# Research Papers — 2025-10 · bundle page 175 of 217 > Full-text markdown of papers #8701–8750 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/175 [← Previous bundle](https://aipod.dev/papers/archive/2025-10/page/174.md) [Next bundle →](https://aipod.dev/papers/archive/2025-10/page/176.md) ### Verifier-free Test-Time Sampling for Vision-Language-Action Models Authors: Suhyeok Jang, Dongyoung Kim, Changyeon Kim, Youngsuk Kim, Jinwoo Shin Published: 2025-10-07 Categories: cs.RO, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2510.05681v2 Vision-Language-Action models (VLAs) have demonstrated remarkable performance in robot control. However, they remain fundamentally limited in tasks that require high precision due to their single-inference paradigm. While test-time scaling approaches using external verifiers have shown promise, they require additional training and fail to generalize to unseen conditions. We propose Masking Distribution Guided Selection (MG-Select), a novel test-time scaling framework for VLAs that leverages the model's internal properties without requiring additional training or external modules. Our approach utilizes KL divergence from a reference action token distribution as a confidence metric for selecting the optimal action from multiple candidates. We introduce a reference distribution generated by the same VLA but with randomly masked states and language conditions as inputs, providing action uncertainty while remaining aligned with the target task distribution. Additionally, we propose a joint training strategy that enables the model to learn both conditional and unconditional distributions by applying dropout to state and language conditions, thereby further improving the quality of the reference distribution. Our experiments demonstrate that MG-Select provides a reliable reference for action selection through task-relevant condition masking and consistently improves base models across diverse simulation and real-world benchmarks. --- ### Gradual Code-Switching as Inference-Time Cross-Lingual Representational Alignment for LLMs Authors: Haneul Yoo, Jiho Jin, Kyunghyun Cho, Alice Oh Published: 2025-10-07 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2510.05678v2 While large language models (LLMs) have achieved notable progress in multilingual settings, their performance remains uneven across languages as LLMs often rely on English-centric latent representations. In this work, we introduce code-switching in-context learning (CSICL), an inference-time mechanism for cross-lingual representational alignment. Rather than relying on an abrupt translation pivot, CSICL explicitly scaffolds the reasoning trajectory by gradually transitioning from a target language to English, aligning non-English inputs with an English-centric reasoning space. Across 4 LLMs, 6 datasets, and 10 languages, CSICL consistently outperforms cross-lingual in-context learning baselines, yielding average gains of 6.0pp and 4.8pp in target and unseen languages, respectively. The improvements generalize across language families and are even more pronounced in low-resource settings, with gains of 14.7pp in target and 5.3pp in unseen languages. These findings establish code-switching as a robust and effective approach for reducing cross-lingual misalignment during inference, moving LLMs towards more equitable and effective multilingual systems. --- ### Inductive inference of gradient-boosted decision trees on graphs for insurance fraud detection Authors: Félix Vandervorst, Bruno Deprez, Wouter Verbeke, Tim Verdonck Published: 2025-10-07 Categories: cs.LG, cs.SI arXiv: https://arxiv.org/abs/2510.05676v2 Graph-based methods are becoming increasingly popular in machine learning due to their ability to model complex data and relations. Insurance fraud is a prime use case, since fraudulent claims are often the result of organised criminals that stage accidents or the same persons filing erroneous claims on multiple policies. One challenge is that graph-based approaches struggle to find meaningful representations of the data because of the high class imbalance present in fraud data. In addition, insurance graphs are heterogeneous and dynamic, given the changing relations among people, companies and policies. As a result, gradient-boosted tree approaches on tabular data still dominate the field. Therefore, we present a novel inductive graph gradient boosting machine (G-GBM) for supervised learning on heterogeneous and dynamic graphs. G-GBM combines the class-imbalance robustness of gradient boosting with heterogeneous graph information encoded through interpretable path-level feature concatenations, while preserving access to the original tabular feature space. In addition, the explicit representation of neighbourhood information enables transparent SHAP-based explanations at the metapath and feature level. We demonstrate G-GBM for insurance fraud detection on an open-source and a real-world, proprietary dataset, and find that G-GBM performs on par or better than the state-of-the-art. The associated insurance fraud dataset is publicly released to facilitate reproducibility. --- ### Context Matters: Learning Global Semantics via Object-Centric Representation Authors: Jike Zhong, Yuxiang Lai, Xiaofeng Yang, Konstantinos Psounis Published: 2025-10-07 Categories: cs.CV arXiv: https://arxiv.org/abs/2510.05674v2 Recent advances in language modeling have witnessed the rise of highly desirable emergent capabilities, such as reasoning and in-context learning. However, vision models have yet to exhibit comparable progress in these areas. In this paper, we argue that this gap could stem from the lack of semantic and contextual guidance in current vision transformer (ViT) training schemes, and such a gap can be narrowed through the design of a semantic-grounded objective. Specifically, we notice that individual words in natural language are inherently semantic, and modeling directly on word tokens naturally learns a realistic distribution. In contrast, ViTs rely on spatial patchification, which inevitably lacks semantic information. To bridge this gap, we propose to directly model "object" as the visual equivalence of "word," pushing the model to learn the global context and semantics among visual elements. We investigate our hypotheses via masked image modeling (MIM), a framework where our approach can be readily tested by applying masks to visual objects rather than random patches. Considerable evidence from qualitative and quantitative evaluations reveals a key finding: object-level representation alone helps to learn a real-world distribution, whereas pixel-averaging shortcuts are often learned without it. Moreover, further evaluations with multimodal LLMs (MLLM) on visual question answering (VQA, GQA, ScienceQA) tasks demonstrate the strong reasoning and contextual understanding gained with this simple objective. We hope our study highlights the effectiveness of object-level encoding and provides a plausible direction for developing stronger vision encoders and tokenizers. Code and model will be publicly released. Keywords: Semantic Visual Tokenizer, Vision Reasoning, In-context Learning, Multimodal Reasoning --- ### Quantifying the Accuracy-Interpretability Trade-Off in Concept-Based Sidechannel Models Authors: David Debot, Giuseppe Marra Published: 2025-10-07 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2510.05670v2 Concept Bottleneck Models (CBNMs) are deep learning models that provide interpretability by enforcing a bottleneck layer where predictions are based exclusively on human-understandable concepts. However, this constraint also restricts information flow and often results in reduced predictive accuracy. Concept Sidechannel Models (CSMs) address this limitation by introducing a sidechannel that bypasses the bottleneck and carry additional task-relevant information. While this improves accuracy, it simultaneously compromises interpretability, as predictions may rely on uninterpretable representations transmitted through sidechannels. Currently, there exists no principled technique to control this fundamental trade-off. In this paper, we close this gap. First, we present a unified probabilistic concept sidechannel meta-model that subsumes existing CSMs as special cases. Building on this framework, we introduce the Sidechannel Independence Score (SIS), a metric that quantifies a CSM's reliance on its sidechannel by contrasting predictions made with and without sidechannel information. We propose SIS regularization, which explicitly penalizes sidechannel reliance to improve interpretability. Finally, we analyze how the expressivity of the predictor and the reliance of the sidechannel jointly shape interpretability, revealing inherent trade-offs across different CSM architectures. Empirical results show that state-of-the-art CSMs, when trained solely for accuracy, exhibit low representation interpretability, and that SIS regularization substantially improves their interpretability, intervenability, and the quality of learned interpretable task predictors. Our work provides both theoretical and practical tools for developing CSMs that balance accuracy and interpretability in a principled manner. --- ### Development and Validation of a Low-Cost Imaging System for Seedling Germination Kinetics through Time-Cumulative Analysis Authors: M. Torrente, A. Follador, A. Calcante, P. Casati, R. Oberti Published: 2025-10-07 Categories: cs.CV arXiv: https://arxiv.org/abs/2510.05668v1 The study investigates the effects of R. solani inoculation on the germination and early development of Lactuca sativa L. seeds using a low-cost, image-based monitoring system. Multiple cameras were deployed to continuously capture images of the germination process in both infected and control groups. The objective was to assess the impact of the pathogen by analyzing germination dynamics and growth over time. To achieve this, a novel image analysis pipeline was developed. The algorithm integrates both morphological and spatial features to identify and quantify individual seedlings, even under complex conditions where traditional image analyses fails. A key innovation of the method lies in its temporal integration: each analysis step considers not only the current status but also their developmental across prior time points. This approach enables robust discrimination of individual seedlings, especially when overlapping leaves significantly hinder object separation. The method demonstrated high accuracy in seedling counting and vigor assessment, even in challenging scenarios characterized by dense and intertwined growth. Results confirm that R. solani infection significantly reduces germination rates and early seedling vigor. The study also validates the feasibility of combining low-cost imaging hardware with advanced computational tools to obtain phenotyping data in a non-destructive and scalable manner. The temporal integration enabled accurate quantification of germinated seeds and precise determination of seedling emergence timing. This approach proved particularly effective in later stages of the experiment, where conventional segmentation techniques failed due to overlapping or intertwined seedlings, making accurate counting. The method achieved a coefficient of determination of 0.98 and a root mean square error (RMSE) of 1.12, demonstrating its robustness and reliability. --- ### Large Language Model-Based Uncertainty-Adjusted Label Extraction for Artificial Intelligence Model Development in Upper Extremity Radiography Authors: Hanna Kreutzer, Anne-Sophie Caselitz, Thomas Dratsch, Daniel Pinto dos Santos, Christiane Kuhl, Daniel Truhn, Sven Nebelung Published: 2025-10-07 Categories: cs.AI arXiv: https://arxiv.org/abs/2510.05664v1 Objectives: To evaluate GPT-4o's ability to extract diagnostic labels (with uncertainty) from free-text radiology reports and to test how these labels affect multi-label image classification of musculoskeletal radiographs. Methods: This retrospective study included radiography series of the clavicle (n=1,170), elbow (n=3,755), and thumb (n=1,978). After anonymization, GPT-4o filled out structured templates by indicating imaging findings as present ("true"), absent ("false"), or "uncertain." To assess the impact of label uncertainty, "uncertain" labels of the training and validation sets were automatically reassigned to "true" (inclusive) or "false" (exclusive). Label-image-pairs were used for multi-label classification using ResNet50. Label extraction accuracy was manually verified on internal (clavicle: n=233, elbow: n=745, thumb: n=393) and external test sets (n=300 for each). Performance was assessed using macro-averaged receiver operating characteristic (ROC) area under the curve (AUC), precision recall curves, sensitivity, specificity, and accuracy. AUCs were compared with the DeLong test. Results: Automatic extraction was correct in 98.6% (60,618 of 61,488) of labels in the test sets. Across anatomic regions, label-based model training yielded competitive performance measured by macro-averaged AUC values for inclusive (e.g., elbow: AUC=0.80 (range, 0.62-0.87)) and exclusive models (elbow: AUC=0.80 (range, 0.61-0.88)). Models generalized well on external datasets (elbow (inclusive): AUC=0.79 (range, 0.61-0.87); elbow (exclusive): AUC=0.79 (range, 0.63-0.89)). No significant differences were observed across labeling strategies or datasets (p>=0.15). Conclusion: GPT-4o extracted labels from radiologic reports to train competitive multi-label classification models with high accuracy. Detected uncertainty in the radiologic reports did not influence the performance of these models. --- ### DeLTa: Demonstration and Language-Guided Novel Transparent Object Manipulation Authors: Taeyeop Lee, Gyuree Kang, Bowen Wen, Youngho Kim, Seunghyeok Back, In So Kweon, David Hyunchul Shim, Kuk-Jin Yoon Published: 2025-10-07 Categories: cs.RO, cs.CV arXiv: https://arxiv.org/abs/2510.05662v2 Despite the prevalence of transparent object interactions in human everyday life, transparent robotic manipulation research remains limited to short-horizon tasks and basic grasping capabilities. Although some methods have partially addressed these issues, most of them have limitations in generalization to novel objects and are insufficient for precise long-horizon robot manipulation. To address this limitation, we propose DeLTa (Demonstration and Language-Guided Novel Transparent Object Manipulation), a novel framework that integrates depth estimation, 6D pose estimation, and vision-language planning for precise long-horizon manipulation of transparent objects guided by natural language task instructions. A key advantage of our method is its single-demonstration approach, which generalizes 6D trajectories to novel transparent objects without requiring category-level priors or additional training. Additionally, we present a task planner that refines the VLM-generated plan to account for the constraints of a single-arm, eye-in-hand robot for long-horizon object manipulation tasks. Through comprehensive evaluation, we demonstrate that our method significantly outperforms existing transparent object manipulation approaches, particularly in long-horizon scenarios requiring precise manipulation capabilities. Project page: https://sites.google.com/view/DeLTa25/ --- ### When and How to Cut Classical Concerts? A Multimodal Automated Video Editing Approach Authors: Daniel Gonzálbez-Biosca, Josep Cabacas-Maso, Carles Ventura, Ismael Benito-Altamirano Published: 2025-10-07 Categories: cs.CV, cs.MM arXiv: https://arxiv.org/abs/2510.05661v1 Automated video editing remains an underexplored task in the computer vision and multimedia domains, especially when contrasted with the growing interest in video generation and scene understanding. In this work, we address the specific challenge of editing multicamera recordings of classical music concerts by decomposing the problem into two key sub-tasks: when to cut and how to cut. Building on recent literature, we propose a novel multimodal architecture for the temporal segmentation task (when to cut), which integrates log-mel spectrograms from the audio signals, plus an optional image embedding, and scalar temporal features through a lightweight convolutional-transformer pipeline. For the spatial selection task (how to cut), we improve the literature by updating from old backbones, e.g. ResNet, with a CLIP-based encoder and constraining distractor selection to segments from the same concert. Our dataset was constructed following a pseudo-labeling approach, in which raw video data was automatically clustered into coherent shot segments. We show that our models outperformed previous baselines in detecting cut points and provide competitive visual shot selection, advancing the state of the art in multimodal automated video editing. --- ### Teleportraits: Training-Free People Insertion into Any Scene Authors: Jialu Gao, K J Joseph, Fernando De La Torre Published: 2025-10-07 Categories: cs.CV arXiv: https://arxiv.org/abs/2510.05660v1 The task of realistically inserting a human from a reference image into a background scene is highly challenging, requiring the model to (1) determine the correct location and poses of the person and (2) perform high-quality personalization conditioned on the background. Previous approaches often treat them as separate problems, overlooking their interconnections, and typically rely on training to achieve high performance. In this work, we introduce a unified training-free pipeline that leverages pre-trained text-to-image diffusion models. We show that diffusion models inherently possess the knowledge to place people in complex scenes without requiring task-specific training. By combining inversion techniques with classifier-free guidance, our method achieves affordance-aware global editing, seamlessly inserting people into scenes. Furthermore, our proposed mask-guided self-attention mechanism ensures high-quality personalization, preserving the subject's identity, clothing, and body features from just a single reference image. To the best of our knowledge, we are the first to perform realistic human insertions into scenes in a training-free manner and achieve state-of-the-art results in diverse composite scene images with excellent identity preservation in backgrounds and subjects. --- ### A Hierarchical Geometry-guided Transformer for Histological Subtyping of Primary Liver Cancer Authors: Anwen Lu, Mingxin Liu, Yiping Jiao, Hongyi Gong, Geyang Xu, Jun Chen, Jun Xu Published: 2025-10-07 Categories: cs.CV arXiv: https://arxiv.org/abs/2510.05657v1 Primary liver malignancies are widely recognized as the most heterogeneous and prognostically diverse cancers of the digestive system. Among these, hepatocellular carcinoma (HCC) and intrahepatic cholangiocarcinoma (ICC) emerge as the two principal histological subtypes, demonstrating significantly greater complexity in tissue morphology and cellular architecture than other common tumors. The intricate representation of features in Whole Slide Images (WSIs) encompasses abundant crucial information for liver cancer histological subtyping, regarding hierarchical pyramid structure, tumor microenvironment (TME), and geometric representation. However, recent approaches have not adequately exploited these indispensable effective descriptors, resulting in a limited understanding of histological representation and suboptimal subtyping performance. To mitigate these limitations, ARGUS is proposed to advance histological subtyping in liver cancer by capturing the macro-meso-micro hierarchical information within the TME. Specifically, we first construct a micro-geometry feature to represent fine-grained cell-level pattern via a geometric structure across nuclei, thereby providing a more refined and precise perspective for delineating pathological images. Then, a Hierarchical Field-of-Views (FoVs) Alignment module is designed to model macro- and meso-level hierarchical interactions inherent in WSIs. Finally, the augmented micro-geometry and FoVs features are fused into a joint representation via present Geometry Prior Guided Fusion strategy for modeling holistic phenotype interactions. Extensive experiments on public and private cohorts demonstrate that our ARGUS achieves state-of-the-art (SOTA) performance in histological subtyping of liver cancer, which provide an effective diagnostic tool for primary liver malignancies in clinical practice. --- ### SD-MVSum: Script-Driven Multimodal Video Summarization Method and Datasets Authors: Manolis Mylonas, Charalampia Zerva, Evlampios Apostolidis, Vasileios Mezaris Published: 2025-10-07 Categories: cs.CV arXiv: https://arxiv.org/abs/2510.05652v2 In this work, we present a method and two large-scale datasets for Script-Driven Multimodal Video Summarization. The proposed method, SD-MVSum, builds on our earlier SD-VSum method for script-driven video summarization, which considered just the visual content of the video. SD-MVSum takes into account, in addition to the visual modality, the relevance of the user-provided script with the spoken content (i.e., audio transcript) of the video. The dependence between each considered pair of data modalities, i.e., script-video and script-transcript, is modeled using a new weighted cross-modal attention mechanism. This mechanism explicitly exploits the semantic similarity between the paired modalities in order to promote the parts of the full-length video with the highest relevance to the user-provided script. Furthermore, we extend two large-scale datasets for script-driven (S-VideoXum) and generic (MrHiSum) video summarization, to make them suitable for training and evaluation of script-driven multimodal video summarization methods. Experimental comparisons document the competitiveness of the proposed SD-MVSum method against other SotA approaches for script-driven and generic video summarization. Our new method and extended datasets are available at: https://github.com/IDT-ITI/SD-MVSum. --- ### EduVerse: A User-Defined Multi-Agent Simulation Space for Education Scenario Authors: Yiping Ma, Shiyu Hu, Buyuan Zhu, Yipei Wang, Yaxuan Kang, Shiqing Liu, Kang Hao Cheong Published: 2025-10-07 Categories: cs.CV, cs.CY arXiv: https://arxiv.org/abs/2510.05650v1 Reproducing cognitive development, group interaction, and long-term evolution in virtual classrooms remains a core challenge for educational AI, as real classrooms integrate open-ended cognition, dynamic social interaction, affective factors, and multi-session development rarely captured together. Existing approaches mostly focus on short-term or single-agent settings, limiting systematic study of classroom complexity and cross-task reuse. We present EduVerse, the first user-defined multi-agent simulation space that supports environment, agent, and session customization. A distinctive human-in-the-loop interface further allows real users to join the space. Built on a layered CIE (Cognition-Interaction-Evolution) architecture, EduVerse ensures individual consistency, authentic interaction, and longitudinal adaptation in cognition, emotion, and behavior-reproducing realistic classroom dynamics with seamless human-agent integration. We validate EduVerse in middle-school Chinese classes across three text genres, environments, and multiple sessions. Results show: (1) Instructional alignment: simulated IRF rates (0.28-0.64) closely match real classrooms (0.37-0.49), indicating pedagogical realism; (2) Group interaction and role differentiation: network density (0.27-0.40) with about one-third of peer links realized, while human-agent tasks indicate a balance between individual variability and instructional stability; (3) Cross-session evolution: the positive transition rate R+ increase by 11.7% on average, capturing longitudinal shifts in behavior, emotion, and cognition and revealing structured learning trajectories. Overall, EduVerse balances realism, reproducibility, and interpretability, providing a scalable platform for educational AI. The system will be open-sourced to foster cross-disciplinary research. --- ### Ocular-Induced Abnormal Head Posture: Diagnosis and Missing Data Imputation Authors: Saja Al-Dabet, Sherzod Turaev, Nazar Zaki, Arif O. Khan, Luai Eldweik Published: 2025-10-07 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2510.05649v1 Ocular-induced abnormal head posture (AHP) is a compensatory mechanism that arises from ocular misalignment conditions, such as strabismus, enabling patients to reduce diplopia and preserve binocular vision. Early diagnosis minimizes morbidity and secondary complications such as facial asymmetry; however, current clinical assessments remain largely subjective and are further complicated by incomplete medical records. This study addresses both challenges through two complementary deep learning frameworks. First, AHP-CADNet is a multi-level attention fusion framework for automated diagnosis that integrates ocular landmarks, head pose features, and structured clinical attributes to generate interpretable predictions. Second, a curriculum learning-based imputation framework is designed to mitigate missing data by progressively leveraging structured variables and unstructured clinical notes to enhance diagnostic robustness under realistic data conditions. Evaluation on the PoseGaze-AHP dataset demonstrates robust diagnostic performance. AHP-CADNet achieves 96.9-99.0 percent accuracy across classification tasks and low prediction errors for continuous variables, with MAE ranging from 0.103 to 0.199 and R2 exceeding 0.93. The imputation framework maintains high accuracy across all clinical variables (93.46-99.78 percent with PubMedBERT), with clinical dependency modeling yielding significant improvements (p < 0.001). These findings confirm the effectiveness of both frameworks for automated diagnosis and recovery from missing data in clinical settings. --- ### The African Languages Lab: A Collaborative Approach to Advancing Low-Resource African NLP Authors: Sheriff Issaka, Keyi Wang, Yinka Ajibola, Oluwatumininu Samuel-Ipaye, Zhaoyi Zhang, Nicte Aguillon Jimenez, Evans Kofi Agyei, Abraham Lin, Rohan Ramachandran, Sadick Abdul Mumin, Faith Nchifor, Mohammed Shuraim et al. Published: 2025-10-07 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2510.05644v1 Despite representing nearly one-third of the world's languages, African languages remain critically underserved by modern NLP technologies, with 88\% classified as severely underrepresented or completely ignored in computational linguistics. We present the African Languages Lab (All Lab), a comprehensive research initiative that addresses this technological gap through systematic data collection, model development, and capacity building. Our contributions include: (1) a quality-controlled data collection pipeline, yielding the largest validated African multi-modal speech and text dataset spanning 40 languages with 19 billion tokens of monolingual text and 12,628 hours of aligned speech data; (2) extensive experimental validation demonstrating that our dataset, combined with fine-tuning, achieves substantial improvements over baseline models, averaging +23.69 ChrF++, +0.33 COMET, and +15.34 BLEU points across 31 evaluated languages; and (3) a structured research program that has successfully mentored fifteen early-career researchers, establishing sustainable local capacity. Our comparative evaluation against Google Translate reveals competitive performance in several languages while identifying areas that require continued development. --- ### Combined Hyperbolic and Euclidean Soft Triple Loss Beyond the Single Space Deep Metric Learning Authors: Shozo Saeki, Minoru Kawahara, Hirohisa Aman Published: 2025-10-07 Categories: cs.CV arXiv: https://arxiv.org/abs/2510.05643v2 Deep metric learning (DML) aims to learn a neural network mapping data to an embedding space, which can represent semantic similarity between data points. Hyperbolic space is attractive for DML since it can represent richer structures, such as tree structures. DML in hyperbolic space is based on pair-based loss or unsupervised regularization loss. On the other hand, supervised proxy-based losses in hyperbolic space have not been reported yet due to some issues in applying proxy-based losses in a hyperbolic space. However, proxy-based losses are attractive for large-scale datasets since they have less training complexity. To address these, this paper proposes the Combined Hyperbolic and Euclidean Soft Triple (CHEST) loss. CHEST loss is composed of the proxy-based losses in hyperbolic and Euclidean spaces and the regularization loss based on hyperbolic hierarchical clustering. We find that the combination of hyperbolic and Euclidean spaces improves DML accuracy and learning stability for both spaces. Finally, we evaluate the CHEST loss on four benchmark datasets, achieving a new state-of-the-art performance. --- ### From Neural Activity to Computation: Biological Reservoirs for Pattern Recognition in Digit Classification Authors: Ludovico Iannello, Luca Ciampi, Fabrizio Tonelli, Gabriele Lagani, Lucio Maria Calcagnile, Federico Cremisi, Angelo Di Garbo, Giuseppe Amato Published: 2025-10-07 Categories: cs.NE, cs.AI, cs.CV arXiv: https://arxiv.org/abs/2510.05637v1 In this paper, we present a biologically grounded approach to reservoir computing (RC), in which a network of cultured biological neurons serves as the reservoir substrate. This system, referred to as biological reservoir computing (BRC), replaces artificial recurrent units with the spontaneous and evoked activity of living neurons. A multi-electrode array (MEA) enables simultaneous stimulation and readout across multiple sites: inputs are delivered through a subset of electrodes, while the remaining ones capture the resulting neural responses, mapping input patterns into a high-dimensional biological feature space. We evaluate the system through a case study on digit classification using a custom dataset. Input images are encoded and delivered to the biological reservoir via electrical stimulation, and the corresponding neural activity is used to train a simple linear classifier. To contextualize the performance of the biological system, we also include a comparison with a standard artificial reservoir trained on the same task. The results indicate that the biological reservoir can effectively support classification, highlighting its potential as a viable and interpretable computational substrate. We believe this work contributes to the broader effort of integrating biological principles into machine learning and aligns with the goals of human-inspired vision by exploring how living neural systems can inform the design of efficient and biologically plausible models. --- ### NEO: No-Optimization Test-Time Adaptation through Latent Re-Centering Authors: Alexander Murphy, Michal Danilowski, Soumyajit Chatterjee, Abhirup Ghosh Published: 2025-10-07 Categories: cs.LG, cs.CV arXiv: https://arxiv.org/abs/2510.05635v2 Test-Time Adaptation (TTA) methods are often computationally expensive, require a large amount of data for effective adaptation, or are brittle to hyperparameters. Based on a theoretical foundation of the geometry of the latent space, we are able to significantly improve the alignment between source and distribution-shifted samples by re-centering target data embeddings at the origin. This insight motivates NEO -- a hyperparameter-free fully TTA method, that adds no significant compute compared to vanilla inference. NEO is able to improve the classification accuracy of ViT-Base on ImageNet-C from 55.6% to 59.2% after adapting on just one batch of 64 samples. When adapting on 512 samples NEO beats all 7 TTA methods we compare against on ImageNet-C, ImageNet-R and ImageNet-S and beats 6/7 on CIFAR-10-C, while using the least amount of compute. NEO performs well on model calibration metrics and additionally is able to adapt from 1 class to improve accuracy on 999 other classes in ImageNet-C. On Raspberry Pi and Jetson Orin Nano devices, NEO reduces inference time by 63% and memory usage by 9% compared to baselines. Our results based on 3 ViT architectures and 4 datasets show that NEO can be used efficiently and effectively for TTA. --- ### Beyond Spectral Peaks: Interpreting the Cues Behind Synthetic Image Detection Authors: Sara Mandelli, Diego Vila-Portela, David Vázquez-Padín, Paolo Bestagini, Fernando Pérez-González Published: 2025-10-07 Categories: cs.CV, cs.AI, cs.CR arXiv: https://arxiv.org/abs/2510.05633v1 Over the years, the forensics community has proposed several deep learning-based detectors to mitigate the risks of generative AI. Recently, frequency-domain artifacts (particularly periodic peaks in the magnitude spectrum), have received significant attention, as they have been often considered a strong indicator of synthetic image generation. However, state-of-the-art detectors are typically used as black-boxes, and it still remains unclear whether they truly rely on these peaks. This limits their interpretability and trust. In this work, we conduct a systematic study to address this question. We propose a strategy to remove spectral peaks from images and analyze the impact of this operation on several detectors. In addition, we introduce a simple linear detector that relies exclusively on frequency peaks, providing a fully interpretable baseline free from the confounding influence of deep learning. Our findings reveal that most detectors are not fundamentally dependent on spectral peaks, challenging a widespread assumption in the field and paving the way for more transparent and reliable forensic tools. --- ### Systematic Exploration of Multi-core Architectures for Efficient LLM Serving using WaferAI-SIM Authors: Tianhao Zhu, Dahu Feng, Erhu Feng, Yubin Xia Published: 2025-10-07 Categories: cs.AR, cs.LG arXiv: https://arxiv.org/abs/2510.05632v2 With the widespread adoption of Large Language Models (LLMs), the demand for high-performance LLM inference services continues to grow. Multi-core AI accelerators, such as Groq, Graphcore IPU, and Cerebras WSE, provide promising platforms for LLM serving, but their distributed memory systems require careful coordination between hardware configuration and serving policies. Otherwise, mismatched tensor partitioning, data placement, and memory management can substantially underutilize compute and communication resources. To address these challenges, we present WaferAI-SIM, a multi-level simulation framework that combines transaction-level simulation with an analytical performance model. WaferAI-SIM enables simulator-driven co-design of LLM serving strategies and multi-core accelerator architectures, targeting the early design stage in which emerging platforms are not yet broadly available for empirical serving studies. It captures how LLM serving policies interact with compute-core count, memory hierarchy, and interconnect topology, enabling architecture-aware exploration beyond GPU-centric assumptions. We evaluate representative LLMs across a range of chip configurations and serving scenarios. Across the evaluation, WaferAI-SIM reports 1.32$\times$--6.03$\times$ latency improvements, where the lower endpoint comes from ring-based placement at TP=16 over the placement baselines, and the upper endpoint comes from K-dimension TP over MN-dimension TP for Qwen3\_4B at TP=4 with sequence length 256. For LLM serving, our findings provide guidance for co-designing hardware architectures and serving strategies for multi-core AI accelerators across diverse LLM workloads. --- ### Generative AI-Driven Hierarchical Multi-Agent Framework for Zero-Touch Optical Networks Authors: Yao Zhang, Yuchen Song, Shengnan Li, Yan Shi, Shikui Shen, Xiongyan Tang, Min Zhang, Danshi Wang Published: 2025-10-07 Categories: cs.NI, cs.AI, cs.CL, cs.MA, eess.SY arXiv: https://arxiv.org/abs/2510.05625v1 The rapid development of Generative Artificial Intelligence (GenAI) has catalyzed a transformative technological revolution across all walks of life. As the backbone of wideband communication, optical networks are expecting high-level autonomous operation and zero-touch management to accommodate their expanding network scales and escalating transmission bandwidth. The integration of GenAI is deemed as the pivotal solution for realizing zero-touch optical networks. However, the lifecycle management of optical networks involves a multitude of tasks and necessitates seamless collaboration across multiple layers, which poses significant challenges to the existing single-agent GenAI systems. In this paper, we propose a GenAI-driven hierarchical multi-agent framework designed to streamline multi-task autonomous execution for zero-touch optical networks. We present the architecture, implementation, and applications of this framework. A field-deployed mesh network is utilized to demonstrate three typical scenarios throughout the lifecycle of optical network: quality of transmission estimation in the planning stage, dynamic channel adding/dropping in the operation stage, and system capacity increase in the upgrade stage. The case studies, illustrate the capabilities of multi-agent framework in multi-task allocation, coordination, execution, evaluation, and summarization. This work provides a promising approach for the future development of intelligent, efficient, and collaborative network management solutions, paving the way for more specialized and adaptive zero-touch optical networks. --- ### Monte Carlo-Type Neural Operator for Differential Equations Authors: Salah Eddine Choutri, Prajwal Chauhan, Othmane Mazhar, Saif Eddin Jabari Published: 2025-10-07 Categories: cs.LG, cs.AI, math.NA, stat.ML arXiv: https://arxiv.org/abs/2510.05620v2 The Monte Carlo-type Neural Operator (MCNO) introduces a framework for learning solution operators of one-dimensional partial differential equations (PDEs) by directly learning the kernel function and approximating the associated integral operator using a Monte Carlo-type approach. Unlike Fourier Neural Operators (FNOs), which rely on spectral representations and assume translation-invariant kernels, MCNO makes no such assumptions. The kernel is represented as a learnable tensor over sampled input-output pairs, and sampling is performed once, uniformly at random from a discretized grid. This design enables generalization across multiple grid resolutions without relying on fixed global basis functions or repeated sampling during training, while an interpolation step maps between arbitrary input and output grids to further enhance flexibility. Experiments on standard 1D PDE benchmarks show that MCNO achieves competitive accuracy with efficient computational cost. We also provide a theoretical analysis proving that the Monte Carlo estimator yields a bounded bias and variance under mild regularity assumptions. This result holds in any spatial dimension, suggesting that MCNO may extend naturally beyond one-dimensional problems. More broadly, this work explores how Monte Carlo-type integration can be incorporated into neural operator frameworks for continuous-domain PDEs, providing a theoretically supported alternative to spectral methods (such as FNO) and to graph-based Monte Carlo approaches (such as the Graph Kernel Neural Operator, GNO). --- ### InstaGeo: Compute-Efficient Geospatial Machine Learning from Data to Deployment Authors: Ibrahim Salihu Yusuf, Iffanice Houndayi, Rym Oualha, Mohamed Aziz Cherif, Kobby Panford-Quainoo, Arnu Pretorius Published: 2025-10-07 Categories: cs.CV, cs.CY, cs.LG arXiv: https://arxiv.org/abs/2510.05617v1 Open-access multispectral imagery from missions like Landsat 8-9 and Sentinel-2 has fueled the development of geospatial foundation models (GFMs) for humanitarian and environmental applications. Yet, their deployment remains limited by (i) the absence of automated geospatial data pipelines and (ii) the large size of fine-tuned models. Existing GFMs lack workflows for processing raw satellite imagery, and downstream adaptations often retain the full complexity of the original encoder. We present InstaGeo, an open-source, end-to-end framework that addresses these challenges by integrating: (1) automated data curation to transform raw imagery into model-ready datasets; (2) task-specific model distillation to derive compact, compute-efficient models; and (3) seamless deployment as interactive web-map applications. Using InstaGeo, we reproduced datasets from three published studies and trained models with marginal mIoU differences of -0.73 pp for flood mapping, -0.20 pp for crop segmentation, and +1.79 pp for desert locust prediction. The distilled models are up to 8x smaller than standard fine-tuned counterparts, reducing FLOPs and CO2 emissions with minimal accuracy loss. Leveraging InstaGeo's streamlined data pipeline, we also curated a larger crop segmentation dataset, achieving a state-of-the-art mIoU of 60.65%, a 12 pp improvement over prior baselines. Moreover, InstaGeo enables users to progress from raw data to model deployment within a single working day. By unifying data preparation, model compression, and deployment, InstaGeo transforms research-grade GFMs into practical, low-carbon tools for real-time, large-scale Earth observation. This approach shifts geospatial AI toward data quality and application-driven innovation. Source code, datasets, and model checkpoints are available at: https://github.com/instadeepai/InstaGeo-E2E-Geospatial-ML.git --- ### TFM Dataset: A Novel Multi-task Dataset and Integrated Pipeline for Automated Tear Film Break-Up Segmentation Authors: Guangrong Wan, Jun liu, Qiyang Zhou, Tang tang, Lianghao Shi, Wenjun Luo, TingTing Xu Published: 2025-10-07 Categories: cs.CV arXiv: https://arxiv.org/abs/2510.05615v2 Tear film break-up (TFBU) analysis is critical for diagnosing dry eye syndrome, but automated TFBU segmentation remains challenging due to the lack of annotated datasets and integrated solutions. This paper introduces the Tear Film Multi-task (TFM) Dataset, the first comprehensive dataset for multi-task tear film analysis, comprising 15 high-resolution videos (totaling 6,247 frames) annotated with three vision tasks: frame-level classification ('clear', 'closed', 'broken', 'blur'), Placido Ring detection, and pixel-wise TFBU area segmentation. Leveraging this dataset, we first propose TF-Net, a novel and efficient baseline segmentation model. TF-Net incorporates a MobileOne-mini backbone with re-parameterization techniques and an enhanced feature pyramid network to achieve a favorable balance between accuracy and computational efficiency for real-time clinical applications. We further establish benchmark performance on the TFM segmentation subset by comparing TF-Net against several state-of-the-art medical image segmentation models. Furthermore, we design TF-Collab, a novel integrated real-time pipeline that synergistically leverages models trained on all three tasks of the TFM dataset. By sequentially orchestrating frame classification for BUT determination, pupil region localization for input standardization, and TFBU segmentation, TF-Collab fully automates the analysis. Experimental results demonstrate the effectiveness of the proposed TF-Net and TF-Collab, providing a foundation for future research in ocular surface diagnostics. Our code and the TFM datasets are available at https://github.com/glory-wan/TF-Net --- ### PointNSP: Autoregressive 3D Point Cloud Generation with Next-Scale Level-of-Detail Prediction Authors: Ziqiao Meng, Qichao Wang, Zhiyang Dou, Zixing Song, Zhipeng Zhou, Irwin King, Peilin Zhao Published: 2025-10-07 Categories: cs.CV, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2510.05613v2 Autoregressive point cloud generation has long lagged behind diffusion-based approaches in quality. The performance gap stems from the fact that autoregressive models impose an artificial ordering on inherently unordered point sets, forcing shape generation to proceed as a sequence of local predictions. This sequential bias emphasizes short-range continuity but undermines the model's capacity to capture long-range dependencies, hindering its ability to enforce global structural properties such as symmetry, consistent topology, and large-scale geometric regularities. Inspired by the level-of-detail (LOD) principle in shape modeling, we propose PointNSP, a coarse-to-fine generative framework that preserves global shape structure at low resolutions and progressively refines fine-grained geometry at higher scales through a next-scale prediction paradigm. This multi-scale factorization aligns the autoregressive objective with the permutation-invariant nature of point sets, enabling rich intra-scale interactions while avoiding brittle fixed orderings. Experiments on ShapeNet show that PointNSP establishes state-of-the-art (SOTA) generation quality for the first time within the autoregressive paradigm. In addition, it surpasses strong diffusion-based baselines in parameter, training, and inference efficiency. Finally, in dense generation with 8,192 points, PointNSP's advantages become even more pronounced, underscoring its scalability potential. --- ### MADIAVE: Multi-Agent Debate for Implicit Attribute Value Extraction Authors: Wei-Chieh Huang, Cornelia Caragea Published: 2025-10-07 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2510.05611v2 Implicit Attribute Value Extraction (AVE) is essential for accurately representing products in e-commerce, as it infers latent attributes from multimodal data. Despite advances in multimodal large language models (MLLMs), implicit AVE remains challenging due to the complexity of multidimensional data and gaps in vision-text understanding. In this work, we introduce MADIAVE, a multi-agent debate framework that employs multiple MLLM agents to iteratively refine inferences. Through a series of debate rounds, agents verify and update each other's responses, thereby improving inference performance and robustness. Experiments on the ImplicitAVE dataset demonstrate that even a few rounds of debate significantly boost accuracy, especially for attributes with initially low performance. We systematically evaluate various debate configurations, including identical or different MLLM agents, and analyze how debate rounds affect convergence dynamics. Our findings highlight the potential of multi-agent debate strategies to address the limitations of single-agent approaches and offer a scalable solution for implicit AVE in multimodal e-commerce. --- ### Efficient Conditional Generation on Scale-based Visual Autoregressive Models Authors: Jiaqi Liu, Tao Huang, Chang Xu Published: 2025-10-07 Categories: cs.CV arXiv: https://arxiv.org/abs/2510.05610v1 Recent advances in autoregressive (AR) models have demonstrated their potential to rival diffusion models in image synthesis. However, for complex spatially-conditioned generation, current AR approaches rely on fine-tuning the pre-trained model, leading to significant training costs. In this paper, we propose the Efficient Control Model (ECM), a plug-and-play framework featuring a lightweight control module that introduces control signals via a distributed architecture. This architecture consists of context-aware attention layers that refine conditional features using real-time generated tokens, and a shared gated feed-forward network (FFN) designed to maximize the utilization of its limited capacity and ensure coherent control feature learning. Furthermore, recognizing the critical role of early-stage generation in determining semantic structure, we introduce an early-centric sampling strategy that prioritizes learning early control sequences. This approach reduces computational cost by lowering the number of training tokens per iteration, while a complementary temperature scheduling during inference compensates for the resulting insufficient training of late-stage tokens. Extensive experiments on scale-based AR models validate that our method achieves high-fidelity and diverse control over image generation, surpassing existing baselines while significantly improving both training and inference efficiency. --- ### HOI-R1: Exploring the Potential of Multimodal Large Language Models for Human-Object Interaction Detection Authors: Junwen Chen, Peilin Xiong, Keiji Yanai Published: 2025-10-07 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2510.05609v2 Recent human-object interaction detection (HOID) methods highly require prior knowledge from vision-language models (VLMs) to enhance the interaction recognition capabilities. The training strategies and model architectures for connecting the knowledge from VLMs to the HOI instance representations from the object detector are challenging, and the whole framework is complex for further development or application. On the other hand, the inherent reasoning abilities of multimodal large language models (MLLMs) on human-object interaction detection are under-explored. Inspired by the recent success of training MLLMs with reinforcement learning (RL) methods, we propose HOI-R1 and first explore the potential of the language model on the HOID task without any additional detection modules. We introduce an HOI reasoning process and HOID reward functions to solve the HOID task by pure text. Experiments on HICO-DET across multiple open-source MLLMs, including the Qwen-VL family (Qwen2.5-VL and Qwen3-VL) and Rex-Omni, show consistent improvements. Especially, HOI-R1 boosts Qwen2.5-VL-3B 2$\times$ accuracy with great generalization ability. The source code is available at https://github.com/cjw2021/HOI-R1. --- ### A Goal Without a Plan Is Just a Wish: Efficient and Effective Global Planner Training for Long-Horizon Agent Tasks Authors: Shuzheng Si, Haozhe Zhao, Kangyang Luo, Gang Chen, Fanchao Qi, Minjia Zhang, Baobao Chang, Maosong Sun Published: 2025-10-07 Categories: cs.CL arXiv: https://arxiv.org/abs/2510.05608v3 Agents based on large language models (LLMs) struggle with brainless trial-and-error and generating hallucinatory actions due to a lack of global planning in long-horizon tasks. In this paper, we introduce a plan-and-execute framework and propose EAGLET, an efficient and effective planner training method to enhance the executor agent's planning abilities without human effort. Specifically, we train a plug-and-play global planner through a two-step process: we first synthesize high-quality plans from an advanced LLM using our proposed homologous consensus filtering strategy, and apply fine-tuning as a cold start. Moreover, we further improve the planner with a rule-based reinforcement learning stage using a novel executor capability gain reward, ensuring it can handle task instructions of varying difficulty. Experiments on three long-horizon agent tasks show that executor agents equipped with our planner outperform existing methods, achieving new state-of-the-art performance. Meanwhile, EAGLET reduces training costs by 8x compared to RL-based baselines, and it does not require manual effort or extra training data, offering an efficient and effective solution. --- ### Fractal basin geometry and the limits to predictability and reproducibility in deep learning Authors: Andrew Ly, Pulin Gong Published: 2025-10-07 Categories: cs.LG arXiv: https://arxiv.org/abs/2510.05606v2 Fundamental limits to predictability are central to our understanding of many physical and computational systems. In deep learning, training outcomes are notoriously difficult to predict and reproduce, yet the mechanism underlying these difficulties has remained unclear. Here we propose fractal, riddled basin geometry as a unifying dynamical-systems mechanism: every initialization leading to one solution lies arbitrarily close to initializations leading to different solutions, forming a fat fractal with structure at arbitrarily fine scales. Using a minimal neural network model, we analytically establish the classical route to riddling and derive a near-zero uncertainty exponent: even sharp increases in initialization precision yield only marginal gains in outcome predictability. In deep networks, intertwined training destinations and near-zero parameter-space exponents provide robust signatures consistent with riddling across tested architectures, optimizers and training protocols, offering a common dynamical account of otherwise disparate empirical observations. We further introduce functional uncertainty exponents, which extend these diagnostics of scale-insensitive irreproducibility from parameter space to the learned functions themselves. Together, our results support riddled basin geometry as a candidate organizing principle for irreproducibility in deep learning, with practical implications for optimization and the reliable deployment of artificial intelligence. --- ### AutoPentester: An LLM Agent-based Framework for Automated Pentesting Authors: Yasod Ginige, Akila Niroshan, Sajal Jain, Suranga Seneviratne Published: 2025-10-07 Categories: cs.CR, cs.AI arXiv: https://arxiv.org/abs/2510.05605v1 Penetration testing and vulnerability assessment are essential industry practices for safeguarding computer systems. As cyber threats grow in scale and complexity, the demand for pentesting has surged, surpassing the capacity of human professionals to meet it effectively. With advances in AI, particularly Large Language Models (LLMs), there have been attempts to automate the pentesting process. However, existing tools such as PentestGPT are still semi-manual, requiring significant professional human interaction to conduct pentests. To this end, we propose a novel LLM agent-based framework, AutoPentester, which automates the pentesting process. Given a target IP, AutoPentester automatically conducts pentesting steps using common security tools in an iterative process. It can dynamically generate attack strategies based on the tool outputs from the previous iteration, mimicking the human pentester approach. We evaluate AutoPentester using Hack The Box and custom-made VMs, comparing the results with the state-of-the-art PentestGPT. Results show that AutoPentester achieves a 27.0% better subtask completion rate and 39.5% more vulnerability coverage with fewer steps. Most importantly, it requires significantly fewer human interactions and interventions compared to PentestGPT. Furthermore, we recruit a group of security industry professional volunteers for a user survey and perform a qualitative analysis to evaluate AutoPentester against industry practices and compare it with PentestGPT. On average, AutoPentester received a score of 3.93 out of 5 based on user reviews, which was 19.8% higher than PentestGPT. --- ### AgentDR: Dynamic Recommendation with Implicit Item-Item Relations via LLM-based Agents Authors: Mingdai Yang, Nurendra Choudhary, Jiangshu Du, Edward W. Huang, Philip S. Yu, Karthik Subbian, Danai Koutra Published: 2025-10-07 Categories: cs.IR, cs.AI arXiv: https://arxiv.org/abs/2510.05598v3 Recent agent-based recommendation frameworks aim to simulate user behaviors by incorporating memory mechanisms and prompting strategies, but they struggle with hallucinating non-existent items and full-catalog ranking. Besides, a largely underexplored opportunity lies in leveraging LLMs'commonsense reasoning to capture user intent through substitute and complement relationships between items, which are usually implicit in datasets and difficult for traditional ID-based recommenders to capture. In this work, we propose a novel LLM-agent framework, AgenDR, which bridges LLM reasoning with scalable recommendation tools. Our approach delegates full-ranking tasks to traditional models while utilizing LLMs to (i) integrate multiple recommendation outputs based on personalized tool suitability and (ii) reason over substitute and complement relationships grounded in user history. This design mitigates hallucination, scales to large catalogs, and enhances recommendation relevance through relational reasoning. Through extensive experiments on three public grocery datasets, we show that our framework achieves superior full-ranking performance, yielding on average a twofold improvement over its underlying tools. We also introduce a new LLM-based evaluation metric that jointly measures semantic alignment and ranking correctness. --- ### From Agentification to Self-Evolving Agentic AI for Wireless Networks: Concepts, Approaches, and Future Research Directions Authors: Changyuan Zhao, Ruichen Zhang, Jiacheng Wang, Dusit Niyato, Geng Sun, Xianbin Wang, Shiwen Mao, Abbas Jamalipour Published: 2025-10-07 Categories: cs.AI arXiv: https://arxiv.org/abs/2510.05596v1 Self-evolving agentic artificial intelligence (AI) offers a new paradigm for future wireless systems by enabling autonomous agents to continually adapt and improve without human intervention. Unlike static AI models, self-evolving agents embed an autonomous evolution cycle that updates models, tools, and workflows in response to environmental dynamics. This paper presents a comprehensive overview of self-evolving agentic AI, highlighting its layered architecture, life cycle, and key techniques, including tool intelligence, workflow optimization, self-reflection, and evolutionary learning. We further propose a multi-agent cooperative self-evolving agentic AI framework, where multiple large language models (LLMs) are assigned role-specialized prompts under the coordination of a supervisor agent. Through structured dialogue, iterative feedback, and systematic validation, the system autonomously executes the entire life cycle without human intervention. A case study on antenna evolution in low-altitude wireless networks (LAWNs) demonstrates how the framework autonomously upgrades fixed antenna optimization into movable antenna optimization. Experimental results show that the proposed self-evolving agentic AI autonomously improves beam gain and restores degraded performance by up to 52.02%, consistently surpassing the fixed baseline with little to no human intervention and validating its adaptability and robustness for next-generation wireless intelligence. --- ### Improving Chain-of-Thought Efficiency for Autoregressive Image Generation Authors: Zeqi Gu, Markos Georgopoulos, Xiaoliang Dai, Marjan Ghazvininejad, Chu Wang, Felix Juefei-Xu, Kunpeng Li, Yujun Shi, Zecheng He, Zijian He, Jiawei Zhou, Abe Davis et al. Published: 2025-10-07 Categories: cs.CV, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2510.05593v1 Autoregressive multimodal large language models have recently gained popularity for image generation, driven by advances in foundation models. To enhance alignment and detail, newer approaches employ chain-of-thought (CoT) reasoning, expanding user inputs into elaborated prompts prior to image synthesis. However, this strategy can introduce unnecessary redundancy -- a phenomenon we call visual overthinking -- which increases computational costs and can introduce details that contradict the original prompt. In this work, we explore how to generate more concise CoT sequences for more efficient image generation. We introduce ShortCoTI, a lightweight optimization framework that encourages more concise CoT while preserving output image quality. ShortCoTI rewards more concise prompts with an adaptive function that scales according to an estimated difficulty for each task. Incorporating this reward into a reinforcement learning paradigm reduces prompt reasoning length by 54% while maintaining or slightly improving quality metrics across multiple benchmarks (T2I-CompBench, GenEval). Qualitative analysis shows that our method eliminates verbose explanations and repetitive refinements, producing reasoning prompts that are both concise and semantically rich. As a result, ShortCoTI improves computational efficiency without compromising the fidelity or visual appeal of generated images. --- ### In-the-Flow Agentic System Optimization for Effective Planning and Tool Use Authors: Zhuofeng Li, Haoxiang Zhang, Seungju Han, Sheng Liu, Jianwen Xie, Yu Zhang, Yejin Choi, James Zou, Pan Lu Published: 2025-10-07 Categories: cs.AI, cs.CL, cs.LG, cs.MA arXiv: https://arxiv.org/abs/2510.05592v2 Outcome-driven reinforcement learning has advanced reasoning in large language models (LLMs), but prevailing tool-augmented approaches train a single, monolithic policy that interleaves thoughts and tool calls under full context; this scales poorly with long horizons and diverse tools and generalizes weakly to new scenarios. Agentic systems offer a promising alternative by decomposing work across specialized modules, yet most remain training-free or rely on offline training decoupled from the live dynamics of multi-turn interaction. We introduce AgentFlow, a trainable, in-the-flow agentic framework that coordinates four modules (planner, executor, verifier, generator) through an evolving memory and directly optimizes its planner inside the multi-turn loop. To train on-policy in live environments, we propose Flow-based Group Refined Policy Optimization (Flow-GRPO), which tackles long-horizon, sparse-reward credit assignment by converting multi-turn optimization into a sequence of tractable single-turn policy updates. It broadcasts a single, verifiable trajectory-level outcome to every turn to align local planner decisions with global success and stabilizes learning with group-normalized advantages. Across ten benchmarks, AgentFlow with a 7B-scale backbone outperforms top-performing baselines with average accuracy gains of 14.9% on search, 14.0% on agentic, 14.5% on mathematical, and 4.1% on scientific tasks, even surpassing larger proprietary models like GPT-4o. Further analyses confirm the benefits of in-the-flow optimization, showing improved planning, enhanced tool-calling reliability, and positive scaling with model size and reasoning turns. --- ### Invariant Representation Learning for Source-Free Time Series Forecasting with LLM-Centric Proxy Denoising Authors: Kangjia Yan, Chenxi Liu, Hao Miao, Xinle Wu, Yan Zhao, Chenjuan Guo, Bin Yang Published: 2025-10-07 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2510.05589v3 Effective time series forecasting enables various real-world applications, benefiting from the proliferation of mobile devices. However, the volume of time series data may vary significantly across domains due to high data acquisition costs and data regulations. To maximally create value from sparse data, this study focuses on a new problem of source-free time series forecasting, aiming to adapt a pretrained model from sufficient source time series to the sparse target time series without access to the source data, enabling data protection. To achieve this, we propose TimeID, a novel source-free time series forecasting framework with a large language model (LLM) centric proxy denoising inspired by the powerful generalization capabilities of LLMs. Specifically, TimeID consists of three key components: (1) dual-branch invariant disentangled feature learning that enforces representation- and gradient-wise invariance by means of season-trend decomposition; (2) lightweight, parameter-free proxy denoising that dynamically calibrates systematic biases of LLMs; and (3) knowledge distillation that bidirectionally aligns the denoised prediction and the original target prediction. Extensive experiments on real-world datasets demonstrate that TimeID outperforms state-of-the-art baselines, improving MSE and MAE by 10.7% and 9.3% on average. The code is available at https://github.com/decisionintelligence/TimeID. --- ### CalibCLIP: Contextual Calibration of Dominant Semantics for Text-Driven Image Retrieval Authors: Bin Kang, Bin Chen, Junjie Wang, Yulin Li, Junzhi Zhao, Zhuotao Tian Published: 2025-10-07 Categories: cs.CV arXiv: https://arxiv.org/abs/2510.05586v1 Existing Visual Language Models (VLMs) suffer structural limitations where a few low contribution tokens may excessively capture global semantics, dominating the information aggregation process and suppressing the discriminative features in text-driven image retrieval tasks. To address this, we introduce \textbf{CalibCLIP}, a training-free method designed to calibrate the suppressive effect of dominant tokens. Specifically, in the visual space, we propose the Contrastive Visual Enhancer (CVE), which decouples visual features into target and low information regions. Subsequently, it identifies dominant tokens and dynamically suppresses their representations.In the textual space, we introduce the Discriminative Concept Calibrator (DCC), which aims to differentiate between general and discriminative concepts within the text query. By mitigating the challenges posed by generic concepts and improving the representations of discriminative concepts, DCC strengthens the differentiation among similar samples. Finally, extensive experiments demonstrate consistent improvements across seven benchmarks spanning three image retrieval tasks, underscoring the effectiveness of CalibCLIP. Code is available at: https://github.com/kangbin98/CalibCLIP --- ### When Does Global Attention Help? A Unified Empirical Study on Atomistic Graph Learning Authors: Arindam Chowdhury, Massimiliano Lupo Pasini Published: 2025-10-07 Categories: cs.LG, cs.DC arXiv: https://arxiv.org/abs/2510.05583v2 Graph neural networks (GNNs) are widely used as surrogates for costly experiments and first-principles simulations to study the behavior of compounds at atomistic scale, and their architectural complexity is constantly increasing to enable the modeling of complex physics. While most recent GNNs combine more traditional message passing neural networks (MPNNs) layers to model short-range interactions with more advanced graph transformers (GTs) with global attention mechanisms to model long-range interactions, it is still unclear when global attention mechanisms provide real benefits over well-tuned MPNN layers due to inconsistent implementations, features, or hyperparameter tuning. We introduce the first unified, reproducible benchmarking framework - built on HydraGNN - that enables seamless switching among four controlled model classes: MPNN, MPNN with chemistry/topology encoders, GPS-style hybrids of MPNN with global attention, and fully fused local-global models with encoders. Using seven diverse open-source datasets for benchmarking across regression and classification tasks, we systematically isolate the contributions of message passing, global attention, and encoder-based feature augmentation. Our study shows that encoder-augmented MPNNs form a robust baseline, while fused local-global models yield the clearest benefits for properties governed by long-range interaction effects. We further quantify the accuracy-compute trade-offs of attention, reporting its overhead in memory. Together, these results establish the first controlled evaluation of global attention in atomistic graph learning and provide a reproducible testbed for future model development. --- ### (Token-Level) InfoRMIA: Stronger Membership Inference and Memorization Assessment for LLMs Authors: Jiashu Tao, Reza Shokri Published: 2025-10-07 Categories: cs.LG arXiv: https://arxiv.org/abs/2510.05582v2 Machine learning models are known to leak sensitive information, as they inevitably memorize (parts of) their training data. More alarmingly, large language models (LLMs) are now trained on nearly all available data, which amplifies the magnitude of information leakage and raises serious privacy risks. Hence, it is more crucial than ever to quantify privacy risk before the release of LLMs. The standard method to quantify privacy is via membership inference attacks, where the state-of-the-art approach is the Robust Membership Inference Attack (RMIA). In this paper, we present InfoRMIA, a principled information-theoretic formulation of membership inference. Our method consistently outperforms RMIA across benchmarks while also offering improved computational efficiency. In the second part of the paper, we identify the limitations of treating sequence-level membership inference as the gold standard for measuring leakage. We propose a new perspective for studying membership and memorization in LLMs: token-level signals and analyses. We show that a simple token-based InfoRMIA can pinpoint which tokens are memorized within generated outputs, thereby localizing leakage from the sequence level down to individual tokens, while achieving stronger sequence-level inference power on LLMs. This new scope rethinks privacy in LLMs and can lead to more targeted mitigation, such as exact unlearning. --- ### Power Mechanism: Private Tabular Representation Release for Model Agnostic Consumption Authors: Praneeth Vepakomma, Kaustubh Ponkshe Published: 2025-10-07 Categories: cs.LG, cs.CR arXiv: https://arxiv.org/abs/2510.05581v1 Traditional collaborative learning approaches are based on sharing of model weights between clients and a server. However, there are advantages to resource efficiency through schemes based on sharing of embeddings (activations) created from the data. Several differentially private methods were developed for sharing of weights while such mechanisms do not exist so far for sharing of embeddings. We propose Ours to learn a privacy encoding network in conjunction with a small utility generation network such that the final embeddings generated from it are equipped with formal differential privacy guarantees. These privatized embeddings are then shared with a more powerful server, that learns a post-processing that results in a higher accuracy for machine learning tasks. We show that our co-design of collaborative and private learning results in requiring only one round of privatized communication and lesser compute on the client than traditional methods. The privatized embeddings that we share from the client are agnostic to the type of model (deep learning, random forests or XGBoost) used on the server in order to process these activations to complete a task. --- ### MetaVLA: Unified Meta Co-training For Efficient Embodied Adaption Authors: Chen Li, Zhantao Yang, Han Zhang, Fangyi Chen, Chenchen Zhu, Anudeepsekhar Bolimera, Marios Savvides Published: 2025-10-07 Categories: cs.AI, cs.RO arXiv: https://arxiv.org/abs/2510.05580v3 Vision-Language-Action (VLA) models show promise in embodied reasoning, yet remain far from true generalists-they often require task-specific fine-tuning, incur high compute costs, and generalize poorly to unseen tasks. We propose MetaVLA, a unified, backbone-agnostic post-training framework for efficient and scalable alignment. MetaVLA introduces Context-Aware Meta Co-Training, which consolidates diverse target tasks into a single fine-tuning stage while leveraging structurally diverse auxiliary tasks to improve in-domain generalization. Unlike naive multi-task SFT, MetaVLA integrates a lightweight meta-learning mechanism-derived from Attentive Neural Processes-to enable rapid adaptation from diverse contexts with minimal architectural change or inference overhead. On the LIBERO benchmark, MetaVLA with six auxiliary tasks outperforms OpenVLA by up to 8.0% on long-horizon tasks, reduces training steps from 240K to 75K, and cuts GPU time by ~76%. These results show that scalable, low-resource post-training is achievable-paving the way toward general-purpose embodied agents. Code will be available. --- ### Mission Impossible: Feedback-Guided Dynamic Interactive Planning for Improving Reasoning on LLMs Authors: Dong Yan, Gaochen Wu, Bowen Zhou Published: 2025-10-07 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2510.05577v1 Recent advancements in language agents have led to significant improvements in multi-hop reasoning tasks. However, existing approaches often struggle with handling open-domain problems, which require massive information retrieval due to their reliance on a fixed sequence of actions. To address this, we propose Feedback-Guided Dynamic Interactive Planning (FGDIP), a novel framework tailored to enhance reasoning in LLMs by utilizing dynamic and adaptive strategies for information exploration in open-domain multi-hop reasoning tasks. Our approach begins by identifying key entities relevant to the problem, which serve as the initial nodes in the reasoning process. From these initial nodes, we then generate reasoning child nodes with the process being refined through a combination of historical error analysis and real-time feedback, which allows the framework to dynamically adjust and optimize its reasoning strategies. By integrating depth-first search with an innovative node generation technique, our framework adapts based on both prior error paths and concurrently generated nodes at the same hierarchical level. This dynamic strategy effectively expands the search space while ensuring the reasoning process systematically converges toward accurate solutions. Experimental results show that FGDIP achieved up to 54.47% F1 score on the HotpotQA dataset and 70.05% on the StrategyQA dataset, surpassing the best baseline by 5.03% and 7.25% respectively, highlighting its versatility and potential to enhance language agents in multi-hop reasoning tasks. --- ### On the Theory of Continual Learning with Gradient Descent for Neural Networks Authors: Hossein Taheri, Avishek Ghosh, Arya Mazumdar Published: 2025-10-07 Categories: stat.ML, cs.IT, cs.LG arXiv: https://arxiv.org/abs/2510.05573v2 Continual learning, the ability of a model to adapt to an ongoing sequence of tasks without forgetting earlier ones, is a central goal of artificial intelligence. To better understand its underlying mechanisms, we study the limitations of continual learning in a tractable yet representative setting. Specifically, we analyze one-hidden-layer quadratic neural networks trained by gradient descent on a sequence of XOR-cluster datasets with Gaussian noise, where different tasks correspond to clusters with orthogonal means. Our analysis is based on a tight characterization of gradient descent dynamics for the training loss, which yields explicit bounds on the rate of train-time forgetting as functions of the number of iterations, sample size, number of tasks, and hidden-layer width. We then leverage an algorithmic stability framework to bound the generalization gap, leading to corresponding guarantees on test-time forgetting. Together, our results provide the first closed-form guarantees for forgetting in continual learning with neural networks and show how key problem parameters jointly govern forgetting dynamics. Numerical experiments corroborate our theoretical results. --- ### Presenting a Paper is an Art: Self-Improvement Aesthetic Agents for Academic Presentations Authors: Chengzhi Liu, Yuzhe Yang, Kaiwen Zhou, Zhen Zhang, Yue Fan, Yanan Xie, Peng Qi, Xin Eric Wang Published: 2025-10-07 Categories: cs.CL arXiv: https://arxiv.org/abs/2510.05571v2 The promotion of academic papers has become an important means of enhancing research visibility. However, existing automated methods struggle limited storytelling, insufficient aesthetic quality, and constrained self-adjustment, making it difficult to achieve efficient and engaging dissemination. At the heart of those challenges is a simple principle: \emph{there is no way to improve it when you cannot evaluate it right}. To address this, we introduce \textbf{EvoPresent}, a self-improvement agent framework that unifies coherent narratives, aesthetic-aware designs, and realistic presentation delivery via virtual characters. Central to EvoPresent is \textbf{PresAesth}, a multi-task reinforcement learning (RL) aesthetic model that provides reliable aesthetic scoring, defect adjustment, and comparative feedback, enabling iterative self-improvement even under limited aesthetic training data. To systematically evaluate the methods, we introduce \textbf{EvoPresent Benchmark}, a comprehensive benchmark comprising: \textit{Presentation Generation Quality}, built on 650 top-tier AI conference papers with multimodal resources (slides, videos and scripts) to assess both content and design; and \textit{Aesthetic Awareness}, consisting of 2,000 slide pairs with varying aesthetic levels, supporting joint training and evaluation on scoring, defect adjustment, and comparison. Our findings highlight that (i) High-quality feedback is essential for agent self-improvement, while initial capability alone does not guarantee effective self-correction. (ii) Automated generation pipelines exhibit a trade-off between visual design and content construction. (iii) Multi-task RL training shows stronger generalization in aesthetic awareness tasks. --- ### Efficient Learning-based Graph Simulation for Temporal Graphs Authors: Sheng Xiang, Chenhao Xu, Dawei Cheng, Xiaoyang Wang, Ying Zhang Published: 2025-10-07 Categories: cs.LG arXiv: https://arxiv.org/abs/2510.05569v1 Graph simulation has recently received a surge of attention in graph processing and analytics. In real-life applications, e.g. social science, biology, and chemistry, many graphs are composed of a series of evolving graphs (i.e., temporal graphs). While most of the existing graph generators focus on static graphs, the temporal information of the graphs is ignored. In this paper, we focus on simulating temporal graphs, which aim to reproduce the structural and temporal properties of the observed real-life temporal graphs. In this paper, we first give an overview of the existing temporal graph generators, including recently emerged learning-based approaches. Most of these learning-based methods suffer from one of the limitations: low efficiency in training or slow generating, especially for temporal random walk-based methods. Therefore, we propose an efficient learning-based approach to generate graph snapshots, namely temporal graph autoencoder (TGAE). Specifically, we propose an attention-based graph encoder to encode temporal and structural characteristics on sampled ego-graphs. And we proposed an ego-graph decoder that can achieve a good trade-off between simulation quality and efficiency in temporal graph generation. Finally, the experimental evaluation is conducted among our proposed TGAE and representative temporal graph generators on real-life temporal graphs and synthesized graphs. It is reported that our proposed approach outperforms the state-of-the-art temporal graph generators by means of simulation quality and efficiency. --- ### Bilevel optimization for learning hyperparameters: Application to solving PDEs and inverse problems with Gaussian processes Authors: Nicholas H. Nelsen, Houman Owhadi, Andrew M. Stuart, Xianjin Yang, Zongren Zou Published: 2025-10-07 Categories: stat.ML, cs.LG, physics.comp-ph arXiv: https://arxiv.org/abs/2510.05568v1 Methods for solving scientific computing and inference problems, such as kernel- and neural network-based approaches for partial differential equations (PDEs), inverse problems, and supervised learning tasks, depend crucially on the choice of hyperparameters. Specifically, the efficacy of such methods, and in particular their accuracy, stability, and generalization properties, strongly depends on the choice of hyperparameters. While bilevel optimization offers a principled framework for hyperparameter tuning, its nested optimization structure can be computationally demanding, especially in PDE-constrained contexts. In this paper, we propose an efficient strategy for hyperparameter optimization within the bilevel framework by employing a Gauss-Newton linearization of the inner optimization step. Our approach provides closed-form updates, eliminating the need for repeated costly PDE solves. As a result, each iteration of the outer loop reduces to a single linearized PDE solve, followed by explicit gradient-based hyperparameter updates. We demonstrate the effectiveness of the proposed method through Gaussian process models applied to nonlinear PDEs and to PDE inverse problems. Extensive numerical experiments highlight substantial improvements in accuracy and robustness compared to conventional random hyperparameter initialization. In particular, experiments with additive kernels and neural network-parameterized deep kernels demonstrate the method's scalability and effectiveness for high-dimensional hyperparameter optimization. --- ### Domain-Shift-Aware Conformal Prediction for Large Language Models Authors: Zhexiao Lin, Yuanyuan Li, Neeraj Sarna, Yuanyuan Gao, Michael von Gablenz Published: 2025-10-07 Categories: stat.ML, cs.AI, cs.CL, cs.LG, stat.AP arXiv: https://arxiv.org/abs/2510.05566v2 Large language models have achieved impressive performance across diverse tasks. However, their tendency to produce overconfident and factually incorrect outputs, known as hallucinations, poses risks in real-world applications. Conformal prediction provides finite-sample, distribution-free coverage guarantees, but standard conformal prediction breaks down under domain shift, often leading to under-coverage and unreliable prediction sets. We propose a new framework called Domain-Shift-Aware Conformal Prediction (DS-CP). Our framework adapts conformal prediction to large language models under domain shift, by systematically reweighting calibration samples based on their proximity to the test prompt, thereby preserving validity while enhancing adaptivity. Our theoretical analysis and experiments on the MMLU benchmark demonstrate that the proposed method delivers more reliable coverage than standard conformal prediction, especially under substantial distribution shifts, while maintaining efficiency. This provides a practical step toward trustworthy uncertainty quantification for large language models in real-world deployment. --- ### Generative Dynamic Graph Representation Learning for Conspiracy Spoofing Detection Authors: Sheng Xiang, Yidong Jiang, Yunting Chen, Dawei Cheng, Guoping Zhao, Changjun Jiang Published: 2025-10-07 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2510.05562v1 Spoofing detection in financial trading is crucial, especially for identifying complex behaviors such as conspiracy spoofing. Traditional machine-learning approaches primarily focus on isolated node features, often overlooking the broader context of interconnected nodes. Graph-based techniques, particularly Graph Neural Networks (GNNs), have advanced the field by leveraging relational information effectively. However, in real-world spoofing detection datasets, trading behaviors exhibit dynamic, irregular patterns. Existing spoofing detection methods, though effective in some scenarios, struggle to capture the complexity of dynamic and diverse, evolving inter-node relationships. To address these challenges, we propose a novel framework called the Generative Dynamic Graph Model (GDGM), which models dynamic trading behaviors and the relationships among nodes to learn representations for conspiracy spoofing detection. Specifically, our approach incorporates the generative dynamic latent space to capture the temporal patterns and evolving market conditions. Raw trading data is first converted into time-stamped sequences. Then we model trading behaviors using the neural ordinary differential equations and gated recurrent units, to generate the representation incorporating temporal dynamics of spoofing patterns. Furthermore, pseudo-label generation and heterogeneous aggregation techniques are employed to gather relevant information and enhance the detection performance for conspiratorial spoofing behaviors. Experiments conducted on spoofing detection datasets demonstrate that our approach outperforms state-of-the-art models in detection accuracy. Additionally, our spoofing detection system has been successfully deployed in one of the largest global trading markets, further validating the practical applicability and performance of the proposed method. --- ### HoloScene: Simulation-Ready Interactive 3D Worlds from a Single Video Authors: Hongchi Xia, Chih-Hao Lin, Hao-Yu Hsu, Quentin Leboutet, Katelyn Gao, Michael Paulitsch, Benjamin Ummenhofer, Shenlong Wang Published: 2025-10-07 Categories: cs.CV arXiv: https://arxiv.org/abs/2510.05560v1 Digitizing the physical world into accurate simulation-ready virtual environments offers significant opportunities in a variety of fields such as augmented and virtual reality, gaming, and robotics. However, current 3D reconstruction and scene-understanding methods commonly fall short in one or more critical aspects, such as geometry completeness, object interactivity, physical plausibility, photorealistic rendering, or realistic physical properties for reliable dynamic simulation. To address these limitations, we introduce HoloScene, a novel interactive 3D reconstruction framework that simultaneously achieves these requirements. HoloScene leverages a comprehensive interactive scene-graph representation, encoding object geometry, appearance, and physical properties alongside hierarchical and inter-object relationships. Reconstruction is formulated as an energy-based optimization problem, integrating observational data, physical constraints, and generative priors into a unified, coherent objective. Optimization is efficiently performed via a hybrid approach combining sampling-based exploration with gradient-based refinement. The resulting digital twins exhibit complete and precise geometry, physical stability, and realistic rendering from novel viewpoints. Evaluations conducted on multiple benchmark datasets demonstrate superior performance, while practical use-cases in interactive gaming and real-time digital-twin manipulation illustrate HoloScene's broad applicability and effectiveness. Project page: https://xiahongchi.github.io/HoloScene. --- ### Midway Network: Learning Representations for Recognition and Motion from Latent Dynamics Authors: Christopher Hoang, Mengye Ren Published: 2025-10-07 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2510.05558v1 Object recognition and motion understanding are key components of perception that complement each other. While self-supervised learning methods have shown promise in their ability to learn from unlabeled data, they have primarily focused on obtaining rich representations for either recognition or motion rather than both in tandem. On the other hand, latent dynamics modeling has been used in decision making to learn latent representations of observations and their transformations over time for control and planning tasks. In this work, we present Midway Network, a new self-supervised learning architecture that is the first to learn strong visual representations for both object recognition and motion understanding solely from natural videos, by extending latent dynamics modeling to this domain. Midway Network leverages a midway top-down path to infer motion latents between video frames, as well as a dense forward prediction objective and hierarchical structure to tackle the complex, multi-object scenes of natural videos. We demonstrate that after pretraining on two large-scale natural video datasets, Midway Network achieves strong performance on both semantic segmentation and optical flow tasks relative to prior self-supervised learning methods. We also show that Midway Network's learned dynamics can capture high-level correspondence via a novel analysis method based on forward feature perturbation. ---