# Research Papers — 2025-10 · bundle page 200 of 217 > Full-text markdown of papers #9951–10000 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/200 [← Previous bundle](https://aipod.dev/papers/archive/2025-10/page/199.md) [Next bundle →](https://aipod.dev/papers/archive/2025-10/page/201.md) ### Cluster-based Adaptive Retrieval: Dynamic Context Selection for RAG Applications Authors: Yifan Xu, Vipul Gupta, Rohit Aggarwal, Varsha Mahadevan, Bhaskar Krishnamachari Published: 2025-10-02 Categories: cs.IR, cs.AI, cs.CL, cs.LG arXiv: https://arxiv.org/abs/2511.14769v1 Retrieval-Augmented Generation (RAG) enhances large language models (LLMs) by pulling in external material, document, code, manuals, from vast and ever-growing corpora, to effectively answer user queries. The effectiveness of RAG depends significantly on aligning the number of retrieved documents with query characteristics: narrowly focused queries typically require fewer, highly relevant documents, whereas broader or ambiguous queries benefit from retrieving more extensive supporting information. However, the common static top-k retrieval approach fails to adapt to this variability, resulting in either insufficient context from too few documents or redundant information from too many. Motivated by these challenges, we introduce Cluster-based Adaptive Retrieval (CAR), an algorithm that dynamically determines the optimal number of documents by analyzing the clustering patterns of ordered query-document similarity distances. CAR detects the transition point within similarity distances, where tightly clustered, highly relevant documents shift toward less pertinent candidates, establishing an adaptive cut-off that scales with query complexity. On Coinbase's CDP corpus and the public MultiHop-RAG benchmark, CAR consistently picks the optimal retrieval depth and achieves the highest TES score, outperforming every fixed top-k baseline. In downstream RAG evaluations, CAR cuts LLM token usage by 60%, trims end-to-end latency by 22%, and reduces hallucinations by 10% while fully preserving answer relevance. Since integrating CAR into Coinbase's virtual assistant, we've seen user engagement jump by 200%. --- ### Diagnosing Bottlenecks in Data Visualization Understanding by Vision-Language Models Authors: Alexa R. Tartaglini, Satchel Grant, Daniel Wurgaft, Christopher Potts, Judith E. Fan Published: 2025-10-02 Categories: cs.CV, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2510.21740v1 Data visualizations are vital components of many scientific articles and news stories. Current vision-language models (VLMs) still struggle on basic data visualization understanding tasks, but the causes of failure remain unclear. Are VLM failures attributable to limitations in how visual information in the data visualization is encoded, how information is transferred between the vision and language modules, or how information is processed within the language module? We developed FUGU, a suite of data visualization understanding tasks, to precisely characterize potential sources of difficulty (e.g., extracting the position of data points, distances between them, and other summary statistics). We used FUGU to investigate three widely used VLMs. To diagnose the sources of errors produced by these models, we used activation patching and linear probes to trace information flow through models across a variety of prompting strategies. We found that some models fail to generate the coordinates of individual data points correctly, and these initial errors often lead to erroneous final responses. When these models are provided with the correct coordinates, performance improves substantially. Moreover, even when the model generates an incorrect response, the correct coordinates can be successfully read out from the latent representations in the vision encoder, suggesting that the source of these errors lies in the vision-language handoff. We further found that while providing correct coordinates helps with tasks involving one or a small number of data points, it generally worsens performance for tasks that require extracting statistical relationships across many data points. Fine-tuning models on FUGU also fails to yield ceiling performance. These findings point to architectural constraints in current VLMs that might pose significant challenges for reliable data visualization understanding. --- ### Next-Generation LLM for UAV: From Natural Language to Autonomous Flight Authors: Liangqi Yuan, Chuhao Deng, Dong-Jun Han, Inseok Hwang, Sabine Brunswicker, Christopher G. Brinton Published: 2025-10-02 Categories: cs.RO, cs.AI, cs.CL, eess.SY arXiv: https://arxiv.org/abs/2510.21739v1 With the rapid advancement of Large Language Models (LLMs), their capabilities in various automation domains, particularly Unmanned Aerial Vehicle (UAV) operations, have garnered increasing attention. Current research remains predominantly constrained to small-scale UAV applications, with most studies focusing on isolated components such as path planning for toy drones, while lacking comprehensive investigation of medium- and long-range UAV systems in real-world operational contexts. Larger UAV platforms introduce distinct challenges, including stringent requirements for airport-based take-off and landing procedures, adherence to complex regulatory frameworks, and specialized operational capabilities with elevated mission expectations. This position paper presents the Next-Generation LLM for UAV (NeLV) system -- a comprehensive demonstration and automation roadmap for integrating LLMs into multi-scale UAV operations. The NeLV system processes natural language instructions to orchestrate short-, medium-, and long-range UAV missions through five key technical components: (i) LLM-as-Parser for instruction interpretation, (ii) Route Planner for Points of Interest (POI) determination, (iii) Path Planner for waypoint generation, (iv) Control Platform for executable trajectory implementation, and (v) UAV monitoring. We demonstrate the system's feasibility through three representative use cases spanning different operational scales: multi-UAV patrol, multi-POI delivery, and multi-hop relocation. Beyond the current implementation, we establish a five-level automation taxonomy that charts the evolution from current LLM-as-Parser capabilities (Level 1) to fully autonomous LLM-as-Autopilot systems (Level 5), identifying technical prerequisites and research challenges at each stage. --- ### Neurotremor: A wearable Supportive Device for Supporting Upper Limb Muscle Function Authors: Aueaphum Aueawattthanaphisut, Thanyanee Srichaisak, Arissa Ieochai Published: 2025-10-02 Categories: q-bio.NC, cs.ET, cs.LG, q-bio.TO arXiv: https://arxiv.org/abs/2510.19826v1 A sensor-fused wearable assistance prototype for upper-limb function (triceps brachii and extensor pollicis brevis) is presented. The device integrates surface electromyography (sEMG), an inertial measurement unit (IMU), and flex/force sensors on an M5StickC plus an ESP32-S3 compute hub. Signals are band-pass and notch filtered; features (RMS, MAV, zero-crossings, and 4-12 Hz tremor-band power) are computed in 250 ms windows and fed to an INT8 TensorFlow Lite Micro model. Control commands are bounded by a control-barrier-function safety envelope and delivered within game-based tasks with lightweight personalization. In a pilot technical feasibility evaluation with healthy volunteers (n = 12) performing three ADL-oriented tasks, tremor prominence decreased (Delta TI = -0.092, 95% CI (-0.102, -0.079)), range of motion increased (+12.65%, 95% CI (+8.43, +13.89)), repetitions rose (+2.99 min^-1, 95% CI (+2.61, +3.35)), and the EMG median-frequency slope became less negative (Delta = +0.100 Hz/min, 95% CI (+0.083, +0.127)). The sensing-to-assist loop ran at 100 Hz with 8.7 ms median on-device latency, 100% session completion, and 0 device-related adverse events. These results demonstrate technical feasibility of embedded, sensor-fused assistance for upper-limb function; formal patient studies under IRB oversight are planned. --- ### Multi-Agent Design Assistant for the Simulation of Inertial Fusion Energy Authors: Meir H. Shachar, Dane M. Sterbentz, Harshitha Menon, Charles F. Jekel, M. Giselle Fernández-Godino, Nathan K. Brown, Ismael D. Boureima, Yue Hao, Kevin Korner, Robert Rieben, Daniel A. White, William J. Schill et al. Published: 2025-10-02 Categories: physics.app-ph, cs.AI arXiv: https://arxiv.org/abs/2510.17830v3 Inertial fusion energy promises nearly unlimited, clean power if it can be achieved. However, the design and engineering of fusion systems requires controlling and manipulating matter at extreme energies and timescales; the shock physics and radiation transport governing the physical behavior under these conditions are complex requiring the development, calibration, and use of predictive multiphysics codes to navigate the highly nonlinear and multi-faceted design landscape. We hypothesize that artificial intelligence reasoning models can be combined with physics codes and emulators to autonomously design fusion fuel capsules. In this article, we construct a multi-agent system where natural language is utilized to explore the complex physics regimes around fusion energy. The agentic system is capable of executing a high-order multiphysics inertial fusion computational code. We demonstrate the capacity of the multi-agent design assistant to translate natural-language design goals into simulation execution, physics-emulator construction, analysis, and iterative refinement of capsule-geometry parameters, ultimately achieving simulated ignition within the computational design workflow. --- ### Dynamic Stress Detection: A Study of Temporal Progression Modelling of Stress in Speech Authors: Vishakha Lall, Yisi Liu Published: 2025-10-02 Categories: eess.AS, cs.AI, cs.CL, cs.SD arXiv: https://arxiv.org/abs/2510.08586v2 Detecting psychological stress from speech is critical in high-pressure settings. While prior work has leveraged acoustic features for stress detection, most treat stress as a static label. In this work, we model stress as a temporally evolving phenomenon influenced by historical emotional state. We propose a dynamic labelling strategy that derives fine-grained stress annotations from emotional labels and introduce cross-attention-based sequential models, a Unidirectional LSTM and a Transformer Encoder, to capture temporal stress progression. Our approach achieves notable accuracy gains on MuSE (+5%) and StressID (+18%) over existing baselines, and generalises well to a custom real-world dataset. These results highlight the value of modelling stress as a dynamic construct in speech. --- ### A Digital Pheromone-Based Approach for In-Control/Out-of-Control Classification Authors: Pedro Pestana, M. Fátima Brilhante Published: 2025-10-02 Categories: cs.NE, eess.SY arXiv: https://arxiv.org/abs/2510.07329v4 In complex production lines, it is essential to have strict, fast-acting rules to determine whether the system is In Control (InC) or Out of Control (OutC). This study explores a bio-inspired method that digitally mimics ant colony behavior to classify InC/OutC states and forecast imminent transitions requiring maintenance. A case study on industrial potato chip frying provides the application context. During each two-minute frying cycle, sequences of eight temperature readings are collected. Each sequence is treated as a digital ant depositing virtual pheromones, generating a Base Score. New sequences, representing new ants, can either reinforce or weaken this score, leading to a Modified Base Score that reflects the system's evolving condition. Signals such as extreme temperatures, large variations within a sequence, or the detection of change-points contribute to a Threat Score, which is added to the Modified Base Score. Since pheromones naturally decay over time unless reinforced, an Environmental Score is incorporated to reflect recent system dynamics, imitating real ant behavior. This score is calculated from the Modified Base Scores collected over the past hour. The resulting Total Score, obtained as the sum of the Modified Base Score, Threat Score, and Environmental Score, is used as the main indicator for real-time system classification and forecasting of transitions from InC to OutC. This ant colony optimization-inspired approach provides an adaptive and interpretable framework for process monitoring and predictive maintenance in industrial environments. --- ### Stacked Regression using Off-the-shelf, Stimulus-tuned and Fine-tuned Neural Networks for Predicting fMRI Brain Responses to Movies (Algonauts 2025 Report) Authors: Robert Scholz, Kunal Bagga, Christine Ahrends, Carlo Alberto Barbano Published: 2025-10-02 Categories: eess.IV, cs.AI, cs.CV, q-bio.NC arXiv: https://arxiv.org/abs/2510.06235v1 We present our submission to the Algonauts 2025 Challenge, where the goal is to predict fMRI brain responses to movie stimuli. Our approach integrates multimodal representations from large language models, video encoders, audio models, and vision-language models, combining both off-the-shelf and fine-tuned variants. To improve performance, we enhanced textual inputs with detailed transcripts and summaries, and we explored stimulus-tuning and fine-tuning strategies for language and vision models. Predictions from individual models were combined using stacked regression, yielding solid results. Our submission, under the team name Seinfeld, ranked 10th. We make all code and resources publicly available, contributing to ongoing efforts in developing multimodal encoding models for brain activity. --- ### User to Video: A Model for Spammer Detection Inspired by Video Classification Technology Authors: Haoyang Zhang, Zhou Yang, Yucai Pang Published: 2025-10-02 Categories: cs.CV arXiv: https://arxiv.org/abs/2510.06233v1 This article is inspired by video classification technology. If the user behavior subspace is viewed as a frame image, consecutive frame images are viewed as a video. Following this novel idea, a model for spammer detection based on user videoization, called UVSD, is proposed. Firstly, a user2piexl algorithm for user pixelization is proposed. Considering the adversarial behavior of user stances, the user is viewed as a pixel, and the stance is quantified as the pixel's RGB. Secondly, a behavior2image algorithm is proposed for transforming user behavior subspace into frame images. Low-rank dense vectorization of subspace user relations is performed using representation learning, while cutting and diffusion algorithms are introduced to complete the frame imageization. Finally, user behavior videos are constructed based on temporal features. Subsequently, a video classification algorithm is combined to identify the spammers. Experiments using publicly available datasets, i.e., WEIBO and TWITTER, show an advantage of the UVSD model over state-of-the-art methods. --- ### Can AI Truly Represent Your Voice in Deliberations? A Comprehensive Study of Large-Scale Opinion Aggregation with LLMs Authors: Shenzhe Zhu, Shu Yang, Michiel A. Bakker, Alex Pentland, Jiaxin Pei Published: 2025-10-02 Categories: cs.CL arXiv: https://arxiv.org/abs/2510.05154v5 Large-scale public deliberations generate thousands of free-form contributions that must be synthesized into representative and neutral summaries for policy use. While LLMs have been shown as a promising tool to generate summaries for large-scale deliberations, they also risk underrepresenting minority perspectives and exhibiting bias with respect to the input order, raising fairness concerns in high-stakes contexts. Studying and fixing these issues requires a comprehensive evaluation at a large scale, yet current practice often relies on LLMs as judges, which show weak alignment with human judgments. To address this, we present DeliberationBank, a large-scale human-grounded dataset with (1) opinion data spanning ten deliberation questions created by 3,000 participants and (2) summary judgment data annotated by 4,500 participants across four dimensions (representativeness, informativeness, neutrality, policy approval). Using these datasets, we train DeliberationJudge, a fine-tuned DeBERTa model that can rate deliberation summaries from individual perspectives. DeliberationJudge is more efficient and more aligned with human judgements compared to a wide range of LLM judges. With DeliberationJudge, we evaluate 18 LLMs and reveal persistent weaknesses in deliberation summarization, especially underrepresentation of minority positions. Our framework provides a scalable and reliable way to evaluate deliberation summarization, helping ensure AI systems are more representative and equitable for policymaking. --- ### An Algorithmic Information-Theoretic Perspective on the Symbol Grounding Problem Authors: Zhangchi Liu Published: 2025-10-02 Categories: cs.AI, cs.IT arXiv: https://arxiv.org/abs/2510.05153v1 This paper provides a definitive, unifying framework for the Symbol Grounding Problem (SGP) by reformulating it within Algorithmic Information Theory (AIT). We demonstrate that the grounding of meaning is a process fundamentally constrained by information-theoretic limits, thereby unifying the Gödelian (self-reference) and No Free Lunch (statistical) perspectives. We model a symbolic system as a universal Turing machine and define grounding as an act of information compression. The argument proceeds in four stages. First, we prove that a purely symbolic system cannot ground almost all possible "worlds" (data strings), as they are algorithmically random and thus incompressible. Second, we show that any statically grounded system, specialized for compressing a specific world, is inherently incomplete because an adversarial, incompressible world relative to the system can always be constructed. Third, the "grounding act" of adapting to a new world is proven to be non-inferable, as it requires the input of new information (a shorter program) that cannot be deduced from the system's existing code. Finally, we use Chaitin's Incompleteness Theorem to prove that any algorithmic learning process is itself a finite system that cannot comprehend or model worlds whose complexity provably exceeds its own. This establishes that meaning is the open-ended process of a system perpetually attempting to overcome its own information-theoretic limitations. --- ### A Single Character can Make or Break Your LLM Evals Authors: Jingtong Su, Jianyu Zhang, Karen Ullrich, Léon Bottou, Mark Ibrahim Published: 2025-10-02 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2510.05152v1 Common Large Language model (LLM) evaluations rely on demonstration examples to steer models' responses to the desired style. While the number of examples used has been studied and standardized, the choice of how to format examples is less investigated. In evaluation protocols and real world usage, users face the choice how to separate in-context examples: use a comma? new line? semi-colon? hashtag? etc.? Surprisingly, we find this seemingly minor choice can dramatically alter model response quality. Across leading model families (Llama, Qwen, Gemma), performance on MMLU for example can vary by $\pm 23\%$ depending on the choice of delimiter. In fact, one can manipulate model rankings to put any model in the lead by only modifying the single character separating examples. We find LLMs' brittleness pervades topics, model families, and doesn't improve with scale. By probing attention head scores, we find that good-performing delimiters steer attention towards key tokens in the input. Finally, we explore methods to improve LLMs' robustness to the choice of delimiter. We find specifying the selected delimiter in the prompt boosts robustness and offer practical recommendations for the best-performing delimiters to select. --- ### Exploring Large Language Models for Financial Applications: Techniques, Performance, and Challenges with FinMA Authors: Prudence Djagba, Abdelkader Y. Saley Published: 2025-10-02 Categories: cs.CL, cs.LG arXiv: https://arxiv.org/abs/2510.05151v1 This research explores the strengths and weaknesses of domain-adapted Large Language Models (LLMs) in the context of financial natural language processing (NLP). The analysis centers on FinMA, a model created within the PIXIU framework, which is evaluated for its performance in specialized financial tasks. Recognizing the critical demands of accuracy, reliability, and domain adaptation in financial applications, this study examines FinMA's model architecture, its instruction tuning process utilizing the Financial Instruction Tuning (FIT) dataset, and its evaluation under the FLARE benchmark. Findings indicate that FinMA performs well in sentiment analysis and classification, but faces notable challenges in tasks involving numerical reasoning, entity recognition, and summarization. This work aims to advance the understanding of how financial LLMs can be effectively designed and evaluated to assist in finance-related decision-making processes. --- ### Chronological Thinking in Full-Duplex Spoken Dialogue Language Models Authors: Donghang Wu, Haoyang Zhang, Chen Chen, Tianyu Zhang, Fei Tian, Xuerui Yang, Gang Yu, Hexin Liu, Nana Hou, Yuchen Hu, Eng Siong Chng Published: 2025-10-02 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2510.05150v3 Recent advances in spoken dialogue language models (SDLMs) reflect growing interest in shifting from turn-based to full-duplex systems, where the models continuously perceive user speech streams while generating responses. This simultaneous listening and speaking design enables real-time interaction and the agent can handle dynamic conversational behaviors like user barge-in. However, during the listening phase, existing systems keep the agent idle by repeatedly predicting the silence token, which departs from human behavior: we usually engage in lightweight thinking during conversation rather than remaining absent-minded. Inspired by this, we propose Chronological Thinking, an on-the-fly conversational thinking mechanism that aims to improve response quality in full-duplex SDLMs. Specifically, chronological thinking presents a paradigm shift from conventional LLM thinking approaches, such as Chain-of-Thought, purpose-built for streaming acoustic input. (1) Strictly causal: the agent reasons incrementally while listening, updating internal hypotheses only from past audio with no lookahead. (2) No additional latency: reasoning is amortized during the listening window; once the user stops speaking, the agent halts thinking and begins speaking without further delay. Experiments demonstrate the effectiveness of chronological thinking through both objective metrics and human evaluations show consistent improvements in response quality. Furthermore, chronological thinking robustly handles conversational dynamics and attains competitive performance on full-duplex interaction metrics. --- ### Percepta: High Performance Stream Processing at the Edge Authors: Clarisse Sousa, Tiago Fonseca, Luis Lino Ferreira, Ricardo Venâncio, Ricardo Severino Published: 2025-10-02 Categories: cs.DC, cs.AI arXiv: https://arxiv.org/abs/2510.05149v1 The rise of real-time data and the proliferation of Internet of Things (IoT) devices have highlighted the limitations of cloud-centric solutions, particularly regarding latency, bandwidth, and privacy. These challenges have driven the growth of Edge Computing. Associated with IoT appears a set of other problems, like: data rate harmonization between multiple sources, protocol conversion, handling the loss of data and the integration with Artificial Intelligence (AI) models. This paper presents Percepta, a lightweight Data Stream Processing (DSP) system tailored to support AI workloads at the edge, with a particular focus on such as Reinforcement Learning (RL). It introduces specialized features such as reward function computation, data storage for model retraining, and real-time data preparation to support continuous decision-making. Additional functionalities include data normalization, harmonization across heterogeneous protocols and sampling rates, and robust handling of missing or incomplete data, making it well suited for the challenges of edge-based AI deployment. --- ### Every Step Counts: Decoding Trajectories as Authorship Fingerprints of dLLMs Authors: Qi Li, Runpeng Yu, Haiquan Lu, Xinchao Wang Published: 2025-10-02 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2510.05148v1 Discrete Diffusion Large Language Models (dLLMs) have recently emerged as a competitive paradigm for non-autoregressive language modeling. Their distinctive decoding mechanism enables faster inference speed and strong performance in code generation and mathematical tasks. In this work, we show that the decoding mechanism of dLLMs not only enhances model utility but also can be used as a powerful tool for model attribution. A key challenge in this problem lies in the diversity of attribution scenarios, including distinguishing between different models as well as between different checkpoints or backups of the same model. To ensure broad applicability, we identify two fundamental problems: what information to extract from the decoding trajectory, and how to utilize it effectively. We first observe that relying directly on per-step model confidence yields poor performance. This is mainly due to the bidirectional decoding nature of dLLMs: each newly decoded token influences the confidence of other decoded tokens, making model confidence highly redundant and washing out structural signal regarding decoding order or dependencies. To overcome this, we propose a novel information extraction scheme called the Directed Decoding Map (DDM), which captures structural relationships between decoding steps and better reveals model-specific behaviors. Furthermore, to make full use of the extracted structural information during attribution, we propose Gaussian-Trajectory Attribution (GTA), where we fit a cell-wise Gaussian distribution at each decoding position for each target model, and define the likelihood of a trajectory as the attribution score: if a trajectory exhibits higher log-likelihood under the distribution of a specific model, it is more likely to have been generated by that model. Extensive experiments under different settings validate the utility of our methods. --- ### Adaptive Reinforcement Learning for Dynamic Configuration Allocation in Pre-Production Testing Authors: Yu Zhu Published: 2025-10-02 Categories: cs.SE, cs.LG, stat.ML arXiv: https://arxiv.org/abs/2510.05147v1 Ensuring reliability in modern software systems requires rigorous pre-production testing across highly heterogeneous and evolving environments. Because exhaustive evaluation is infeasible, practitioners must decide how to allocate limited testing resources across configurations where failure probabilities may drift over time. Existing combinatorial optimization approaches are static, ad hoc, and poorly suited to such non-stationary settings. We introduce a novel reinforcement learning (RL) framework that recasts configuration allocation as a sequential decision-making problem. Our method is the first to integrate Q-learning with a hybrid reward design that fuses simulated outcomes and real-time feedback, enabling both sample efficiency and robustness. In addition, we develop an adaptive online-offline training scheme that allows the agent to quickly track abrupt probability shifts while maintaining long-run stability. Extensive simulation studies demonstrate that our approach consistently outperforms static and optimization-based baselines, approaching oracle performance. This work establishes RL as a powerful new paradigm for adaptive configuration allocation, advancing beyond traditional methods and offering broad applicability to dynamic testing and resource scheduling domains. --- ### Efficient Tree-Structured Deep Research with Adaptive Resource Allocation Authors: Lunyiu Nie, Nedim Lipka, Ryan A. Rossi, Swarat Chaudhuri Published: 2025-10-02 Categories: cs.DC, cs.AI, cs.MA arXiv: https://arxiv.org/abs/2510.05145v2 Deep research agents, which synthesize information across diverse sources, are significantly constrained by the sequential nature of reasoning. This bottleneck results in high latency, poor runtime adaptability, and inefficient resource allocation, making today's deep research systems impractical for interactive applications. To overcome this, we introduce ParallelResearch, a novel framework for efficient deep research that transforms sequential processing into parallel, runtime orchestration by dynamically decomposing complex queries into tree-structured sub-tasks. Our core contributions are threefold: (1) an adaptive planner that dynamically allocates computational resources based on query complexity; (2) a runtime orchestration layer that prunes redundant paths to reallocate resources and enables speculative execution; and (3) a fully-asynchronous execution infrastructure that enables concurrency across both research breadth and depth. Experiments on two benchmarks show up to 5x speedups with comparable final report quality, and consistent quality improvements with the same time budgets. --- ### Understanding Transformers for Time Series: Rank Structure, Flow-of-ranks, and Compressibility Authors: Annan Yu, Danielle C. Maddix, Boran Han, Xiyuan Zhang, Abdul Fatir Ansari, Oleksandr Shchur, Christos Faloutsos, Andrew Gordon Wilson, Michael W. Mahoney, Yuyang Wang Published: 2025-10-02 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2510.03358v1 Transformers are widely used across data modalities, and yet the principles distilled from text models often transfer imperfectly to models trained to other modalities. In this paper, we analyze Transformers through the lens of rank structure. Our focus is on the time series setting, where the structural properties of the data differ remarkably from those of text or vision. We show that time-series embeddings, unlike text or vision, exhibit sharply decaying singular value spectra: small patch sizes and smooth continuous mappings concentrate the data into low-rank subspaces. From this, we prove that the associated $Q/K/V$ projections admit accurate low-rank approximations, and that attention layers become compressible in proportion to the decay of the embedding spectrum. We introduce the concept of flow-of-ranks, a phenomenon by which nonlinear mixing across depth inflates the rank, explaining why early layers are most amenable to compression and why ranks grow with depth. Guided by these theoretical and empirical results, we use these insights to compress Chronos, a large time series foundation model, achieving a reduction of $65\%$ in inference time and $81\%$ in memory, without loss of accuracy. Our findings provide principled guidance for allocating width, depth, and heads in time series foundation models, and for exploiting their inherent compressibility. --- ### Learned Display Radiance Fields with Lensless Cameras Authors: Ziyang Chen, Yuta Itoh, Kaan Akşit Published: 2025-10-02 Categories: cs.CV, cs.ET arXiv: https://arxiv.org/abs/2510.03356v2 Calibrating displays is a basic and regular task that content creators must perform to maintain optimal visual experience, yet it remains a troublesome issue. Measuring display characteristics from different viewpoints often requires specialized equipment and a dark room, making it inaccessible to most users. To avoid specialized hardware requirements in display calibrations, our work co-designs a lensless camera and an Implicit Neural Representation based algorithm for capturing display characteristics from various viewpoints. More specifically, our pipeline enables efficient reconstruction of light fields emitted from a display from a viewing cone of 46.6° X 37.6°. Our emerging pipeline paves the initial steps towards effortless display calibration and characterization. --- ### High Cycle S-N curve prediction for Al 7075-T6 alloy using Recurrent Neural Networks (RNNs) Authors: Aryan Patel Published: 2025-10-02 Categories: cs.LG, cond-mat.mtrl-sci, physics.app-ph arXiv: https://arxiv.org/abs/2510.03355v1 Aluminum is a widely used alloy, which is susceptible to fatigue failure. Characterizing fatigue performance for materials is extremely time and cost demanding, especially for high cycle data. To help mitigate this, a transfer learning based framework has been developed using Long short-term memory networks (LSTMs) in which a source LSTM model is trained based on pure axial fatigue data for Aluminum 7075-T6 alloy which is then transferred to predict high cycle torsional S-N curves. The framework was able to accurately predict Al torsional S-N curves for a much higher cycle range. It is the belief that this framework will help to drastically mitigate the cost of gathering fatigue characteristics for different materials and help prioritize tests with better cost and time constraints. --- ### Sonar Image Datasets: A Comprehensive Survey of Resources, Challenges, and Applications Authors: Larissa S. Gomes, Gustavo P. Almeida, Bryan U. Moreira, Marco Quiroz, Breno Xavier, Lucas Soares, Stephanie L. Brião, Felipe G. Oliveira, Paulo L. J. Drews-Jr Published: 2025-10-02 Categories: cs.CV arXiv: https://arxiv.org/abs/2510.03353v1 Sonar images are relevant for advancing underwater exploration, autonomous navigation, and ecosystem monitoring. However, the progress depends on data availability. The scarcity of publicly available, well-annotated sonar image datasets creates a significant bottleneck for the development of robust machine learning models. This paper presents a comprehensive and concise review of the current landscape of sonar image datasets, seeking not only to catalog existing resources but also to contextualize them, identify gaps, and provide a clear roadmap, serving as a base guide for researchers of any kind who wish to start or advance in the field of underwater acoustic data analysis. We mapped publicly accessible datasets across various sonar modalities, including Side Scan Sonar (SSS), Forward-Looking Sonar (FLS), Synthetic Aperture Sonar (SAS), Multibeam Echo Sounder (MBES), and Dual-Frequency Identification Sonar (DIDSON). An analysis was conducted on applications such as classification, detection, segmentation, and 3D reconstruction. This work focuses on state-of-the-art advancements, incorporating newly released datasets. The findings are synthesized into a master table and a chronological timeline, offering a clear and accessible comparison of characteristics, sizes, and annotation details datasets. --- ### Inference-Time Search Using Side Information for Diffusion-Based Image Reconstruction Authors: Mahdi Farahbakhsh, Vishnu Teja Kunde, Dileep Kalathil, Krishna Narayanan, Jean-Francois Chamberland Published: 2025-10-02 Categories: cs.CV, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2510.03352v3 Diffusion models have been used as priors for solving inverse problems. However, existing approaches typically overlook side information that could significantly improve reconstruction quality, especially in severely ill-posed settings. In this work, we propose a novel framework that incorporates side information into existing diffusion-based inverse problem solvers via inference-time search, in a plug-and-play, training-free manner. Through extensive experiments across a range of inverse problems, including inpainting, super-resolution, and several deblurring tasks, and across multiple diffusion-based inverse problem solvers (DPS, DAPS, and MPGD), we show that augmenting each solver with our framework consistently improves the quality of the reconstructions over the corresponding original method. To demonstrate the generality of our approach, we consider diverse forms of side information, including reference images, textual descriptions, and anatomical MRI scans. The code is available at this \href{https://github.com/mahdi-farahbakhsh/DISS}{repository}\footnote{https://github.com/mahdi-farahbakhsh/DISS}. --- ### Interpretable Neuropsychiatric Diagnosis via Concept-Guided Graph Neural Networks Authors: Song Wang, Zhenyu Lei, Zhen Tan, Jundong Li, Javier Rasero, Aiying Zhang, Chirag Agarwal Published: 2025-10-02 Categories: cs.LG, cs.AI, eess.IV arXiv: https://arxiv.org/abs/2510.03351v1 Nearly one in five adolescents currently live with a diagnosed mental or behavioral health condition, such as anxiety, depression, or conduct disorder, underscoring the urgency of developing accurate and interpretable diagnostic tools. Resting-state functional magnetic resonance imaging (rs-fMRI) provides a powerful lens into large-scale functional connectivity, where brain regions are modeled as nodes and inter-regional synchrony as edges, offering clinically relevant biomarkers for psychiatric disorders. While prior works use graph neural network (GNN) approaches for disorder prediction, they remain complex black-boxes, limiting their reliability and clinical translation. In this work, we propose CONCEPTNEURO, a concept-based diagnosis framework that leverages large language models (LLMs) and neurobiological domain knowledge to automatically generate, filter, and encode interpretable functional connectivity concepts. Each concept is represented as a structured subgraph linking specific brain regions, which are then passed through a concept classifier. Our design ensures predictions through clinically meaningful connectivity patterns, enabling both interpretability and strong predictive performance. Extensive experiments across multiple psychiatric disorder datasets demonstrate that CONCEPTNEURO-augmented GNNs consistently outperform their vanilla counterparts, improving accuracy while providing transparent, clinically aligned explanations. Furthermore, concept analyses highlight disorder-specific connectivity patterns that align with expert knowledge and suggest new hypotheses for future investigation, establishing CONCEPTNEURO as an interpretable, domain-informed framework for psychiatric disorder diagnosis. --- ### AgentCaster: Reasoning-Guided Tornado Forecasting Authors: Michael Chen Published: 2025-10-02 Categories: cs.LG, cs.AI, cs.CL, physics.ao-ph arXiv: https://arxiv.org/abs/2510.03349v1 There is a growing need to evaluate Large Language Models (LLMs) on complex, high-impact, real-world tasks to assess their true readiness as reasoning agents. To address this gap, we introduce AgentCaster, a contamination-free framework employing multimodal LLMs end-to-end for the challenging, long-horizon task of tornado forecasting. Within AgentCaster, models interpret heterogeneous spatiotemporal data from a high-resolution convection-allowing forecast archive. We assess model performance over a 40-day period featuring diverse historical data, spanning several major tornado outbreaks and including over 500 tornado reports. Each day, models query interactively from a pool of 3,625 forecast maps and 40,125 forecast soundings for a forecast horizon of 12-36 hours. Probabilistic tornado-risk polygon predictions are verified against ground truths derived from geometric comparisons across disjoint risk bands in projected coordinate space. To quantify accuracy, we propose domain-specific TornadoBench and TornadoHallucination metrics, with TornadoBench highly challenging for both LLMs and domain expert human forecasters. Notably, human experts significantly outperform state-of-the-art models, which demonstrate a strong tendency to hallucinate and overpredict risk intensity, struggle with precise geographic placement, and exhibit poor spatiotemporal reasoning in complex, dynamically evolving systems. AgentCaster aims to advance research on improving LLM agents for challenging reasoning tasks in critical domains. --- ### Monocular Visual Odometry without Calibration or Test-time Optimization Authors: Vladimir Yugay, Duy-Kien Nguyen, Theo Gevers, Cees G. M. Snoek, Martin R. Oswald Published: 2025-10-02 Categories: cs.CV arXiv: https://arxiv.org/abs/2510.03348v5 The most accurate monocular visual odometry systems require known camera intrinsics, refine their estimates with test-time optimization, and recover trajectories only up to an unknown factor. Systems built on large 3D models need no intrinsics, but they remain considerably less accurate and slower for odometry. Direct pose regression avoids all these requirements, yet it has not matched either approach's accuracy. We revisit this formulation with a transformer that predicts relative camera poses together with separate rotation and translation confidences over overlapping image windows, supervised by camera poses alone. A confidence-weighted module then aggregates the overlapping predictions into a single trajectory. The resulting method, CalfVO, needs no intrinsics, no bundle adjustment, and no loop closure, and it recovers scale from learned priors, accurately enough that it is evaluated without any alignment to the ground truth. Across five benchmarks, it is the most accurate calibration-free method on every metric we report, and it runs at 53 FPS, faster than every baseline. --- ### KVComm: Enabling Efficient LLM Communication through Selective KV Sharing Authors: Xiangyu Shi, Marco Chiesa, Gerald Q. Maguire, Dejan Kostic Published: 2025-10-02 Categories: cs.LG, cs.AI, cs.MA arXiv: https://arxiv.org/abs/2510.03346v3 Large Language Models (LLMs) are increasingly deployed in multi-agent systems, where effective inter-model communication is crucial. Existing communication protocols either rely on natural language, incurring high inference costs and information loss, or on hidden states, which suffer from information concentration bias and inefficiency. To address these limitations, we propose KVComm, a novel communication framework that enables efficient communication between LLMs through selective sharing of KV pairs. KVComm leverages the rich information encoded in the KV pairs while avoiding the pitfalls of hidden states. We introduce a KV layer-wise selection strategy based on attention importance scores with a Gaussian prior to identify the most informative KV pairs for communication. Extensive experiments across diverse tasks and model pairs demonstrate that KVComm achieves comparable performance to the upper-bound method, which directly merges inputs to one model without any communication, while transmitting as few as 30\% of layers' KV pairs. Our study highlights the potential of KV pairs as an effective medium for inter-LLM communication, paving the way for scalable and efficient multi-agent systems. --- ### Pilot selection in the era of Virtual reality: algorithms for accurate and interpretable machine learning models Authors: Luoma Ke, Guangpeng Zhang, Jibo He, Yajing Li, Yan Li, Xufeng Liu, Peng Fang Published: 2025-10-02 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2510.03345v1 With the rapid growth of the aviation industry, there is a need for a large number of flight crew. How to select the right pilots in a cost-efficient manner has become an important research question. In the current study, twenty-three pilots were recruited from China Eastern Airlines, and 23 novices were from the community of Tsinghua University. A novel approach incorporating machine learning and virtual reality technology was applied to distinguish features between these participants with different flight skills. Results indicate that SVM with the MIC feature selection method consistently achieved the highest prediction performance on all metrics with an Accuracy of 0.93, an AUC of 0.96, and an F1 of 0.93, which outperforms four other classifier algorithms and two other feature selection methods. From the perspective of feature selection methods, the MIC method can select features with a nonlinear relationship to sampling labels, instead of a simple filter-out. Our new implementation of the SVM + MIC algorithm outperforms all existing pilot selection algorithms and perhaps provides the first implementation based on eye tracking and flight dynamics data. This study's VR simulation platforms and algorithms can be used for pilot selection and training. --- ### Assessing the impact of contact time on leachate chemistry from recycled concrete aggregates Authors: Morgan D. Sanger, Gabrielle Campagnola, Robin Ritchey, Tuncer B. Edil, Matthew Ginder-Vogel Published: 2025-10-02 Categories: physics.chem-ph, cond-mat.mtrl-sci, cs.LG arXiv: https://arxiv.org/abs/2510.03344v1 Recycled concrete aggregate (RCA) is recognized as a readily available, mechanically sufficient construction and demolition waste product that is suitable as a base course substitute for natural, virgin aggregate in pavement construction. Environmentally responsible applications of RCA must consider the high alkalinity, high pH leachate, and heavy metal leaching risks reported in the literature. The existing body of literature does not address discrepancies between field and laboratory measurements of RCA leachate pH, nor are there any existing studies of aged RCA leachate composition as a function of time. To consider the influence of contact time on RCA leachate, the present study evaluates recovered RCA base course samples from the Minnesota Road Research highway construction study site using modified batch test methodology. Leachate pH, alkalinity, and calcium ion (Ca2+) concentration were monitored for 24 hours to understand RCA leachate chemistry during the initial contact period. Leachate pH is high upon initial contact with water (pH > 10) and decreases over time as it reacts with atmospheric carbon dioxide. Calcium ion concentration increases rapidly in the initial contact period, then more gradually as calcium saturation is reached. Alkalinity stabilizes (50-65 mg CaCO3/L) after a dramatic increase during the initial contact period. --- ### Defining a Strategic Action Plan for AI in Higher Education Authors: Nikolaos Avouris Published: 2025-10-02 Categories: cs.CY, cs.AI arXiv: https://arxiv.org/abs/2510.03343v1 This paper discusses key challenges of Artificial Intelligence in Education, with main focus on higher education institutions. We start with reviewing normative actions of international organizations and concerns expressed about the current technical landscape. Then we proceed with proposing a framework that comprises five key dimensions relating to the main challenges relating to AI in higher education institutions, followed by five key strategic actions that the main stakeholders need to take in order to address the current developments. We map these actions to the main stakeholders of higher education and propose a deployment plan. This defines a framework along the dimensions: Challenges, Actions, Stakeholders, Deployment CASD. Examples of AI specific actions at the institutional and individual course level are also provided and discussed. --- ### OpusAnimation: Code-Based Dynamic Chart Generation Authors: Bozheng Li, Miao Yang, Zhenhan Chen, Jiawang Cao, Mushui Liu, Yi Lu, Yongliang Wu, Bin Zhang, Yangguang Ji, Licheng Tang, Jay Wu, Wenbo Zhu Published: 2025-10-02 Categories: cs.CV arXiv: https://arxiv.org/abs/2510.03341v1 Dynamic Chart Generation (DCG) involves producing code-rendered animated visualizations as charts. While recent advances in multi-modal large language models (MLLMs) have significantly improved their capability on static chart generation and comprehension, MLLMs' potential for handling dynamic chart generation and understanding remains underexplored. To bridge this research gap, we introduce DCG-Bench (Dynamic Chart Generation Benchmark), the first benchmark evaluating MLLM's capability on dynamic chart generation tasks from three dimensions: Simple Text-to-Chart, Detailed Text-to-Chart, and Video-to-Chart tasks. We construct DCG-8K, a high-quality DCG dataset with annotations covering instruction-code-video triplets and QA pairs for both code and video evaluation. Based on DCG-8K, we explored a two-stage training recipe, proposing Joint-Code-Visual Reward for group relative policy optimization to construct expert MLLM Qwen2.5-VL-DCG-3B for the DCG task. Our benchmarking result reveals shortcomings of existing MLLMs in the visual-to-chart task, and our model beats the best open-sourced MLLM with an average 8.31% performance gain across three tasks, and shows on par performance against proprietary models with only 3B parameters, proving the effectiveness of our training recipe. Our code and dataset will be publicly available. --- ### Learning Pareto-Optimal Pandemic Intervention Policies with MORL Authors: Marian Chen, Miri Zilka Published: 2025-10-02 Categories: cs.LG, cs.AI, cs.CY, q-bio.PE arXiv: https://arxiv.org/abs/2510.03340v1 The COVID-19 pandemic underscored a critical need for intervention strategies that balance disease containment with socioeconomic stability. We approach this challenge by designing a framework for modeling and evaluating disease-spread prevention strategies. Our framework leverages multi-objective reinforcement learning (MORL) - a formulation necessitated by competing objectives - combined with a new stochastic differential equation (SDE) pandemic simulator, calibrated and validated against global COVID-19 data. Our simulator reproduces national-scale pandemic dynamics with orders of magnitude higher fidelity than other models commonly used in reinforcement learning (RL) approaches to pandemic intervention. Training a Pareto-Conditioned Network (PCN) agent on this simulator, we illustrate the direct policy trade-offs between epidemiological control and economic stability for COVID-19. Furthermore, we demonstrate the framework's generality by extending it to pathogens with different epidemiological profiles, such as polio and influenza, and show how these profiles lead the agent to discover fundamentally different intervention policies. To ground our work in contemporary policymaking challenges, we apply the model to measles outbreaks, quantifying how a modest 5% drop in vaccination coverage necessitates significantly more stringent and costly interventions to curb disease spread. This work provides a robust and adaptable framework to support transparent, evidence-based policymaking for mitigating public health crises. --- ### Pool Me Wisely: On the Effect of Pooling in Transformer-Based Models Authors: Sofiane Ennadir, Levente Zólyomi, Oleg Smirnov, Tianze Wang, John Pertoft, Filip Cornell, Lele Cao Published: 2025-10-02 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2510.03339v1 Transformer models have become the dominant backbone for sequence modeling, leveraging self-attention to produce contextualized token representations. These are typically aggregated into fixed-size vectors via pooling operations for downstream tasks. While much of the literature has focused on attention mechanisms, the role of pooling remains underexplored despite its critical impact on model behavior. In this paper, we introduce a theoretical framework that rigorously characterizes the expressivity of Transformer-based models equipped with widely used pooling methods by deriving closed-form bounds on their representational capacity and the ability to distinguish similar inputs. Our analysis extends to different variations of attention formulations, demonstrating that these bounds hold across diverse architectural variants. We empirically evaluate pooling strategies across tasks requiring both global and local contextual understanding, spanning three major modalities: computer vision, natural language processing, and time-series analysis. Results reveal consistent trends in how pooling choices affect accuracy, sensitivity, and optimization behavior. Our findings unify theoretical and empirical perspectives, providing practical guidance for selecting or designing pooling mechanisms suited to specific tasks. This work positions pooling as a key architectural component in Transformer models and lays the foundation for more principled model design beyond attention alone. --- ### Error correction in multiclass image classification of facial emotion on unbalanced samples Authors: Andrey A. Lebedev, Victor B. Kazantsev, Sergey V. Stasenko Published: 2025-10-02 Categories: cs.CV, q-bio.NC arXiv: https://arxiv.org/abs/2510.03337v1 This paper considers the problem of error correction in multi-class classification of face images on unbalanced samples. The study is based on the analysis of a data frame containing images labeled by seven different emotional states of people of different ages. Particular attention is paid to the problem of class imbalance, in which some emotions significantly prevail over others. To solve the classification problem, a neural network model based on LSTM with an attention mechanism focusing on key areas of the face that are informative for emotion recognition is used. As part of the experiments, the model is trained on all possible configurations of subsets of six classes with subsequent error correction for the seventh class, excluded at the training stage. The results show that correction is possible for all classes, although the degree of success varies: some classes are better restored, others are worse. In addition, on the test sample, when correcting some classes, an increase in key quality metrics for small classes was recorded, which indicates the promise of the proposed approach in solving applied problems related to the search for rare events, for example, in anti-fraud systems. Thus, the proposed method can be effectively applied in facial expression analysis systems and in tasks requiring stable classification under skewed class distribution. --- ### Linguistic and Audio Embedding-Based Machine Learning for Alzheimer's Dementia and Mild Cognitive Impairment Detection: Insights from the PROCESS Challenge Authors: Adharsha Sam Edwin Sam Devahi, Sohail Singh Sangha, Prachee Priyadarshinee, Jithin Thilakan, Ivan Fu Xing Tan, Christopher Johann Clarke, Sou Ka Lon, Balamurali B T, Yow Wei Quin, Chen Jer-Ming Published: 2025-10-02 Categories: cs.SD, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2510.03336v1 Early detection of Alzheimer's Dementia (AD) and Mild Cognitive Impairment (MCI) is critical for timely intervention, yet current diagnostic approaches remain resource-intensive and invasive. Speech, encompassing both acoustic and linguistic dimensions, offers a promising non-invasive biomarker for cognitive decline. In this study, we present a machine learning framework for the PROCESS Challenge, leveraging both audio embeddings and linguistic features derived from spontaneous speech recordings. Audio representations were extracted using Whisper embeddings from the Cookie Theft description task, while linguistic features-spanning pronoun usage, syntactic complexity, filler words, and clause structure-were obtained from transcriptions across Semantic Fluency, Phonemic Fluency, and Cookie Theft picture description. Classification models aimed to distinguish between Healthy Controls (HC), MCI, and AD participants, while regression models predicted Mini-Mental State Examination (MMSE) scores. Results demonstrated that voted ensemble models trained on concatenated linguistic features achieved the best classification performance (F1 = 0.497), while Whisper embedding-based ensemble regressors yielded the lowest MMSE prediction error (RMSE = 2.843). Comparative evaluation within the PROCESS Challenge placed our models among the top submissions in regression task, and mid-range for classification, highlighting the complementary strengths of linguistic and audio embeddings. These findings reinforce the potential of multimodal speech-based approaches for scalable, non-invasive cognitive assessment and underline the importance of integrating task-specific linguistic and acoustic markers in dementia detection. --- ### Matching the Optimal Denoiser in Point Cloud Diffusion with (Improved) Rotational Alignment Authors: Ameya Daigavane, YuQing Xie, Bodhi P. Vani, Saeed Saremi, Joseph Kleinhenz, Tess Smidt Published: 2025-10-02 Categories: cs.LG, eess.IV arXiv: https://arxiv.org/abs/2510.03335v1 Diffusion models are a popular class of generative models trained to reverse a noising process starting from a target data distribution. Training a diffusion model consists of learning how to denoise noisy samples at different noise levels. When training diffusion models for point clouds such as molecules and proteins, there is often no canonical orientation that can be assigned. To capture this symmetry, the true data samples are often augmented by transforming them with random rotations sampled uniformly over $SO(3)$. Then, the denoised predictions are often rotationally aligned via the Kabsch-Umeyama algorithm to the ground truth samples before computing the loss. However, the effect of this alignment step has not been well studied. Here, we show that the optimal denoiser can be expressed in terms of a matrix Fisher distribution over $SO(3)$. Alignment corresponds to sampling the mode of this distribution, and turns out to be the zeroth order approximation for small noise levels, explaining its effectiveness. We build on this perspective to derive better approximators to the optimal denoiser in the limit of small noise. Our experiments highlight that alignment is often a `good enough' approximation for the noise levels that matter most for training diffusion models. --- ### Semantic-Aware Scheduling for GPU Clusters with Large Language Models Authors: Zerui Wang, Qinghao Hu, Ana Klimovic, Tianwei Zhang, Yonggang Wen, Peng Sun, Dahua Lin Published: 2025-10-02 Categories: cs.LG, cs.DC arXiv: https://arxiv.org/abs/2510.03334v1 Deep learning (DL) schedulers are pivotal in optimizing resource allocation in GPU clusters, but operate with a critical limitation: they are largely blind to the semantic context of the jobs they manage. This forces them to rely on limited metadata, leading to high profiling overhead, unreliable duration estimation, inadequate failure handling, and poor observability. To this end, we propose SchedMate, a framework that bridges this semantic gap by systematically extracting deep insights from overlooked, unstructured data sources: source code, runtime logs, and historical jobs. SchedMate enhances existing schedulers non-intrusively through three LLM-based components. Our implementation integrates seamlessly with existing deep learning schedulers. Evaluations on a 128-GPU physical cluster and extensive simulations on production traces show SchedMate reduces average job completion times by up to 1.91x, substantially enhancing the scheduling performance, demonstrating the critical role of semantic-awareness in modern DL scheduling. --- ### Input-Aware Sparse Attention for Real-Time Co-Speech Video Generation Authors: Beijia Lu, Ziyi Chen, Jing Xiao, Jun-Yan Zhu Published: 2025-10-02 Categories: cs.CV arXiv: https://arxiv.org/abs/2510.02617v1 Diffusion models can synthesize realistic co-speech video from audio for various applications, such as video creation and virtual agents. However, existing diffusion-based methods are slow due to numerous denoising steps and costly attention mechanisms, preventing real-time deployment. In this work, we distill a many-step diffusion video model into a few-step student model. Unfortunately, directly applying recent diffusion distillation methods degrades video quality and falls short of real-time performance. To address these issues, our new video distillation method leverages input human pose conditioning for both attention and loss functions. We first propose using accurate correspondence between input human pose keypoints to guide attention to relevant regions, such as the speaker's face, hands, and upper body. This input-aware sparse attention reduces redundant computations and strengthens temporal correspondences of body parts, improving inference efficiency and motion coherence. To further enhance visual quality, we introduce an input-aware distillation loss that improves lip synchronization and hand motion realism. By integrating our input-aware sparse attention and distillation loss, our method achieves real-time performance with improved visual quality compared to recent audio-driven and input-driven methods. We also conduct extensive experiments showing the effectiveness of our algorithmic design choices. --- ### On the Role of Temperature Sampling in Test-Time Scaling Authors: Yuheng Wu, Azalia Mirhoseini, Thierry Tambe Published: 2025-10-02 Categories: cs.AI, cs.CL, cs.LG arXiv: https://arxiv.org/abs/2510.02611v1 Large language models (LLMs) can improve reasoning at inference time through test-time scaling (TTS), where multiple reasoning traces are generated and the best one is selected. Prior work shows that increasing the number of samples K steadily improves accuracy. In this paper, we demonstrate that this trend does not hold indefinitely: at large K, further scaling yields no gains, and certain hard questions remain unsolved regardless of the number of traces. Interestingly, we find that different sampling temperatures solve different subsets of problems, implying that single-temperature scaling explores only part of a model's potential. We therefore propose scaling along the temperature dimension, which enlarges the reasoning boundary of LLMs. Averaged over Qwen3 (0.6B, 1.7B, 4B, 8B) and five representative reasoning benchmarks (AIME 2024/2025, MATH500, LiveCodeBench, Hi-ToM), temperature scaling yields an additional 7.3 points over single-temperature TTS. Temperature scaling also enables base models to reach performance comparable to reinforcement learning (RL)-trained counterparts, without additional post-training. We further provide a comprehensive analysis of this phenomenon and design a multi-temperature voting method that reduces the overhead of temperature scaling. Overall, our findings suggest that TTS is more powerful than previously thought, and that temperature scaling offers a simple and effective way to unlock the latent potential of base models. --- ### MINERVA: Mutual Information Neural Estimation for Supervised Feature Selection Authors: Taurai Muvunza, Egor Kraev, Pere Planell-Morell, Alexander Y. Shestopaloff Published: 2025-10-02 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2510.02610v2 Existing feature filters rely on statistical pair-wise dependence metrics to model feature-target relationships, but this approach may fail when the target depends on higher-order feature interactions rather than individual contributions. We introduce Mutual Information Neural Estimation Regularized Vetting Algorithm (MINERVA), a novel approach to supervised feature selection based on neural estimation of mutual information between features and targets. We paramaterize the approximation of mutual information with neural networks and perform feature selection using a carefully designed loss function augmented with sparsity-inducing regularizers. Our method is implemented in a two-stage process to decouple representation learning from feature selection, ensuring better generalization and a more accurate expression of feature importance. We present examples of ubiquitous dependency structures that are rarely captured in literature and show that our proposed method effectively captures these complex feature-target relationships by evaluating feature subsets as an ensemble. Experimental results on synthetic and real-life fraud datasets demonstrate the efficacy of our method and its ability to perform exact solutions. --- ### Mitigating Modal Imbalance in Multimodal Reasoning Authors: Chen Henry Wu, Neil Kale, Aditi Raghunathan Published: 2025-10-02 Categories: cs.AI arXiv: https://arxiv.org/abs/2510.02608v2 Foundation models (FMs) deployed in real-world tasks such as computer-use agents must integrate diverse modalities. How good are FMs at performing joint reasoning, simultaneously reasoning over multiple modalities, especially when the modalities interact and relate to each other to form cross-modal context? To better understand this problem, we study FMs on cross-modal conflicts: scenarios where conflicting evidence is presented across modalities. This allows us to examine whether FMs prioritize one modality over another or reason jointly to reconcile the conflict. Our experiments reveal that FMs can recognize conflicts in unimodal contexts, composed of a single modality, 90% of the time, but the ratio falls as low as 3% when evidence is split across modalities -- similar observations hold in cross-lingual contexts, composed of multiple languages. We trace this failure to cross-modal attention imbalance, showing that FMs exhibit extreme asymmetry in attention scores, disproportionately prioritizing certain modalities. We show that cross-modal attention imbalance does not go away by simply scaling up multimodal or multilingual datasets blindly, since they lack training examples that explicitly require cross-modal reasoning. We demonstrate that even a simple and scalable method of explicitly combining multiple modalities within each training instance significantly reduces attention imbalance. Reduced attention imbalance directly translates to improved downstream performance on several vision-language benchmarks. Our findings underscore the importance of systematically addressing cross-modal contexts to build reliable foundation models. --- ### Towards CONUS-Wide ML-Augmented Conceptually-Interpretable Modeling of Catchment-Scale Precipitation-Storage-Runoff Dynamics Authors: Yuan-Heng Wang, Yang Yang, Fabio Ciulla, Hoshin V. Gupta, Charuleka Varadharajan Published: 2025-10-02 Categories: cs.LG arXiv: https://arxiv.org/abs/2510.02605v2 While many modern studies are dedicated to ML-based large-sample hydrologic modeling, these efforts have not necessarily translated into predictive improvements that are grounded in enhanced physical-conceptual understanding. Here, we report on a CONUS-wide large-sample study (spanning diverse hydro-geo-climatic conditions) using ML-augmented physically-interpretable catchment-scale models of varying complexity based in the Mass-Conserving Perceptron (MCP). Results were evaluated using attribute masks such as snow regime, forest cover, and climate zone. Our results indicate the importance of selecting model architectures of appropriate model complexity based on how process dominance varies with hydrological regime. Benchmark comparisons show that physically-interpretable mass-conserving MCP-based models can achieve performance comparable to data-based models based in the Long Short-Term Memory network (LSTM) architecture. Overall, this study highlights the potential of a theory-informed, physically grounded approach to large-sample hydrology, with emphasis on mechanistic understanding and the development of parsimonious and interpretable model architectures, thereby laying the foundation for future models of everywhere that architecturally encode information about spatially- and temporally-varying process dominance. --- ### Ego-Exo 3D Hand Tracking in the Wild with a Mobile Multi-Camera Rig Authors: Patrick Rim, Kun He, Kevin Harris, Braden Copple, Shangchen Han, Sizhe An, Ivan Shugurov, Tomas Hodan, He Wen, Xu Xie Published: 2025-10-02 Categories: cs.CV arXiv: https://arxiv.org/abs/2510.02601v1 Accurate 3D tracking of hands and their interactions with the world in unconstrained settings remains a significant challenge for egocentric computer vision. With few exceptions, existing datasets are predominantly captured in controlled lab setups, limiting environmental diversity and model generalization. To address this, we introduce a novel marker-less multi-camera system designed to capture precise 3D hands and objects, which allows for nearly unconstrained mobility in genuinely in-the-wild conditions. We combine a lightweight, back-mounted capture rig with eight exocentric cameras, and a user-worn Meta Quest 3 headset, which contributes two egocentric views. We design an ego-exo tracking pipeline to generate accurate 3D hand pose ground truth from this system, and rigorously evaluate its quality. By collecting an annotated dataset featuring synchronized multi-view images and precise 3D hand poses, we demonstrate the capability of our approach to significantly reduce the trade-off between environmental realism and 3D annotation accuracy. --- ### PEO: Training-Free Aesthetic Quality Enhancement in Pre-Trained Text-to-Image Diffusion Models with Prompt Embedding Optimization Authors: Hovhannes Margaryan, Bo Wan, Tinne Tuytelaars Published: 2025-10-02 Categories: cs.CV arXiv: https://arxiv.org/abs/2510.02599v1 This paper introduces a novel approach to aesthetic quality improvement in pre-trained text-to-image diffusion models when given a simple prompt. Our method, dubbed Prompt Embedding Optimization (PEO), leverages a pre-trained text-to-image diffusion model as a backbone and optimizes the text embedding of a given simple and uncurated prompt to enhance the visual quality of the generated image. We achieve this by a tripartite objective function that improves the aesthetic fidelity of the generated image, ensures adherence to the optimized text embedding, and minimal divergence from the initial prompt. The latter is accomplished through a prompt preservation term. Additionally, PEO is training-free and backbone-independent. Quantitative and qualitative evaluations confirm the effectiveness of the proposed method, exceeding or equating the performance of state-of-the-art text-to-image and prompt adaptation methods. --- ### Multimodal Large Language Model Framework for Safe and Interpretable Grid-Integrated EVs Authors: Jean Douglas Carvalho, Hugo Kenji, Ahmad Mohammad Saber, Glaucia Melo, Max Mauro Dias Santos, Deepa Kundur Published: 2025-10-02 Categories: cs.AI arXiv: https://arxiv.org/abs/2510.02592v1 The integration of electric vehicles (EVs) into smart grids presents unique opportunities to enhance both transportation systems and energy networks. However, ensuring safe and interpretable interactions between drivers, vehicles, and the surrounding environment remains a critical challenge. This paper presents a multi-modal large language model (LLM)-based framework to process multimodal sensor data - such as object detection, semantic segmentation, and vehicular telemetry - and generate natural-language alerts for drivers. The framework is validated using real-world data collected from instrumented vehicles driving on urban roads, ensuring its applicability to real-world scenarios. By combining visual perception (YOLOv8), geocoded positioning, and CAN bus telemetry, the framework bridges raw sensor data and driver comprehension, enabling safer and more informed decision-making in urban driving scenarios. Case studies using real data demonstrate the framework's effectiveness in generating context-aware alerts for critical situations, such as proximity to pedestrians, cyclists, and other vehicles. This paper highlights the potential of LLMs as assistive tools in e-mobility, benefiting both transportation systems and electric networks by enabling scalable fleet coordination, EV load forecasting, and traffic-aware energy planning. Index Terms - Electric vehicles, visual perception, large language models, YOLOv8, semantic segmentation, CAN bus, prompt engineering, smart grid. --- ### Use the Online Network If You Can: Towards Fast and Stable Reinforcement Learning Authors: Ahmed Hendawy, Henrik Metternich, Théo Vincent, Mahdi Kallel, Jan Peters, Carlo D'Eramo Published: 2025-10-02 Categories: cs.LG arXiv: https://arxiv.org/abs/2510.02590v2 The use of target networks is a popular approach for estimating value functions in deep Reinforcement Learning (RL). While effective, the target network remains a compromise solution that preserves stability at the cost of slowly moving targets, thus delaying learning. Conversely, using the online network as a bootstrapped target is intuitively appealing, albeit well-known to lead to unstable learning. In this work, we aim to obtain the best out of both worlds by introducing a novel update rule that computes the target using the MINimum estimate between the Target and Online network, giving rise to our method, MINTO. Through this simple, yet effective modification, we show that MINTO enables faster and stable value function learning, by mitigating the potential overestimation bias of using the online network for bootstrapping. Notably, MINTO can be seamlessly integrated into a wide range of value-based and actor-critic algorithms with a negligible cost. We evaluate MINTO extensively across diverse benchmarks, spanning online and offline RL, as well as discrete and continuous action spaces. Across all benchmarks, MINTO consistently improves performance, demonstrating its broad applicability and effectiveness. --- ### A Benchmark Study of Deep Reinforcement Learning Algorithms for the Container Stowage Planning Problem Authors: Yunqi Huang, Nishith Chennakeshava, Alexis Carras, Vladislav Neverov, Wei Liu, Aske Plaat, Yingjie Fan Published: 2025-10-02 Categories: cs.AI arXiv: https://arxiv.org/abs/2510.02589v1 Container stowage planning (CSPP) is a critical component of maritime transportation and terminal operations, directly affecting supply chain efficiency. Owing to its complexity, CSPP has traditionally relied on human expertise. While reinforcement learning (RL) has recently been applied to CSPP, systematic benchmark comparisons across different algorithms remain limited. To address this gap, we develop a Gym environment that captures the fundamental features of CSPP and extend it to include crane scheduling in both multi-agent and single-agent formulations. Within this framework, we evaluate five RL algorithms: DQN, QR-DQN, A2C, PPO, and TRPO under multiple scenarios of varying complexity. The results reveal distinct performance gaps with increasing complexity, underscoring the importance of algorithm choice and problem formulation for CSPP. Overall, this paper benchmarks multiple RL methods for CSPP while providing a reusable Gym environment with crane scheduling, thus offering a foundation for future research and practical deployment in maritime logistics. --- ### FLOWR.root: A flow matching based foundation model for joint multi-purpose structure-aware 3D ligand generation and affinity prediction Authors: Julian Cremer, Tuan Le, Mohammad M. Ghahremanpour, Emilia Sługocka, Filipe Menezes, Djork-Arné Clevert Published: 2025-10-02 Categories: q-bio.BM, cs.LG arXiv: https://arxiv.org/abs/2510.02578v6 We present FLOWR.root, an SE(3)-equivariant flow-matching model for pocket-aware 3D ligand generation with joint potency and binding affinity prediction and confidence estimation. The model supports de novo generation, interaction- and pharmacophore-conditional sampling, fragment elaboration and replacement, and multi-endpoint affinity prediction (pIC50, pKi, pKd, pEC50). Training combines large-scale ligand libraries with mixed-fidelity protein-ligand complexes, refined on curated co-crystal datasets and adapted to project-specific data through parameter-efficient finetuning. The base FLOWR.root model achieves state-of-the-art performance in unconditional 3D molecule and pocket-conditional ligand generation. On HiQBind, the pre-trained and finetuned model demonstrates highly accurate affinity predictions, and outperforms recent state-of-the-art methods such as Boltz-2 on the FEP+/OpenFE benchmark with substantial speed advantages. However, we show that addressing unseen structure-activity landscapes requires domain adaptation; parameter-efficient LoRA finetuning yields marked improvements on diverse proprietary datasets and PDE10A. Joint generation and affinity prediction enable inference-time scaling through importance sampling, steering design toward higher-affinity compounds. Case studies validate this: selective CK2$α$ ligand generation against CLK3 shows significant correlation between predicted and quantum-mechanical binding energies. Scaffold elaboration on ER$α$, TYK2, and BACE1 demonstrates strong agreement between predicted affinities and QM calculations while confirming geometric fidelity. By integrating structure-aware generation, affinity estimation, property-guided sampling, and efficient domain adaptation, FLOWR.root provides a comprehensive foundation for structure-based drug design from hit identification through lead optimization. --- ### Geospatial Machine Learning Libraries Authors: Adam J. Stewart, Caleb Robinson, Arindam Banerjee Published: 2025-10-02 Categories: cs.LG, cs.SE arXiv: https://arxiv.org/abs/2510.02572v2 Recent advances in machine learning have been supported by the emergence of domain-specific software libraries, enabling streamlined workflows and increased reproducibility. For geospatial machine learning (GeoML), the availability of Earth observation data has outpaced the development of domain libraries to handle its unique challenges, such as varying spatial resolutions, spectral properties, temporal cadence, data coverage, coordinate systems, and file formats. This chapter presents a comprehensive overview of GeoML libraries, analyzing their evolution, core functionalities, and the current ecosystem. It also introduces popular GeoML libraries such as TorchGeo, eo-learn, and Raster Vision, detailing their architecture, supported data types, and integration with ML frameworks. Additionally, it discusses common methodologies for data preprocessing, spatial--temporal joins, benchmarking, and the use of pretrained models. Through a case study in crop type mapping, it demonstrates practical applications of these tools. Best practices in software design, licensing, and testing are highlighted, along with open challenges and future directions, particularly the rise of foundation models and the need for governance in open-source geospatial software. Our aim is to guide practitioners, developers, and researchers in navigating and contributing to the rapidly evolving GeoML landscape. --- ### How Confident are Video Models? Empowering Video Models to Express their Uncertainty Authors: Zhiting Mei, Ola Shorinwa, Anirudha Majumdar Published: 2025-10-02 Categories: cs.CV, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2510.02571v1 Generative video models demonstrate impressive text-to-video capabilities, spurring widespread adoption in many real-world applications. However, like large language models (LLMs), video generation models tend to hallucinate, producing plausible videos even when they are factually wrong. Although uncertainty quantification (UQ) of LLMs has been extensively studied in prior work, no UQ method for video models exists, raising critical safety concerns. To our knowledge, this paper represents the first work towards quantifying the uncertainty of video models. We present a framework for uncertainty quantification of generative video models, consisting of: (i) a metric for evaluating the calibration of video models based on robust rank correlation estimation with no stringent modeling assumptions; (ii) a black-box UQ method for video models (termed S-QUBED), which leverages latent modeling to rigorously decompose predictive uncertainty into its aleatoric and epistemic components; and (iii) a UQ dataset to facilitate benchmarking calibration in video models. By conditioning the generation task in the latent space, we disentangle uncertainty arising due to vague task specifications from that arising from lack of knowledge. Through extensive experiments on benchmark video datasets, we demonstrate that S-QUBED computes calibrated total uncertainty estimates that are negatively correlated with the task accuracy and effectively computes the aleatoric and epistemic constituents. ---