# Research Papers — 2025-10 · bundle page 169 of 217

> Full-text markdown of papers #8401–8450 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/169
[← Previous bundle](https://aipod.dev/papers/archive/2025-10/page/168.md)
[Next bundle →](https://aipod.dev/papers/archive/2025-10/page/170.md)

### Low-Latency Neural Inference on an Edge Device for Real-Time Handwriting Recognition from EEG Signals

Authors: Ovishake Sen, Raghav Soni, Darpan Virmani, Akshar Parekh, Patrick Lehman, Sarthak Jena, Adithi Katikhaneni, Adam Khalifa, Baibhab Chatterjee
Published: 2025-10-07
Categories: eess.SP, cs.LG
arXiv: https://arxiv.org/abs/2510.19832v1

Brain-computer interfaces (BCIs) offer a pathway to restore communication for individuals with severe motor or speech impairments. Imagined handwriting provides an intuitive paradigm for character-level neural decoding, bridging the gap between human intention and digital communication. While invasive approaches such as electrocorticography (ECoG) achieve high accuracy, their surgical risks limit widespread adoption. Non-invasive electroencephalography (EEG) offers safer and more scalable alternatives but suffers from low signal-to-noise ratio and spatial resolution, constraining its decoding precision. This work demonstrates that advanced machine learning combined with informative EEG feature extraction can overcome these barriers, enabling real-time, high-accuracy neural decoding on portable edge devices. A 32-channel EEG dataset was collected from fifteen participants performing imagined handwriting. Signals were preprocessed with bandpass filtering and artifact subspace reconstruction, followed by extraction of 85 time-, frequency-, and graphical-domain features. A hybrid architecture, EEdGeNet, integrates a Temporal Convolutional Network with a multilayer perceptron trained on the extracted features. When deployed on an NVIDIA Jetson TX2, the system achieved 89.83 percent accuracy with 914.18 ms per-character latency. Selecting only ten key features reduced latency by 4.5 times to 202.6 ms with less than 1 percent loss in accuracy. These results establish a pathway for accurate, low-latency, and fully portable non-invasive BCIs supporting real-time communication.

---

### SSL-SE-EEG: A Framework for Robust Learning from Unlabeled EEG Data with Self-Supervised Learning and Squeeze-Excitation Networks

Authors: Meghna Roy Chowdhury, Yi Ding, Shreyas Sen
Published: 2025-10-07
Categories: eess.SP, cs.AI, cs.LG
arXiv: https://arxiv.org/abs/2510.19829v1

Electroencephalography (EEG) plays a crucial role in brain-computer interfaces (BCIs) and neurological diagnostics, but its real-world deployment faces challenges due to noise artifacts, missing data, and high annotation costs. We introduce SSL-SE-EEG, a framework that integrates Self-Supervised Learning (SSL) with Squeeze-and-Excitation Networks (SE-Nets) to enhance feature extraction, improve noise robustness, and reduce reliance on labeled data. Unlike conventional EEG processing techniques, SSL-SE-EEG} transforms EEG signals into structured 2D image representations, suitable for deep learning. Experimental validation on MindBigData, TUH-AB, SEED-IV and BCI-IV datasets demonstrates state-of-the-art accuracy (91% in MindBigData, 85% in TUH-AB), making it well-suited for real-time BCI applications. By enabling low-power, scalable EEG processing, SSL-SE-EEG presents a promising solution for biomedical signal analysis, neural engineering, and next-generation BCIs.

---

### GAZE:Governance-Aware pre-annotation for Zero-shot World Model Environments

Authors: Leela Krishna, Mengyang Zhao, Saicharithreddy Pasula, Harshit Rajgarhia, Abhishek Mukherji
Published: 2025-10-07
Categories: cs.CV, cs.AI
arXiv: https://arxiv.org/abs/2510.14992v1

Training robust world models requires large-scale, precisely labeled multimodal datasets, a process historically bottlenecked by slow and expensive manual annotation. We present a production-tested GAZE pipeline that automates the conversion of raw, long-form video into rich, task-ready supervision for world-model training. Our system (i) normalizes proprietary 360-degree formats into standard views and shards them for parallel processing; (ii) applies a suite of AI models (scene understanding, object tracking, audio transcription, PII/NSFW/minor detection) for dense, multimodal pre-annotation; and (iii) consolidates signals into a structured output specification for rapid human validation. The GAZE workflow demonstrably yields efficiency gains (~19 minutes saved per review hour) and reduces human review volume by >80% through conservative auto-skipping of low-salience segments. By increasing label density and consistency while integrating privacy safeguards and chain-of-custody metadata, our method generates high-fidelity, privacy-aware datasets directly consumable for learning cross-modal dynamics and action-conditioned prediction. We detail our orchestration, model choices, and data dictionary to provide a scalable blueprint for generating high-quality world model training data without sacrificing throughput or governance.

---

### The Role of Federated Learning in Improving Financial Security: A Survey

Authors: Cade Houston Kennedy, Amr Hilal, Morteza Momeni
Published: 2025-10-07
Categories: cs.CR, cs.AI
arXiv: https://arxiv.org/abs/2510.14991v1

With the growth of digital financial systems, robust security and privacy have become a concern for financial institutions. Even though traditional machine learning models have shown to be effective in fraud detections, they often compromise user data by requiring centralized access to sensitive information. In IoT-enabled financial endpoints such as ATMs and POS Systems that regularly produce sensitive data that is sent over the network. Federated Learning (FL) offers a privacy-preserving, decentralized model training across institutions without sharing raw data. FL enables cross-silo collaboration among banks while also using cross-device learning on IoT endpoints. This survey explores the role of FL in enhancing financial security and introduces a novel classification of its applications based on regulatory and compliance exposure levels ranging from low-exposure tasks such as collaborative portfolio optimization to high-exposure tasks like real-time fraud detection. Unlike prior surveys, this work reviews FL's practical use within financial systems, discussing its regulatory compliance and recent successes in fraud prevention and blockchain-integrated frameworks. However, FL deployment in finance is not without challenges. Data heterogeneity, adversarial attacks, and regulatory compliance make implementation far from easy. This survey reviews current defense mechanisms and discusses future directions, including blockchain integration, differential privacy, secure multi-party computation, and quantum-secure frameworks. Ultimately, this work aims to be a resource for researchers exploring FL's potential to advance secure, privacy-compliant financial systems.

---

### Applying Graph Analysis for Unsupervised Fast Malware Fingerprinting

Authors: ElMouatez Billah Karbab, Mourad Debbabi
Published: 2025-10-07
Categories: cs.CR, cs.LG
arXiv: https://arxiv.org/abs/2510.12811v2

Malware proliferation is increasing at a tremendous rate, with hundreds of thousands of new samples identified daily. Manual investigation of such a vast amount of malware is an unrealistic, time-consuming, and overwhelming task. To cope with this volume, there is a clear need to develop specialized techniques and efficient tools for preliminary filtering that can group malware based on semantic similarity. In this paper, we propose TrapNet, a novel, scalable, and unsupervised framework for malware fingerprinting and grouping. TrapNet employs graph community detection techniques for malware fingerprinting and family attribution based on static analysis, as follows: (1) TrapNet detects packed binaries and unpacks them using known generic packer tools. (2) From each malware sample, it generates a digest that captures the underlying semantics. Since the digest must be dense, efficient, and suitable for similarity checking, we designed FloatHash (FH), a novel numerical fuzzy hashing technique that produces a short real-valued vector summarizing the underlying assembly items and their order. FH is based on applying Principal Component Analysis (PCA) to ordered assembly items (e.g., opcodes, function calls) extracted from the malware's assembly code. (3) Representing malware with short numerical vectors enables high-performance, large-scale similarity computation, which allows TrapNet to build a malware similarity network. (4) Finally, TrapNet employs state-of-the-art community detection algorithms to identify dense communities, which represent groups of malware with similar semantics. Our extensive evaluation of TrapNet demonstrates its effectiveness in terms of the coverage and purity of the detected communities, while also highlighting its runtime efficiency, which outperforms other state-of-the-art solutions.

---

### Semantic-Cohesive Knowledge Distillation for Deep Cross-modal Hashing

Authors: Changchang Sun, Vickie Chen, Yan Yan
Published: 2025-10-07
Categories: cs.LG, cs.CV, cs.IR
arXiv: https://arxiv.org/abs/2510.09664v1

Recently, deep supervised cross-modal hashing methods have achieve compelling success by learning semantic information in a self-supervised way. However, they still suffer from the key limitation that the multi-label semantic extraction process fail to explicitly interact with raw multimodal data, making the learned representation-level semantic information not compatible with the heterogeneous multimodal data and hindering the performance of bridging modality gap. To address this limitation, in this paper, we propose a novel semantic cohesive knowledge distillation scheme for deep cross-modal hashing, dubbed as SODA. Specifically, the multi-label information is introduced as a new textual modality and reformulated as a set of ground-truth label prompt, depicting the semantics presented in the image like the text modality. Then, a cross-modal teacher network is devised to effectively distill cross-modal semantic characteristics between image and label modalities and thus learn a well-mapped Hamming space for image modality. In a sense, such Hamming space can be regarded as a kind of prior knowledge to guide the learning of cross-modal student network and comprehensively preserve the semantic similarities between image and text modality. Extensive experiments on two benchmark datasets demonstrate the superiority of our model over the state-of-the-art methods.

---

### Adversarial-Resilient RF Fingerprinting: A CNN-GAN Framework for Rogue Transmitter Detection

Authors: Raju Dhakal, Prashant Shekhar, Laxima Niure Kandel
Published: 2025-10-07
Categories: cs.CR, cs.AI
arXiv: https://arxiv.org/abs/2510.09663v1

Radio Frequency Fingerprinting (RFF) has evolved as an effective solution for authenticating devices by leveraging the unique imperfections in hardware components involved in the signal generation process. In this work, we propose a Convolutional Neural Network (CNN) based framework for detecting rogue devices and identifying genuine ones using softmax probability thresholding. We emulate an attack scenario in which adversaries attempt to mimic the RF characteristics of genuine devices by training a Generative Adversarial Network (GAN) using In-phase and Quadrature (IQ) samples from genuine devices. The proposed approach is verified using IQ samples collected from ten different ADALM-PLUTO Software Defined Radios (SDRs), with seven devices considered genuine, two as rogue, and one used for validation to determine the threshold.

---

### Assessment of different loss functions for fitting equivalent circuit models to electrochemical impedance spectroscopy data

Authors: Ali Jaberi, Amin Sadeghi, Runze Zhang, Zhaoyang Zhao, Qiuyu Shi, Robert Black, Zoya Sadighi, Jason Hattrick-Simpers
Published: 2025-10-07
Categories: cs.LG, cond-mat.mtrl-sci
arXiv: https://arxiv.org/abs/2510.09662v1

Electrochemical impedance spectroscopy (EIS) data is typically modeled using an equivalent circuit model (ECM), with parameters obtained by minimizing a loss function via nonlinear least squares fitting. This paper introduces two new loss functions, log-B and log-BW, derived from the Bode representation of EIS. Using a large dataset of generated EIS data, the performance of proposed loss functions was evaluated alongside existing ones in terms of R2 scores, chi-squared, computational efficiency, and the mean absolute percentage error (MAPE) between the predicted component values and the original values. Statistical comparisons revealed that the choice of loss function impacts convergence, computational efficiency, quality of fit, and MAPE. Our analysis showed that X2 loss function (squared sum of residuals with proportional weighting) achieved the highest performance across multiple quality of fit metrics, making it the preferred choice when the quality of fit is the primary goal. On the other hand, log-B offered a slightly lower quality of fit while being approximately 1.4 times faster and producing lower MAPE for most circuit components, making log-B as a strong alternative. This is a critical factor for large-scale least squares fitting in data-driven applications, such as training machine learning models on extensive datasets or iterations.

---

### Learning What Matters: Steering Diffusion via Spectrally Anisotropic Forward Noise

Authors: Luca Scimeca, Thomas Jiralerspong, Berton Earnshaw, Jason Hartford, Yoshua Bengio
Published: 2025-10-07
Categories: cs.LG, cs.AI, cs.CV
arXiv: https://arxiv.org/abs/2510.09660v4

Diffusion Probabilistic Models (DPMs) have achieved strong generative performance, yet their inductive biases remain largely implicit. In this work, we aim to build inductive biases into the training and sampling of diffusion models to better accommodate the target distribution of the data to model. We introduce an anisotropic noise operator that shapes these biases by replacing the isotropic forward covariance with a structured, frequency-diagonal covariance. This operator unifies band-pass masks and power-law weightings, allowing us to emphasize or suppress designated frequency bands, while keeping the forward process Gaussian. We refer to this as Spectrally Anisotropic Gaussian Diffusion (SAGD). In this work, we derive the score relation for anisotropic forward covariances and show that, under full support, the learned score converges to the true data score as $t\!\to\!0$, while anisotropy reshapes the probability-flow path from noise to data. Empirically, we show the induced anisotropy outperforms standard diffusion across several vision datasets, and enables selective omission: learning while ignoring known corruptions confined to specific bands. Together, these results demonstrate that carefully designed anisotropic forward noise provides a simple, yet principled, handle to tailor inductive bias in DPMs.

---

### Heterogeneous Point Set Transformers for Segmentation of Multiple View Particle Detectors

Authors: Edgar E. Robles, Dikshant Sagar, Alejandro Yankelevich, Jianming Bian, Pierre Baldi, NOvA Collaboration
Published: 2025-10-07
Categories: cs.LG, hep-ex
arXiv: https://arxiv.org/abs/2510.09659v2

NOvA is a long-baseline neutrino oscillation experiment that detects neutrino particles from the NuMI beam at Fermilab. Before data from this experiment can be used in analyses, raw hits in the detector must be matched to their source particles, and the type of each particle must be identified. This task has commonly been done using a mix of traditional clustering approaches and convolutional neural networks (CNNs). Due to the construction of the detector, the data is presented as two sparse 2D images: an XZ and a YZ view of the detector, rather than a 3D representation. We propose a point set neural network that operates on the sparse matrices with an operation that mixes information from both views. Our model uses less than 10% of the memory required using previous methods while achieving a 96.8% AUC score, a higher score than obtained when both views are processed independently (85.4%).

---

### Gradient-Sign Masking for Task Vector Transport Across Pre-Trained Models

Authors: Filippo Rinaldi, Aniello Panariello, Giacomo Salici, Fengyuan Liu, Marco Ciccone, Angelo Porrello, Simone Calderara
Published: 2025-10-07
Categories: cs.LG, cs.AI, cs.CV
arXiv: https://arxiv.org/abs/2510.09658v3

When a new release of a foundation model is published, practitioners typically need to repeat fine-tuning, even if the same task was already tackled in the previous version. A promising alternative is to reuse the parameter changes (i.e., task vectors) that capture how a model adapts to a specific task. However, these vectors often fail to transfer across different pre-trained models because their parameter spaces are misaligned. In this work, we show that successful transfer depends strongly on the gradient-sign structure of the new model. Based on this insight, we propose GradFix, which approximates the ideal sign structure and leverages it to transfer knowledge using only a handful of labeled samples. Notably, this requires no additional fine-tuning: we only compute a few target-model gradients without parameter updates and mask the source task vector accordingly. This yields an update that is locally aligned with the target loss landscape, effectively rebasing the task vector onto the new pre-training. We provide a theoretical guarantee that our method ensures first-order descent. Empirically, we demonstrate significant performance gains on vision and language benchmarks, consistently outperforming naive task vector addition and few-shot fine-tuning. We further show that transporting task vectors improves multi-task and multi-source model merging. Code is available at https://github.com/fillo-rinaldi/GradFix.

---

### Generative Models for Helmholtz Equation Solutions: A Dataset of Acoustic Materials

Authors: Riccardo Fosco Gramaccioni, Christian Marinoni, Fabrizio Frezza, Aurelio Uncini, Danilo Comminiello
Published: 2025-10-07
Categories: cs.LG, cs.AI, eess.SP, math.NA
arXiv: https://arxiv.org/abs/2510.09657v1

Accurate simulation of wave propagation in complex acoustic materials is crucial for applications in sound design, noise control, and material engineering. Traditional numerical solvers, such as finite element methods, are computationally expensive, especially when dealing with large-scale or real-time scenarios. In this work, we introduce a dataset of 31,000 acoustic materials, named HA30K, designed and simulated solving the Helmholtz equations. For each material, we provide the geometric configuration and the corresponding pressure field solution, enabling data-driven approaches to learn Helmholtz equation solutions. As a baseline, we explore a deep learning approach based on Stable Diffusion with ControlNet, a state-of-the-art model for image generation. Unlike classical solvers, our approach leverages GPU parallelization to process multiple simulations simultaneously, drastically reducing computation time. By representing solutions as images, we bypass the need for complex simulation software and explicit equation-solving. Additionally, the number of diffusion steps can be adjusted at inference time, balancing speed and quality. We aim to demonstrate that deep learning-based methods are particularly useful in early-stage research, where rapid exploration is more critical than absolute accuracy.

---

### Data Provenance Auditing of Fine-Tuned Large Language Models with a Text-Preserving Technique

Authors: Yanming Li, Cédric Eichler, Nicolas Anciaux, Alexandra Bensamoun, Lorena Gonzalez Manzano, Seifeddine Ghozzi
Published: 2025-10-07
Categories: cs.CR, cs.AI
arXiv: https://arxiv.org/abs/2510.09655v3

We propose a system for marking sensitive or copyrighted texts to detect their use in fine-tuning large language models under black-box access with statistical guarantees. Our method builds digital ``marks'' using invisible Unicode characters organized into (``cue'', ``reply'') pairs. During an audit, prompts containing only ``cue'' fragments are issued to trigger regurgitation of the corresponding ``reply'', indicating document usage. To control false positives, we compare against held-out counterfactual marks and apply a ranking test, yielding a verifiable bound on the false positive rate. Empirically, we obtain a true positive rate of 96.7% at 0% false positive rate and reply regurgitation rates exceeding 28% per document with only 40 (4%) watermarked documents. The approach is minimally invasive, scalable across many sources, robust to standard processing pipelines, and achieves high detection power even when marked data is a small fraction of the fine-tuning corpus.

---

### TreeNet: Layered Decision Ensembles

Authors: Zeshan Khan
Published: 2025-10-07
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.09654v1

Within the domain of medical image analysis, three distinct methodologies have demonstrated commendable accuracy: Neural Networks, Decision Trees, and Ensemble-Based Learning Algorithms, particularly in the specialized context of genstro institutional track abnormalities detection. These approaches exhibit efficacy in disease detection scenarios where a substantial volume of data is available. However, the prevalent challenge in medical image analysis pertains to limited data availability and data confidence. This paper introduces TreeNet, a novel layered decision ensemble learning methodology tailored for medical image analysis. Constructed by integrating pivotal features from neural networks, ensemble learning, and tree-based decision models, TreeNet emerges as a potent and adaptable model capable of delivering superior performance across diverse and intricate machine learning tasks. Furthermore, its interpretability and insightful decision-making process enhance its applicability in complex medical scenarios. Evaluation of the proposed approach encompasses key metrics including Accuracy, Precision, Recall, and training and evaluation time. The methodology resulted in an F1-score of up to 0.85 when using the complete training data, with an F1-score of 0.77 when utilizing 50\% of the training data. This shows a reduction of F1-score of 0.08 while in the reduction of 50\% of the training data and training time. The evaluation of the methodology resulted in the 32 Frame per Second which is usable for the realtime applications. This comprehensive assessment underscores the efficiency and usability of TreeNet in the demanding landscape of medical image analysis specially in the realtime analysis.

---

### GraphGhost: Tracing Structures Behind Large Language Models

Authors: Xinnan Dai, Xianxuan Long, Chung-Hsiang Lo, Kai Guo, Shenglai Zeng, Dongsheng Luo, Jiliang Tang
Published: 2025-10-07
Categories: cs.CL
arXiv: https://arxiv.org/abs/2510.08613v2

Large Language Models (LLMs) exhibit strong reasoning capabilities on structured tasks, yet the internal mechanisms underlying such behaviors remain poorly understood. Existing interpretation methods mainly focus on token-level attributions, which provide limited insight into multi-step reasoning inside the model. We propose GraphGhost, a graph-based framework that models internal token interactions and neuron activations in LLMs as graphs. By aggregating token dependencies traced across layers, GraphGhost captures global information flow underlying model predictions. We formalize GraphGhost from two complementary perspectives: a sample view, which traces token dependencies for individual predictions, and a dataset view, which aggregates recurring structural patterns learned during training. Through graph analytics and quantitative experiments, we show that graph structural properties are closely associated with influential tokens and neuron nodes, and that perturbations to structurally critical nodes lead to measurable changes in reasoning behavior. These results indicate that the structural patterns captured by GraphGhost reflect meaningful internal organization of LLM reasoning. The codes are available at software part. Artifacts will be made available for research use only.

---

### Impact of LLMs on Team Collaboration in Software Development

Authors: Devang Dhanuka
Published: 2025-10-07
Categories: cs.SE, cs.AI
arXiv: https://arxiv.org/abs/2510.08612v1

Large Language Models (LLMs) are increasingly being integrated into software development processes, with the potential to transform team workflows and productivity. This paper investigates how LLMs affect team collaboration throughout the Software Development Life Cycle (SDLC). We reframe and update a prior study with recent developments as of 2025, incorporating new literature and case studies. We outline the problem of collaboration hurdles in SDLC and explore how LLMs can enhance productivity, communication, and decision-making in a team context. Through literature review, industry examples, a team survey, and two case studies, we assess the impact of LLM-assisted tools (such as code generation assistants and AI-powered project management agents) on collaborative software engineering practices. Our findings indicate that LLMs can significantly improve efficiency (by automating repetitive tasks and documentation), enhance communication clarity, and aid cross-functional collaboration, while also introducing new challenges like model limitations and privacy concerns. We discuss these benefits and challenges, present research questions guiding the investigation, evaluate threats to validity, and suggest future research directions including domain-specific model customization, improved integration into development tools, and robust strategies for ensuring trust and security.

---

### Relative Positioning Based Code Chunking Method For Rich Context Retrieval In Repository Level Code Completion Task With Code Language Model

Authors: Imranur Rahman, Md Rayhanur Rahman
Published: 2025-10-07
Categories: cs.SE, cs.AI, cs.LG
arXiv: https://arxiv.org/abs/2510.08610v1

Code completion can help developers improve efficiency and ease the development lifecycle. Although code completion is available in modern integrated development environments (IDEs), research lacks in determining what makes a good context for code completion based on the information available to the IDEs for the large language models (LLMs) to perform better. In this paper, we describe an effective context collection strategy to assist the LLMs in performing better at code completion tasks. The key idea of our strategy is to preprocess the repository into smaller code chunks and later use syntactic and semantic similarity-based code chunk retrieval with relative positioning. We found that code chunking and relative positioning of the chunks in the final context improve the performance of code completion tasks.

---

### Which Is Better For Reducing Outdated and Vulnerable Dependencies: Pinning or Floating?

Authors: Imranur Rahman, Jill Marley, William Enck, Laurie Williams
Published: 2025-10-07
Categories: cs.SE, cs.CR, cs.LG, cs.PL
arXiv: https://arxiv.org/abs/2510.08609v3

Developers consistently use version constraints to specify acceptable versions of the dependencies for their project. Pinning dependencies can reduce the likelihood of breaking changes, but comes with a cost of manually managing the replacement of outdated and vulnerable dependencies. On the other hand, floating can be used to automatically get bug fixes and security fixes, but comes with the risk of breaking changes. Security practitioners advocate pinning dependencies to prevent against software supply chain attacks, e.g., malicious package updates. However, since pinning is the tightest version constraint, pinning is the most likely to result in outdated dependencies. Nevertheless, how the likelihood of becoming outdated or vulnerable dependencies changes across version constraint types is unknown. The goal of this study is to aid developers in making an informed dependency version constraint choice by empirically evaluating the likelihood of dependencies becoming outdated or vulnerable across version constraint types at scale. In this study, we first identify the trends in dependency version constraint usage and the patterns of version constraint type changes made by developers in the npm, PyPI, and Cargo ecosystems. We then modeled the dependency state transitions using survival analysis and estimated how the likelihood of becoming outdated or vulnerable changes when using pinning as opposed to the rest of the version constraint types. We observe that among outdated and vulnerable dependencies, the most commonly used version constraint type is floating-minor, with pinning being the next most common. We also find that floating-major is the least likely to result in outdated and floating-minor is the least likely to result in vulnerable dependencies.

---

### MMA-ASIA: A Multilingual and Multimodal Alignment Framework for Culturally-Grounded Evaluation

Authors: Weihua Zheng, Zhengyuan Liu, Tanmoy Chakraborty, Weiwen Xu, Xiaoxue Gao, Bryan Chen Zhengyu Tan, Bowei Zou, Chang Liu, Yujia Hu, Xing Xie, Xiaoyuan Yi, Jing Yao et al.
Published: 2025-10-07
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2510.08608v1

Large language models (LLMs) are now used worldwide, yet their multimodal understanding and reasoning often degrade outside Western, high-resource settings. We propose MMA-ASIA, a comprehensive framework to evaluate LLMs' cultural awareness with a focus on Asian contexts. MMA-ASIA centers on a human-curated, multilingual, and multimodally aligned multiple-choice benchmark covering 8 Asian countries and 10 languages, comprising 27,000 questions; over 79 percent require multi-step reasoning grounded in cultural context, moving beyond simple memorization. To our knowledge, this is the first dataset aligned at the input level across three modalities: text, image (visual question answering), and speech. This enables direct tests of cross-modal transfer. Building on this benchmark, we propose a five-dimensional evaluation protocol that measures: (i) cultural-awareness disparities across countries, (ii) cross-lingual consistency, (iii) cross-modal consistency, (iv) cultural knowledge generalization, and (v) grounding validity. To ensure rigorous assessment, a Cultural Awareness Grounding Validation Module detects "shortcut learning" by checking whether the requisite cultural knowledge supports correct answers. Finally, through comparative model analysis, attention tracing, and an innovative Vision-ablated Prefix Replay (VPR) method, we probe why models diverge across languages and modalities, offering actionable insights for building culturally reliable multimodal LLMs.

---

### Centering Emotion Hotspots: Multimodal Local-Global Fusion and Cross-Modal Alignment for Emotion Recognition in Conversations

Authors: Yu Liu, Hanlei Shi, Haoxun Li, Yuqing Sun, Yuxuan Ding, Linlin Gong, Leyuan Qu, Taihao Li
Published: 2025-10-07
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2510.08606v1

Emotion Recognition in Conversations (ERC) is hard because discriminative evidence is sparse, localized, and often asynchronous across modalities. We center ERC on emotion hotspots and present a unified model that detects per-utterance hotspots in text, audio, and video, fuses them with global features via Hotspot-Gated Fusion, and aligns modalities using a routed Mixture-of-Aligners; a cross-modal graph encodes conversational structure. This design focuses modeling on salient spans, mitigates misalignment, and preserves context. Experiments on standard ERC benchmarks show consistent gains over strong baselines, with ablations confirming the contributions of HGF and MoA. Our results point to a hotspot-centric view that can inform future multimodal learning, offering a new perspective on modality fusion in ERC.

---

### Toward a Safer Web: Multilingual Multi-Agent LLMs for Mitigating Adversarial Misinformation Attacks

Authors: Nouar Aldahoul, Yasir Zaki
Published: 2025-10-07
Categories: cs.CL, cs.AI, cs.CR, cs.LG
arXiv: https://arxiv.org/abs/2510.08605v1

The rapid spread of misinformation on digital platforms threatens public discourse, emotional stability, and decision-making. While prior work has explored various adversarial attacks in misinformation detection, the specific transformations examined in this paper have not been systematically studied. In particular, we investigate language-switching across English, French, Spanish, Arabic, Hindi, and Chinese, followed by translation. We also study query length inflation preceding summarization and structural reformatting into multiple-choice questions. In this paper, we present a multilingual, multi-agent large language model framework with retrieval-augmented generation that can be deployed as a web plugin into online platforms. Our work underscores the importance of AI-driven misinformation detection in safeguarding online factual integrity against diverse attacks, while showcasing the feasibility of plugin-based deployment for real-world web applications.

---

### LatentBreak: Jailbreaking Large Language Models through Latent Space Feedback

Authors: Raffaele Mura, Giorgio Piras, Kamilė Lukošiūtė, Maura Pintor, Amin Karbasi, Battista Biggio
Published: 2025-10-07
Categories: cs.CL, cs.AI, cs.LG
arXiv: https://arxiv.org/abs/2510.08604v2

Jailbreaks are adversarial attacks designed to bypass the built-in safety mechanisms of large language models. Automated jailbreaks typically optimize an adversarial suffix or adapt long prompt templates by forcing the model to generate the initial part of a restricted or harmful response. In this work, we show that existing jailbreak attacks that leverage such mechanisms to unlock the model response can be detected by a straightforward perplexity-based filtering on the input prompt. To overcome this issue, we propose LatentBreak, a white-box jailbreak attack that generates natural adversarial prompts with low perplexity capable of evading such defenses. LatentBreak substitutes words in the input prompt with semantically-equivalent ones, preserving the initial intent of the prompt, instead of adding high-perplexity adversarial suffixes or long templates. These words are chosen by minimizing the distance in the latent space between the representation of the adversarial prompt and that of harmless requests. Our extensive evaluation shows that LatentBreak leads to shorter and low-perplexity prompts, thus outperforming competing jailbreak algorithms against perplexity-based filters on multiple safety-aligned models.

---

### YpathRAG:A Retrieval-Augmented Generation Framework and Benchmark for Pathology

Authors: Deshui Yu, Yizhi Wang, Saihui Jin, Taojie Zhu, Fanyi Zeng, Wen Qian, Zirui Huang, Jingli Ouyang, Jiameng Li, Zhen Song, Tian Guan, Yonghong He
Published: 2025-10-07
Categories: cs.CL
arXiv: https://arxiv.org/abs/2510.08603v1

Large language models (LLMs) excel on general tasks yet still hallucinate in high-barrier domains such as pathology. Prior work often relies on domain fine-tuning, which neither expands the knowledge boundary nor enforces evidence-grounded constraints. We therefore build a pathology vector database covering 28 subfields and 1.53 million paragraphs, and present YpathRAG, a pathology-oriented RAG framework with dual-channel hybrid retrieval (BGE-M3 dense retrieval coupled with vocabulary-guided sparse retrieval) and an LLM-based supportive-evidence judgment module that closes the retrieval-judgment-generation loop. We also release two evaluation benchmarks, YpathR and YpathQA-M. On YpathR, YpathRAG attains Recall@5 of 98.64%, a gain of 23 percentage points over the baseline; on YpathQA-M, a set of the 300 most challenging questions, it increases the accuracies of both general and medical LLMs by 9.0% on average and up to 15.6%. These results demonstrate improved retrieval quality and factual reliability, providing a scalable construction paradigm and interpretable evaluation for pathology-oriented RAG.

---

### Human Texts Are Outliers: Detecting LLM-generated Texts via Out-of-distribution Detection

Authors: Cong Zeng, Shengkun Tang, Yuanzhou Chen, Zhiqiang Shen, Wenchao Yu, Xujiang Zhao, Haifeng Chen, Wei Cheng, Zhiqiang Xu
Published: 2025-10-07
Categories: cs.CL, cs.LG
arXiv: https://arxiv.org/abs/2510.08602v1

The rapid advancement of large language models (LLMs) such as ChatGPT, DeepSeek, and Claude has significantly increased the presence of AI-generated text in digital communication. This trend has heightened the need for reliable detection methods to distinguish between human-authored and machine-generated content. Existing approaches both zero-shot methods and supervised classifiers largely conceptualize this task as a binary classification problem, often leading to poor generalization across domains and models. In this paper, we argue that such a binary formulation fundamentally mischaracterizes the detection task by assuming a coherent representation of human-written texts. In reality, human texts do not constitute a unified distribution, and their diversity cannot be effectively captured through limited sampling. This causes previous classifiers to memorize observed OOD characteristics rather than learn the essence of `non-ID' behavior, limiting generalization to unseen human-authored inputs. Based on this observation, we propose reframing the detection task as an out-of-distribution (OOD) detection problem, treating human-written texts as distributional outliers while machine-generated texts are in-distribution (ID) samples. To this end, we develop a detection framework using one-class learning method including DeepSVDD and HRN, and score-based learning techniques such as energy-based method, enabling robust and generalizable performance. Extensive experiments across multiple datasets validate the effectiveness of our OOD-based approach. Specifically, the OOD-based method achieves 98.3% AUROC and AUPR with only 8.9% FPR95 on DeepFake dataset. Moreover, we test our detection framework on multilingual, attacked, and unseen-model and -domain text settings, demonstrating the robustness and generalizability of our framework. Code, pretrained weights, and demo will be released.

---

### Mnemosyne: An Unsupervised, Human-Inspired Long-Term Memory Architecture for Edge-Based LLMs

Authors: Aneesh Jonelagadda, Christina Hahn, Haoze Zheng, Salvatore Penachio
Published: 2025-10-07
Categories: cs.CL, cs.AI, cs.LG, cs.MA
arXiv: https://arxiv.org/abs/2510.08601v1

Long-term memory is essential for natural, realistic dialogue. However, current large language model (LLM) memory systems rely on either brute-force context expansion or static retrieval pipelines that fail on edge-constrained devices. We introduce Mnemosyne, an unsupervised, human-inspired long-term memory architecture designed for edge-based LLMs. Our approach uses graph-structured storage, modular substance and redundancy filters, memory committing and pruning mechanisms, and probabilistic recall with temporal decay and refresh processes modeled after human memory. Mnemosyne also introduces a concentrated "core summary" efficiently derived from a fixed-length subset of the memory graph to capture the user's personality and other domain-specific long-term details such as, using healthcare application as an example, post-recovery ambitions and attitude towards care. Unlike existing retrieval-augmented methods, Mnemosyne is designed for use in longitudinal healthcare assistants, where repetitive and semantically similar but temporally distinct conversations are limited by naive retrieval. In experiments with longitudinal healthcare dialogues, Mnemosyne demonstrates the highest win rate of 65.8% in blind human evaluations of realism and long-term memory capability compared to a baseline RAG win rate of 31.1%. Mnemosyne also achieves current highest LoCoMo benchmark scores in temporal reasoning and single-hop retrieval compared to other same-backboned techniques. Further, the average overall score of 54.6% was second highest across all methods, beating commonly used Mem0 and OpenAI baselines among others. This demonstrates that improved factual recall, enhanced temporal reasoning, and much more natural user-facing responses can be feasible with an edge-compatible and easily transferable unsupervised memory architecture.

---

### Enhancing Maritime Object Detection in Real-Time with RT-DETR and Data Augmentation

Authors: Nader Nemati
Published: 2025-10-07
Categories: cs.CV, cs.LG
arXiv: https://arxiv.org/abs/2510.07346v1

Maritime object detection faces essential challenges due to the small target size and limitations of labeled real RGB data. This paper will present a real-time object detection system based on RT-DETR, enhanced by employing augmented synthetic images while strictly evaluating on real data. This study employs RT-DETR for the maritime environment by combining multi-scale feature fusion, uncertainty-minimizing query selection, and smart weight between synthetic and real training samples. The fusion module in DETR enhances the detection of small, low-contrast vessels, query selection focuses on the most reliable proposals, and the weighting strategy helps reduce the visual gap between synthetic and real domains. This design preserves DETR's refined end-to-end set prediction while allowing users to adjust between speed and accuracy at inference time. Data augmentation techniques were also used to balance the different classes of the dataset to improve the robustness and accuracy of the model. Regarding this study, a full Python robust maritime detection pipeline is delivered that maintains real-time performance even under practical limits. It also verifies how each module contributes, and how the system handles failures in extreme lighting or sea conditions. This study also includes a component analysis to quantify the contribution of each architectural module and explore its interactions.

---

### Mitigating Surgical Data Imbalance with Dual-Prediction Video Diffusion Model

Authors: Danush Kumar Venkatesh, Adam Schmidt, Muhammad Abdullah Jamal, Omid Mohareri
Published: 2025-10-07
Categories: q-bio.QM, cs.AI, eess.IV
arXiv: https://arxiv.org/abs/2510.07345v1

Surgical video datasets are essential for scene understanding, enabling procedural modeling and intra-operative support. However, these datasets are often heavily imbalanced, with rare actions and tools under-represented, which limits the robustness of downstream models. We address this challenge with $SurgiFlowVid$, a sparse and controllable video diffusion framework for generating surgical videos of under-represented classes. Our approach introduces a dual-prediction diffusion module that jointly denoises RGB frames and optical flow, providing temporal inductive biases to improve motion modeling from limited samples. In addition, a sparse visual encoder conditions the generation process on lightweight signals (e.g., sparse segmentation masks or RGB frames), enabling controllability without dense annotations. We validate our approach on three surgical datasets across tasks including action recognition, tool presence detection, and laparoscope motion prediction. Synthetic data generated by our method yields consistent gains of 10-20% over competitive baselines, establishing $SurgiFlowVid$ as a promising strategy to mitigate data imbalance and advance surgical video understanding methods.

---

### Local MAP Sampling for Diffusion Models

Authors: Shaorong Zhang, Rob Brekelmans, Greg Ver Steeg
Published: 2025-10-07
Categories: cs.GR, cs.AI, eess.IV
arXiv: https://arxiv.org/abs/2510.07343v3

Diffusion Posterior Sampling (DPS) provides a principled Bayesian approach to inverse problems by sampling from $p(x_0 \mid y)$. While posterior sampling is valuable for capturing uncertainty and multi-modality, many classical and practical inverse problem settings ultimately prioritize accurate point estimation -- most notably the MAP estimator, which has long served as a standard reconstruction objective in imaging and scientific applications. We introduce Local MAP Sampling (LMAPS), a new inference framework that iteratively solves local MAP subproblems along the diffusion trajectory. This perspective clarifies their connection to global MAP and DPS, offering a unified probabilistic interpretation for optimization-based methods. Building on this foundation, we develop practical algorithms with a covariance approximation motivated by a Gaussian prior assumption, and a reformulated objective for stability and interpretability. Across a broad set of image restoration and scientific tasks, LMAPS achieves state-of-the-art performance.

---

### Beyond Grid-Locked Voxels: Neural Response Functions for Continuous Brain Encoding

Authors: Haomiao Chen, Keith W Jamison, Mert R. Sabuncu, Amy Kuceyeski
Published: 2025-10-07
Categories: q-bio.NC, cs.LG, eess.IV
arXiv: https://arxiv.org/abs/2510.07342v2

Neural encoding models aim to predict fMRI-measured brain responses to natural images. fMRI data is acquired as a 3D volume of voxels, where each voxel has a defined spatial location in the brain. However, conventional encoding models often flatten this volume into a 1D vector and treat voxel responses as independent outputs. This removes spatial context, discards anatomical information, and ties each model to a subject-specific voxel grid. We introduce the Neural Response Function (NRF), a framework that models fMRI activity as a continuous function over anatomical space rather than a flat vector of voxels. NRF represents brain activity as a continuous implicit function: given an image and a spatial coordinate (x, y, z) in standardized MNI space, the model predicts the response at that location. This formulation decouples predictions from the training grid, supports querying at arbitrary spatial resolutions, and enables resolution-agnostic analyses. By grounding the model in anatomical space, NRF exploits two key properties of brain responses: (1) local smoothness -- neighboring voxels exhibit similar response patterns; modeling responses continuously captures these correlations and improves data efficiency, and (2) cross-subject alignment -- MNI coordinates unify data across individuals, allowing a model pretrained on one subject to be fine-tuned on new subjects. In experiments, NRF outperformed baseline models in both intrasubject encoding and cross-subject adaptation, achieving high performance while reducing the data size needed by orders of magnitude. To our knowledge, NRF is the first anatomically aware encoding model to move beyond flattened voxels, learning a continuous mapping from images to brain responses in 3D space.

---

### Learning Neuron Dynamics within Deep Spiking Neural Networks

Authors: Eric Jahns, Davi Moreno, Michel A. Kinsy
Published: 2025-10-07
Categories: cs.NE
arXiv: https://arxiv.org/abs/2510.07341v1

Spiking Neural Networks (SNNs) offer a promising energy-efficient alternative to Artificial Neural Networks (ANNs) by utilizing sparse and asynchronous processing through discrete spike-based computation. However, the performance of deep SNNs remains limited by their reliance on simple neuron models, such as the Leaky Integrate-and-Fire (LIF) model, which cannot capture rich temporal dynamics. While more expressive neuron models exist, they require careful manual tuning of hyperparameters and are difficult to scale effectively. This difficulty is evident in the lack of successful implementations of complex neuron models in high-performance deep SNNs. In this work, we address this limitation by introducing Learnable Neuron Models (LNMs). LNMs are a general, parametric formulation for non-linear integrate-and-fire dynamics that learn neuron dynamics during training. By learning neuron dynamics directly from data, LNMs enhance the performance of deep SNNs. We instantiate LNMs using low-degree polynomial parameterizations, enabling efficient and stable training. We demonstrate state-of-the-art performance in a variety of datasets, including CIFAR-10, CIFAR-100, ImageNet, and CIFAR-10 DVS. LNMs offer a promising path toward more scalable and high-performing spiking architectures.

---

### SpotDiff: Spotting and Disentangling Interference in Feature Space for Subject-Preserving Image Generation

Authors: Yongzhi Li, Saining Zhang, Yibing Chen, Boying Li, Yanxin Zhang, Xiaoyu Du
Published: 2025-10-07
Categories: cs.GR, cs.LG
arXiv: https://arxiv.org/abs/2510.07340v1

Personalized image generation aims to faithfully preserve a reference subject's identity while adapting to diverse text prompts. Existing optimization-based methods ensure high fidelity but are computationally expensive, while learning-based approaches offer efficiency at the cost of entangled representations influenced by nuisance factors. We introduce SpotDiff, a novel learning-based method that extracts subject-specific features by spotting and disentangling interference. Leveraging a pre-trained CLIP image encoder and specialized expert networks for pose and background, SpotDiff isolates subject identity through orthogonality constraints in the feature space. To enable principled training, we introduce SpotDiff10k, a curated dataset with consistent pose and background variations. Experiments demonstrate that SpotDiff achieves more robust subject preservation and controllable editing than prior methods, while attaining competitive performance with only 10k training samples.

---

### Decoding the dark proteome: Deep learning-enabled discovery of druggable enzymes in Wuchereria bancrofti

Authors: Shawnak Shivakumar, Jefferson Hernandez
Published: 2025-10-07
Categories: q-bio.QM, cs.LG, q-bio.MN
arXiv: https://arxiv.org/abs/2510.07337v1

Wuchereria bancrofti, the parasitic roundworm responsible for lymphatic filariasis, permanently disables over 36 million people and places 657 million at risk across 39 countries. A major bottleneck for drug discovery is the lack of functional annotation for more than 90 percent of the W. bancrofti dark proteome, leaving many potential targets unidentified. In this work, we present a novel computational pipeline that converts W. bancrofti's unannotated amino acid sequence data into precise four-level Enzyme Commission (EC) numbers and drug candidates. We utilized a DEtection TRansformer to estimate the probability of enzymatic function, fine-tuned a hierarchical nearest neighbor EC predictor on 4,476 labeled parasite proteins, and applied rejection sampling to retain only four-level EC classifications at 100 percent confidence. This pipeline assigned precise EC numbers to 14,772 previously uncharacterized proteins and discovered 543 EC classes not previously known in W. bancrofti. A qualitative triage emphasizing parasite-specific targets, chemical tractability, biochemical importance, and biological plausibility prioritized six enzymes across five separate strategies: anti-Wolbachia cell-wall inhibition, proteolysis blockade, transmission disruption, purinergic immune interference, and cGMP-signaling destabilization. We curated a 43-compound library from ChEMBL and BindingDB and co-folded across multiple protein conformers with Boltz-2. All six targets exhibited at least moderately strong predicted binding affinities below 1 micromolar, with moenomycin analogs against peptidoglycan glycosyltransferase and NTPase inhibitors showing promising nanomolar hits and well-defined binding pockets. While experimental validation remains essential, our results provide the first large-scale functional map of the W. bancrofti dark proteome and accelerate early-stage drug development for the species.

---

### Text-to-Image Models Leave Identifiable Signatures: Implications for Leaderboard Security

Authors: Ali Naseh, Anshuman Suri, Yuefeng Peng, Harsh Chaudhari, Alina Oprea, Amir Houmansadr
Published: 2025-10-07
Categories: cs.LG, cs.CR
arXiv: https://arxiv.org/abs/2510.06525v1

Generative AI leaderboards are central to evaluating model capabilities, but remain vulnerable to manipulation. Among key adversarial objectives is rank manipulation, where an attacker must first deanonymize the models behind displayed outputs -- a threat previously demonstrated and explored for large language models (LLMs). We show that this problem can be even more severe for text-to-image leaderboards, where deanonymization is markedly easier. Using over 150,000 generated images from 280 prompts and 19 diverse models spanning multiple organizations, architectures, and sizes, we demonstrate that simple real-time classification in CLIP embedding space identifies the generating model with high accuracy, even without prompt control or historical data. We further introduce a prompt-level separability metric and identify prompts that enable near-perfect deanonymization. Our results indicate that rank manipulation in text-to-image leaderboards is easier than previously recognized, underscoring the need for stronger defenses.

---

### Real-Time Glass Detection and Reprojection using Sensor Fusion Onboard Aerial Robots

Authors: Malakhi Hopkins, Varun Murali, Vijay Kumar, Camillo J Taylor
Published: 2025-10-07
Categories: cs.RO, cs.CV, eess.SY
arXiv: https://arxiv.org/abs/2510.06518v1

Autonomous aerial robots are increasingly being deployed in real-world scenarios, where transparent obstacles present significant challenges to reliable navigation and mapping. These materials pose a unique problem for traditional perception systems because they lack discernible features and can cause conventional depth sensors to fail, leading to inaccurate maps and potential collisions. To ensure safe navigation, robots must be able to accurately detect and map these transparent obstacles. Existing methods often rely on large, expensive sensors or algorithms that impose high computational burdens, making them unsuitable for low Size, Weight, and Power (SWaP) robots. In this work, we propose a novel and computationally efficient framework for detecting and mapping transparent obstacles onboard a sub-300g quadrotor. Our method fuses data from a Time-of-Flight (ToF) camera and an ultrasonic sensor with a custom, lightweight 2D convolution model. This specialized approach accurately detects specular reflections and propagates their depth into corresponding empty regions of the depth map, effectively rendering transparent obstacles visible. The entire pipeline operates in real-time, utilizing only a small fraction of a CPU core on an embedded processor. We validate our system through a series of experiments in both controlled and real-world environments, demonstrating the utility of our method through experiments where the robot maps indoor environments containing glass. Our work is, to our knowledge, the first of its kind to demonstrate a real-time, onboard transparent obstacle mapping system on a low-SWaP quadrotor using only the CPU.

---

### Visualizing Multimodality in Combinatorial Search Landscapes

Authors: Xavier F. C. Sánchez-Díaz, Ole Jakob Mengshoel
Published: 2025-10-07
Categories: cs.GR, cs.AI, cs.NE
arXiv: https://arxiv.org/abs/2510.06517v1

This work walks through different visualization techniques for combinatorial search landscapes, focusing on multimodality. We discuss different techniques from the landscape analysis literature, and how they can be combined to provide a more comprehensive view of the search landscape. We also include examples and discuss relevant work to show how others have used these techniques in practice, based on the geometric and aesthetic elements of the Grammar of Graphics. We conclude that there is no free lunch in visualization, and provide recommendations for future work as there are several paths to continue the work in this field.

---

### Limited-Angle Tomography Reconstruction via Projector Guided 3D Diffusion

Authors: Zhantao Deng, Mériem Er-Rafik, Anna Sushko, Cécile Hébert, Pascal Fua
Published: 2025-10-07
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.06516v2

Limited-angle electron tomography aims to reconstruct 3D shapes from 2D projections of Transmission Electron Microscopy (TEM) within a restricted range and number of tilting angles, but it suffers from the missing-wedge problem that causes severe reconstruction artifacts. Deep learning approaches have shown promising results in alleviating these artifacts, yet they typically require large high-quality training datasets with known 3D ground truth which are difficult to obtain in electron microscopy. To address these challenges, we propose TEMDiff, a novel 3D diffusion-based iterative reconstruction framework. Our method is trained on readily available volumetric FIB-SEM data using a simulator that maps them to TEM tilt series, enabling the model to learn realistic structural priors without requiring clean TEM ground truth. By operating directly on 3D volumes, TEMDiff implicitly enforces consistency across slices without the need for additional regularization. On simulated electron tomography datasets with limited angular coverage, TEMDiff outperforms state-of-the-art methods in reconstruction quality. We further demonstrate that a trained TEMDiff model generalizes well to real-world TEM tilts obtained under different conditions and can recover accurate structures from tilt ranges as narrow as 8 degrees, with 2-degree increments, without any retraining or fine-tuning.

---

### Online Matching via Reinforcement Learning: An Expert Policy Orchestration Strategy

Authors: Chiara Mignacco, Matthieu Jonckheere, Gilles Stoltz
Published: 2025-10-07
Categories: stat.ML, cs.LG
arXiv: https://arxiv.org/abs/2510.06515v1

Online matching problems arise in many complex systems, from cloud services and online marketplaces to organ exchange networks, where timely, principled decisions are critical for maintaining high system performance. Traditional heuristics in these settings are simple and interpretable but typically tailored to specific operating regimes, which can lead to inefficiencies when conditions change. We propose a reinforcement learning (RL) approach that learns to orchestrate a set of such expert policies, leveraging their complementary strengths in a data-driven, adaptive manner. Building on the Adv2 framework (Jonckheere et al., 2024), our method combines expert decisions through advantage-based weight updates and extends naturally to settings where only estimated value functions are available. We establish both expectation and high-probability regret guarantees and derive a novel finite-time bias bound for temporal-difference learning, enabling reliable advantage estimation even under constant step size and non-stationary dynamics. To support scalability, we introduce a neural actor-critic architecture that generalizes across large state spaces while preserving interpretability. Simulations on stochastic matching models, including an organ exchange scenario, show that the orchestrated policy converges faster and yields higher system level efficiency than both individual experts and conventional RL baselines. Our results highlight how structured, adaptive learning can improve the modeling and management of complex resource allocation and decision-making processes.

---

### LogSTOP: Temporal Scores over Prediction Sequences for Matching and Retrieval

Authors: Avishree Khare, Hideki Okamoto, Bardh Hoxha, Georgios Fainekos, Rajeev Alur
Published: 2025-10-07
Categories: cs.CV, cs.AI
arXiv: https://arxiv.org/abs/2510.06512v1

Neural models such as YOLO and HuBERT can be used to detect local properties such as objects ("car") and emotions ("angry") in individual frames of videos and audio clips respectively. The likelihood of these detections is indicated by scores in (0, 1). Lifting these scores to temporal properties over sequences can be useful for several downstream applications such as query matching (e.g., "does the speaker eventually sound happy in this audio clip?"), and ranked retrieval (e.g., "retrieve top 5 videos with a 10 second scene where a car is detected until a pedestrian is detected"). In this work, we formalize this problem of assigning Scores for TempOral Properties (STOPs) over sequences, given potentially noisy score predictors for local properties. We then propose a scoring function called LogSTOP that can efficiently compute these scores for temporal properties represented in Linear Temporal Logic. Empirically, LogSTOP, with YOLO and HuBERT, outperforms Large Vision / Audio Language Models and other Temporal Logic-based baselines by at least 16% on query matching with temporal properties over objects-in-videos and emotions-in-speech respectively. Similarly, on ranked retrieval with temporal properties over objects and actions in videos, LogSTOP with Grounding DINO and SlowR50 reports at least a 19% and 16% increase in mean average precision and recall over zero-shot text-to-video retrieval baselines respectively.

---

### From Captions to Keyframes: KeyScore for Multimodal Frame Scoring and Video-Language Understanding

Authors: Shih-Yao Lin, Sibendu Paul, Caren Chen
Published: 2025-10-07
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.06509v2

Selecting informative keyframes is critical for efficient video understanding, yet existing approaches often rely on heuristics, ignore semantics, or produce redundant frames. We propose KeyScore, a caption-aware frame scoring method that combines three complementary signals: semantic similarity to captions, temporal representativeness, and contextual drop impact. Applied to large-scale video-caption datasets, KeyScore generates frame-level importance scores that enable training keyframe extractors or guiding video-language models. To support this, we also propose STACFP, a Spatio-Temporal Adaptive Clustering method that generates diverse and compact frame proposals across long videos. Together, KeyScore and STACFP reduce uninformative frames while preserving critical content, resulting in faster and more accurate inference. Our experiments on three standard video-language benchmarks (MSRVTT, MSVD, DiDeMo) show that combining STACFP and KeyScore enables up to 99% frame reduction compared to full-frame processing, while outperforming uniform 8-frame encoders in video-text retrieval, keyframe extraction, and action recognition tasks. By focusing on semantically relevant frames, our method enhances both efficiency and performance, enabling scalable and caption-grounded video understanding.

---

### Medix: Out-of-Distribution Detection from Unlabeled Wild Data via Robust Gradient Statistics

Authors: Momin Abbas, Ali Falahati, Hossein Goli, Mohammad Mohammadi Amiri
Published: 2025-10-07
Categories: cs.LG, cs.AI, math.OC, stat.ML
arXiv: https://arxiv.org/abs/2510.06505v2

Out-of-distribution (OOD) detection plays a crucial role in ensuring the robustness of machine learning systems deployed in real-world applications. Recent approaches have explored the use of unlabeled data, showing potential for enhancing OOD detection capabilities. However, effectively utilizing unlabeled in-the-wild data remains challenging due to the mixed nature of both in-distribution (InD) and OOD samples. The lack of a distinct set of OOD samples complicates the task of training an optimal OOD classifier. In this work, we introduce Medix, a novel framework designed to identify potential outliers from unlabeled data using the median-based robust gradient statistics. We use the median because it provides a stable estimate of the central tendency, as an OOD detection mechanism, due to its robustness against noise and outliers. Using these identified outliers, along with labeled InD data, we train a robust OOD classifier. From a theoretical perspective, we derive error bounds that demonstrate Medix achieves a low error rate. Empirical results further substantiate our claims, as Medix outperforms existing methods across the board in open-world settings.

---

### Text2Interact: High-Fidelity and Diverse Text-to-Two-Person Interaction Generation

Authors: Qingxuan Wu, Zhiyang Dou, Chuan Guo, Yiming Huang, Qiao Feng, Bing Zhou, Jian Wang, Lingjie Liu
Published: 2025-10-07
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.06504v1

Modeling human-human interactions from text remains challenging because it requires not only realistic individual dynamics but also precise, text-consistent spatiotemporal coupling between agents. Currently, progress is hindered by 1) limited two-person training data, inadequate to capture the diverse intricacies of two-person interactions; and 2) insufficiently fine-grained text-to-interaction modeling, where language conditioning collapses rich, structured prompts into a single sentence embedding. To address these limitations, we propose our Text2Interact framework, designed to generate realistic, text-aligned human-human interactions through a scalable high-fidelity interaction data synthesizer and an effective spatiotemporal coordination pipeline. First, we present InterCompose, a scalable synthesis-by-composition pipeline that aligns LLM-generated interaction descriptions with strong single-person motion priors. Given a prompt and a motion for an agent, InterCompose retrieves candidate single-person motions, trains a conditional reaction generator for another agent, and uses a neural motion evaluator to filter weak or misaligned samples-expanding interaction coverage without extra capture. Second, we propose InterActor, a text-to-interaction model with word-level conditioning that preserves token-level cues (initiation, response, contact ordering) and an adaptive interaction loss that emphasizes contextually relevant inter-person joint pairs, improving coupling and physical plausibility for fine-grained interaction modeling. Extensive experiments show consistent gains in motion diversity, fidelity, and generalization, including out-of-distribution scenarios and user studies. We will release code and models to facilitate reproducibility.

---

### ATLO-ML: Adaptive Time-Length Optimizer for Machine Learning -- Insights from Air Quality Forecasting

Authors: I-Hsi Kao, Kanji Uchino
Published: 2025-10-07
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2510.06503v1

Accurate time-series predictions in machine learning are heavily influenced by the selection of appropriate input time length and sampling rate. This paper introduces ATLO-ML, an adaptive time-length optimization system that automatically determines the optimal input time length and sampling rate based on user-defined output time length. The system provides a flexible approach to time-series data pre-processing, dynamically adjusting these parameters to enhance predictive performance. ATLO-ML is validated using air quality datasets, including both GAMS-dataset and proprietary data collected from a data center, both in time series format. Results demonstrate that utilizing the optimized time length and sampling rate significantly improves the accuracy of machine learning models compared to fixed time lengths. ATLO-ML shows potential for generalization across various time-sensitive applications, offering a robust solution for optimizing temporal input parameters in machine learning workflows.

---

### GUIDE: Guided Initialization and Distillation of Embeddings

Authors: Khoa Trinh, Gaurav Menghani, Erik Vee
Published: 2025-10-07
Categories: cs.LG
arXiv: https://arxiv.org/abs/2510.06502v1

Algorithmic efficiency techniques such as distillation (\cite{hinton2015distillation}) are useful in improving model quality without increasing serving costs, provided a larger teacher model is available for a smaller student model to learn from during training. Standard distillation methods are limited to only forcing the student to match the teacher's outputs. Given the costs associated with training a large model, we believe we should be extracting more useful information from a teacher model than by just making the student match the teacher's outputs. In this paper, we introduce \guide (Guided Initialization and Distillation of Embeddings). \guide can be considered a distillation technique that forces the student to match the teacher in the parameter space. Using \guide we show 25-26\% reduction in the teacher-student quality gap when using large student models (400M - 1B parameters) trained on $\approx$ 20B tokens. We also present a thorough analysis demonstrating that \guide can be combined with knowledge distillation with near additive improvements. Furthermore, we show that applying \guide alone leads to substantially better model quality than applying knowledge distillation by itself. Most importantly, \guide introduces no training or inference overhead and hence any model quality gains from our method are virtually free.

---

### Webscale-RL: Automated Data Pipeline for Scaling RL Data to Pretraining Levels

Authors: Zhepeng Cen, Haolin Chen, Shiyu Wang, Zuxin Liu, Zhiwei Liu, Jielin Qiu, Ding Zhao, Silvio Savarese, Caiming Xiong, Huan Wang, Weiran Yao
Published: 2025-10-07
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2510.06499v2

Large Language Models (LLMs) have achieved remarkable success through imitation learning on vast text corpora, but this paradigm creates a training-generation gap and limits robust reasoning. Reinforcement learning (RL) offers a more data-efficient solution capable of bridging this gap, yet its application has been constrained by a critical data bottleneck: existing RL datasets are orders of magnitude smaller and less diverse than web-scale pre-training corpora. To address this, we introduce the Webscale-RL pipeline, a scalable data engine that systematically converts large-scale pre-training documents into millions of diverse, verifiable question-answer pairs for RL. Using this pipeline, we construct the Webscale-RL dataset, containing 1.2 million examples across more than 9 domains. Our experiments show that the model trained on this dataset significantly outperforms continual pretraining and strong data refinement baselines across a suite of benchmarks. Notably, RL training with our dataset proves substantially more efficient, achieving the performance of continual pre-training with up to 100$\times$ fewer tokens. Our work presents a viable path toward scaling RL to pre-training levels, enabling more capable and efficient language models.

---

### Superpixel Integrated Grids for Fast Image Segmentation

Authors: Jack Roberts, Jeova Farias Sales Rocha Neto
Published: 2025-10-07
Categories: cs.CV
arXiv: https://arxiv.org/abs/2510.06487v1

Superpixels have long been used in image simplification to enable more efficient data processing and storage. However, despite their computational potential, their irregular spatial distribution has often forced deep learning approaches to rely on specialized training algorithms and architectures, undermining the original motivation for superpixelations. In this work, we introduce a new superpixel-based data structure, SIGRID (Superpixel-Integrated Grid), as an alternative to full-resolution images in segmentation tasks. By leveraging classical shape descriptors, SIGRID encodes both color and shape information of superpixels while substantially reducing input dimensionality. We evaluate SIGRIDs on four benchmark datasets using two popular convolutional segmentation architectures. Our results show that, despite compressing the original data, SIGRIDs not only match but in some cases surpass the performance of pixel-level representations, all while significantly accelerating model training. This demonstrates that SIGRIDs achieve a favorable balance between accuracy and computational efficiency.

---

### Active Next-Best-View Optimization for Risk-Averse Path Planning

Authors: Amirhossein Mollaei Khass, Guangyi Liu, Vivek Pandey, Wen Jiang, Boshu Lei, Kostas Daniilidis, Nader Motee
Published: 2025-10-07
Categories: cs.RO, cs.CV
arXiv: https://arxiv.org/abs/2510.06481v1

Safe navigation in uncertain environments requires planning methods that integrate risk aversion with active perception. In this work, we present a unified framework that refines a coarse reference path by constructing tail-sensitive risk maps from Average Value-at-Risk statistics on an online-updated 3D Gaussian-splat Radiance Field. These maps enable the generation of locally safe and feasible trajectories. In parallel, we formulate Next-Best-View (NBV) selection as an optimization problem on the SE(3) pose manifold, where Riemannian gradient descent maximizes an expected information gain objective to reduce uncertainty most critical for imminent motion. Our approach advances the state-of-the-art by coupling risk-averse path refinement with NBV planning, while introducing scalable gradient decompositions that support efficient online updates in complex environments. We demonstrate the effectiveness of the proposed framework through extensive computational studies.

---

### Anytime-Valid Answer Sufficiency Certificates for LLM Generation via Sequential Information Lift

Authors: Sanjeda Akter, Ibne Farabi Shihab, Anuj Sharma
Published: 2025-10-07
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2510.06478v2

We introduce Sequential-EDFL (Empirical Dynamic Formal Lift), which applies anytime-valid sequential testing to language model generation stopping. Our approach tracks information lift, defined as the log-likelihood ratio between the full model and deliberately weakened "skeleton" baselines, using self-normalized empirical-Bernstein e-processes that provide formal delta-level error control regardless of stopping time. This delta guarantee controls premature stopping when information lift is insufficient relative to the skeleton, and it does not imply delta control of factual incorrectness or hallucinations. We handle unknown centering through online mean estimation, combine multiple parameters via mixture e-processes, and support adaptive resets under distributional drift. On six benchmarks, Sequential-EDFL reduces generation length by 22 to 28 percent relative to sequential baselines while maintaining delta-level control with 12 percent computational overhead. We introduce automated skeletons (distilled submodels and randomized logits) and show robustness across skeleton families. Composing EDFL with a lightweight correctness gate (sentence boundaries plus a verifier) improves end-task correctness while preserving anytime-valid guarantees by only delaying stopping. Our certificates control information sufficiency, not factual correctness. Specifically, 10.9 percent of stopped sequences remain incorrect even with the gate (13.2 to 22.7 percent without it). EDFL serves as a first-stage filter that can reduce verification burden: when applied to stopped sequences, the gate validates 83 percent of stops, requiring full verification only for the remaining 17 percent, plus all non-stopped sequences. EDFL is not a standalone solution for safety-critical domains.

---

### Attention Sinks and Compression Valleys in LLMs are Two Sides of the Same Coin

Authors: Enrique Queipo-de-Llano, Álvaro Arroyo, Federico Barbero, Xiaowen Dong, Michael Bronstein, Yann LeCun, Ravid Shwartz-Ziv
Published: 2025-10-07
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2510.06477v2

Attention sinks and compression valleys have attracted significant attention as two puzzling phenomena in large language models, but have been studied in isolation. In this work, we present a surprising connection between attention sinks and compression valleys, tracing both to the formation of massive activations in the residual stream. We prove theoretically that massive activations necessarily produce representational compression and establish bounds on the resulting entropy reduction. Through experiments across several models (410M-120B parameters), we confirm that when the beginning-of-sequence token develops extreme activation norms in the middle layers, both compression valleys and attention sinks emerge simultaneously. Targeted ablation studies validate our theoretical predictions. This unified view motivates us to propose the Mix-Compress-Refine theory of information flow, as an attempt to explain how LLMs organize their computation in depth by controlling attention and representational compression via massive activations. Specifically, we posit that Transformer-based LLMs process tokens in three distinct phases: (1) broad mixing in the early layers, (2) compressed computation with limited mixing in the middle layers, and (3) selective refinement in the late layers. Our framework helps explain why embedding tasks perform best at intermediate layers, whereas generation tasks benefit from full-depth processing, clarifying differences in task-dependent representations.

---

### PuzzlePlex: Benchmarking Foundation Models on Reasoning and Planning with Puzzles

Authors: Yitao Long, Yuru Jiang, Hongjun Liu, Yilun Zhao, Jingchen Sun, Yiqiu Shen, Chen Zhao, Arman Cohan, Dennis Shasha
Published: 2025-10-07
Categories: cs.AI, cs.CL
arXiv: https://arxiv.org/abs/2510.06475v1

This work investigates the reasoning and planning capabilities of foundation models and their scalability in complex, dynamic environments. We introduce PuzzlePlex, a benchmark designed to assess these capabilities through a diverse set of puzzles. PuzzlePlex consists of 15 types of puzzles, including deterministic and stochastic games of varying difficulty, as well as single-player and two-player scenarios. The PuzzlePlex framework provides a comprehensive environment for each game, and supports extensibility to generate more challenging instances as foundation models evolve. Additionally, we implement customized game-playing strategies for comparison. Building on this benchmark, we develop fine-grained metrics to measure performance and conduct an in-depth analysis of frontier foundation models across two settings: instruction-based and code-based. Furthermore, we systematically investigate their scaling limits. Our findings show that reasoning models outperform others in instruction-based settings, while code-based execution presents greater challenges but offers a scalable and efficient alternative. PuzzlePlex enables targeted evaluation and guides future improvements in reasoning, planning, and generalization for foundation models.

---

### Deep Generative Model for Human Mobility Behavior

Authors: Ye Hong, Yatao Zhang, Konrad Schindler, Martin Raubal
Published: 2025-10-07
Categories: physics.soc-ph, cs.AI, cs.SI
arXiv: https://arxiv.org/abs/2510.06473v4

Understanding and modeling human mobility is central to challenges in transport planning, sustainable urban design, and public health. Despite decades of effort, simulating individual mobility remains challenging because of its complex, context-dependent, and exploratory nature. Here, building on the activity-based view of daily mobility, we propose MobilityGen, a diffusion-based generative framework for simulating multi-attribute activity-travel sequences over days to weeks at large spatial scales. By linking behavioral attributes with environmental context, MobilityGen reproduces key patterns such as scaling laws for location visits, activity time allocation, and the coupled evolution of travel mode and destination choices. It reflects spatio-temporal variability and generates diverse and plausible mobility patterns consistent with the built environment. Beyond standard validation, MobilityGen enables analyses that have been difficult with earlier models, including how access to urban space varies across travel modes and how co-presence dynamics shape social exposure and segregation. Together, these results support an integrated, data-driven basis for fine-grained studies of human mobility behavior and its societal implications.

---