# Research Papers — 2026-09 · bundle page 144 of 266 > Full-text markdown of papers #7151–7200 of 13298 published in 2026-09, newest first. AI-readable mirrors of the full 13298-paper 2026-09 archive (50 papers per file). Canonical page: https://aipod.dev/papers/archive/2026-09/page/144 [← Previous bundle](https://aipod.dev/papers/archive/2026-09/page/143.md) [Next bundle →](https://aipod.dev/papers/archive/2026-09/page/145.md) ### Opinion Dynamics-based Coalition Formation for Federated Learning in Heterogeneous IoT Systems Authors: Mohammed El Hanjri, Anas Abouaomar, Hamidou Tembine, Abdellatif Kobbane Published: 2026-09-17 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.19695v1 Federated learning (FL) enables privacy-preserving, on-device training across heterogeneous Internet-of-Things (IoT) deployments such as smart-city water-metering networks, where each smart meter observes a household-specific consumption time series. Under such statistical heterogeneity, the standard Federated Averaging (FedAvg) aggregation averages dissimilar local models into a single global model that may fail to capture client-specific patterns. We address this by forming client coalitions directly in the local-weight space and aggregating at the coalition level. Extending a prior weight-driven coalition-formation scheme, we model coalition formation as a Hegselmann-Krause (HK) bounded-confidence opinion-dynamics process acting on the local weights, and develop variants of the HK interaction based on Euclidean-distance and cosine-similarity confidence criteria. The framework is applied to short-term water-consumption forecasting with local Long Short-Term Memory (LSTM) models and evaluated against FedAvg, Per-FedAvg, FedProx, and FedAvg with Euclidean-distance or cosine-similarity coalition formation. Experiments on a real smart-metering dataset of water consumption show that the proposed HK-based coalition formation produces stable, endogenous coalition structures within at most ten inner iterations, incurs no additional client-side computation or communication compared to FedAvg, and reduces the average MAE by up to 54% relative to FedAvg, 39% relative to FedProx, and 24% relative to Per-FedAvg, while achieving the highest global accuracy (83-85%). --- ### IMFD: End-to-end Multi-Face Forgery Detection through Instruction-based Large Vision-Language Models Authors: Dasom Choi, Sangjun Moon, Hyeongchan Im, Jaeeon Park, Jingun Kwon, Hidetaka Kamigaito, Taro Watanabe, Manabu Okumura Published: 2026-09-17 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.19693v2 The rapid increase of deepfakes has raised significant concerns due to their spread on social media. Traditional multi-face forgery detectors crop and verify each face independently, ignoring background context and inter-face relationships, which often yields suboptimal performance. To overcome these limitations, we leverage instruction-based Large Vision-Language Models (LVLMs), which can interpret entire images and follow complex textual instructions. We propose a simple yet effective single-stage multi-face forgery detector, called IMFD (Instruction-based Multi-face Forgery Detector), which is trained end-to-end to jointly localize faces and predict per-face forgery labels. Rather than treating face box prediction only as a joint objective, IMFD explicitly integrates predicted face bounding boxes into the instruction as visual cues that enhance instruction grounding and forgery detection. To support the training and evaluation of IMFD, we convert existing multi-face forgery datasets into an instruction-based format. Experimental results and analyses show that IMFD improves multi-face forgery detection by integrating face bounding boxes into the instruction, and consistently outperforms various state-of-the-art methods. --- ### UniExo: Unified Multi-Skill Policies for Musculoskeletal Locomotion and Co-Adaptive Exoskeleton Control Authors: Yifei Yuan, Jakob Wolf, Ghaith Androwis, Xianlian Zhou Published: 2026-09-17 Categories: cs.RO, cs.LG arXiv: https://arxiv.org/abs/2609.19690v1 Daily locomotion encompasses diverse activities and frequent transitions between them, yet most exoskeleton controllers are designed for a single activity or a narrow set of related movements. Changes in activity therefore typically require explicit mode switching and separately tuned or retrained controllers. Simulation-based learning reduces the need for hardware-based tuning but generally retains this limitation. Here we present UniExo, a framework that first constructs a multi-skill musculoskeletal human policy and then jointly trains an exoskeleton control policy with it. Four single-skill imitation experts for walking, turning, running and backward walking are distilled into a single network structured by a skill latent and subsequently fine-tuned through reinforcement learning on transition sequences. The resultant unified human policy achieves a mean tracking success rate of 94.7% on unseen clips of the four skills and exhibits greater robustness to perturbations than its constituent experts. A single hip exoskeleton controller (UniExo) is initialized from hip moment prediction of the human policy and co-adapted with it through multi-agent reinforcement learning across the four skills. This co-adaptation shifts the timing of the assistance torque and raises the fraction of positive work delivered to the hip. When deployed on a custom hip exoskeleton, the controller generalizes across four treadmill speeds in six participants and assists one participant through a continuous route of all four skills and their transitions, without skill labels or explicit mode switching. UniExo thus provides a step towards replacing activity-specific controllers with unified, user-specific controllers that support diverse locomotor activities and the transitions between them. --- ### MiX: Micro-Inverted-Scaling for End-to-End Low-Bit Vision-Language Model Acceleration Authors: Yuan Liao, Jae-sun Seo Published: 2026-09-17 Categories: cs.AR, cs.CV arXiv: https://arxiv.org/abs/2609.19683v1 The deployment of Vision-Language Models (VLMs) on edge devices is severely bottlenecked by memory bandwidth, necessitating aggressive sub-8-bit quantization. Since edge accelerators are strictly constrained by area and power, they require end-to-end quantized models. However, the extreme dynamic range gap between multi-modal tokens causes standard block formats to suffer "microscaling collapse," where a single massive outlier hijacks the shared exponent, underflowing surrounding elements and destroying attention maps. To break this bottleneck, we propose Micro-Inverted-Scaling (MiX), a novel format that mathematically inverts the microscaling paradigm: rather than grouping multiple mantissas under one shared exponent, MiX groups private, per-element exponents under a single shared mantissa. To handle asymmetric VLM outlier topologies, we introduce an adaptive dual-format (MiX-MX) inference framework. By algebraically factoring out the shared MiX mantissa, this framework maps to a custom accelerator, replacing multipliers with efficient shifters. Evaluated end-to-end on multiple VLMs, our 4.5-bit MiX formulation exhibits equivalent or superior accuracy on multi-modal benchmarks compared to NVFP4. Simultaneously, the MiX accelerator delivers a 25% improvement in area efficiency over the NVFP4 baseline and a 2.3-4.5x speedup with 1.4-2.9x energy reduction across models compared to the state-of-the-art accelerator Focus, proving the inverted-scaling datapath is physically superior for efficient VLM deployment. --- ### FINSKILLOPS: A Self-Evolving Multi-Agent System for SEC Filing QA Authors: Yanzhang Ma, Zhenghan Tai, Hanwei Wu, Sizhe Guan, Jianliang Lei, Hailin He, Chaolong Jiang, Jijun Chi, Tung Sum Thomas Kwok, Bohuai Xiao, Jingrui Tian, Xinlu Wu et al. Published: 2026-09-17 Categories: cs.AI, cs.IR, cs.MA, cs.SE arXiv: https://arxiv.org/abs/2609.19680v1 Financial QA systems are typically improved before deployment through better retrieval, prompting, or agent coordination, leaving their reliability behavior fixed thereafter. In practice, new SEC-filing questions repeatedly expose heterogeneous errors in period, entity, evidence use, and calculation. Existing self-improvement methods can turn failures into new behaviors, but offer limited control over where a correction should apply or which previously correct answers it may break. We therefore frame post-deployment improvement as controlled behavioral maintenance: recurring failures should become scoped skill patches, and each patch should earn deployment with- out introducing regressions. We instantiate this view in FINSKILLOPS, a multi-agent system for SEC filing QA. FINSKILLOPS derives reusable skills from evidence-grounded, typed failure diagnoses and governs them through targeted validation, protected-case regression checks, negative controls, and versioned replacement or retirement. Across six financial QA benchmarks, a single frozen skill registry achieves the highest verdict-weighted correctness and reference consistency among the evaluated systems. Evolved skills raise correctness from 3.70 to 4.55 on our enhanced benchmark. In a separate 12-round operational study, only six of 33 proposed skills are promoted, while the monitoring non-correct rate falls from 20.0% to 12.5%. These results establish controlled skill scope, admission, and lifecycle management as the foundation for reliable self-improvement. --- ### Conservation Buys Stability and Factoring Buys Counterfactuals in Physical World Models Authors: Yufeng Wang, Parivesh Priye, Lu Wei, Haibin Ling Published: 2026-09-17 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.19674v2 A learned simulator can reproduce its training conditions accurately yet fail in two distinct ways once those conditions change. Over long rollouts, small errors accumulate until the trajectory drifts away from physically plausible behavior; under an intervention on a physical parameter, the model may continue to follow the law seen during training rather than the intervened one. We show that these two failures require different structural remedies. Evolving a learned energy with a symplectic integrator preserves the geometry of the conservative dynamics and keeps rollouts bounded and physically meaningful for up to $100\times$ the training horizon, while equal-capacity predictors, an energy-regularized predictor, and a tuned neural ODE diverge. By contrast, encoding the physical coupling through an explicit linear factorization enables the model to follow a never-seen sign of that coupling, whereas an unrestricted parameterization remains locked to the training law. Crucially, the two mechanisms are separable: removing the structure responsible for long-horizon stability leaves counterfactual transfer intact, while removing the factorized coupling destroys counterfactual transfer without eliminating stability. This double dissociation, established with matched controls that remove or replace one structural component at a time, persists beyond the headline three-body system and remains visible when the physical state must be inferred from pixels rather than provided directly. The result is a concrete design principle for physical world models: long-horizon stability and changed-law generalization arise from distinct structural commitments, and each can be imposed deliberately without requiring the other. --- ### When2Think: Learning When and How Much to Reason Authors: Jaejun Shim, HyunJin Kim, Young Jin Kim, JinYeong Bak Published: 2026-09-17 Categories: cs.AI, cs.LG arXiv: https://arxiv.org/abs/2609.19671v2 Large Reasoning Models (LRMs) often overthink easy problems and underthink hard ones, leading to inefficient computation allocation. Existing methods regulate generated computation or select between direct answering and explicit reasoning, but do not jointly control whether}to reason and how much computation to allocate within reasoning. We call the resulting difficulty-dependent loss in accuracy under computation reduction the efficiency tax. We propose When2Think, an RLVR-based post-training framework for instance-adaptive computation allocation. Its core mechanism, Instance-level Difficulty-Aware Control (IDAC), uses cached reference statistics of success and token cost to modulate a correctness-gated efficiency bonus based on generated token count. Importance sampling supports exploration of Think and NoThink, while Batch-Wise Standardization constructs standardized advantages for critic-free optimization. The framework requires neither a learned reward model nor a learned critic, and offline reference caching avoids online reference-model queries during policy updates. On AIME24, When2Think improves Pass@3 by 10.0 percentage points while reducing token usage by 27.9% relative to the backbone. --- ### CoRe: Coherence and Relational Alignment for Multivariate Time Series Forecasting Authors: Xiaoyu Lin, Huiran Duan, Yining Liu, Zhixiang Wu, Chu Lin, Lin Lu Published: 2026-09-17 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.19670v1 Direct forecasting has become a standard paradigm for multivariate time-series forecasting because it predicts the full future horizon in a single pass. However, its training objective is often still decomposed into pointwise errors such as MSE. Such objectives provide stable supervision, but they do not explicitly preserve the structure of the future trajectory: temporal coherence within each variable and relational consistency across variables can both be weakened. We propose CoRe, a model-agnostic learning objective for direct multivariate forecasting. CoRe replaces pointwise supervision with two output-space constraints: a frequency coherence loss that aligns predicted and target spectra, and a low-rank relational graph loss that matches sampled pairwise differences in a target-derived PCA subspace. The resulting objective introduces no trainable parameters and can be applied to existing forecasting backbones by changing only the loss. Experiments on standard benchmarks show that CoRe improves strong baselines, compares favorably with recent forecasting objectives, and remains effective across different backbones, datasets, and hyperparameter settings overall consistently. --- ### Beyond Patch Removal: Persistent Adversarial Effects in Vision-Language-Action Policies Authors: Enhao Wu, Fusen Guo, Yuxin Cao, Ziyang Lyu, Lin Li, Wei Song Published: 2026-09-17 Categories: cs.CV, cs.RO arXiv: https://arxiv.org/abs/2609.19669v1 Adversarial patches to Vision-Language-Action (VLA) policies can cause both immediate action corruption and persistent state effects that remain after the patch is removed. Existing evaluations largely focus on continuous attacks and do not separate these two effects. We introduce a state-restoration protocol that removes the patch at matched action-chunk boundaries and measures subsequent recoverability under the same remaining step budget. Clean, random-patch, deviation-matched, and fixed-direction controls distinguish adversarial effects from occlusion, action-error magnitude, and directional persistence. We also evaluate a recovery adapter trained on attack-induced states under controlled intervention latency. On OpenVLA-OFT with EDPA attacks, only 36.2% of LIBERO-Long episodes remain recoverable after five chunks, compared with 89.9% and 87.0% for the deviation-matched and fixed-direction controls. Similar persistent effects are observed on autoregressive OpenVLA. The recovery adapter improves recovery from 7.7% to 47.4% at one-chunk latency, but its benefit decreases substantially with delayed intervention. These results show that adversarial effects can persist after patch removal and that timely intervention is critical for recovery. --- ### VideoResearcher: Self-Improving Tool Design for Long-Video Understanding Authors: Dingqiang Ye, Dongdi Zhao, Kaishen Wang, Qingqiao Hu, Jingchen Sun, Yijun Liang, Yuqi Jia, Yiqiao Huang, Yunjie Tian, Jiaxing Zhang, Chuanyang Jin, Ke Zhang et al. Published: 2026-09-17 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.19664v1 Video agents have made substantial progress in long-video understanding. Yet effective video-agent systems require costly, time-consuming manual design and trial and error. Current self-improvement methods either refine low-impact prompts, recombine predefined micro-tools, or struggle with convergence in harness optimization. To bridge this gap, we target high-impact video-tool with VideoResearcher, a training-free multi-agent framework that autonomously designs, tests, and refines tools for video understanding, like a human researcher. VideoResearcher operates through dual Solving and Evolving loops: it analyzes tool-use trajectories to identify capability gaps, coordinates specialized agents to develop and validate executable tools, and reuses evolved tools to strengthen evidence acquisition in subsequent video reasoning. Through iterative tool refinement and validation, it progressively strengthens evidence acquisition without updating model parameters. VideoResearcher achieves state-of-the-art performance among self-improving agents and approaches the human-designed upper bound, demonstrating a training-free paradigm for long-video understanding that expands agent capabilities through autonomous tool development while reducing costly manual engineering. --- ### Towards Active Cross-View Object Geo-Localization Authors: Shunyu Yao, Xiaohan Zhang, Zhuoran Yang, Haoqi Lai, Qi Ming, Xiaoxi Hu, Hui-Liang Shen, Si-Yuan Cao Published: 2026-09-17 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.19662v1 Cross-view object geo-localization (CVOGL) typically assumes a fixed query image, overlooking the ability of mobile agents to actively acquire more informative observations. To address this limitation, we introduce Active Cross-View Object Geo-Localization (ActiveGeo), where an agent sequentially selects new viewpoints and determines when to stop, aiming to improve localization with minimal observations. We further propose ActiveMoPT, an ActiveGeo framework with three-stage training. First, Multi-View Prompt-Preserving Adaptation enables the model to aggregate multiple query views while reusing the initial prompt. Second, Trajectory-Guided Policy Initialization uses supervised agent trajectories to learn viewpoint selection and initial stopping behavior. Third, Cost-Aware Policy Refinement employs GRPO with a gain-cost reward to jointly optimize localization accuracy and observation efficiency. We also construct ActiveGeo-858, a zero-shot test set containing 858 scenes and 1,716 target annotations. Experiments show that ActiveMoPT achieves state-of-the-art performance on MoP-UAV using only 1.45 query views on average, and substantially outperforms previous CVOGL approaches under zero-shot evaluation on ActiveGeo-858. --- ### EmbodiedMind: Adaptive Data Curation and Prefix-Tree Reinforcement Learning for Efficient Embodied Intelligence Authors: Feifan Wang, Zongbing Zhang, Yu Zhang, Lingfeng Wang, Yurui Zhu, Jin Deng, Mingliang Zhang, Zhengguang Gao, Yongcheng Wang, Jin Xu, Ri Yang Published: 2026-09-17 Categories: cs.RO, cs.LG arXiv: https://arxiv.org/abs/2609.19659v1 Training embodied foundation models typically requires massive-scale datasets and extensive computational resources, yet often suffers from three critical limitations: (1) inefficient sample utilization due to low-informative samples; (2) imbalanced gradient contributions across heterogeneous tasks; and (3) severe credit assignment problem in long-horizon planning, where trajectory-level rewards indiscriminately penalize all tokens. To address these issues, we propose an efficient training paradigm that achieves state-of-the-art average performance through strategic data selection and hierarchical policy optimization. Our approach consists of three synergistic stages. First, Rejection Sampling-based Fine-Tuning (RSFT) filters out low-informative samples to establish robust behavioral priors while preventing distributional collapse. Second, Iterative Rejection GRPO (IR-GRPO) employs task-specific queues stratified by difficulty to keep datasets balanced across reinforcement learning iterations, coupled with a hybrid reward mechanism for precise cross-task feedback. Third, to enhance long-horizon task planning, we introduce Trie-GRPO, a novel reinforcement learning algorithm based on action prefix trees, which enables step-level advantage estimation. This resolves the credit assignment problem by isolating intermediate correct decisions from downstream errors, while effectively balancing exploration efficiency and depth compared to conventional search trees. As a result, EmbodiedMind achieves a state-of-the-art average performance of 70.02% across 18 benchmarks, and significantly outperforms other embodied foundation models in long-horizon task planning accuracy. Our project will be released for reproducibility. --- ### PrefixBench-H100: Characterizing Prefix Reuse and Time-to-First-Token in H100 LLM Serving Authors: Omkar Shewale, Deepak Kumar, Divakar Kumar Yadav Published: 2026-09-17 Categories: cs.PF, cs.LG arXiv: https://arxiv.org/abs/2609.19657v1 Repeated prompt prefixes are increasingly common in LLM serving workloads, appearing in system prompts, templated retrieval-augmented generation pipelines, agent frameworks, and multi-turn conversations. Modern inference runtimes such as vLLM and TensorRT-LLM provide mechanisms for reusing previously computed KV-cache state across requests, yet it remains unclear when prefix reuse materially improves serving performance on contemporary accelerators and when its benefits are limited by scheduling, cache granularity, concurrency, or memory pressure. This paper presents PrefixBench-H100, a reproducible benchmark and measurement framework for characterizing prefix reuse on a single NVIDIA H100. PrefixBench-H100 combines controlled synthetic traces with chat-style and retrieval-style workloads, and evaluates two widely used LLM serving runtimes under matched workload conditions. The benchmark varies shared-prefix length, suffix diversity, request arrival pattern, concurrency, output length, and cache configuration, while collecting time-to-first-token, inter-token latency, end-to-end latency, throughput, cache-hit statistics, GPU memory usage, and selected profiling traces. The goal of PrefixBench-H100 is not to introduce a new caching algorithm, but to expose the practical operating envelope of prefix reuse for H100-class LLM serving. The study identifies the regime where prefix reuse provides substantial first-token latency reductions and the regime where cache pressure erodes them, while showing that cache effectiveness itself is largely insensitive to concurrency and output length; the cross-runtime differences that remain arise above the cache, in the scheduling layer. --- ### Self-Evolving Search Index Authors: Sangam Lee, Wonjae Lee, Sunghwan Kim, Deogyong Kim, Jaehoon Kim, Daye Nam, SeongKu Kang, Dongha Lee Published: 2026-09-17 Categories: cs.IR, cs.AI arXiv: https://arxiv.org/abs/2609.19656v1 Information retrieval is increasingly important as LLM agents tackle complex tasks involving diverse information needs. Because retrieval relies on an index that represents each document through index keys, retrieval quality depends heavily on how effectively these keys expose the knowledge contained in each document. However, effective index representations vary across retrieval environments, making it difficult for any fixed optimization strategy to perform consistently. Yet evolving an index to its retrieval environment remains largely human-driven, requiring humans to diagnose retrieval failures, refine the optimization strategy, and reprocess the index accordingly. We propose SELF-INDEX, a framework that enables an index to self-evolve without human intervention. Its Optimizer autonomously diagnoses retrieval shortfalls, selectively revises the responsible index keys, and validates each revision before updating the index. Beyond reacting to observed retrieval demands, SELF-INDEX proactively explores additional demands through a Query Simulator, allowing the index to evolve beyond the queries already available for optimization. Across diverse corpora and retrievers, SELF-INDEX consistently improves retrieval performance while outperforming existing index optimization methods. We further show that these benefits extend to downstream applications, improving the effectiveness and efficiency of search agents and helping agent memory systems retrieve useful past interactions. --- ### Replan, Repair, or Edit? A Unified Empirical Evaluation of Travel Agents for Itinerary Revision under Resource Disruptions Authors: Xiaofei Yuan, Yan Zhang, Shaobo Qiao, Huangleshuai He, Leyan Ni, Mingchen Ju, Lujia Yang, Sijia Xu, Yifu Tang, Zhengyi Yang Published: 2026-09-17 Categories: cs.AI, cs.MA arXiv: https://arxiv.org/abs/2609.19654v1 Travel-planning agents generate itineraries that may become infeasible after acceptance because of flight cancellations, hotel unavailability, or attraction closures. Revising these itineraries involves full replanning, classical plan repair, and LLM-based travel-agent revision, whose differing task formulations and evaluation protocols hinder comparison. We conduct a systematic empirical study using two TREK-derived benchmark sets: 500 single-disruption cases, including feasible and infeasible instances, and 200 feasible simultaneous compound-disruption cases. We compare LLM-Z3 full replanning, IPyHOPPER hierarchical repair, and an iTIMO local-revision adapter across effectiveness, plan stability, and computational cost. LLM-Z3 with Gemini achieved the highest observed compound-disruption success. IPyHOPPER nearly matched that configuration's single-disruption overall success, while preserving substantially more of the accepted itinerary on successful repairs. Successful hierarchical and local repairs made fewer edits and retained more accepted commitments than full replanning. Computational profiles differed: IPyHOPPER used no LLM inference, the evaluated LLM-Z3 adapter used compact one-call inference, and the iTIMO adapter consumed substantially more tokens. The study provides practical guidelines for balancing feasibility recovery, commitment preservation, and computational cost within evaluated settings. --- ### Improving Cross-Lingual Transfer for Sequential Sentence Classification in Research Papers via Structural Similarity Authors: Kazuhiro Yamauchi, Marie Katsurai Published: 2026-09-17 Categories: cs.CL, cs.DL arXiv: https://arxiv.org/abs/2609.19650v1 Sequential sentence classification (SSC) is an essential task for structuring scientific publications, and extending SSC research to languages other than English can improve accessibility to scientific knowledge in multilingual digital libraries. Cross-lingual transfer is a promising approach to address the scarcity of training data in non-English languages. Prior work on other natural language processing tasks has shown the benefits of capturing linguistic similarity between source and target languages. However, SSC inherently depends on patterns at the discourse level, such as label sequences and positional regularities, which appear consistently across languages regardless of linguistic differences. To examine the factors that determine transfer success in SSC, we constructed a multilingual SSC dataset covering 13 non-English languages collected from five academic databases. Our cross-lingual transfer experiments, using both encoder-based and generative models, show that linguistic proximity has no consistent predictive power for transfer performance, whereas structural similarity in rhetorical organization shows a weak but consistent positive correlation across models. After controlling for source-language performance, the similarity of label distributions is the most consistent predictor. Building on this finding, we propose a set of three methods that explicitly leverage structural information using generative models. In the in-domain evaluation, the best combination reaches parity with the strongest encoder baselines, and in transfer to languages unseen during training, it outperforms the strongest encoder baseline. --- ### Well-posedness of neural turbulence closures and tangent dissipation Authors: Zhen Zhang, George Em Karniadakis Published: 2026-09-17 Categories: physics.flu-dyn, cs.LG arXiv: https://arxiv.org/abs/2609.19647v1 A neural turbulence closure defines a new boundary-value problem, $R(U)=N(U)+F(U)=0$, with a coupled Jacobian $J(U)=N'(U)+F'(U)$, where $N$ is the original mean-flow operator and $F$ the learned closure. We establish two consequences of global tangent dissipation. For a monotone original operator, a positive uniform margin supplied by the original operator and closure together guarantees existence, uniqueness and a global inverse-sensitivity bound relating a posteriori solution error to the a priori residual. For a general original operator, a dissipative closure cannot worsen tangent dissipation, but this alone does not guarantee uniqueness. Tangent dissipation depends on both diffusion and reaction. We study two complementary ways to promote it: (1) an exact-integral construction enforcing non-negative tangent diffusion while leaving reaction unconstrained, and (2) a penalty on tangent-reaction violations at sampled states. Tangent diffusion enters the Jacobian, and non-negative secant eddy viscosity alone does not control its coercivity. We conduct tests with channel flow at $Re_τ=180$--$5200$, which provides a strongly monotone baseline. Both constrained closures reach accurate solutions in all 50 training-seed/Reynolds-number cases. At $Re_τ=1000$, we conduct tests with 10,000 starts for one fixed network per closure and we find one root for each constrained closure and multiple roots for the other closures. Although this does not prove uniqueness, it provides strong empirical evidence for uniqueness of the tested constrained closures. At $Re_τ=5200$, the construction and penalty reduce the reported inverse sensitivity relative to the original operator by approximately $372\times$ and $11\times$, respectively. --- ### ScientistTwo: Pioneering the Human Knowledge Frontier with Autonomous AI Authors: Jaehyun Nam, Jinsung Yoon, Yanzhou Pan, Yubo Wang, Rui Meng, Parthasarathy Ranganathan, Tomas Pfister Published: 2026-09-17 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.19644v1 Scientific discovery is defined by the ability to identify the boundaries of existing knowledge and venture into unexplored territory. The ultimate vision for AI in science is problem-driven autonomous research: given a fundamental challenge by a human expert, the AI independently navigates the scientific landscape, uncovers theoretical and empirical bottlenecks, and systematically expands the frontier of knowledge. In this paper, we introduce ScientistTwo, a fully autonomous multi-agent framework designed to realize this vision. Specifically, ScientistTwo takes an initial problem as input, establishes state-of-the-art baselines, formulates novel hypotheses, and coordinates specialized agents to orchestrate an end-to-end discovery cycle without human intervention. Moreover, the framework rigorously conducts experiments using diverse datasets and metrics, refines methodologies through automated ablation studies, and validates research findings via a closed-loop simulated peer-review rebuttal engine. To evaluate ScientistTwo's capabilities against the highest standards of human scientific achievement, we benchmark it across papers accepted at top-tier conferences such as ICLR, ICML, and NeurIPS. As a result, ScientistTwo autonomously generates expert-level, publishable papers and fully verified, executable codebases. Its solutions consistently outperform human state-of-the-art models, and achieve higher average review ratings than human-authored papers under automated AI review agents. These results show that ScientistTwo is not merely an assistive tool but an autonomous scientific pioneer capable of pushing the frontiers of human discovery. Project website: https://scientist-two.github.io/ --- ### Improving Sample Efficiency in Peptide-HLA Binding Prediction with Hybrid Quantum-Classical Neural Networks Authors: Chenyan Jia, Cong Guo, Siyue Chen, Pengpeng Ye, Xiaochun Chen Published: 2026-09-17 Categories: quant-ph, stat.ML arXiv: https://arxiv.org/abs/2609.19642v1 Peptide-HLA binding prediction is a critical step in neoantigen identification for personalized cancer immunotherapy and holds significant clinical value. However, the training data available for many HLA alleles are extremely limited, which severely constrains the performance of conventional methods on this task. Parameterized quantum circuits are hypothesized to induce inductive biases beneficial for learning from small datasets, yet their application to biological sequence prediction remains underexplored. To address this, we propose a hybrid quantum-classical neural network (HQNN) specifically designed for peptide-HLA binding prediction. HQNN integrates multi-source biological feature encoding with parallel quantum feature extractors and a quantum-enhanced classifier. On two HLA alleles (A*02:01 and B*07:02), HQNN outperforms a parameter-matched classical CNN baseline across all training sizes, with the performance gap widening as training data decreases. Ablation studies confirm the respective contributions of the quantum feature extraction module and the quantum classifier. In noise-aware simulations, performance degrades only mildly, and such degradation is reasonable and acceptable under realistic quantum hardware noise levels. These results suggest that hybrid quantum-classical architectures can provide practical sample-efficiency gains for immunoinformatics tasks in low-data regimes. --- ### A Policy Profile for Croissant: Refusal as a Property of the Dataset Authors: Alexander Chernov Published: 2026-09-17 Categories: cs.LG, cs.CR, cs.DB arXiv: https://arxiv.org/abs/2609.19640v1 Croissant is the de facto machine-readable descriptor for ML datasets: JSON-LD over schema.org. Since version 1.1 it also carries data use conditions, recommending DUO and ODRL for them. What no version specifies is how any of them is evaluated: no decision procedure, no bound on evaluation cost, no outcome for a condition an implementation cannot evaluate, no record of what was checked, and nothing on composition with caller-side authority. We supply that half. An additive profile lets a dataset declare the operations it admits and the conditions under which it admits them, over a closed set of five operators whose decision procedure is given in full, so a gate decides from the descriptor alone and records what it checked. Two corpora evaluate it and their evidence is kept apart. Three descriptors that gated a real nf-core pipeline give the deployment result: decisions from a profile document match the gate's native descriptor record for record, stripping the layer leaves a valid Croissant document, and the added cost is 11.7 $μ$s against a 119 $μ$s decision. A corpus generated from the profile's grammar gives the breadth, covering every operator, refusal class and conformance clause. Across its valid cases, 552 complete decision records agree three ways -- native descriptor, profile terms, and the same policy as ODRL in usageInfo. The carrier is therefore not the contribution; the evaluation semantics is. Finally, caller-bound and data-bound policies range over non-overlapping state spaces, so neither permit set contains the other. --- ### Reach or Solve? Attributing Agentic RL Gains with Checkpoint Handoffs Authors: Xuan Liu, Jingbin Qian Published: 2026-09-17 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.19636v1 Reinforcement learning now trains language-model agents that act over dozens of steps in live environments. The gains are large, and they are read as better decision-making. An agent in a closed loop writes its own inputs. Each observation follows from its own earlier actions, so the states it meets late in an episode are partly of its own making. An SFT checkpoint and an RL checkpoint are then scored from different states, even on identical tasks. Endpoint success mixes two changes: where the agent arrives, and what it does once it is there. Restricting the comparison to states both policies reach does not separate them. That restriction selects on an outcome, and in our data it flips the sign of the effect. We introduce checkpoint handoff, an evaluation protocol that clones a state one released checkpoint reached and hands it to another, with no retraining. Crossing a reacher role and a solver role over SFT and RL splits an endpoint gain into REACH and SOLVE. REACH is how often a policy arrives at a state the environment confirms is a fixed number of actions from success. SOLVE is how often it finishes from an identical cloned state. Across two benchmarks and two independently released pipelines, the reacher by solver interaction is positive in all five conditions. An RL history is worth more to an RL solver than the same history is to an SFT solver. On ALFWorld, RL improves both terms, and the SFT solver never succeeds where the RL solver fails. Independent REACH and SOLVE gaps predict the aggregate interaction. Handoff asks only that one checkpoint's history can be replayed under another, so long-horizon evaluation can report arrival and completion beside endpoint success. --- ### Scientific Image Quality Assessment via Multi-modal Retrieval-Augmented Generation Authors: Yinuo Zhang, Bingshuo Liu, Zhiying Tu, Dianhui Chu, Qingbin Liu, Xi Chen, Jiang Bian, Xiaoyan Yu, Dianbo Sui Published: 2026-09-17 Categories: cs.CV, cs.CL arXiv: https://arxiv.org/abs/2609.19634v1 This paper proposes a Retrieval-Augmented Generation (RAG) framework for scientific image quality assessment, designed to simultaneously address both the understanding track (SIQA-U) and the scoring track (SIQA-S) of the SIQA challenge. We construct a multimodal index that integrates textual semantics with fine-grained visual features, and develop a multi-route retrieval and fusion mechanism to provide large language models with highly relevant reference cases, thereby enhancing their capability to evaluate complex scientific images. Experimental results demonstrate that the proposed framework effectively aligns with the judgment criteria of human experts. Ultimately, our method achieves 1st place in the SIQA-U track of the SIQA challenge at the ICME 2026 Grand Challenges. --- ### Instance Segmentation and Fine-grained Classification for Urban Buildings with Adaptive Region Dividing and Spatially-Supervised Contrastive Learning Authors: Weiyuan Zhang, Qi Zhang, Hui Huang Published: 2026-09-17 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.19631v1 Accurate instance-level and functional understanding of urban buildings in large-scale point clouds is essential for digital city modeling and urban analysis. However, the extensive spatial coverage of urban scenes leads most existing methods to rely on predefined blocks for training and evaluation, although such partitions are rarely available in real-world applications and introduce additional preprocessing while fragmenting complete building structures. To address this issue, we propose an adaptive region-dividing strategy with unified scene-level evaluation. Specifically, the 3D point cloud is projected onto a bird's-eye-view (BEV) plane, where a pretrained segmentation model is used to detect building regions. The detected bounding boxes are then back-projected to the original point cloud to construct structure-aligned adaptive training blocks, enabling semantically guided dynamic partitioning without manual design. Furthermore, beyond instance-level understanding, few methods have explored fine-grained classification for urban buildings, and thus we also put forward a fine-grained classification model for urban buildings with a spatially-supervised contrastive loss. First, for each segmented building, a point transformer classifier jointly encodes its body and local context using geometric, color, and core-context information. Then, the class-balanced weighted cross-entropy is used to alleviate severe class imbalance. The proposed spatially-supervised contrastive loss further enhances inter-class discriminability by assigning greater weight to spatially proximate, same-category buildings, encouraging compact functional representations while separating easily confused categories. Extensive experiments on UrbanBIS and STPLS3D demonstrate the advantages of the proposed method in building instance segmentation and fine-grained classification compared to existing SOTA methods. --- ### From Intent to Action: Benchmarking LLM Safety in Vehicle Voice Command Authorization Authors: Diba Afroze, Xingli Zhang, Yazhou Tu, Xiali Hei Published: 2026-09-17 Categories: cs.AI, cs.CL, cs.RO arXiv: https://arxiv.org/abs/2609.19630v1 Large language models (LLMs) are increasingly integrated into vehicle voice assistants. But linking natural-language requests to vehicle functions creates a safety-critical authorization problem. Before executing a command, the system must choose whether to execute, refuse, clarify, require confirmation, defer to manual control, trigger an emergency response, or make no tool call. To our knowledge, prior evaluations do not isolate this pre-action decision across speaker role, authentication status, vehicle state, and tool availability. We introduce a 202-scenario benchmark with Reference Decisions under a seven-class taxonomy. We evaluate two local open-weight models and three API-based LLMs using Decision Alignment and safety-specific error metrics. Alignment ranges from 40.1% for Llama 3.2 3B to 89.1% for Gemini 3.1 Pro Preview. The API-based models score between 83.2% and 89.1%, with no statistically significant differences among them. Even these models produce two to three False Executes among 161 non-execution scenarios, and persistent errors remain in confirmation and manual-control decisions. A controlled Llama 3.2 3B ablation increases alignment to 40.1% under the structured authorization policy, versus 28.2-29.2% under schema-only and generic-safety baselines, but it does not eliminate False Executes. Structured LLM decisions are therefore insufficient as a standalone safety mechanism, and deployment requires an independent enforcement layer that verifies tool permissions and vehicle-state constraints before invoking any vehicle function. --- ### VGGT-GS SLAM: Uncalibrated Monocular Gaussian Splatting SLAM with Feed-Forward Priors Authors: Yuhang Han, Hao Wang, Jiaxi Cao, Xingyu Liu Published: 2026-09-17 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.19628v1 We present VGGT-GS SLAM, a monocular 3D Gaussian Splatting SLAM system designed for uncalibrated videos. Starting from feed-forward VGGT pose and depth priors, our system performs submap differentiable bundle adjustment that jointly refines camera poses and a 3D Gaussian map, while optimizing submap-shared intrinsics and radial--tangential distortion through analytic calibration Jacobians. To improve global consistency, we introduce Gaussian-native alignment (GNA) for camera-anchored scale refinement between sequential submaps and verification of loop-closure candidates. Extensive experiments on standard indoor benchmarks show consistent improvements in localization accuracy and strong rendering quality under uncalibrated settings, establishing a strong baseline for uncalibrated Gaussian SLAM. --- ### DataCanvas-EDU: An Agentic Framework for Instructor-Guided Synthetic Data Generation in Business Analytics Education Authors: Bang An, Maria Hamdani, Joseph Fox Published: 2026-09-17 Categories: cs.HC, cs.AI arXiv: https://arxiv.org/abs/2609.19617v1 Business analytics education requires diverse datasets to support different learning objectives, student backgrounds, and analytical tasks. Real-world data can be difficult to obtain and offer limited flexibility for adapting a case to a particular course. Even when suitable data are available, instructors must investigate the patterns, verify the results, and prepare assignments and reference solutions, requiring substantial time and effort. The use of large language models (LLMs) introduces an additional concern about training data contamination. Widely used public datasets often have extensive tutorials and worked analyses that models may have encountered during training. Students may therefore receive explanations drawn from existing analyses without practicing how to investigate unfamiliar data in collaboration with AI. This paper presents DataCanvas-EDU, an agentic framework for instructor-guided synthetic data generation in business analytics education. Instructors specify teaching goals and intended patterns through conversation, while an AI agent writes generation code, checks the resulting data, and prepares assignments, reference analyses, and rubrics. Four phases, Plan, Create, Verify / Test Analysis, and Evaluate, organize the process and support instructor review and revision. The framework is intended to simplify case preparation while creating opportunities for students to investigate newly designed patterns with AI. We illustrate the approach with WindowDash, a food delivery case containing 15,000 orders and nine designed patterns. DataCanvas-EDU is packaged as a reusable AI Agent Skill for compatible agent environments, with the package and installation instructions available at https://github.com/BANG23333/datacanvas-edu --- ### The Complexity Kink: A Prompt-Side Structural Complexity Index for Code-Generation Reliability Authors: Michael Hernandez, Tian Zhao Published: 2026-09-17 Categories: cs.LG, cs.SE arXiv: https://arxiv.org/abs/2609.19616v2 Complexity measured from generated code is failure-dependent: a difficult prompt can yield a short failing program and be assigned low output complexity. We introduce a six-dimension prompt-side structural-complexity index scored before generation and kept separate from correctness. We select 5,000 Python prompts across six bands of a preliminary single-rater rubric. Four out-of-panel LLM raters rescore the locked prompts, giving 19,997 score rows; composite inter-rater reliability is ICC = 0.872 on the 4,998 prompts with all four ratings. We evaluate 21 models per prompt, yielding 105,000 generations. In the unadjusted mean-pooled analysis, pass rate has a nonmonotone breakpoint at composite 13.75, with 79.9% at or below and 87.6% above. This is not a universal failure cutoff. Task-type fixed effects shift the breakpoint to 10.75 and cut the regime gap from 7.6 to 2.1 points. A construction-frame control shifts it to 8.50 with a raw gap of -3.5 points, and neither frame alone reproduces the pooled +7.6-point change. Model-specific fits include 16 upward and five downward changes. A 365-prompt audit-clean extension matches the original five-model estimates at bins 15 and 16 but adds only 14 prompts above bin 16. Among zero-pass generations with computable Lizard complexity, 28.5% pair a prompt composite above 8 with output complexity at most 10. Human agreement is moderate and rater-dependent on a disagreement-enriched calibration set; paraphrase and cross-language rescoring preserve score ordering. Overidentification tests reject the joint restrictions on the six dimensions, so we treat the composite as an index and make no causal interpretation of the 2SLS estimates. The contribution is a pre-generation measurement framework and a bounded observational analysis of reliability regimes. --- ### Semantic Layer Induction from Raw Telemetry via Hierarchical LLM and RAG Abstraction Authors: Yuanzhe Jia, Ali Anaissi Published: 2026-09-17 Categories: cs.CL, cs.AI, cs.IR, cs.SE arXiv: https://arxiv.org/abs/2609.19615v1 Modern applications generate massive volumes of raw telemetry data, but translating those noisy, heterogeneous event streams into actionable business insights remains a fundamental challenge. Data engineers and analysts expend substantial effort reconciling semantic discrepancies, hand-crafting parsing logics, and maintaining fragile mappings between raw data and business KPIs. In this paper, we present an end-to-end framework that fully automates the construction of a business semantic layer from application raw logs. Our approach introduces a two-stage semantic abstraction: first, high-level business features are identified via LLM inference augmented with domain-specific industry knowledge; second, fine-grained business nodes are derived through a structured pipeline comprising data refinement, hybrid retrieval, multi-stage filtering, semantic clustering, and canonical naming. Evaluation on production-scale telemetry demonstrates that our system improves human-assessed semantic quality from 50 to 80+ on a 100-point scale, reduces maintenance effort by 80%, filters out 74% of noise, and achieves 0.87 Cohen's kappa via an integrated LLM-as-Judge evaluation, enabling continuous, scalable quality assurance. Overall, our work distinguishes itself from prior work by addressing the novel problem of business semantic layer induction from raw telemetry, operating without labeled training data or manual rule engineering. --- ### TacSushi: Tactile-Grounded World-Action Modeling for Dexterous Sushi Manipulation Authors: Haodi Hu, Kaen Kogashi, Toshiaki Koike-Akino Published: 2026-09-17 Categories: cs.RO, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2609.19613v2 Dexterous food manipulation requires control under deformation, occlusion, and uncertain contact. We present TacSushi, a tactile-grounded, Cosmos3-based world-action policy that learns from recorded future consequences while acting on current observations. The backbone encodes current RGB, language, and hand state, and feature-wise gated fusion incorporates fingertip tactile features into the action representation. During training, a decoder conditioned on demonstrated action chunks predicts logged future visual observations, task progress, relative contact risk, and tactile summaries; this decoder is removed at deployment. Failed trials provide consequence supervision, but their actions are excluded from imitation. We train TacSushi on 340 successful and 50 failed real-robot trials and compare six methods in 600 separate rollouts across three in-distribution tasks and two out-of-distribution ingredient variants. To assess food quality beyond a single geometric threshold, we score terminal outcomes using an anchored visual-quality protocol that equally weights five human ratings and three vision-language-model ratings per rollout. Full TacSushi achieves 68.3% average in-distribution success and 37.5% out-of-distribution success, compared with 36.7%/10.0% without future-consequence supervision and 25.0%/17.5% with direct tactile concatenation in place of gated fusion. These comparisons support complementary benefits of feature-wise gated tactile fusion and training-only predictive supervision. --- ### The Output-Space Hypothesis: Enumerative Equivalence Checking for Tensor Programs Authors: Paul Biberstein, Joseph Devietti, Mayur Naik Published: 2026-09-17 Categories: cs.PL, cs.LG, cs.SE arXiv: https://arxiv.org/abs/2609.19611v1 Tensor programs, as used in deep learning models, are a prime target for optimization, as small performance improvements can have a large impact across training or inference workloads. However, such optimizations are complicated and can produce subtle bugs. Traditionally, correctness is assumed when differential testing against a reference on random inputs fails to reveal bugs. However, the inputs to these programs are massive tensors, and finding bugs can require generating extremely low likelihood inputs with precise relationships among their values. We propose a novel way to find bugs more consistently by flipping the quantifiers. Rather than generating a single input and checking all output tensor locations for equivalence, what if you could check a single output tensor location's equivalence for all inputs? We implement this idea in a system, \dirigo, by using a novel symbolic execution strategy. We demonstrate that \dirigo can find bugs effectively in a public dataset of 6,988 AI-written CUDA kernels that are all marked correct by differential testing. Of these, \dirigo finds 600 kernels that are actually buggy, and finds 97.3\% of those bugs within two minutes. --- ### SIMLIFE: Pattern Understanding for Long-Horizon Human-Agent Partnership Authors: Run Peng, Zinnia Nie, Jing Ding, Yinpei Dai, Yichi Zhang, Zengqing Wu, Yao Fu, Ziqiao Ma, Jiayuan Mao, Joyce Chai Published: 2026-09-17 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.19610v1 Understanding humans over long horizons requires agents to infer not only what people need in the moment, but also how routines form, why they repeat, and when they change. We introduce SimLife, a scalable platform for simulating long-term household life with rich visual observations, ground-truth action logs, and synthetic dialogues with audio. Built on SimLife, SimLife-BP evaluates long-context pattern understanding: the ability to infer latent behavioral rules from weeks or months of everyday observations. The benchmark contains 106 episodes averaging 15.49 hours and 38.57 in-game days, and 1,439 question-answer pairs. Each task probes direct, counterfactual, noisy, and inverse reasoning under different levels of rule hints. Evaluating frontier models and architectures, we find that current models often achieve surface-level prediction without comprehensive rule understanding, rely on frequency-based heuristics rather than if-then reasoning over evidence, and struggle to adapt when behavioral patterns change. These findings suggest that long-context pattern understanding remains a major bottleneck for future embodied agents, while SimLife opens a broader space for studying memory, personalization, adaptation, and long-horizon planning in everyday human-AI interaction. --- ### DeltaSelect: Affordable A/B Testing for Coding Agents Authors: Nicholas J. Conn Published: 2026-09-17 Categories: cs.SE, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2609.19607v1 Coding-agent benchmarks are built for broad and comprehensive comparisons, not frequent development decisions. Individual runs vary, full suites are expensive, and the benchmark harness may differ from the harness used in practice. In a resampling analysis of DeepSWE's published trials, only 19.5% of tasks (22 of 113) had a fifth-percentile Pearson correlation of at least 0.50 with full-benchmark performance. The paper presents DeltaSelect, an open-source method that identifies tasks whose one-run results consistently track full-benchmark performance using Pearson correlation, maps fractional verifier results to a common score using linear regression, and selects a fixed task set within a dollar budget. DeltaSelect is intended for repeated baseline-versus-candidate comparisons during development, not model rankings. In a gpt-5.6-luna low-reasoning case study, DeltaSelect was used to revise custom skills and instructions. Across 13 evaluations, the recorded cost was USD 27.86 at rates published August 16, 2026. The adopted version cost 58.1% less than the initial version (USD 1.75 versus USD 4.18; p=0.008), while the calibrated score was higher (42.36% versus 36.46%; published-analog variance p=0.326). --- ### Chain-of-Thought Entropy as a Reliability Signal: A Preregistered Reproduction Authors: Theodore O. Cochran Published: 2026-09-17 Categories: cs.CL, cs.LG arXiv: https://arxiv.org/abs/2609.19606v1 This empirical study is an independent reproduction of the dissociation Zhao reported in 2026. The shape of a large language model's chain-of-thought entropy trajectory predicts whether the final answer is correct, while the magnitude of its total entropy drop does not. The dissociation merits reproduction because the magnitude half rests on a single 300-problem run with one model at one seed, while the shape half was reported at full scale on both benchmarks and on a second model family. Registered at OSF before any confirmatory run, the reproduction crosses the complete GSM8K and MATH-500 benchmark test sets with four open-weight models including one reasoning-distilled model of a kind the original did not test. The shape signal replicates. The magnitude signal divides by setting. On the anchor model the accuracy gap between monotone and non-monotone chains is +9.6 percentage points on GSM8K and +27.5 on MATH-500, while the rank correlation of the total entropy drop with correctness is -0.018 on GSM8K and +0.414 on MATH-500. On the reasoning-distilled model the binary form of the shape signal fires on about one chain in a hundred, too few to estimate the registered contrast, while the graded violation count remains predictive there. In an exploratory comparison the final-step entropy alone outperforms the binary shape flag in all eight model-by-benchmark cells by ROC area, and in six or seven by the risk-coverage area the original reports, depending on an integration range the original does not state. The study contributes a reproduction of the shape signal at full test-set scale under seven documented protocol differences, a map of the settings where the magnitude signal holds and fails, and measurements of four protocol dependencies the original does not report. --- ### Full-Duplex Speech Models Take the Floor When Asked, Not When Needed Authors: Linkai Peng, Baorian Nuchged, Kaiqi Fu, Yuyang Yao Published: 2026-09-17 Categories: cs.CL, cs.HC arXiv: https://arxiv.org/abs/2609.19596v1 Full-duplex speech models listen and speak at once, promising always-on assistants. Yet they must also decide when they should speak. Human listeners speak when addressed or when the speaker stops, but also self-select to correct a false claim, supply a missing word, or warn of danger. We ask whether full-duplex models do the same. To separate the reason to speak from the opportunity, we construct context-matched English monologues in which only the trigger utterance varies within a topic, define 10 conditions from turn-allocation rules, and compress inter-word pauses to limit opportunities created by silence. Across five model families, being addressed and silence are far more reliable triggers than false facts or hazards. Frame-level text-token probabilities in Moshi and PersonaPlex are lower for false facts than for Neutral when averaged over the first 2\,s after trigger end. Pauses or permission to interrupt do not close this gap either. Given the floor, Moshi and PersonaPlex answer most direct questions, yet the proportion of non-empty false-fact replies that challenge the claim is only .14--.15, and the proportion of hazard replies that warn of danger is .04--.07. This paper thus identifies a gap in both speech initiation and response content. Closing it requires genuine content understanding and intervention decisions grounded in it. --- ### Selective Cotton Boll Localization for Robotic Harvesting: Evaluation of Deep Learning Vision Models Under Field Conditions Authors: Thevathayarajh Thayananthan, Xin Zhang, Isuru Laddusinghe Badu, Jonathan Harjono, Glen C. Rains, Beiwen Li, Leonardo M. Bastos, Nuwan K. Wijewardane, Vitor S. Martins Published: 2026-09-17 Categories: cs.CV, cs.RO arXiv: https://arxiv.org/abs/2609.19592v1 This study developed and evaluated a deep-learning-based perception framework for selective robotic cotton picking. The dataset contained 1,008 annotated field images collected using three cameras under varying natural lighting and weather conditions. Object-detection models from the YOLOv8 through YOLOv13 families were evaluated using their default configurations, while segmentation performance was assessed using YOLOv8-seg, YOLOv11-seg, YOLOv12-seg, the Segment Anything Model (SAM), SAMv2.1, FastSAM, and Grounded-SAM with the Recognize Anything Model (RAM). Among the detection models, GELAN-s achieved the most favorable balance between mean average precision (mAP) and inference speed, obtaining an mAP of 86.1%, precision of 81.6%, recall of 76.6%, and an F1-score of 79.0%, with an average inference time of 42.3 ms per image. Among the direct segmentation models, YOLOv12-m-seg provided the most favorable balance between AP@0.5 and FPS, achieving a segmentation AP@0.5 of 83.7% with an inference time of 20.4 ms per image. In the detection-prompted segmentation approach, bounding-box prompts generated by GELAN-s improved the localization of cotton bolls for SAM and SAMv2.1, while SAMv2.1 Tiny consistently outperformed FastSAM and Grounded-SAM with RAM. In the area-based evaluation against manually annotated segmentation masks, YOLOv12-m-seg achieved an $R^2$ value of 0.966, compared with 0.860 for GELAN-s + SAMv2.1 Tiny. Field experiments conducted using a UR5e robotic manipulator, a custom end-effector, and a ZED2i stereo camera further validated the effectiveness of the YOLOv12-m-seg model for real-time cotton boll detection, segmentation, and selective picking under varying confidence levels. These results demonstrate that YOLOv12-m-seg provides an efficient perception model for robotic cotton harvesting and has strong potential for field deployment. --- ### Form Over Content In Gradient-Based Data Attribution Methods Authors: Sunwoo Kim, Seokwon Jung, Sohyung Kim, Seong Joon Oh, Alice Oh Published: 2026-09-17 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2609.19589v1 Data attribution methods using gradient similarity are widely used to analyze and select training data for large language models, but what gradient similarity actually measures is debated. Some interpret it as identifying task-relevant skills, while other work reports that surface form is the main factor. We resolve this debate for supervised fine-tuning examples by varying task and answer format independently. Specifically, we render benchmarks in different answer formats, such that datasets can share a task without a format or a format without a task. We find that gradient alignment follows the answer format, as benchmark pairs sharing an answer format align strongly (disattenuated cosine near 0.4), while same benchmarks rendered with different answer format classes show no alignment (near 0.0). We demonstrate that this ordering holds from the earliest pretraining checkpoints through post-training, and across model scales and families. We then analyze the released selections of LESS, a gradient-based data selection method for instruction tuning, and find that each target's selections over-represent the target's own answer format. Hence, we demonstrate that gradient-based attribution methods track format similarity more than task semantics, meaning that such methods, as well as the semantic interpretation of the gradient, should be tested on data where answer format and task vary independently for greater robustness and reliability. --- ### Red-Teaming Auto Mode: Improving Blocking Classifiers Against Malign Coding Agents Authors: Alex Remedios, Simon Storf, Fabien Roger, John Hughes Published: 2026-09-17 Categories: cs.CR, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2609.19587v2 To keep coding agents from going off the rails, production systems now review each proposed action with a blocking monitor that can reject it before it runs (Auto Mode in Claude Code, Guardian in OpenAI's Codex). Prior evaluations of such monitors largely measure robustness to accidental harm or prompt injections from untrusted sources looking to hijack the agent. Less understood is how they hold up when the agent they monitor is persistently misaligned. To understand this risk, we task an adversarial agent with evading production blocking monitors and causing catastrophic harm, e.g. by exfiltrating its own weights. We find that when instructed with high-level attack strategies, adversarial agents can succeed through several distinct mechanisms, such as agent-generated prompt injection against the monitor, multi-agent attacks, and malicious compaction. In particular we find that in 79% of trials, the agent can use an injection attack against Auto Mode and Guardian to run arbitrary bash commands. We also find that it is possible to greatly improve Auto Mode through design changes like enhancements to tool coverage, changes in transcript formatting, and the addition of an agentic monitor stage. Even with these improvements, preventing multi-context attacks at an acceptable cost remains an open problem. By detailing our red-teaming methodology and highlighting new attack vectors, we aim to help defenders evaluate their mitigations against the possibility of persistent malign coding agents. Code is available at https://github.com/safety-research/red-teaming-auto-mode. --- ### CliniCIRCA: A Modular LLM Framework for Constructing Longitudinal Mental Health Patient Journeys from Raw EHR Narratives Authors: Aiwei Ivy Zhang, Nimra Ishfaq, Mohit Chandra, Santiago Alvarez Lesmes, Adam Coscia, Khatiya Chelidze Moon, Xiaohan Ding, Munmun De Choudhury Published: 2026-09-17 Categories: cs.CL, cs.AI, cs.IR, cs.LG arXiv: https://arxiv.org/abs/2609.19585v1 In mental health care, reasoning over patient journeys is a key task for clinicians. Yet these journeys, encompassing a longitudinal progression of biological, psychological, and social events, are often spread across disparate unstructured text narratives, making temporal recovery challenging. We present CliniCIRCA, a multi-stage LLM framework for Calendar-anchored, Imprecision-aware Reconstruction of Clinical Annals. To our knowledge, CliniCIRCA is the first to temporally classify clinical events across unstructured discharge summaries without event-level timestamps. From 14,882 MIMIC-III mental health admissions, we first construct a benchmark of 52 discharge summaries on which CliniCIRCA produces 15,891 temporally tagged events. After correcting 629 errors based on a clinician-in-the-loop evaluation, we produce verified gold-standard labels. Finally, the corrected timelines drive a temporally grounded summarization stage that compresses each source 1.52 times into a date-grouped chronological record. We then scale the framework to generate 1,000 silver-standard timelines and evaluate them as training data. Compared with zero- and few-shot prompting, instruction tuning generally improves five open-weight models on event extraction, temporal tagging, and summarization across silver and clinician-verified evaluations. --- ### Large Language Model Agents for Evidence Based Genetic Disease Severity Classification Authors: Tohid Ghasemnejad, Ahmadreza Argha, Mark Grosser, John Wang, Min Yang, Thantrira Porntaveetus, Tony Roscioli, Nigel H. Lovell, Mahmoud Aarabi, Hamid Alinejad-Rokny Published: 2026-09-17 Categories: q-bio.GN, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2609.19569v2 Disease severity classification for genetic conditions is subjective and labor-intensive, creating bottlenecks in genomic screening, where commercial panels vary widely in size and overlap. We developed an autonomous AI agent integrating Reasoning and Acting (ReAct) with Retrieval-Augmented Generation (RAG) to classify 10,211 Human Phenotype Ontology terms. It uses American College of Medical Genetics (ACMG)-endorsed severity guidelines and American College of Obstetricians and Gynecologists (ACOG) quality-of-life criteria to retrieve PubMed literature, generate interpretable reasoning chains, and independently verify claims. At the phenotype level, using expert-curated cohorts, the agent achieved 93.55% accuracy (MCC 0.9237) with 82.6% to 91.4% of claims supported by direct evidence or valid inferences. Gene-level severity was aggregated across 8,738 pairs, identifying 3,283 autosomal recessive pairs with severe or profound presentations. External validation showed 95.2% concordance with Mackenzie's Mission gene list. This system enables standardized panel design by providing reliable, automated classification supported by direct evidence. --- ### FedFIbOS: Fisher Importance based Optimal Submodelling for Heterogeneous Federated Learning Authors: Yasmeen Afzal, Jeremiah D. Deng, Haibo Zhang Published: 2026-09-17 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.19559v1 Heterogeneous federated learning requires clients with diverse computational capacities to collaboratively train a global model, where each client trains a capacity-constrained submodel. Existing methods select submodel parameters using heuristic importance measures---most prominently parameter magnitude---without theoretical justification for why these measures support convergence. We identify a fundamental gap: existing parameter selection criteria lack theoretical grounding in the convergence framework, partial client participation introduces additional estimation effects in the Fisher scores. We propose \textbf{FedFIbOS}: Fisher Importance-based Optimal Submodelling for heterogeneous federated learning, using Fisher Information in a principled criterion derived from minimizing submodel masking error. %We formally establish when magnitude selection is equivalent to Fisher selection fail under non-IID heterogeneous federated learning. We theoretically formulate submodel selection through a Fisher-weighted quadratic masking surrogate and show that the raw Fisher top-$k$ rule implemented by FedFIbOS solves this surrogate under a Fisher-dominant ranking condition. The resulting method retains the convergence structure of the underlying masked federated optimization bound. Fisher scores are efficiently estimated from empirical diagonal Fisher information using squared gradients, enabling stable and adaptive parameter selection without additional optimization overhead. Experiments on CIFAR-10, CIFAR-100, and AGNews under pathological and Dirichlet non-IID settings show FedFIbOS achieves ${\approx}10\%$ higher accuracy than the state of the art, with improvements becoming more pronounced under stronger heterogeneity. --- ### A Multi-Modal Generative Model for Tomato Disease Leaves Understanding Authors: Khang Nguyen Quoc, Minh-Phuoc Tran, Gia-Han Truong, Luyl-Da Quach Published: 2026-09-17 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2609.19555v1 Artificial intelligence for plant disease analysis has advanced from task-specific classifiers to multi-modal models capable of jointly interpreting visual and textual information. However, practical deployment in precision agriculture remains limited because most existing approaches treat disease understanding as isolated prediction tasks, failing to capture the complementary relationships among symptom recognition, severity assessment, and question-driven diagnostic reasoning. In tomato pathology, accurate interpretation of diseased leaves requires more than label prediction; it demands integrating visual symptoms with semantic context to support a comprehensive and explainable understanding. Here, we present SOLAR, a multimodal generative model that understands tomato disease spanning six question-answering tasks. SOLAR learns to align visual features with task-aware language representations by Fusion Expert module based on mixture-of-expert, enabling it to generate contextually relevant answers across diverse diagnostic tasks. By formulating tomato disease analysis as a generative Visual Question Answering (VQA) task, SOLAR provides a flexible framework that supports multi-task inference within a single model while improving performance and cross-task knowledge sharing. We evaluate SOLAR on $41,677$ images, including $216,209$ Question-Answering (QA) pairs to understand tomato leaf disease under both closed and open-ended QA settings. Experimental results show that SOLAR consistently outperforms state-of-the-art vision-only, vision-language, and task-specific models across all tasks, demonstrating superior accuracy, robustness, and multimodal reasoning. These findings highlight the potential of generative multimodal modeling as an effective direction for understanding of plant disease. The code for this study is available at https://github.com/EnalisUs/SOLAR. --- ### VABench: Measuring Embodied Spatial Intelligence through Visual Demonstrations, Active Perception, and Metric Control Authors: Zhongbo Zhang, Jiayi Jin, Yifan Wang, Zaibin Zhang, Haiwen Diao, Lijun Wang, Huchuan Lu Published: 2026-09-17 Categories: cs.RO, cs.CV arXiv: https://arxiv.org/abs/2609.19554v1 Spatial intelligence requires more than describing object locations. Under incomplete observation, models must identify and acquire missing evidence, interpret it in a common spatial frame, and act on it. We introduce VA-Bench to evaluate the complete observe-reason-act-revise loop. General-purpose MLLMs learn procedural context from RGB-only demonstrations, actively select camera viewpoints, issue metric Cartesian commands, and revise them from execution feedback. Models receive no privileged object poses, oracle trajectories, or learned action heads. A fixed model-agnostic controller executes only model-specified targets. VA-Bench contains 14 base task families (11 single-arm and three dual-arm), seven held-out geometry/layout variants, and a long-horizon five-object composition track. We evaluate 12 primary model conditions in three independent runs over the same 20 physically verified seeds per base task, reporting terminal success, nine trajectory-level behavioral diagnostics, and subtask progress. First, the best-performing model scores 100.0% on target localization and 78.9% on spatial relations in the annotated run. Its three-run macro-average task success is only 53.93+/-3.17%. Second, active camera control significantly improves task success over passive multi-view observation. In one matched comparison, success rises from 27.86% to 57.50%. Third, held-out geometric transfer can reduce task success by over 30 percentage points. No model completes a strict long-horizon episode, despite substantial partial progress. VA-Bench thus tests whether general-purpose MLLMs can turn visual demonstrations and actively acquired evidence into successful embodied action. --- ### From Parameters to Behaviors: A Survey of Model Fusion for Large Language Models Authors: Shuo Cai, Yanggan Gu, Zihao Wang, Yuanyi Wang, Yibo Yan, Wenjun Wang, Yuhang Liu, Guanghao Zhu, Sirui Huang, Ming Li, Hongxia Yang Published: 2026-09-17 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.19553v1 Model fusion integrates the capabilities from source models into a single target model. As of June 2026, Hugging Face hosts more than 2M models. This growing pool provides a rich base for model reuse and capability integration. Yet existing surveys often cover only separate parts of this space, and they do not provide a unified definition or a systematic taxonomy. This survey defines model fusion and organizes prior work into three levels: parameter-level, representation-level, and behavior-level fusion. We also review related metrics, benchmarks, and applications, summarize current challenges, and identify future directions. Our goal is to provide a clear map of this area and support future work on model fusion. A comprehensive list of papers about model fusion is available at https://github.com/Baicaihaochi/Awesome-Model-Fusion-Survey. --- ### Continual Enterprise World Model Discovery in Dynamic Systems Authors: Shambhavi Mishra, David Vazquez, Perouz Taslakian, Marco Pedersoli, Jose Dolz, Issam H. Laradji Published: 2026-09-17 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.19551v1 In an enterprise system, updating one field can set another, create a record, or start an approval. These effects are produced by business rules that are not built into the platform but written by each organization and revised over time. An agent working in such a system cannot predict the result of its own actions without knowing these rules. We study continual enterprise world model discovery, where an agent starts without knowledge of these business rules and discovers them by interacting with records and observing the outcomes. From those observations it builds a world model, which it revises as the rules change. To evaluate this, we introduce EnterpriseWorldShift, built on a live ServiceNow environment with nine tables, 25 hidden rules and 600 evaluation actions. It presents four versions of the same enterprise world, with the tables and records held fixed while a rule is modified, then added, then removed, so that discovery, revision, extension and retirement are each tested in turn. Our Continual Discovery Agent (CDA) builds such a model and carries it from one world to the next. It predicts the effects of the hidden rules more accurately than looking them up for each question, the approach taken by prior work, by up to 8.98 IoU points, and it answers from its own model without querying the running system. --- ### Portfolio-Based Constrained Multi-Objective Bayesian Optimization for Materials Design Authors: Sushant Sinha, Christofer Hardcastle, Robert Robinson, Shakti Prasad Padhy, Brent Vela, Douglas Allaire, Raymundo Arroyave Published: 2026-09-17 Categories: cond-mat.mtrl-sci, stat.ML arXiv: https://arxiv.org/abs/2609.19550v1 Materials discovery and design campaigns can be formulated as constrained multi-objective Bayesian optimization (CMOBO) problems, within which each experimental decision negotiates between two coupled but competing goals: discovering feasible candidates and refining the underlying Pareto front. Here we recast acquisition-function choice as an adaptive policy-selection problem over a portfolio of conventional and feasibility-focused acquisition functions. This was done using two controllers: UCB-Bandit, a modified UCB multi-armed bandit, and Agentic-Switch, a multi-agent decision system driven by a large language model (LLM). Both were evaluated against fixed-policy baselines in silico across five synthetic benchmark functions and two materials design case studies. The adaptive policies performed competitively in terms of both cumulative feasibility count and feasible hypervolume improvement, while each individual acquisition function performed well for only one metric, suggesting that adaptive policies are better suited for constrained materials science problems. --- ### Finding Common Ground: Graded Communal Knowledge in Bluesky Starter Packs Authors: Sagar Kumar, Lawrence Swaminathan Xavier Prince, Julia Mendelsohn, Brooke Foucault Welles, Nicholas W. Landry Published: 2026-09-17 Categories: cs.SI, cs.CL arXiv: https://arxiv.org/abs/2609.19549v1 Communication is made possible by common ground---the unspoken knowledge that people share and presuppose of one another, whether that be online or offline. In his conception of common ground, Clark (1996) distinguishes between personal and communal common ground, and asserts that the latter is graded: the more community affiliations two people share, the more common ground they share as well. Social media research has invoked this mechanism to explain how users connect, but it has gone largely untested because community memberships are rarely visible and, where they are, they are coupled to user interactions in a way that leads to conflating effects. To circumvent these challenges, this study repurposes Bluesky starter packs (SPs) as user-curated community affiliation labels. Across 191,648 pairs of users, we show that shared lexical repertoire---our proxy for common ground---grows monotonically with the number of SPs that users share, with users sharing a single pack being roughly twice as similar as equally connected strangers. A semantic renormalization of SP co-membership shows furthermore that it is more so the number of topically \emph{distinct} communities, rather than the raw count, in which common ground is graded. Finally, we show that community co-membership adds to common ground independently of proximity in the Bluesky follow network. These results lead to the conclusion that community membership is a measurable, separable, and semantically structured carrier of common ground. Reading it as such makes common ground observable before an exchange rather than inferred from it, and thus opens the door for large-scale observational approaches to a set of questions that have so far only been posed in the laboratory. --- ### PerSeM: Persistent Semantic Memory for Long-Horizon Open-Vocabulary UAV Mapping Authors: Saurbh Singh Jamwal, Ganesh Ramakrishnan Published: 2026-09-17 Categories: cs.CV, cs.RO arXiv: https://arxiv.org/abs/2609.19542v2 Open-vocabulary segmentation enables rich semantic perception for UAVs, but frame-wise predictions can remain temporally inconsistent across repeated observations and changing viewpoints. We present PerSeM, a training-free persistent semantic memory framework for long-horizon open-vocabulary UAV mapping. PerSeM associates frame-wise semantic observations with persistent world-space voxels and constructs a majority-based semantic memory, which is conservatively refined through history-preserving spatial refinement, trust-aware replay, and context-guided verification. Experiments on the Forest and UAVScenes benchmarks show that persistent 3D memory provides substantial gains in semantic correctness and temporal stability over frame-wise predictions. Beyond this strong persistent-memory baseline, PerSeM provides consistent additional improvements, improving both semantic accuracy and temporal stability across all five evaluated UAVScenes sequences. Analysis using regions identified independently of the final PerSeM predictions further shows that these gains are concentrated in semantically difficult and temporally unstable regions, where majority-based memory is most likely to remain uncertain. These results demonstrate that persistent 3D aggregation provides a strong foundation for long-horizon semantic mapping, while conservative refinement of uncertain memory states can provide additional improvements without retraining or additional neural-network inference. --- ### Compressed Active Subspaces for Scalable Bayesian Inference Authors: Thomas Flynn, Sanket Jantre, Byung-Jun Yoon, Kibaek Kim Published: 2026-09-17 Categories: cs.LG, cs.AI, stat.ME, stat.ML arXiv: https://arxiv.org/abs/2609.19539v1 Active subspace methods provide a framework for quantifying predictive uncertainty in high-dimensional models by identifying and performing inference along parameter directions that have the greatest influence on the model output. However, the construction of active subspaces requires storing many full-dimensional model gradients, which becomes prohibitive as model size increases. We address this limitation by proposing Compressed Active Subspaces (CAS), a scalable approach that first maps the model parameters to a compressed space using a structured isometric embedding and then constructs the active subspace within this reduced parameterization. Our approach substantially reduces the memory required for active subspace construction and enables Bayesian inference for large models where standard active subspace methods become impractical. We demonstrate the scalability of CAS on neural networks of increasing size while maintaining predictive performance and robust uncertainty estimates. --- ### Agentic AI Networking for Heterogeneous Unmanned Aerial Systems in Low-Altitude Wireless Networks Authors: Nguyen Duc Minh Quang, Chang Liu, Shuangyang Li, Derrick Wing Kwan Ng Published: 2026-09-17 Categories: cs.AI, cs.MA arXiv: https://arxiv.org/abs/2609.19538v1 Low-altitude wireless networks (LAWNs) are emerging as a key infrastructure for heterogeneous unmanned aerial systems that support concurrent services within a shared three-dimensional airspace. Their coexistence creates strong coupling among mobility, connectivity, and shared network resources, while heterogeneous services impose distinct and time-varying requirements. These interactions naturally form a dynamic non-cooperative game in which both operating conditions and coordination objectives evolve over time. Conventional optimization and learning-based controllers typically rely on predefined objectives, limiting their ability to adapt autonomously to changing service requirements and resource priorities. To address this challenge, we propose a hierarchical hybrid large language model (LLM)- multi-agent reinforcement learning (MARL) architecture organized as a dual-loop structure. Specifically, an outer adaptation loop employs LLM-assisted game orchestration to interpret service requirements and operator intent, and reconfigure objectives and resource priorities, while an inner loop executes decentralized, parameter-conditioned MARL policies under the configured game. A logistics-monitoring case study illustrates how the proposed framework facilitates coordinated coexistence among heterogeneous services, adapting to evolving operating conditions without retraining the underlying MARL policies. Finally, we discuss key challenges and research directions toward scalable, trustworthy, and adaptive agentic LAWNs. --- ### Detecting Soft Errors in Parallel Software with LLM-tuned Instruction Duplication Authors: Yafan Huang, Guanpeng Li Published: 2026-09-17 Categories: cs.DC, cs.AI, cs.PL arXiv: https://arxiv.org/abs/2609.19531v1 We propose PaRID (PaRallel Instruction Duplication), a software-directed soft error detection framework that requires only compile-time effort for multithreading parallel programs. PaRID addresses two key challenges: supporting parallel programs with mixed serial and parallel regions and minimizing performance overhead without relying on costly dynamic profiling. It combines parallel-aware code transformation with LLM-tuned performance modeling, guided by eight generalizable findings from an offline characterization study, to enable fast soft error detection in parallel applications. Evaluation on NPB benchmarks shows that PaRID reduces protection overhead from 162.79% to 59.84% on average and achieves up to 5x speedup while maintaining full error detection effectiveness. ---