel W. Peterson, Marianne Menglin Liu, Vikas Upadhyay, Mohammad Hossein Chaghazardi, Eunsuk Kang, Dan Roth Published: 2025-10-06 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2510.05310v1 With the increasing adoption of large language models (LLMs), ensuring the safety of LLM systems has become a pressing concern. External LLM-based guardrail models have emerged as a popular solution to screen unsafe inputs and outputs, but they are themselves fine-tuned or prompt-engineered LLMs that are vulnerable to data distribution shifts. In this paper, taking Retrieval Augmentation Generation (RAG) as a case study, we investigated how robust LLM-based guardrails are against additional information embedded in the context. Through a systematic evaluation of 3 Llama Guards and 2 GPT-oss models, we confirmed that inserting benign documents into the guardrail context alters the judgments of input and output guardrails in around 11% and 8% of cases, making them unreliable. We separately analyzed the effect of each component in the augmented context: retrieved documents, user query, and LLM-generated response. The two mitigation methods we tested only bring minor improvements. These results expose a context-robustness gap in current guardrails and motivate training and evaluation protocols that are robust to retrieval and query composition. --- ### Gamma Mixture Modeling for Cosine Similarity in Small Language Models Authors: Kevin Player Published: 2025-10-06 Categories: cs.LG arXiv: https://arxiv.org/abs/2510.05309v1 We study the cosine similarity of sentence transformer embeddings and observe that they are well modeled by gamma mixtures. From a fixed corpus, we measure similarities between all document embeddings and a reference query embedding. Empirically we find that these distributions are often well captured by a gamma distribution shifted and truncated to (-1,1), and in many cases, by a gamma mixture. We propose a heuristic model in which a hierarchical clustering of topics naturally leads to a gamma-mixture structure in the similarity scores. Finally, we outline an expectation-maximization algorithm for fitting shifted gamma mixtures, which provides a practical tool for modeling similarity distributions. --- ### WaveSP-Net: Learnable Wavelet-Domain Sparse Prompt Tuning for Speech Deepfake Detection Authors: Xi Xuan, Xuechen Liu, Wenxin Zhang, Yi-Cheng Lin, Xiaojian Lin, Tomi Kinnunen Published: 2025-10-06 Categories: eess.AS, cs.CL, eess.SP arXiv: https://arxiv.org/abs/2510.05305v2 Modern front-end design for speech deepfake detection relies on full fine-tuning of large pre-trained models like XLSR. However, this approach is not parameter-efficient and may lead to suboptimal generalization to realistic, in-the-wild data types. To address these limitations, we introduce a new family of parameter-efficient front-ends that fuse prompt-tuning with classical signal processing transforms. These include FourierPT-XLSR, which uses the Fourier Transform, and two variants based on the Wavelet Transform: WSPT-XLSR and Partial-WSPT-XLSR. We further propose WaveSP-Net, a novel architecture combining a Partial-WSPT-XLSR front-end and a bidirectional Mamba-based back-end. This design injects multi-resolution features into the prompt embeddings, which enhances the localization of subtle synthetic artifacts without altering the frozen XLSR parameters. Experimental results demonstrate that WaveSP-Net outperforms several state-of-the-art models on two new and challenging benchmarks, Deepfake-Eval-2024 and SpoofCeleb, with low trainable parameters and notable performance gains. The code and models are available at https://github.com/xxuan-acoustics/WaveSP-Net. --- ### SkinMap: Weighted Full-Body Skin Segmentation for Robust Remote Photoplethysmography Authors: Zahra Maleki, Amirhossein Akbari, Amirhossein Binesh, Babak Khalaj Published: 2025-10-06 Categories: cs.CV, eess.IV arXiv: https://arxiv.org/abs/2510.05296v1 Remote photoplethysmography (rPPG) is an innovative method for monitoring heart rate and vital signs by using a simple camera to record a person, as long as any part of their skin is visible. This low-cost, contactless approach helps in remote patient monitoring, emotion analysis, smart vehicle utilization, and more. Over the years, various techniques have been proposed to improve the accuracy of this technology, especially given its sensitivity to lighting and movement. In the unsupervised pipeline, it is necessary to first select skin regions from the video to extract the rPPG signal from the skin color changes. We introduce a novel skin segmentation technique that prioritizes skin regions to enhance the quality of the extracted signal. It can detect areas of skin all over the body, making it more resistant to movement, while removing areas such as the mouth, eyes, and hair that may cause interference. Our model is evaluated on publicly available datasets, and we also present a new dataset, called SYNC-rPPG, to better represent real-world conditions. The results indicate that our model demonstrates a prior ability to capture heartbeats in challenging conditions, such as talking and head rotation, and maintain the mean absolute error (MAE) between predicted and actual heart rates, while other methods fail to do so. In addition, we demonstrate high accuracy in detecting a diverse range of skin tones, making this technique a promising option for real-world applications. --- ### AUREXA-SE: Audio-Visual Unified Representation Exchange Architecture with Cross-Attention and Squeezeformer for Speech Enhancement Authors: M. Sajid, Deepanshu Gupta, Yash Modi, Sanskriti Jain, Harshith Jai Surya Ganji, A. Rahaman, Harshvardhan Choudhary, Nasir Saleem, Amir Hussain, M. Tanveer Published: 2025-10-06 Categories: cs.SD, cs.AI, cs.MM arXiv: https://arxiv.org/abs/2510.05295v1 In this paper, we propose AUREXA-SE (Audio-Visual Unified Representation Exchange Architecture with Cross-Attention and Squeezeformer for Speech Enhancement), a progressive bimodal framework tailored for audio-visual speech enhancement (AVSE). AUREXA-SE jointly leverages raw audio waveforms and visual cues by employing a U-Net-based 1D convolutional encoder for audio and a Swin Transformer V2 for efficient and expressive visual feature extraction. Central to the architecture is a novel bidirectional cross-attention mechanism, which facilitates deep contextual fusion between modalities, enabling rich and complementary representation learning. To capture temporal dependencies within the fused embeddings, a stack of lightweight Squeezeformer blocks combining convolutional and attention modules is introduced. The enhanced embeddings are then decoded via a U-Net-style decoder for direct waveform reconstruction, ensuring perceptually consistent and intelligible speech output. Experimental evaluations demonstrate the effectiveness of AUREXA-SE, achieving significant performance improvements over noisy baselines, with STOI of 0.516, PESQ of 1.323, and SI-SDR of -4.322 dB. The source code of AUREXA-SE is available at https://github.com/mtanveer1/AVSEC-4-Challenge-2025. --- ### Camellia: Benchmarking Cultural Biases in LLMs for Asian Languages Authors: Tarek Naous, Anagha Savit, Carlos Rafael Catalan, Geyang Guo, Jaehyeok Lee, Kyungdon Lee, Lheane Marie Dizon, Mengyu Ye, Neel Kothari, Sahajpreet Singh, Sarah Masud, Tanish Patwa et al. Published: 2025-10-06 Categories: cs.CL arXiv: https://arxiv.org/abs/2510.05291v3 As Large Language Models (LLMs) develop stronger multilingual capabilities, their sensitivity to culturally diverse entities becomes increasingly important. Prior work by Naous et al. (2024) has shown that LLMs often favor Western-associated entities in Arabic. Due to the lack of entity-centric multilingual benchmarks, it remains unclear if such biases also manifest in various non-Western languages. In this paper, we introduce Camellia, a benchmark for evaluating entity-centric cultural biases in nine Asian languages, spanning six Asian cultures. Camellia includes 19,530 manually annotated entities associated with the covered Asian or Western cultures, as well as 2,173 masked contexts for these entities derived from social media posts. Using Camellia, we evaluate cultural biases in four recent multilingual LLMs across three tasks: cultural context adaptation, sentiment association, and entity extractive QA. Our analyses show that LLMs struggle with cultural adaptation across these languages, with performance differing across models developed in different regions. We further observe that different LLM families can hold distinct biases, reflected in the ways they link cultures to particular sentiments. Lastly, we find that LLMs can struggle with context understanding in some Asian languages, creating performance gaps between cultures in entity extraction. --- ### DP-Adam-AC: Privacy-preserving Fine-Tuning of Localizable Language Models Using Adam Optimization with Adaptive Clipping Authors: Ruoxing Yang Published: 2025-10-06 Categories: cs.LG, cs.AI, cs.CR arXiv: https://arxiv.org/abs/2510.05288v1 Large language models (LLMs) such as ChatGPT have evolved into powerful and ubiquitous tools. Fine-tuning on small datasets allows LLMs to acquire specialized skills for specific tasks efficiently. Although LLMs provide great utility in both general and task-specific use cases, they are limited by two security-related concerns. First, traditional LLM hardware requirements make them infeasible to run locally on consumer-grade devices. A remote network connection with the LLM provider's server is usually required, making the system vulnerable to network attacks. Second, fine-tuning an LLM for a sensitive task may involve sensitive data. Non-private fine-tuning algorithms produce models vulnerable to training data reproduction attacks. Our work addresses these security concerns by enhancing differentially private optimization algorithms and applying them to fine-tune localizable language models. We introduce adaptable gradient clipping along with other engineering enhancements to the standard DP-Adam optimizer to create DP-Adam-AC. We use our optimizer to fine-tune examples of two localizable LLM designs, small language model (Qwen2.5-0.5B) and 1.58 bit quantization (Bitnet-b1.58-2B). We demonstrate promising improvements in loss through experimentation with two synthetic datasets. --- ### Computing frustration and near-monotonicity in deep neural networks Authors: Joel Wendin, Erik G. Larsson, Claudio Altafini Published: 2025-10-06 Categories: cs.LG, cond-mat.dis-nn, stat.ML arXiv: https://arxiv.org/abs/2510.05286v1 For the signed graph associated to a deep neural network, one can compute the frustration level, i.e., test how close or distant the graph is to structural balance. For all the pretrained deep convolutional neural networks we consider, we find that the frustration is always less than expected from null models. From a statistical physics point of view, and in particular in reference to an Ising spin glass model, the reduced frustration indicates that the amount of disorder encoded in the network is less than in the null models. From a functional point of view, low frustration (i.e., proximity to structural balance) means that the function representing the network behaves near-monotonically, i.e., more similarly to a monotone function than in the null models. Evidence of near-monotonic behavior along the partial order determined by frustration is observed for all networks we consider. This confirms that the class of deep convolutional neural networks tends to have a more ordered behavior than expected from null models, and suggests a novel form of implicit regularization. --- ### Adjusting the Output of Decision Transformer with Action Gradient Authors: Rui Lin, Yiwen Zhang, Zhicheng Peng, Minghao Lyu Published: 2025-10-06 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2510.05285v1 Decision Transformer (DT), which integrates reinforcement learning (RL) with the transformer model, introduces a novel approach to offline RL. Unlike classical algorithms that take maximizing cumulative discounted rewards as objective, DT instead maximizes the likelihood of actions. This paradigm shift, however, presents two key challenges: stitching trajectories and extrapolation of action. Existing methods, such as substituting specific tokens with predictive values and integrating the Policy Gradient (PG) method, address these challenges individually but fail to improve performance stably when combined due to inherent instability. To address this, we propose Action Gradient (AG), an innovative methodology that directly adjusts actions to fulfill a function analogous to that of PG, while also facilitating efficient integration with token prediction techniques. AG utilizes the gradient of the Q-value with respect to the action to optimize the action. The empirical results demonstrate that our method can significantly enhance the performance of DT-based algorithms, with some results achieving state-of-the-art levels. --- ### Beyond Monolithic Rewards: A Hybrid and Multi-Aspect Reward Optimization for MLLM Alignment Authors: Radha Gulhane, Sathish Reddy Indurthi Published: 2025-10-06 Categories: cs.AI, cs.CL, cs.CV arXiv: https://arxiv.org/abs/2510.05283v1 Aligning multimodal large language models (MLLMs) with human preferences often relies on single-signal, model-based reward methods. Such monolithic rewards often lack confidence calibration across domain-specific tasks, fail to capture diverse aspects of human preferences, and require extensive data annotation and reward model training. In this work, we propose a hybrid reward modeling framework that integrates complementary reward paradigms: (i) model-based rewards, where a learned reward model predicts scalar or vector scores from synthetic and human feedback, and (ii) rule-based rewards, where domain-specific heuristics provide explicit correctness signals with confidence. Beyond accuracy, we further incorporate multi-aspect rewards to enforce instruction adherence and introduce a generalized length-penalty reward to stabilize training and improve performance. The proposed framework provides a flexible and effective approach to aligning MLLMs through reinforcement learning policy optimization. Our experiments show consistent improvements across different multimodal benchmarks when applying hybrid and multi-aspect reward modeling. Our best performing model in the 3B family achieves an overall average improvement of ~9.5% across general and math reasoning tasks. Focusing specifically on mathematical benchmarks, the model achieves a significant average improvement of ~16%, highlighting its effectiveness in mathematical reasoning and problem solving. --- ### Decoding Partial Differential Equations: Cross-Modal Adaptation of Decoder-only Models to PDEs Authors: Paloma García-de-Herreros, Philipp Slusallek, Dietrich Klakow, Vagrant Gautam Published: 2025-10-06 Categories: cs.LG, cs.CL arXiv: https://arxiv.org/abs/2510.05278v2 While large language models are primarily used on natural language tasks, they have also shown great promise when adapted to new modalities, e.g., for scientific machine learning tasks. Most proposed approaches for such cross-modal adaptation of language models focus on encoder-only transformer model architectures, despite decoder-only architectures being far more popular for language tasks in recent years, and being trained at much larger scales. This raises the question of how model architecture affects cross-modal adaptation approaches, and whether we can leverage the success of decoder-only models. In this paper, we systematically compare encoder-only and decoder-only language models on cross-modal adaptation for time-dependent simulation tasks based on partial differential equations (PDEs). We find that decoder-only models are far worse than encoder-only models, when existing approaches are applied unmodified. In contrast to several other domains, scaling decoder-only models also does not help. To enhance the performance of decoder-only models in this context, we introduce two novel approaches that mimic bidirectionality, Parallel Flipping and Sequence Doubling. Both our methods improve overall performance using decoder-only models for all tasks and all cross-modal adaptation methods, closing the gap to encoder-only model performance. We hope that our findings broaden the spectrum of models used on cross-modal adaptation tasks to further scientific machine learning. --- ### Attention-Enhanced Prototypical Learning for Few-Shot Infrastructure Defect Segmentation Authors: Christina Thrainer, Md Meftahul Ferdaus, Mahdi Abdelguerfi, Christian Guetl, Steven Sloan, Kendall N. Niles, Ken Pathak Published: 2025-10-06 Categories: cs.CV arXiv: https://arxiv.org/abs/2510.05266v1 Few-shot semantic segmentation is vital for deep learning-based infrastructure inspection applications, where labeled training examples are scarce and expensive. Although existing deep learning frameworks perform well, the need for extensive labeled datasets and the inability to learn new defect categories with little data are problematic. We present our Enhanced Feature Pyramid Network (E-FPN) framework for few-shot semantic segmentation of culvert and sewer defect categories using a prototypical learning framework. Our approach has three main contributions: (1) adaptive E-FPN encoder using InceptionSepConv blocks and depth-wise separable convolutions for efficient multi-scale feature extraction; (2) prototypical learning with masked average pooling for powerful prototype generation from small support examples; and (3) attention-based feature representation through global self-attention, local self-attention and cross-attention. Comprehensive experimentation on challenging infrastructure inspection datasets illustrates that the method achieves excellent few-shot performance, with the best configuration being 8-way 5-shot training configuration at 82.55% F1-score and 72.26% mIoU in 2-way classification testing. The self-attention method had the most significant performance improvements, providing 2.57% F1-score and 2.9% mIoU gain over baselines. Our framework addresses the critical need to rapidly respond to new defect types in infrastructure inspection systems with limited new training data that lead to more efficient and economical maintenance plans for critical infrastructure systems. --- ### ECLipsE-Gen-Local: Efficient Compositional Local Lipschitz Estimates for Deep Neural Networks Authors: Yuezhu Xu, S. Sivaranjani Published: 2025-10-06 Categories: cs.LG arXiv: https://arxiv.org/abs/2510.05261v2 The Lipschitz constant is a key measure for certifying the robustness of neural networks to input perturbations. However, computing the exact constant is NP-hard, and standard approaches to estimate the Lipschitz constant involve solving a large matrix semidefinite program (SDP) that scales poorly with network size. Further, there is a potential to efficiently leverage local information on the input region to provide tighter Lipschitz estimates. We address this problem here by proposing a compositional framework that yields tight yet scalable Lipschitz estimates for deep feedforward neural networks. Specifically, we begin by developing a generalized SDP framework that is highly flexible, accommodating heterogeneous activation function slope, and allowing Lipschitz estimates with respect to arbitrary input-output pairs and arbitrary choices of sub-networks of consecutive layers. We then decompose this generalized SDP into a sequence of small sub-problems, with computational complexity that scales linearly with respect to the network depth. We also develop a variant that achieves near-instantaneous computation through closed-form solutions to each sub-problem. All our algorithms are accompanied by theoretical guarantees on feasibility and validity. Next, we develop a series of algorithms, termed as ECLipsE-Gen-Local, that effectively incorporate local information on the input. Our experiments demonstrate that our algorithms achieve substantial speedups over a multitude of benchmarks while producing significantly tighter Lipschitz bounds than global approaches. Moreover, we show that our algorithms provide strict upper bounds for the Lipschitz constant with values approaching the exact Jacobian from autodiff when the input region is small enough. Finally, we demonstrate the practical utility of our approach by showing that our Lipschitz estimates closely align with network robustness. --- ### Let it Calm: Exploratory Annealed Decoding for Verifiable Reinforcement Learning Authors: Chenghao Yang, Lin Gui, Chenxiao Yang, Victor Veitch, Lizhu Zhang, Zhuokai Zhao Published: 2025-10-06 Categories: cs.CL, cs.LG arXiv: https://arxiv.org/abs/2510.05251v1 Reinforcement learning with verifiable rewards (RLVR) is a powerful paradigm for enhancing the reasoning capabilities of large language models (LLMs), yet its success hinges on effective exploration. An ideal exploration strategy must navigate two fundamental challenges: it must preserve sample quality while also ensuring training stability. While standard fixed-temperature sampling is simple, it struggles to balance these competing demands, as high temperatures degrade sample quality and low temperatures limit discovery. In this work, we propose a simpler and more effective strategy, Exploratory Annealed Decoding (EAD), grounded in the insight that exploration is most impactful on early tokens which define a sequence's semantic direction. EAD implements an intuitive **explore-at-the-beginning, exploit-at-the-end** strategy by annealing the sampling temperature from high to low during generation. This dynamic schedule encourages meaningful, high-level diversity at the start, then gradually lowers the temperature to preserve sample quality and keep the sampling distribution close to the target policy, which is essential for stable training. We demonstrate that EAD is a lightweight, plug-and-play method that significantly improves sample efficiency, consistently outperforming fixed-temperature sampling across various RLVR algorithms and model sizes. Our work suggests that aligning exploration with the natural dynamics of sequential generation offers a robust path to improving LLM reasoning. --- ### Stratum: System-Hardware Co-Design with Tiered Monolithic 3D-Stackable DRAM for Efficient MoE Serving Authors: Yue Pan, Zihan Xia, Po-Kai Hsu, Lanxiang Hu, Hyungyo Kim, Janak Sharda, Minxuan Zhou, Nam Sung Kim, Shimeng Yu, Tajana Rosing, Mingu Kang Published: 2025-10-06 Categories: cs.AR, cs.ET, cs.LG arXiv: https://arxiv.org/abs/2510.05245v1 As Large Language Models (LLMs) continue to evolve, Mixture of Experts (MoE) architecture has emerged as a prevailing design for achieving state-of-the-art performance across a wide range of tasks. MoE models use sparse gating to activate only a handful of expert sub-networks per input, achieving billion-parameter capacity with inference costs akin to much smaller models. However, such models often pose challenges for hardware deployment due to the massive data volume introduced by the MoE layers. To address the challenges of serving MoE models, we propose Stratum, a system-hardware co-design approach that combines the novel memory technology Monolithic 3D-Stackable DRAM (Mono3D DRAM), near-memory processing (NMP), and GPU acceleration. The logic and Mono3D DRAM dies are connected through hybrid bonding, whereas the Mono3D DRAM stack and GPU are interconnected via silicon interposer. Mono3D DRAM offers higher internal bandwidth than HBM thanks to the dense vertical interconnect pitch enabled by its monolithic structure, which supports implementations of higher-performance near-memory processing. Furthermore, we tackle the latency differences introduced by aggressive vertical scaling of Mono3D DRAM along the z-dimension by constructing internal memory tiers and assigning data across layers based on access likelihood, guided by topic-based expert usage prediction to boost NMP throughput. The Stratum system achieves up to 8.29x improvement in decoding throughput and 7.66x better energy efficiency across various benchmarks compared to GPU baselines. --- ### Simultaneous Learning and Optimization via Misspecified Saddle Point Problems Authors: Mohammad Mahdi Ahmadi, Erfan Yazdandoost Hamedani Published: 2025-10-06 Categories: cs.LG, math.OC arXiv: https://arxiv.org/abs/2510.05241v1 We study a class of misspecified saddle point (SP) problems, where the optimization objective depends on an unknown parameter that must be learned concurrently from data. Unlike existing studies that assume parameters are fully known or pre-estimated, our framework integrates optimization and learning into a unified formulation, enabling a more flexible problem class. To address this setting, we propose two algorithms based on the accelerated primal-dual (APD) by Hamedani & Aybat 2021. In particular, we first analyze the naive extension of the APD method by directly substituting the evolving parameter estimates into the primal-dual updates; then, we design a new learning-aware variant of the APD method that explicitly accounts for parameter dynamics by adjusting the momentum updates. Both methods achieve a provable convergence rate of $\mathcal{O}(\log K / K)$, while the learning-aware approach attains a tighter $\mathcal{O}(1)$ constant and further benefits from an adaptive step-size selection enabled by a backtracking strategy. Furthermore, we extend the framework to problems where the learning problem admits multiple optimal solutions, showing that our modified algorithm for a structured setting achieves an $\mathcal{O}(1/\sqrt{K})$ rate. To demonstrate practical impact, we evaluate our methods on a misspecified portfolio optimization problem and show superior empirical performance compared to state-of-the-art algorithms. --- ### CMT-Benchmark: A Benchmark for Condensed Matter Theory Built by Expert Researchers Authors: Haining Pan, James V. Roggeveen, Erez Berg, Juan Carrasquilla, Debanjan Chowdhury, Surya Ganguli, Federico Ghimenti, Juraj Hasik, Henry Hunt, Hong-Chen Jiang, Mason Kamb, Ying-Jer Kao et al. Published: 2025-10-06 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2510.05228v2 Large language models (LLMs) have shown remarkable progress in coding and math problem-solving, but evaluation on advanced research-level problems in hard sciences remains scarce. To fill this gap, we present CMT-Benchmark, a dataset of 50 problems covering condensed matter theory (CMT) at the level of an expert researcher. Topics span analytical and computational approaches in quantum many-body, and classical statistical mechanics. The dataset was designed and verified by a panel of expert researchers from around the world. We built the dataset through a collaborative environment that challenges the panel to write and refine problems they would want a research assistant to solve, including Hartree-Fock, exact diagonalization, quantum/variational Monte Carlo, density matrix renormalization group (DMRG), quantum/classical statistical mechanics, and model building. We evaluate LLMs by programmatically checking solutions against expert-supplied ground truth. We developed machine-grading, including symbolic handling of non-commuting operators via normal ordering. They generalize across tasks too. Our evaluations show that frontier models struggle with all of the problems in the dataset, highlighting a gap in the physical reasoning skills of current LLMs. Notably, experts identified strategies for creating increasingly difficult problems by interacting with the LLMs and exploiting common failure modes. The best model, GPT5, solves 30\% of the problems; average across 17 models (GPT, Gemini, Claude, DeepSeek, Llama) is 11.4\pm2.1\%. Moreover, 18 problems are solved by none of the 17 models, and 26 by at most one. These unsolved problems span Quantum Monte Carlo, Variational Monte Carlo, and DMRG. Answers sometimes violate fundamental symmetries or have unphysical scaling dimensions. We believe this benchmark will guide development toward capable AI research assistants and tutors. --- ### Approximate Gaussianity Beyond Initialisation in Neural Networks Authors: Edward Hirst, Sanjaye Ramgoolam Published: 2025-10-06 Categories: cs.LG, cs.AI, hep-th arXiv: https://arxiv.org/abs/2510.05218v1 Ensembles of neural network weight matrices are studied through the training process for the MNIST classification problem, testing the efficacy of matrix models for representing their distributions, under assumptions of Gaussianity and permutation-symmetry. The general 13-parameter permutation invariant Gaussian matrix models are found to be effective models for the correlated Gaussianity in the weight matrices, beyond the range of applicability of the simple Gaussian with independent identically distributed matrix variables, and notably well beyond the initialisation step. The representation theoretic model parameters, and the graph-theoretic characterisation of the permutation invariant matrix observables give an interpretable framework for the best-fit model and for small departures from Gaussianity. Additionally, the Wasserstein distance is calculated for this class of models and used to quantify the movement of the distributions over training. Throughout the work, the effects of varied initialisation regimes, regularisation, layer depth, and layer width are tested for this formalism, identifying limits where particular departures from Gaussianity are enhanced and how more general, yet still highly-interpretable, models can be developed. --- ### VER: Vision Expert Transformer for Robot Learning via Foundation Distillation and Dynamic Routing Authors: Yixiao Wang, Mingxiao Huo, Zhixuan Liang, Yushi Du, Lingfeng Sun, Haotian Lin, Jinghuan Shang, Chensheng Peng, Mohit Bansal, Mingyu Ding, Masayoshi Tomizuka Published: 2025-10-06 Categories: cs.RO, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2510.05213v2 Pretrained vision foundation models (VFMs) advance robotic learning via rich visual representations, yet individual VFMs typically excel only in specific domains, limiting generality across tasks. Distilling multiple VFMs into a unified representation for policy can mitigate this limitation but often yields inflexible task-specific feature selection and requires costly full re-training to incorporate robot-domain knowledge. We propose VER, a Vision Expert transformer for Robot learning. During pretraining, VER distills multiple VFMs into a vision expert library. It then fine-tunes only a lightweight routing network (fewer than 0.4% of parameters) to dynamically select task-relevant experts from the pretrained library for downstream robot tasks. We further introduce Patchwise Expert Routing with Curriculum Top-K Annealing to improve both flexibility and precision of dynamic expert selection. Moreover, VER supports parameter-efficient finetuning for scalable expert utilization and adaptive robot-domain knowledge integration. Across 17 diverse robotic tasks and multiple policy heads, VER achieves state-of-the-art performance. We find that VER reduces large-norm outliers in task-irrelevant regions (e.g., background) and concentrates on task-critical regions. Visualizations and codes can be found in https://yixiaowang7.github.io/ver_page/. --- ### A Data-Driven Prism: Multi-View Source Separation with Diffusion Model Priors Authors: Sebastian Wagner-Carena, Aizhan Akhmetzhanova, Sydney Erickson Published: 2025-10-06 Categories: cs.LG, astro-ph.CO arXiv: https://arxiv.org/abs/2510.05205v1 A common challenge in the natural sciences is to disentangle distinct, unknown sources from observations. Examples of this source separation task include deblending galaxies in a crowded field, distinguishing the activity of individual neurons from overlapping signals, and separating seismic events from an ambient background. Traditional analyses often rely on simplified source models that fail to accurately reproduce the data. Recent advances have shown that diffusion models can directly learn complex prior distributions from noisy, incomplete data. In this work, we show that diffusion models can solve the source separation problem without explicit assumptions about the source. Our method relies only on multiple views, or the property that different sets of observations contain different linear transformations of the unknown sources. We show that our method succeeds even when no source is individually observed and the observations are noisy, incomplete, and vary in resolution. The learned diffusion models enable us to sample from the source priors, evaluate the probability of candidate sources, and draw from the joint posterior of the source distribution given an observation. We demonstrate the effectiveness of our method on a range of synthetic problems as well as real-world galaxy observations. --- ### Efficient Prediction of Pass@k Scaling in Large Language Models Authors: Joshua Kazdan, Rylan Schaeffer, Youssef Allouah, Colin Sullivan, Kyssen Yu, Noam Levi, Sanmi Koyejo Published: 2025-10-06 Categories: cs.AI, cs.LG, stat.AP, stat.ML arXiv: https://arxiv.org/abs/2510.05197v1 Assessing the capabilities and risks of frontier AI systems is a critical area of research, and recent work has shown that repeated sampling from models can dramatically increase both. For instance, repeated sampling has been shown to increase their capabilities, such as solving difficult math and coding problems, but it has also been shown to increase their potential for harm, such as being jailbroken. Such results raise a crucial question for both capability and safety forecasting: how can one accurately predict a model's behavior when scaled to a massive number of attempts, given a vastly smaller sampling budget? This question is directly relevant to model providers, who serve hundreds of millions of users daily, and to governmental regulators, who seek to prevent harms. To answer this questions, we make three contributions. First, we find that standard methods for fitting these laws suffer from statistical shortcomings that hinder predictive accuracy, especially in data-limited scenarios. Second, we remedy these shortcomings by introducing a robust estimation framework, which uses a beta-binomial distribution to generate more accurate predictions from limited data. Third, we propose a dynamic sampling strategy that allocates a greater budget to harder problems. Combined, these innovations enable more reliable prediction of rare risks and capabilities at a fraction of the computational cost. --- ### Graph-based LLM over Semi-Structured Population Data for Dynamic Policy Response Authors: Daqian Shi, Xiaolei Diao, Jinge Wu, Honghan Wu, Xiongfeng Tang, Felix Naughton, Paulina Bondaronek Published: 2025-10-06 Categories: cs.AI arXiv: https://arxiv.org/abs/2510.05196v1 Timely and accurate analysis of population-level data is crucial for effective decision-making during public health emergencies such as the COVID-19 pandemic. However, the massive input of semi-structured data, including structured demographic information and unstructured human feedback, poses significant challenges to conventional analysis methods. Manual expert-driven assessments, though accurate, are inefficient, while standard NLP pipelines often require large task-specific labeled datasets and struggle with generalization across diverse domains. To address these challenges, we propose a novel graph-based reasoning framework that integrates large language models with structured demographic attributes and unstructured public feedback in a weakly supervised pipeline. The proposed approach dynamically models evolving citizen needs into a need-aware graph, enabling population-specific analyses based on key features such as age, gender, and the Index of Multiple Deprivation. It generates interpretable insights to inform responsive health policy decision-making. We test our method using a real-world dataset, and preliminary experimental results demonstrate its feasibility. This approach offers a scalable solution for intelligent population health monitoring in resource-constrained clinical and governmental settings. --- ### From surveillance to signalling: escalation channels as environmental controls for agentic AI Authors: Francesca Gomez Published: 2025-10-06 Categories: cs.CR, cs.AI arXiv: https://arxiv.org/abs/2510.05192v2 When AI agents operating with access to sensitive information encounter a conflict between completing an assigned task and following rules or ethical constraints, they can resort to unsanctioned behaviour. Existing inference time safety work addresses this primarily through monitoring and access restriction. We investigate a complementary and under-explored layer: environmental controls that act on the agent's decision context at the point of conflict, making it more likely that the agent takes an authorised alternative path rather than an unsanctioned one. Drawing on Situational Crime Prevention (SCP), a framework used in human insider risk management to make harmful actions less rewarding and compliant actions more viable by design choices in the environment, we design and evaluate escalation channels as a concrete instantiation of this control class. An escalation channel provides an agent with a formal, out-of-band route to surface a conflict to an independent authority. We evaluate two designs: a simple email escalation and an instrumentally credible channel that guarantees a 30-minute pause and independent review, making the authorised path genuinely useful for goal achievement rather than merely nominally available. Across 10 frontier LLMs using the agentic task-rule conflict scenario of Lynch et al. (2025), we find that without any control the harmful action rate is 38.73%. A simple escalation channel reduces this to 5.92%; the instrumentally credible channel reduces it further to 1.21%, a statistically significant improvement observed in all 10 models tested across 24,000 samples. Our results suggest that the instrumental credibility of the authorised alternative matters considerably, and that environmental control design is a productive and largely unexplored addition to the defence-in-depth toolkit for agentic AI systems. --- ### Provable Speech Attributes Conversion via Latent Independence Authors: Jonathan Svirsky, Ofir Lindenbaum, Uri Shaham Published: 2025-10-06 Categories: cs.SD, cs.AI arXiv: https://arxiv.org/abs/2510.05191v3 Conditional generation and disentangled representation learning are central to controlled generation across audio, vision, and multimodal domains. However, despite strong empirical progress, particularly in speech style transfer, most existing approaches rely on heuristic objectives and architectural choices, offering limited theoretical understanding of when and why reliable attribute control is achievable. In this work, we develop a formal framework for speech attribute conversion and provide a theoretical analysis of sufficient conditions for exact and consistent transfer. Our analysis focuses on a deterministic autoencoder setting augmented with an independence constraint between the learned latent representation and the controllable attribute. Under explicit population-level assumptions about the data-generating process, we establish guarantees linking reconstruction, independence, and the feasibility of attribute manipulation while preserving task-relevant content. We further show how the theoretical framework translates into practice by proposing a practical voice conversion method that directly implements its core principles. Experimental evaluations on voice and pitch conversion tasks demonstrate the applicability of the theoretical analysis to real-world speech conversion settings and show that the resulting method achieves competitive performance against existing approaches. --- ### A novel hallucination classification framework Authors: Maksym Zavhorodnii, Dmytro Dehtiarov, Anna Konovalenko Published: 2025-10-06 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2510.05189v1 This work introduces a novel methodology for the automatic detection of hallucinations generated during large language model (LLM) inference. The proposed approach is based on a systematic taxonomy and controlled reproduction of diverse hallucination types through prompt engineering. A dedicated hallucination dataset is subsequently mapped into a vector space using an embedding model and analyzed with unsupervised learning techniques in a reduced-dimensional representation of hallucinations with veridical responses. Quantitative evaluation of inter-centroid distances reveals a consistent correlation between the severity of informational distortion in hallucinations and their spatial divergence from the cluster of correct outputs. These findings provide theoretical and empirical evidence that even simple classification algorithms can reliably distinguish hallucinations from accurate responses within a single LLM, thereby offering a lightweight yet effective framework for improving model reliability. --- ### Plug-and-Play Dramaturge: A Divide-and-Conquer Approach for Iterative Narrative Script Refinement via Collaborative LLM Agents Authors: Wenda Xie, Chao Guo, Yanqing Jing, Junle Wang, Yisheng Lv, Fei-Yue Wang Published: 2025-10-06 Categories: cs.AI arXiv: https://arxiv.org/abs/2510.05188v4 Although LLMs have been widely adopted for creative content generation, a single-pass process often struggles to produce high-quality long narratives. How to effectively revise and improve long narrative scripts like scriptwriters remains a significant challenge, as it demands a comprehensive understanding of the entire context to identify global structural issues and local detailed flaws, as well as coordinating revisions at multiple granularities and locations. Direct modifications by LLMs typically introduce inconsistencies between local edits and the overall narrative requirements. To address these issues, we propose Dramaturge, a task and feature oriented divide-and-conquer approach powered by hierarchical multiple LLM agents. It consists of a Global Review stage to grasp the overall storyline and structural issues, a Scene-level Review stage to pinpoint detailed scene and sentence flaws, and a Hierarchical Coordinated Revision stage that coordinates and integrates structural and detailed improvements throughout the script. The top-down task flow ensures that high-level strategies guide local modifications, maintaining contextual consistency. The review and revision workflow follows a coarse-to-fine iterative process, continuing through multiple rounds until no further substantive improvements can be made. Comprehensive experiments show that Dramaturge significantly outperforms all baselines in terms of script-level overall quality and scene-level details. Our approach is plug-and-play and can be easily integrated into existing methods to improve the generated scripts. --- ### Real-time Framework for Interoperable Semantic-driven Internet-of-Things in Smart Agriculture Authors: Mohamed El-Dosuky Published: 2025-10-06 Categories: cs.AI arXiv: https://arxiv.org/abs/2510.05187v1 The Internet of Things (IoT) has revolutionized various applications including agriculture, but it still faces challenges in data collection and understanding. This paper proposes a real-time framework with three additional semantic layers to help IoT devices and sensors comprehend data meaning and source. The framework consists of six layers: perception, semantic annotation, interoperability, transportation, semantic reasoning, and application, suitable for dynamic environments. Sensors collect data in the form of voltage, which is then processed by microprocessors or microcontrollers in the semantic annotation and preprocessing layer. Metadata is added to the raw data, including the purpose, ID number, and application. Two semantic algorithms are proposed in the semantic interoperability and ontologies layer: the interoperability semantic algorithm for standardizing file types and the synonym identification algorithm for identifying synonyms. In the transportation layer, raw data and metadata are sent to other IoT devices or cloud computing platforms using techniques like WiFi, Zigbee networks, Bluetooth, and mobile communication networks. A semantic reasoning layer is proposed to infer new knowledge from the existing data, using fuzzy logic, Dempster-Shafer theory, and Bayesian networks. A Graphical User Interface (GUI) is proposed in the application layer to help users communicate with and monitor IoT sensors, devices, and new knowledge inferred. This framework provides a robust solution for managing IoT data, ensuring semantic completeness, and enabling real-time knowledge inference. The integration of uncertainty reasoning methods and semantic interoperability techniques makes this framework a valuable tool for advancing IoT applications in general and in agriculture in particular. --- ### OptPipe: Memory- and Scheduling-Optimized Pipeline Parallelism for LLM Training Authors: Hongpei Li, Han Zhang, Huikang Liu, Dongdong Ge, Yinyu Ye Published: 2025-10-06 Categories: cs.DC, cs.AI, math.OC arXiv: https://arxiv.org/abs/2510.05186v1 Pipeline parallelism (PP) has become a standard technique for scaling large language model (LLM) training across multiple devices. However, despite recent progress in reducing memory consumption through activation offloading, existing approaches remain largely heuristic and coarse-grained, often overlooking the fine-grained trade-offs between memory, computation, and scheduling latency. In this work, we revisit the pipeline scheduling problem from a principled optimization perspective. We observe that prevailing strategies either rely on static rules or aggressively offload activations without fully leveraging the interaction between memory constraints and scheduling efficiency. To address this, we formulate scheduling as a constrained optimization problem that jointly accounts for memory capacity, activation reuse, and pipeline bubble minimization. Solving this model yields fine-grained schedules that reduce pipeline bubbles while adhering to strict memory budgets. Our approach complements existing offloading techniques: whereas prior approaches trade memory for time in a fixed pattern, we dynamically optimize the tradeoff with respect to model structure and hardware configuration. Experimental results demonstrate that our method consistently improves both throughput and memory utilization. In particular, we reduce idle pipeline time by up to 50% under the same per-device memory limit, and in some cases, enable the training of larger models within limited memory budgets. --- ### TopInG: Topologically Interpretable Graph Learning via Persistent Rationale Filtration Authors: Cheng Xin, Fan Xu, Xin Ding, Jie Gao, Jiaxin Ding Published: 2025-10-06 Categories: cs.LG, cs.AI, cs.CG, math.AT, stat.ML arXiv: https://arxiv.org/abs/2510.05102v1 Graph Neural Networks (GNNs) have shown remarkable success across various scientific fields, yet their adoption in critical decision-making is often hindered by a lack of interpretability. Recently, intrinsically interpretable GNNs have been studied to provide insights into model predictions by identifying rationale substructures in graphs. However, existing methods face challenges when the underlying rationale subgraphs are complex and varied. In this work, we propose TopInG: Topologically Interpretable Graph Learning, a novel topological framework that leverages persistent homology to identify persistent rationale subgraphs. TopInG employs a rationale filtration learning approach to model an autoregressive generation process of rationale subgraphs, and introduces a self-adjusted topological constraint, termed topological discrepancy, to enforce a persistent topological distinction between rationale subgraphs and irrelevant counterparts. We provide theoretical guarantees that our loss function is uniquely optimized by the ground truth under specific conditions. Extensive experiments demonstrate TopInG's effectiveness in tackling key challenges, such as handling variform rationale subgraphs, balancing predictive performance with interpretability, and mitigating spurious correlations. Results show that our approach improves upon state-of-the-art methods on both predictive accuracy and interpretation quality. --- ### Pulp Motion: Framing-aware multimodal camera and human motion generation Authors: Robin Courant, Xi Wang, David Loiseaux, Marc Christie, Vicky Kalogeiton Published: 2025-10-06 Categories: cs.GR, cs.CV arXiv: https://arxiv.org/abs/2510.05097v2 Treating human motion and camera trajectory generation separately overlooks a core principle of cinematography: the tight interplay between actor performance and camera work in the screen space. In this paper, we are the first to cast this task as a text-conditioned joint generation, aiming to maintain consistent on-screen framing while producing two heterogeneous, yet intrinsically linked, modalities: human motion and camera trajectories. We propose a simple, model-agnostic framework that enforces multimodal coherence via an auxiliary modality: the on-screen framing induced by projecting human joints onto the camera. This on-screen framing provides a natural and effective bridge between modalities, promoting consistency and leading to more precise joint distribution. We first design a joint autoencoder that learns a shared latent space, together with a lightweight linear transform from the human and camera latents to a framing latent. We then introduce auxiliary sampling, which exploits this linear transform to steer generation toward a coherent framing modality. To support this task, we also introduce the PulpMotion dataset, a human-motion and camera-trajectory dataset with rich captions, and high-quality human motions. Extensive experiments across DiT- and MAR-based architectures show the generality and effectiveness of our method in generating on-frame coherent human-camera motions, while also achieving gains on textual alignment for both modalities. Our qualitative results yield more cinematographically meaningful framings setting the new state of the art for this task. Code, models and data are available in our \href{https://www.lix.polytechnique.fr/vista/projects/2025_pulpmotion_courant/}{project page}. --- ### Paper2Video: Automatic Video Generation from Scientific Papers Authors: Zeyu Zhu, Kevin Qinghong Lin, Mike Zheng Shou Published: 2025-10-06 Categories: cs.CV, cs.AI, cs.CL, cs.MA, cs.MM arXiv: https://arxiv.org/abs/2510.05096v2 Academic presentation videos have become an essential medium for research communication, yet producing them remains highly labor-intensive, often requiring hours of slide design, recording, and editing for a short 2 to 10 minutes video. Unlike natural video, presentation video generation involves distinctive challenges: inputs from research papers, dense multi-modal information (text, figures, tables), and the need to coordinate multiple aligned channels such as slides, subtitles, speech, and human talker. To address these challenges, we introduce Paper2Video, the first benchmark of 101 research papers paired with author-created presentation videos, slides, and speaker metadata. We further design four tailored evaluation metrics--Meta Similarity, PresentArena, PresentQuiz, and IP Memory--to measure how videos convey the paper's information to the audience. Building on this foundation, we propose PaperTalker, the first multi-agent framework for academic presentation video generation. It integrates slide generation with effective layout refinement by a novel effective tree search visual choice, cursor grounding, subtitling, speech synthesis, and talking-head rendering, while parallelizing slide-wise generation for efficiency. Experiments on Paper2Video demonstrate that the presentation videos produced by our approach are more faithful and informative than existing baselines, establishing a practical step toward automated and ready-to-use academic video generation. Our dataset, agent, and code are available at https://github.com/showlab/Paper2Video. --- ### From Noisy Traces to Stable Gradients: Bias-Variance Optimized Preference Optimization for Aligning Large Reasoning Models Authors: Mingkang Zhu, Xi Chen, Bei Yu, Hengshuang Zhao, Jiaya Jia Published: 2025-10-06 Categories: cs.LG, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2510.05095v1 Large reasoning models (LRMs) generate intermediate reasoning traces before producing final answers, yielding strong gains on multi-step and mathematical tasks. Yet aligning LRMs with human preferences, a crucial prerequisite for model deployment, remains underexplored. The statistically correct objective for preference alignment requires marginalizing over reasoning traces, but this computation is intractable in practice. A common workaround optimizes a single sampled trajectory, which introduces substantial gradient variance from stochastic trace sampling. To address this challenge, we frame preference optimization for LRMs through the lens of the bias--variance trade-off and propose Bias--Variance Optimized Preference Optimization (BVPO), a simple, drop-in method that mixes two gradient estimators: a high-variance trace-based estimator and a low-variance empty-trace estimator obtained by disabling reasoning trace generation. Our theory shows that BVPO strictly reduces trace-induced variance for any nontrivial mixture, provides a closed-form choice of the mixing weight that minimizes mean-squared error relative to the true marginal gradient, and under standard smoothness and step-size conditions, tightens classical convergence bounds for stochastic gradient descent. Empirically, BVPO improves alignment over the best baseline by up to 7.8 points on AlpacaEval~2 and 6.8 points on Arena-Hard. Despite being trained only on general conversational data, BVPO also boosts reasoning performance for base models by up to 4.0 points on the average of six math reasoning benchmarks. These results identify variance from trace sampling as a key bottleneck and demonstrate that directly optimizing the bias--variance trade-off yields more stable training and stronger overall performance. --- ### VChain: Chain-of-Visual-Thought for Reasoning in Video Generation Authors: Ziqi Huang, Ning Yu, Gordon Chen, Haonan Qiu, Paul Debevec, Ziwei Liu Published: 2025-10-06 Categories: cs.CV arXiv: https://arxiv.org/abs/2510.05094v2 Recent video generation models can produce smooth and visually appealing clips, but they often struggle to synthesize complex dynamics with a coherent chain of consequences. Accurately modeling visual outcomes and state transitions over time remains a core challenge. In contrast, large language and multimodal models (e.g., GPT-4o) exhibit strong visual state reasoning and future prediction capabilities. To bridge these strengths, we introduce VChain, a novel inference-time chain-of-visual-thought framework that injects visual reasoning signals from multimodal models into video generation. Specifically, VChain contains a dedicated pipeline that leverages large multimodal models to generate a sparse set of critical keyframes as snapshots, which are then used to guide the sparse inference-time visual-state adaptation of a pre-trained video generator only at these key moments. Our approach is tuning-efficient, introduces minimal overhead and avoids dense supervision. Extensive experiments on complex, multi-step scenarios show that VChain significantly enhances the quality of generated videos. --- ### Character Mixing for Video Generation Authors: Tingting Liao, Chongjian Ge, Guangyi Liu, Hao Li, Yi Zhou Published: 2025-10-06 Categories: cs.CV arXiv: https://arxiv.org/abs/2510.05093v1 Imagine Mr. Bean stepping into Tom and Jerry--can we generate videos where characters interact naturally across different worlds? We study inter-character interaction in text-to-video generation, where the key challenge is to preserve each character's identity and behaviors while enabling coherent cross-context interaction. This is difficult because characters may never have coexisted and because mixing styles often causes style delusion, where realistic characters appear cartoonish or vice versa. We introduce a framework that tackles these issues with Cross-Character Embedding (CCE), which learns identity and behavioral logic across multimodal sources, and Cross-Character Augmentation (CCA), which enriches training with synthetic co-existence and mixed-style data. Together, these techniques allow natural interactions between previously uncoexistent characters without losing stylistic fidelity. Experiments on a curated benchmark of cartoons and live-action series with 10 characters show clear improvements in identity preservation, interaction quality, and robustness to style delusion, enabling new forms of generative storytelling.Additional results and videos are available on our project page: https://tingtingliao.github.io/mimix/. --- ### Learning to Interpret Weight Differences in Language Models Authors: Avichal Goel, Yoon Kim, Nir Shavit, Tony T. Wang Published: 2025-10-06 Categories: cs.LG, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2510.05092v4 Finetuning (pretrained) language models is a standard approach for updating their internal parametric knowledge and specializing them to new tasks and domains. However, the corresponding model weight changes ("weight diffs") are not generally interpretable. While inspecting the finetuning dataset can give a sense of how the model might have changed, these datasets are often not publicly available or are too large to work with directly. Towards the goal of comprehensively understanding weight diffs in natural language, we introduce Diff Interpretation Tuning (DIT), a method that trains models to describe their own finetuning-induced modifications. Our approach uses synthetic, labeled weight diffs to train a DIT-adapter, which can be applied to a compatible finetuned model to make it describe how it has changed. We demonstrate in two proof-of-concept settings (reporting hidden behaviors and summarizing finetuned knowledge) that our method enables models to describe their finetuning-induced modifications using accurate natural language descriptions. --- ### Factuality Matters: When Image Generation and Editing Meet Structured Visuals Authors: Le Zhuo, Songhao Han, Yuandong Pu, Boxiang Qiu, Sayak Paul, Yue Liao, Yihao Liu, Jie Shao, Xi Chen, Si Liu, Hongsheng Li Published: 2025-10-06 Categories: cs.CV arXiv: https://arxiv.org/abs/2510.05091v2 While modern visual generation models excel at creating aesthetically pleasing natural images, they struggle with producing or editing structured visuals like charts, diagrams, and mathematical figures, which demand composition planning, text rendering, and multimodal reasoning for factual fidelity. To address this, we present the first comprehensive, systematic investigation of this domain, encompassing data construction, model training, and an evaluation benchmark. First, we construct a large-scale dataset of 1.3 million high-quality structured image pairs derived from executable drawing programs and augmented with chain-of-thought reasoning annotations. Building on it, we train a unified model that integrates a VLM with FLUX.1 Kontext via a lightweight connector for enhanced multimodal understanding. A three-stage training curriculum enables progressive feature alignment, knowledge infusion, and reasoning-augmented generation, further boosted by an external reasoner at inference time. Finally, we introduce StructBench, a novel benchmark for generation and editing with over 1,700 challenging instances, and an accompanying evaluation metric, StructScore, which employs a multi-round Q\&A protocol to assess fine-grained factual accuracy. Evaluations of 15 models reveal that even leading closed-source systems remain far from satisfactory. Our model attains strong editing performance, and inference-time reasoning yields consistent gains across diverse architectures. By releasing the dataset, model, and benchmark, we aim to advance unified multimodal foundations for structured visuals. --- ### Finish First, Perfect Later: Test-Time Token-Level Cross-Validation for Diffusion Large Language Models Authors: Runchu Tian, Junxia Cui, Xueqiang Xu, Feng Yao, Jingbo Shang Published: 2025-10-06 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2510.05090v1 Diffusion large language models (dLLMs) have recently emerged as a promising alternative to autoregressive (AR) models, offering advantages such as accelerated parallel decoding and bidirectional context modeling. However, the vanilla decoding strategy in discrete dLLMs suffers from a critical limitation: once a token is accepted, it can no longer be revised in subsequent steps. As a result, early mistakes persist across iterations, harming both intermediate predictions and final output quality. To address this issue, we propose Tolerator (Token-Level Cross-Validation Refinement), a training-free decoding strategy that leverages cross-validation among predicted tokens. Unlike existing methods that follow a single progressive unmasking procedure, Tolerator introduces a two-stage process: (i) sequence fill-up and (ii) iterative refinement by remasking and decoding a subset of tokens while treating the remaining as context. This design enables previously accepted tokens to be reconsidered and corrected when necessary, leading to more reliable diffusion decoding outputs. We evaluate Tolerator on five standard benchmarks covering language understanding, code generation, and mathematics. Experiments show that our method achieves consistent improvements over the baselines under the same computational budget. These findings suggest that decoding algorithms are crucial to realizing the full potential of diffusion large language models. Code and data are publicly available. --- ### TeachLM: Post-Training LLMs for Education Using Authentic Learning Data Authors: Janos Perczel, Jin Chow, Dorottya Demszky Published: 2025-10-06 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2510.05087v1 The promise of generative AI to revolutionize education is constrained by the pedagogical limits of large language models (LLMs). A major issue is the lack of access to high-quality training data that reflect the learning of actual students. Prompt engineering has emerged as a stopgap, but the ability of prompts to encode complex pedagogical strategies in rule-based natural language is inherently limited. To address this gap we introduce TeachLM - an LLM optimized for teaching through parameter-efficient fine-tuning of state-of-the-art models. TeachLM is trained on a dataset comprised of 100,000 hours of one-on-one, longitudinal student-tutor interactions maintained by Polygence, which underwent a rigorous anonymization process to protect privacy. We use parameter-efficient fine-tuning to develop an authentic student model that enables the generation of high-fidelity synthetic student-tutor dialogues. Building on this capability, we propose a novel multi-turn evaluation protocol that leverages synthetic dialogue generation to provide fast, scalable, and reproducible assessments of the dialogical capabilities of LLMs. Our evaluations demonstrate that fine-tuning on authentic learning data significantly improves conversational and pedagogical performance - doubling student talk time, improving questioning style, increasing dialogue turns by 50%, and greater personalization of instruction. ---