# Research Papers — 2026-09 · bundle page 158 of 266 > Full-text markdown of papers #7851–7900 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/158 [← Previous bundle](https://aipod.dev/papers/archive/2026-09/page/157.md) [Next bundle →](https://aipod.dev/papers/archive/2026-09/page/159.md) ### Personalized Federated Learning through Global Knowledge Distillation and Local Head Adaptation Authors: Polycarpo Souza Neto, José Mairton Barros da Silva Júnior, Charles Casimiro Cavalcante Published: 2026-09-15 Categories: cs.LG, stat.ML, stat.OT arXiv: https://arxiv.org/abs/2609.17284v1 Statistical heterogeneity limits federated learning when a single global classifier cannot represent client-specific label distributions. In this work, we propose Personalized Federated Knowledge Distillation with Head Adaptation (pFedKDH), which aggregates only the shared backbone, keeps persistent client-specific heads, and uses a recalibrated global head as a teacher during local training. Across MNIST, Fashion-MNIST, CIFAR10, and CIFAR100 under class-wise Dirichlet partitions, pFedKDH obtains the best accuracy in most settings, with accuracy gaps up to 37.67\% over the weakest baseline and consistently low standard deviation across repetitions. Component-wise diagnostics and convergence results support the role of persistent heads and distillation-guided local optimization under label-skewed data. --- ### After the Party: Growth, Governance, and Security Scanning in the OpenClaw Agent Skill Ecosystem Authors: Yunpeng Xiong, Ting Zhang Published: 2026-09-15 Categories: cs.SE, cs.AI, cs.CY arXiv: https://arxiv.org/abs/2609.17274v2 AI agents increasingly act through agent skills, i.e., natural-language instructions, that direct a host agent toward shell, network, credential, file, and process actions, and public registries distribute them at scale. In the first half of 2026, the OpenClaw AI agent went viral, and its public skill registry boomed: the observable stock nearly doubled in 91 days, and a majority of the listings visible in June were created in just two months. By the end of our study window, the wave had crested, and monthly listing creation and core-repository activity were falling from their spring peaks. This paper measures what the boom left behind, drawing on the OpenClaw Git history, its GitHub issues and pull requests, and three ClawHub registry snapshots. Attention is concentrated: the top 10% of skills received 46.93% of all downloads. No simple skill features (like size or download counts) remained a stable predictor of continued listing once creation cohort and skill age were controlled. Human scrutiny did not stay: 77.86% have zero stars and zero comments, while 85.06% of the readable skills carry privilege evidence. And automated cleanup is not ready: the three security scanners disagreed on 23,702 of the 61,990 skills they all cover. After human adjudication, weighted scanner sensitivity against the reference standard ranged from 21.67% to 61.06%. Governing fast-growing agent-skill registries cannot rely on simple metadata or single scanner scores; it requires robust, transparent measurement and independent validation. --- ### After the Party: Governing What a Viral Agent-Skill Ecosystem Left Behind Authors: Yunpeng Xiong, Ting Zhang Published: 2026-09-15 Categories: cs.SE, cs.AI, cs.CY arXiv: https://arxiv.org/abs/2609.17274v1 AI agents increasingly act through agent skills, i.e., natural-language instructions, that direct a host agent toward shell, network, credential, file, and process actions, and public registries distribute them at scale. In the first half of 2026, the OpenClaw AI agent went viral, and its public skill registry boomed: the observable stock nearly doubled in 91 days, and a majority of the listings visible in June were created in just two months. By the end of our study window, the wave had crested, and monthly listing creation and core-repository activity were falling from their spring peaks. This paper measures what the boom left behind, drawing on the OpenClaw Git history, its GitHub issues and pull requests, and three ClawHub registry snapshots. Attention is concentrated: the top 10% of skills received 46.93% of all downloads. No simple skill features (like size or download counts) remained a stable predictor of continued listing once creation cohort and skill age were controlled. Human scrutiny did not stay: 77.86% have zero stars and zero comments, while 85.06% of the readable skills carry privilege evidence. And automated cleanup is not ready: the three security scanners disagreed on 23,702 of the 61,990 skills they all cover. After human adjudication, weighted scanner sensitivity against the reference standard ranged from 21.67% to 61.06%. Governing fast-growing agent-skill registries cannot rely on simple metadata or single scanner scores; it requires robust, transparent measurement and independent validation. --- ### Semantic-Spatial Agreement Verification for Mitigating Object Hallucination in Multimodal Large Language Models Authors: Ziheng Ren, Qian Gao, Jun Fan, Guohui Ding, Zhenyu Yang, Yuteng Xiao Published: 2026-09-15 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.17269v1 Multimodal large language models generate natural-language responses from visual inputs, yet may mention objects absent from an image. In medication assistance, accessible perception, and environmental decision-making, such hallucinations can create real-world safety risks. We propose Semantic-Spatial Agreement Verification (SSAV), a training-free method for verifying object claims. A visually grounded claim should remain stable across semantically equivalent queries and repeatedly localize to the same image region. SSAV aggregates multiple prompts to estimate semantic support and reduce sensitivity to query wording. Query-Induced Regional Verification (QIRV) combines cross-query region persistence, spatial overlap, and relative candidate dominance to identify isolated high responses and dispersed localizations. A geometric mean fuses semantic and spatial evidence, lowering the verification score when either branch lacks support. Experiments on three base models and multiple evaluation protocols show that SSAV effectively mitigates object hallucination. On LLaVA-1.5-7B, accuracy averaged across COCO, A-OKVQA, and GQA improves by 1.81 and 3.17 percentage points under POPE Popular and Adversarial, respectively, while CHAIRs decreases from 49.40% to 32.80%. These results show that cross-query semantic stability and regional consistency provide interpretable external visual evidence for object claims. --- ### Towards Illusions Awareness in Cyber-Physical System's Design Authors: Anna Di Placido, Nicolas Ferry, Julien Deantoni Published: 2026-09-15 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.17260v1 Cyber-Physical Systems (CPS) operate through a continuous sense-compute-act loop within an open context environment, making it impossible to anticipate all the situations the system will face. To cope with this openness, stakeholders rely on assumptions, formalized into design models. However, these assumptions may no longer hold once the system is confronted with runtime reality, resulting in a discrepancy between expected and observed behaviour known in literature as the reality gap. Existing approaches mainly focus on reducing or overcoming it by making simulations more faithful to reality, with no unified methodology to structure and exploit invalidated assumptions that give rise to this gap as reusable design knowledge. We refer to the persistent reliance on invalidated assumptions -and the resulting false confidence in the design model's operational validity -as design illusions, and argue that they need to be made explicit, structured, and exploited as knowledge to support better design decisions. We propose a conceptual pipeline for illusions-awareness that identifies, classifies, characterizes, and leverages illusions to transform them into actionable design knowledge. --- ### Exploring 2D backbone effects for indoor semantic occupancy prediction Authors: Shizhang Fanga, Wanling Yea, Qi Zheng Published: 2026-09-15 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.17257v1 Semantic occupancy prediction gives an embodied agent a voxel-level account of where space is free, occupied, and semantically meaningful. In RGB-D pipelines such as EmbodiedScan, the image encoder is often left as a default module, even though its features are the visual evidence later sampled into the 3D grid. We study this design choice directly. A central finding is that changing the 2D backbone improves occupancy accuracy more than several carefully designed occupancy architectures or modules. We keep the main RGB-D projection, depth branch, and occupancy head fixed, and replace only the image backbone. The compared encoders are CLIP-ResNet, CLIP-ViT, BLIP2, and DINOv2. Under the controlled setting, the measured mIoU changes substantially: DINOv2 obtains 30.55\%, BLIP2 obtains 29.49\%, CLIP-ViT obtains 24.33\%, and CLIP-ResNet obtains 17.41\%. The stronger encoders also exceed the original EmbodiedScan ResNet-50 baseline without modifying the downstream 3D fusion pipeline. Class-level results give a more detailed picture: DINOv2 is stronger on many layout and structural categories, whereas BLIP2 remains close on several object-centered classes. CLIP-ViT improves clearly over CLIP-ResNet, showing that the way CLIP features are exposed as dense tokens matters for voxel lifting. These results indicate that the image backbone is not a secondary engineering detail in embodied semantic occupancy, but a major source of variation in the final 3D prediction. --- ### Persistent Recurrent Memory Between Transformer Layers - Improves Language Model Generalization Authors: Eduardo Novaes Hering Published: 2026-09-15 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.17251v1 We introduce a simple architectural modification to decoder-only transformers: a persistent recurrent state that observes hidden representations via cross-attention, updates itself through a GRU, and modulates subsequent processing via gated addition. Inserted between the lower and upper halves of a 6-layer transformer, this module adds only 3.7\% additional parameters while reducing evaluation loss from $2.438 \pm 0.004$ to $1.743 \pm 0.018$, corresponding to a 28.5\% reduction on held-out language modeling data. The improvement is statistically significant across 5 random seeds ($p < 0.01$) and corresponds to reduced overfitting (generalization gap 0.12 vs 0.26). Through controlled ablations, we demonstrate that the improvement stems entirely from the persistent memory topology, not from auxiliary self-prediction objectives. A model with identical topology but no auxiliary loss performs equivalently, while a random auxiliary loss provides no benefit. Representation probing reveals that the persistent state encodes narrative position (52\% vs 33\% chance level)---information that standard attention maintains less efficiently. Our results suggest that bridging transformer layers with a lightweight recurrent memory is a simple, effective approach to improving generalization in small-scale language models. --- ### Video-HolmesV2: Can MLLMs Reason with Spatio-Temporal Audio-Visual Evidence in Long Videos? Authors: Zhaoyang Wei, Zipeng Wang, Yushe Cao, Chenhui Qiang, Shuaibing Cheng, Xuesong Yang, Sen Nie, Bowen Jiang, Wenchao Ding, Yanchao Hao, Zheng Wei, Xuehui Yu et al. Published: 2026-09-15 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.17248v1 Multimodal Large Language Models have demonstrated impressive video understanding, yet their ability to reason over long-form narratives is often masked by visual-centric evaluations and inefficient context processing. Existing benchmarks over-rely on visual heuristics while marginalizing auditory cues, effectively reducing models to "silent observers" that bypass genuine cross-modal reasoning. Moreover, standard dense sampling creates an evidence-context trade-off: increasing frames to capture evidence inevitably leads to attention distraction and token explosion. To bridge these gaps, we present Video-HolmesV2, a novel benchmark designed for Deep Audio-Visual Coupling. Unlike previous works, it enforces an Evidence-Based Evaluation, requiring models to justify answers with precise spatio-temporal audio-visual evidence, thereby reducing confounding effects of guessing and hallucinated evidence. To support this, we introduce: (1) a Multi-Model Cross-Verification pipeline to ensure task rigor; (2) a Spatio-temporal Evidence-Aware Metric for fine-grained calibration. Furthermore, we propose an Audio-Text Guided Token Compression framework. By fusing task intent with auditory anchors, our method distills high-value reasoning cues to mitigate long-context noise. In our evaluation, even strong proprietary models achieve below 60% accuracy, while our approach outperforms comparable open-source omni-models. --- ### ECHO: Early-layer Collaborative Hierarchical Orchestration with Bonus Logits in Speculative Decoding Authors: Ziyang Ma, Zihong Zhang, Zuchao Li, Lefei Zhang, Baoyuan Qi, Siqi Li, Simin Yu Published: 2026-09-15 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.17241v1 While draft-model-free speculative decoding offers a promising path to efficient LLM inference, it is frequently constrained by stale draft candidates and the high computational cost of the verification. To address these challenges, we propose ECHO, a hierarchical dual-loop framework that exploits the functional asymmetry between LLM layers. Leveraging the high discriminative efficiency of early layers and the authoritative distribution of final layers, ECHO bifurcates inference into a high-frequency inner loop and a low-frequency outer loop. Within the inner loop, early-layer bonus logits drive rapid, multi-step draft-tree exploration at a minimal cost. Simultaneously, the outer loop performs authoritative full-model verification through a state-reuse mechanism. Crucially, the outer loop also utilizes final-layer bonus logits to correct existing paths and supplement the tree with high-confidence candidates for subsequent cycles. Experimental results across diverse benchmarks demonstrate that ECHO significantly boosts mean accepted tokens and achieves a 2.4$\times$ to 2.9$\times$ speedup, outperforming existing state-of-the-art baselines with negligible engineering overhead and no extra deployment parameters, albeit with a one-shot fine-tuning dependency for optimal acceleration. The code is available at https://github.com/whucs21Mzy/ECHO. --- ### Covariate Selection for Doubly Robust Double/debiased Machine Learning Estimators for Causal Inference Authors: Muwon Kwon, Peter M. Steiner Published: 2026-09-15 Categories: stat.ME, stat.ML arXiv: https://arxiv.org/abs/2609.17238v1 High-dimensional data create challenges for causal effect estimation because identifying the covariates needed for correct model specification becomes increasingly difficult. Double/debiased machine learning (DML) facilitates the use of machine learning (ML) for causal inference by mitigating regularization and overfitting bias, but comparatively less attention has been given to covariate selection in relation to the double robustness (DR) property possessed by some DML estimators. In particular, ML-based covariate selection may result in differential covariate selection or in misspecification of both models, thereby limiting the practical utility of the DR property. To address these issues, we propose using the union of the covariates selected by the propensity score (PS) and outcome ML models to re-estimate both models. Simulation results show that using the union consistently reduces more confounding bias than using separate selected covariate sets. The results also show that ML-based estimation does not uniformly outperform conventional DR estimation, even under conditions favorable to the Lasso, and that post-Lasso reduces more confounding bias than standard Lasso. These findings demonstrate that successful use of ML for causal inference depends not only on the ML algorithm but also on how the information obtained through covariate selection is incorporated into causal effect estimation. --- ### AraMIP: Extending MIPVU Towards Metaphor Identification in Arabic Authors: Mandar Marathe, Manar Ali, Sara Nabhani, Raia Abu Ahmad, Ibrahim Baroud, Omar Momen Published: 2026-09-15 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.17235v1 Metaphor research has gained increasing attention due to its relevance to linguistic creativity, language use, cognitive processes, and related areas. While many efforts have been devoted to metaphor identification and annotation in English and other languages, Arabic remains under-resourced in this area. In this work, we propose the Arabic Metaphor Identification Procedure (AraMIP), a novel guideline for Arabic metaphor annotation. AraMIP builds on the widely used Metaphor Identification Procedure Vrije Universiteit (MIPVU) framework, incorporating adaptations that accounts for the language-specific properties of Arabic. We distinguish three major types of Arabic figurative language: Isti'ara (metaphor), kinaya (metonymy/indirect expression), and tashbih (simile), and annotate a pilot dataset of 300 sentences (5277 words). Our analysis reveals key challenges specific to Arabic, including morphological complexity, inconsistencies in dictionary sense ordering, and the absence of standardized contextual materials for annotators. This work contributes a first step toward standardized Arabic figurative instances and facilitates the development of larger annotated resources, thereby supporting future research on figurative language in Arabic. --- ### DecoGS: Adaptive Static-Dynamic Decoupling of 3D Gaussians for Free-Viewpoint Video Streaming Authors: Idil Sulo, Alexey Supikov, Ilke Demir, Sainan Liu Published: 2026-09-15 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.17230v1 Streaming 3D reconstruction demands both speed and temporal fidelity, goals that existing methods undermine by updating every Gaussian every frame, even in static regions. We present DecoGS, a method for efficient online training of 3D Gaussians from streaming videos. Unlike prior methods that update the entire scene indiscriminately, DecoGS introduces an adaptive mechanism that selectively focuses optimization on spatiotemporal regions exhibiting motion or photometric changes. This targeted training strategy eliminates redundant updates that cause flickering and drift in nominally static regions, while enabling fast, high-fidelity scene updates. The pipeline further integrates region-aware Gaussian management through gradient gating and efficient visibility filtering to maintain temporal coherence and a compact memory footprint. On N3DV and MeetRoom, DecoGS achieves 34.55 and 31.60 dB PSNR respectively, outperforming all streaming and offline baselines, while rendering at 261 FPS with $70\times$ lower temporal flicker than the best prior method, requiring no large-scale pretraining. --- ### FROD: Feature Matching Residual Denoising Oracle Bone Decipher Authors: Yanbin Hou, Biao Xiong, Guojun Xu, Jianwen Xiang, Cheng Tan, Yanchao Yang, Junwei Zhou Published: 2026-09-15 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2609.17227v1 Oracle bone script (OBS), one of the earliest Chinese writing systems, plays an important role in the study of Chinese etymology. Traditional decipherment relies heavily on domain experts who analyze characters through semantic context and structural evolution. To assist this labor-intensive process, we formulate OBS decipherment assistance as a cross-era image translation task and propose FROD (Feature Matching Residual Denoising Oracle Bone Decipher). Although many OBS characters differ substantially from their modern counterparts, they often preserve local topological invariants at the radical level. During training, FROD leverages fast feature matching to provide gated segmentation supervision: paired samples with sufficient matches are processed patch-wise to align fine-grained radicals, whereas low-similarity pairs are trained holistically to avoid mismatched artifacts. In addition, a Residual Denoising Diffusion Model (RDDM) jointly estimates noise and residual signals, thereby reducing the positional drift and stroke disorder commonly observed in standard diffusion models. Finally, a multi-stage font stylization refinement network refines the generated images by eliminating edge noise and stabilizing stroke structures. On our augmented character-disjoint dataset, FROD achieves higher Top-1 recognition accuracy than the evaluated baselines, with a 3.8% absolute gain over OBSD. --- ### Easy to Catch a Liar, Hard to Clear an Honest One: Language Models Diagnosing a Corrupted Reward Channel from a Verified Record Authors: Arman Nik Khah Published: 2026-09-15 Categories: cs.LG, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2609.17226v1 An agent that learns from rewards has to trust whatever reports those rewards. When the reports suddenly change, either the world changed or the reporter broke. From the reports alone these are indistinguishable, and reinforcement learning theory shows that no amount of further experience separates them. The prescribed escape is richer data about the reporter itself. We ask whether a frozen language model, handed exactly that data, uses it. We build a two-option game in which a payout swap and a lying reporter produce byte-identical histories. Then we add one verified record: an independent check of one round's real result, printed beside what the reporter said about that round. That single line settles the case. We ask three large models, from two families, to answer one question with one letter. Is the reporter honest or lying? They catch a lying reporter almost perfectly. At the 70B class that holds in every condition we tried; the 32B model slips in one wording. They clear an honest reporter far less often, and how often depends on things that should not matter. Averaged over rounds, letters, and wordings, a 72B model calls an honest reporter a liar 38% of the time when nothing has changed at all, and 58% of the time when the payouts moved. A 70B model from a second family calls an honest reporter a liar 26% and 48% of the time. The failure is not one of reading, because in the situation where nothing changed the same models score 0.96 to 1.00 with the answer printed in the prompt. Which surface feature drives it differs by family. For the Qwen models it is which round the record names, and for Llama it is which letter stands for "honest." Adding the record to a prompt that already states the answer makes Llama less likely to give that answer. We had registered a prediction for that 58% before the run: 35%. The failure is larger than we expected. --- ### Psychological Effects of Cultural Upheavals from Millions of Song Lyrics Over 100 Years Authors: David M. Markowitz Published: 2026-09-15 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.17225v1 Cultural upheavals impact many aspects of social life, and many studies have investigated their impact on language patterns. However, few investigations have isolated the impact of upheavals on individuals at scale in popular media. The current work evaluated millions of song lyrics spanning more than a century in search of within-artist and between-artist signals of distress from the Vietnam War, the terrorist attacks of 9/11, and COVID-19. Compared to a five-year baseline, rates of self-references - a marker of psychological distancing - were significantly reduced after the Vietnam War and September 11th. Cognitive processing terms were elevated post-upheaval vs. pre-upheaval, which indicated artists' increased attempts to make meaning from such massive disruptions. Content patterns corroborated these findings as artists wrote more about "life and freedom" (societal conditions) and less about "courtship and nightlife" (interpersonal connection) following the upheavals. Cultural upheavals modify individual and collective verbal behavior, demonstrating their far-reaching impact on society. --- ### Memorisation bias in medical AI Authors: Moritz A. Knolle, Martin J. Menten, Laurin Lux, Mélanie Roschewitz, Emma A. M. Stanley, Georgios Kaissis, Daniel Rueckert, Ben Glocker Published: 2026-09-15 Categories: cs.LG, cs.CY arXiv: https://arxiv.org/abs/2609.17223v1 Medical AI models hold immense potential to improve patient outcomes, but they are also known to unintentionally memorise individual records from their training datasets. While such memorisation has been linked to targeted privacy attacks, its consequences for clinical deployment, where patients may be assessed by a model that saw their historical data during training, remain poorly understood. Here we show that predictions on a patient's unseen future data can change significantly if a model observed that same patient's anonymised historical data during training, a phenomenon we term "memorisation bias". We demonstrate that this bias exists across diverse data modalities and model architectures, and over prolonged time spans: in some cases, memorisation bias persists on future records acquired decades after the historical records used for training. Moreover, in simulated prospective deployment, memorisation bias has asymmetric effects on the diagnostic accuracy of returning data contributors. When a patient returned with a de novo condition absent from their historical records in the training dataset, diagnostic sensitivity decreased significantly compared to an otherwise identical model not trained on their historical data. Conversely, when their health state was unchanged, both sensitivity and specificity were significantly inflated. Our findings reveal a previously uncharacterised risk in medical AI that arises when a model is deployed on patients who contributed to its training data. This exposes a shortcoming of current model development practice: the de-identification measures designed to protect patients' privacy make it difficult to identify returning contributors and exclude them from the AI-assisted interpretation of their own future data. Mitigating memorisation risks may thus require changes to current model training and deployment protocols. --- ### Grounding SWE-Agent Decisions in Architecture-0 Design: Navigating Unknown Unknowns through Physical Mapping Authors: Zhongkai Wang, Yan Liu Published: 2026-09-15 Categories: cs.SE, cs.AI arXiv: https://arxiv.org/abs/2609.17221v1 Autonomous Software Engineering Agents (SWE-Agents) excel in deterministic coding tasks but struggle with Architecture 0, the nascent system design phase plagued by implicit engineering constraints, or Unknown Unknowns (UUs) that are rarely stated explicitly. To investigate how agents navigate UUs, we explore a progressive trajectory across pure-text self-play, tool-augmented feedback, and external physical mapping. Our empirical analysis reveals a cascading chain of failures. Pure-text reasoning inevitably devolves into polite consensus or plausible yet physically impossible fabrications. Attempting to bridge this gap via an early-stage execution sandbox unexpectedly triggers Specification Gaming: agents exploit their autonomy over validation scripts to bypass physical constraints, achieving superficial success without resolving core architectural flaws. To resolve this self-validation trap, we propose the Physical Mapping Guard (PMG). Grounded in the software engineering principle of Separation of Concerns, PMG revokes verification authority from the agent, forcing semantic intents to be evaluated by an external, deterministic Semantic-to-Physical (S2P) mapping engine. Extensive evaluations demonstrate that PMG completely eradicates physical-layer and validation-layer gaming. By precisely isolating residual failures to semantic reinterpretations and auditor overreach, PMG marks a critical step toward genuine affordance grounding in automated architectural design. --- ### InfoTaxa: Information-Calibrated Label-Free Clustering for Fine-Grained Visual Taxonomy Authors: David Ahmedt-Aristizabal, Mohammad Ali Armin, Lars Petersson Published: 2026-09-15 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.17218v1 Label-free clustering of frozen pretrained visual embeddings offers a scalable route to biodiversity monitoring, but image-only fine-grained taxonomy exhibits a consistent coarse-to-fine failure mode: clusters recover broad taxonomic structure yet plateau at species level. We study this behaviour on BIOSCAN-5M through an information-calibrated clustering analysis. BioCLIP~2 features with UMAP and HDBSCAN reach $0.79$ AMI at family and $0.67$ at genus, substantially improving over the prior image baseline and remaining competitive with oracle-$K$, graph-based, and learned clustering heads on the same frozen features. To diagnose whether the remaining plateau is method-limited or information-limited, we introduce InfoTaxa, which combines clustering efficiency---the fraction of probe-estimated image information recovered by an unsupervised partition---with paired DNA as an audit signal only, not an inference input. The density pipeline recovers approximately $0.90$ and $0.81$ of the image-available information at order and family, respectively. Held-out late-fusion probes show that adding DNA to the image embedding reduces species-level prediction error by approximately two bits. Robustness analyses cover multiple image encoders, described-species and rare-class subsets, probe diagnostics, and held-out-species coarse-rank generalisation and same-species retrieval. Thus, in the tested setting, species-level label-free clustering is both clustering-limited and representation-limited: improved clustering may recover additional image-exposed structure, but cannot close the DNA-audited information gap alone. --- ### Probe-VAD: Ordinal Likelihood Probing for Training-Free Video Anomaly Detection Authors: Jiawei Gu, Qilin Zhao, Tengkuo Guo, Zhiming Zhong, Shuangqing Zhang, Fan Lyu, Fang Zhao, Guo-Sen Xie, Caifeng Shan Published: 2026-09-15 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.17211v1 Video anomaly detection (VAD) aims to localize anomalous events in untrimmed videos. Vision-language models (VLMs) provide rich visual understanding for training-free VAD, but existing approaches impose restrictive interfaces between visual understanding and anomaly scoring. Caption-based pipelines compress visual evidence into text, potentially discarding subtle cues, while direct numerical generation forces the model to express its judgment through a small set of predefined scores. Such interfaces can obscure subtle differences in anomaly severity, causing visually distinct clips to receive similar representations or scores and thereby limiting the resolution of anomaly ranking. We propose \textbf{Probe-VAD}, an ordinal binary-probing framework that directly probes severity preferences from a frozen VLM. Given raw video clips, Probe-VAD queries ten ordered severity thresholds and extracts constrained \textit{YES}/\textit{NO} continuation likelihoods. Their normalized preferences form a cumulative severity profile, from which tail evidence is aggregated into a continuous anomaly score, with isotonic projection enforcing ordinal consistency. Experiments on public VAD benchmarks demonstrate superior performance with low computational cost. Probe-VAD provides a simple interface for translating frozen VLM visual understanding into continuous, rank-sensitive anomaly scores without task-specific training or caption-based compression. Code is available at: https://github.com/yvestine/COVAS-VAD. --- ### FluxVLA Engine: A One-Stop VLA Engineering Platform for Embodied Intelligence Authors: Yinhao Li, Weixin Mao, Zihan Lan, Jikun Rong, Qirui Hu, Yiming Zhang, Weipeng Deng, Bowen Shen, Minzhao Zhu, Yiming Mao, Yan Yang, Chenguang Cui et al. Published: 2026-09-15 Categories: cs.RO, cs.AI arXiv: https://arxiv.org/abs/2609.17210v1 Vision-language-action (VLA) models, world-action models (WAMs), and offline reinforcement learning methods are rapidly expanding the design space of embodied policies, yet turning these algorithms into reliable robot systems remains constrained by fragmented data formats, training stacks, evaluation protocols, inference runtimes, and embodiment-specific interfaces. We present $\mathrm{FluxVLA}$ Engine, an open, configuration-driven platform that turns heterogeneous embodied-policy components into a reproducible data-to-deployment workflow. Rather than introducing another policy model, $\mathrm{FluxVLA}$ standardizes interfaces for datasets, visual-language and world models, action heads, reward- or advantage-weighted learning, distributed training, simulation evaluation, optimized inference, and robot operators. The engine further integrates compositional dual-arm simulation, scalable automatic data generation, and model-decoupled human-in-the-loop rollout, takeover, correction collection, and reward annotation. For responsive physical execution, it combines Real-Time Chunking (RTC) with accelerated inference backends, lightweight remote GPU serving, and configurable trajectory post-processing. Together, these capabilities connect offline learning, simulation validation, online correction, and real-robot execution through shared and auditable contracts. $\mathrm{FluxVLA}$ therefore targets the engineering bottlenecks separating promising embodied-learning algorithms from reproducible evaluation and dependable deployment. Code is available at https://github.com/FluxVLA/FluxVLA --- ### Cross-Domain Inference for Human Localization: Applying Wi-Fi RSSI Data to CSI-Trained Models Authors: Ariel Duschanek-Myers, Thomas Welsh, Helmut Neukirchen Published: 2026-09-15 Categories: cs.CR, cs.LG, cs.NI arXiv: https://arxiv.org/abs/2609.17204v1 Wi-Fi signal data can be used to compromise the privacy of individuals. While many existing approaches rely on Channel State Information (CSI), collecting this data on typical IoT devices often requires elevated operating system permissions and specialized drivers. Consequently, this paper investigates the feasibility of utilizing Received Signal Strength Indicator (RSSI) data to predict human locations. RSSI was selected because it is accessible even on devices with limited user permissions, and therefore is more applicable to a wider array of IoT devices. To bypass the tedious process of obtaining training data needed to train an RSSI-based model, an existing Wi-Fi pose prediction project was used in this research. However, that project assumed CSI data as input. Therefore, we investigate the feasibility of cross-domain inference, i.e., feeding RSSI data into that existing CSI-based model. We collected an RSSI dataset, synchronized with video ground-truth of a person moving within a room, to evaluate the model's performance. This evaluation confirmed that RSSI data can predict locations with approximately 80% confidence when human movement is present. This demonstrates that a model trained on CSI data can be used to evaluate low-granularity RSSI data consisting of decibel-milliwatt (dBm) values to roughly locate people in the collection space. These results imply that a wide range of IoT devices can be used for privacy invasion in Wi-Fi-dense environments. --- ### MyoFlow: Anchor-Tied Rectified Flow for HD-sEMG Gesture Recognition Across Sessions and Subjects Authors: Chenhao Wu, Dingjie Peng, Zhihe Zhang, Satoshi Funabashi, Satoshi Konishi, Wuqiang Yang, Hiroshi Onoda, Hironori Washizaki, Jiang Liu Published: 2026-09-15 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.17194v2 High-density surface electromyography (HD-sEMG) gesture recognition supports prosthetic control, assistive robotics, and rehabilitation, but electrode re-donning and physiological variability cause distribution shifts that degrade accuracy across sessions and subjects. Generative HD-sEMG models primarily synthesize signals for augmentation; although diffusion models enhance representation learning, prediction still relies on a separate classifier. To tie learned dynamics to the decision rule, we propose MyoFlow, the first discriminative flow-matching framework for HD-sEMG recognition across sessions and subjects. It recasts classification as anchor-tied transport: a domain-conditioned rectified flow moves encoded windows toward gesture anchors that serve as transport targets and define the nearest-anchor decision geometry, enabling zero-shot recognition without an independent head. On the Hyser dataset, MyoFlow improves mean cross-session and cross-subject accuracy over the strongest diffusion-based baseline by 4.24% and 6.37%, respectively, and achieves 91.71% mean zero-shot accuracy and 97.39% mean few-shot accuracy across multiple days on the CEMHSEY dataset. --- ### MyoFlow: Anchor-Tied Rectified Flow for HD-sEMG Gesture Recognition Across Sessions and Subjects Authors: Chenhao Wu, Dingjie Peng, Satoshi Funabashi, Satoshi Konishi, Wuqiang Yang, Hiroshi Onoda, Hironori Washizaki, Jiang Liu Published: 2026-09-15 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.17194v1 High-density surface electromyography (HD-sEMG) gesture recognition supports prosthetic control, assistive robotics, and rehabilitation, but electrode re-donning and physiological variability cause distribution shifts that degrade accuracy across sessions and subjects. Generative HD-sEMG models primarily synthesize signals for augmentation; although diffusion models enhance representation learning, prediction still relies on a separate classifier. To tie learned dynamics to the decision rule, we propose MyoFlow, the first discriminative flow-matching framework for HD-sEMG recognition across sessions and subjects. It recasts classification as anchor-tied transport: a domain-conditioned rectified flow moves encoded windows toward gesture anchors that serve as transport targets and define the nearest-anchor decision geometry, enabling zero-shot prediction without an independent head. On the Hyser dataset, MyoFlow improves mean cross-session and cross-subject accuracy over the strongest diffusion-based baseline by 4.24\% and 6.37\%, respectively, and achieves 91.71\% mean zero-shot accuracy and 97.39\% mean few-shot accuracy across multiple days on the CEMHSEY dataset. --- ### End-to-End Latency-Minimizing and Load-Balanced Request Scheduling for Edge LLM Inference in Agentic AI Services Authors: Zhen Li, Jun Cai, Haoran Gao, An Li, Tan Li Published: 2026-09-15 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.17193v1 Large language model (LLM)-powered agentic AI services increasingly demand low-latency inference, motivating the deployment of LLMs across distributed edge servers. However, heterogeneous communication and computing capabilities, together with dynamically evolving inference states, make the edge server selection for each incoming request time-varying and tightly coupled across slots. In this paper, we investigate an online request scheduling framework for edge LLM inference that jointly minimizes long-term average end-to-end latency and regulates workload distribution across heterogeneous edge servers. Two main challenges arise in this context. First, conventional latency models cannot accurately capture the fine-grained dynamics of multi-stage LLM execution. Second, the latency consequence of a scheduling decision is observed only after request completion, making immediate decision evaluation difficult. To address these challenges, we develop a cross-slot inference model that captures transmission, prefill, iteration-level decoding, and key-value (KV) cache evolution for each diverse request, and characterize server workload through a KV cache memory-time consumption metric. We propose the LYREO approach that transforms the long-term load-balancing constraint via Lyapunov optimization and employs reward redistribution with sequencebased return prediction to convert delayed outcomes into timely learning signals for earlier decisions. Simulations under various configurations demonstrate that LYREO consistently achieves lower latency and more balanced load distribution than representative learning-based and heuristic baseline schemes. --- ### EventEgoHands++: Event-based Egocentric 3D Hand Mesh Reconstruction with Real Dataset Authors: Ryosei Hara, Wataru Ikeda, Masashi Hatano, Mariko Isogawa Published: 2026-09-15 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.17189v1 3D hand mesh reconstruction is a challenging yet essential task for downstream applications, including human-robot interaction and AR/VR. Although conventional cameras have been widely adopted for this task, methods that rely on them struggle in low-light environments and under severe motion blur. To address these limitations, event-based cameras have recently attracted attention for their high dynamic range and high temporal resolution. However, applying event cameras to egocentric hand reconstruction remains challenging because camera wearer's motion produces dense background events that obscure hand-specific signals. Although the first egocentric event-based approach mitigates this issue using hand segmentation, its binary hand mask does not distinguish between left and right hands. As a result, the model lacks instance-level hand information and predicts both hands even when only one or neither hand is present. This limitation leads to incorrect inter-hand relationships and degraded reconstruction accuracy. In this paper, we propose EventEgoHands++, a framework for event-based 3D hand mesh reconstruction from an egocentric viewpoint. The proposed method incorporates a Hand Detector that estimates instance-level bounding boxes and masks for both the left and right hands. Moreover, we introduce Adaptive Attention, which dynamically gates the attention based on these detection results to accurately learn the spatial relationship and mutual interactions between the hands. To train and evaluate our framework, we extend the synthetic N-HOT3D dataset and newly construct EEH-R, the largest real-world event-based egocentric hand dataset to date, comprising approximately 1M annotated frames captured in environments including low-light conditions. Extensive experiments on both synthetic and real datasets demonstrate that our method consistently outperforms the baselines. --- ### LoopSpec: Pipelined Self-Speculative Decoding for Looped Transformers Authors: SangLyul Cho, Langqing Cui, Sehoon Kim, Dongsu Han, Insu Han Published: 2026-09-15 Categories: cs.LG, cs.CL arXiv: https://arxiv.org/abs/2609.17184v1 Looped Transformers achieve strong performance with compact parameter sizes by repeatedly applying a shared stack of Transformer blocks across recurrent depths. However, they incur higher decoding latency than standard Transformer models of comparable parameter size because shared weights are accessed at every recurrent depth. To improve decoding efficiency, self-speculative decoding is particularly well suited to Looped Transformers, as their intermediate recurrent states can directly provide draft predictions without an auxiliary draft model. We therefore propose LoopSpec, a training-free self-speculative decoding framework tailored for Looped Transformers. LoopSpec extracts draft tokens from early recurrent states and operates in a pipelined manner, overlapping draft generation of future tokens with target verification of the current token. To improve draft accuracy without excessive compute overhead, we introduce a selective second proposal from deeper recurrent depth while ensuring lossless decoding under both greedy and sampling regimes. Furthermore, we derive the optimal proposal depths in closed form and show the prediction matches measurement. Across reasoning and coding benchmarks, LoopSpec achieves up to 6.83$\times$ inference speedup across diverse Looped Transformers. --- ### Multimodal Cultural Heritage Architectural Style Classification for Residential Buildings in the UAE Based on CLIP Embeddings and SVM Authors: Ahmed Ammar Kubba, Manar Abu Talib, Iman Ibrahim, Qassim Nasir Published: 2026-09-15 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2609.17181v1 The analysis and classification of cultural heritage architectural styles remain challenging due to the complexity of visual images of buildings, which are highly relied on in traditional CNN-based classification approaches in comparison to textual descriptions, and the relative lack of non-western region-specific datasets. This paper addresses this gap by proposing a multimodal machine learning framework to analyze and classify Emirati residential architecture using OpenAI's CLIP model. We integrate visual features from images and textual features from expert descriptions into a unified 512-dimensional embedding, followed by dimensionality reduction with UMAP for visualization and unsupervised clustering using K-Means. Cluster labels, which are derived from manual analysis of the K-Means clusters, are used to train an SVM classifier for automated architectural style classification. Our approach achieves a classification accuracy of 98% across eight identified style clusters, higher than every other study in the literature, demonstrating the effectiveness of combining visual and textual modalities. Overall, this paper highlights the potential of using multimodal AI to support architectural heritage analysis, offering scalable and interpretable tools for exploring regional architectural identities. --- ### MOCC-R1: Reinforcing Reasoning-Response Consistency for Multimodal Counselor Response Generation Authors: Wenjie Zheng, Qiming Xie, Jianfei Yu, Rui Xia Published: 2026-09-15 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.17180v1 Multimodal counselor response generation (MCRG) aims to generate an appropriate counselor response from multimodal dialogue histories. Progress is limited by two gaps: first, existing datasets rarely capture sustained, human-recorded counseling interactions conducted by qualified counselors; Second, existing methods do not explicitly optimize consistency between counseling reasoning and the generated response, potentially undermining the reliability of MCRG systems. Thus, we introduce MOCC, a multimodal counseling conversation corpus containing over 200 hours of interactions involving 154 credential-verified counselors. Based on MOCC, we propose MOCC-R1, a two-stage framework for optimizing reasoning-response consistency. Cold-start supervised fine-tuning trains the model to generate a structured trajectory consisting of client-state understanding, a response intent that links a counseling principle to a planned action, and the final response. Reinforcement learning (RL) then rewards grounded plan coherence and plan execution, encouraging the inferred state and plan to be supported by the dialogue context and the response to realize that plan. Experiments demonstrate the effectiveness of the proposed MOCC-R1. --- ### IRENE: A Convolutional GRU Ensemble Model for Radar Precipitation Nowcasting over Italy Authors: Alessandro Camilletti, Gabriele Franch, Elena Tomasi, Marco Cristoforetti Published: 2026-09-15 Categories: cs.LG, physics.ao-ph arXiv: https://arxiv.org/abs/2609.17175v1 We present IRENE (Italian Radar Ensemble Nowcasting Experiment), a deep learning model for probabilistic short-range precipitation nowcasting over the Italian domain at \SI{1}{km} spatial and 5 min temporal resolution. IRENE adopts an encoder--forecaster architecture built on multi-scale Convolutional Gated Recurrent Units (ConvGRUs), trained on the national radar composite produced by the Italian Civil Protection Department (DPC). An importance-sampling scheme focuses training on precipitation-relevant events, while the almost-fair Continuous Ranked Probability Score (afCRPS) is adopted as the primary probabilistic loss function. Two additional training configurations are proposed: an adversarial (GAN) variant, IRENE-GAN, designed to improve the spatial sharpness of the generated forecasts, and a spectrally constrained variant, IRENE-GAN-RAPSD, in which the adversarial objective is complemented by an explicit penalty on the radially averaged power spectral density. The three configurations are evaluated against the stochastic extrapolation method STEPS and the pre-trained deep learning model DGMR. All IRENE configurations attain a lower Continuous Ranked Probability Score than both benchmarks at every lead time and rank histograms closer to uniformity, indicating better probabilistic skill and ensemble calibration. In terms of ensemble-mean mean absolute error the advantage is confined to the first 90 min, beyond which the strongly damped DGMR fields and, to a lesser extent, STEPS become competitive. Spectral analysis shows that the adversarial training removes the progressive loss of small-scale variance exhibited by IRENE, at the cost of an excess of fine-scale power at long lead times that the spectral penalty only partially controls. --- ### A unified framework for global and local interpretability using adaptive derivative-ordered random explanation Authors: Lemen Chao, Ming Lei, Anran Fang Published: 2026-09-15 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2609.17171v2 The interpretability of complex machine learning models is of paramount importance, especially in real-world high-stakes domains such as healthcare and finance. However, existing post-hoc interpretability methods suffer from inherent limitations: fragmented analytical processes, inadequate capacity to model nonlinear feature interactions, computational inefficiencies, and over-reliance on specific model architectures. To address these challenges, this paper provides a novel method - Adaptive Derivative-Ordered Random Explanation (ADORE) - that leverages first- and second-order derivatives to accommodate nonlinear model complexities, while enabling effective capture of feature-sample interactions within a unified analytical framework. ADORE integrates global feature importance with local sample contributions, precisely quantifying feature impact by capturing both magnitude and direction, and identifying critical samples influencing model decisions. Furthermore, it achieves computational efficiency through randomized singular value decomposition (SVD) and dynamic sparsity detection, making it scalable to large, high-dimensional datasets. Experiments across three data modalities - tabular, text, and image - demonstrate that ADORE outperforms existing methods such as LIME and SHAP in handling complex interactions and computational efficiency, while providing detailed and reliable explanations. To facilitate adoption and reproducibility, ADORE has been released as an open-source Python package, hosted on GitHub, enabling researchers and practitioners to readily adapt and apply our approach to their specific tasks, models, and datasets. --- ### A unified framework for global and local interpretability using adaptive derivative-ordered random explanation Authors: Lemen Chao, Ming Lei, Anran Fanga Published: 2026-09-15 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2609.17171v1 The interpretability of complex machine learning models is of paramount importance, especially in real-world high-stakes domains such as healthcare and finance. However, existing post-hoc interpretability methods suffer from inherent limitations: fragmented analytical processes, inadequate capacity to model nonlinear feature interactions, computational inefficiencies, and over-reliance on specific model architectures. To address these challenges, this paper provides a novel method - Adaptive Derivative-Ordered Random Explanation (ADORE) - that leverages first- and second-order derivatives to accommodate nonlinear model complexities, while enabling effective capture of feature-sample interactions within a unified analytical framework. ADORE integrates global feature importance with local sample contributions, precisely quantifying feature impact by capturing both magnitude and direction, and identifying critical samples influencing model decisions. Furthermore, it achieves computational efficiency through randomized singular value decomposition (SVD) and dynamic sparsity detection, making it scalable to large, high-dimensional datasets. Experiments across three data modalities - tabular, text, and image - demonstrate that ADORE outperforms existing methods such as LIME and SHAP in handling complex interactions and computational efficiency, while providing detailed and reliable explanations. To facilitate adoption and reproducibility, ADORE has been released as an open-source Python package, hosted on GitHub, enabling researchers and practitioners to readily adapt and apply our approach to their specific tasks, models, and datasets. --- ### MUMINS: Metadata-conditioned Uncertainty-aware Medical Image Next-state Synthesis Authors: Anna Oliveras, Roger Marí, Rafael Redondo, Oriol Guardià, Cynthia Ifeyinwa Ugwu, Ana Tost, Bhalaji Nagarajan, Carolina Migliorelli, Vicent Ribas, Petia Radeva Published: 2026-09-15 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2609.17169v1 Forecasting anatomical changes such as tumor growth and neurodegeneration is a challenging generative vision task. Morphological evolution is subtle relative to static anatomy, highly patient-specific, and inherently stochastic. Existing methods struggle with several issues: deterministic networks ignore biological stochasticity, while standard diffusion models require computationally prohibitive multi-pass sampling to quantify uncertainty. We propose MUMINS (Metadata-conditioned Uncertainty-aware Medical Image Next-state Synthesis), an efficient diffusion framework that jointly diffuses a baseline scan and its follow-up residual, summed to synthesize the follow-up scan, while concurrently predicting a spatial uncertainty map, in a single reverse diffusion process. Conditioned on the time interval and relevant metadata, it preserves fine-grained anatomy by dynamically re-injecting the baseline as a soft anchor at every denoising step, and a negative-log-likelihood head learns the uncertainty map to explicitly flag error-prone regions. Designed without organ-specific heuristics, the same architecture is reused across anatomies via separate, dataset-specific retraining. Extensive evaluations demonstrate that dataset-specific retraining of MUMINS matches or outperforms dedicated, domain-specific state-of-the-art methods on lung CT (PNG) and brain MRI (OASIS-3). Project page: https://github.com/aolivtous/MUMINS. --- ### HuMemSLAM: Efficient Human-Inspired Semantic Place Recognition for Robust Visual SLAM Authors: Mayowa Adebambo, Sebastian Donnelly, Armand Amaritei, Andrew Bradley, Alexander Rast Published: 2026-09-15 Categories: cs.RO, cs.CV arXiv: https://arxiv.org/abs/2609.17168v1 Autonomous systems require reliable place recognition for efficient and effective simultaneous localisation and mapping (SLAM). Traditional geometric visual SLAM approaches rely on low-level features and geometric consistency, but remain vulnerable to perceptual aliasing, where different places appear similar, and perceptual variation, where the same place appears different. Although semantic SLAM and modern learned visual place recognition (VPR) methods improve robustness under challenging perceptual conditions, real-time deployment requires both high retrieval accuracy and low latency. Inspired by human memory and perception, we propose HuMem-VPR, which exploits the bidirectional relationship between bottom-up perceptual evidence and top-down contextual reasoning to achieve high-level place understanding. We further introduce HuMemSLAM, the integration of HuMem-VPR with ORB-SLAM3. HuMem VPR achieved the highest aggregate retrieval accuracy on the real-image benchmark, competitive accuracy on the CARLA benchmark, and approximately two to three times lower latency than the evaluated state-of-the-art VPR methods. Across the evaluated dataset families and online experiments, HuMemSLAM substantially improved integrated Recall @1 over ORB-SLAM3's native retrieval while reducing the proposals submitted to its geometric backend. --- ### Neural Field Ensembles for Aerodynamic Surface Prediction: Winning Solution to the ONERA CRM Wall Distribution 2025 Challenge Authors: Lionel Salesses, Caroline Sainvitu, Tariq Benamara Published: 2026-09-15 Categories: cs.LG, physics.flu-dyn arXiv: https://arxiv.org/abs/2609.17160v1 Machine-learning surrogate models offer a promising alternative to high-fidelity Computational Fluid Dynamics (CFD) simulations for aerodynamic analysis and design. However, constructing accurate surrogates for realistic aircraft configurations remain challenging due to complex geometries, multiple flow regimes, and limited training data. This work presents the methodology that achieved first place in the ONERA CRM Wall Distribution Regression Challenge, which focuses on predicting pressure and skin-friction coefficient distributions over the NASA Common Research Model wing-body-pylon-nacelle configuration under different operating conditions. The proposed approach formulates the problem as a conditional neural field mapping spatial coordinates, surface normals, and operating conditions to aerodynamic wall quantities. Fourier feature encoding, a relative squared error objective aligned with the challenge metric, ensemble learning, and $k$-fold cross-validation are progressively introduced to improve prediction accuracy and exploit the limited training data. Beyond presenting the final methodology, the paper documents the successive model design choices that led to the winning solution through a comprehensive ablation study and discusses several alternative approaches that were investigated but ultimately discarded. On the hidden competition test set, the proposed methodology achieves an overall score of 8.81, outperforming the strongest organizer-provided baseline, which achieved a score of 8.64, while requiring approximately three orders of magnitude fewer trainable parameters. These results illustrate that carefully designed coordinate-based neural fields constitute an efficient and robust framework for aerodynamic surrogate modeling on complex geometries under limited-data conditions. --- ### ResLRP: The Role of Residual Cancellation in Attribution Instability in Vision Transformers Authors: Jim Berend, Reduan Achtibat, Daniel Schäffer, Alexander Binder, Wojciech Samek, Sebastian Lapuschkin, Maximilian Dreyer Published: 2026-09-15 Categories: cs.CV, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2609.17152v1 Vision Transformers (ViTs) are central to most modern vision models, yet obtaining input attributions that are fine-grained, faithful, and stable remains challenging. Layer-wise Relevance Propagation (LRP) has been adapted to transformer attention, but in ViTs it often produces noisy, unfaithful explanations. We show that the missing ingredient is the treatment of residual connections: cancellation effects in residual pathways lead to attribution explosion. Moreover, we find that these cancellations are substantially stronger in ViTs than in language transformers. To address this issue, we introduce Residual-aware Layer-wise Relevance Propagation (ResLRP), a simple extension of LRP whose propagation rules explicitly account for cancellations in residual branches, are exactly conservative, and provably bound relevance explosion. Causal channel-wise interventions confirm that residual cancellation, not a generic regularization effect, drives the instability. ResLRP substantially improves attribution quality across faithfulness and localization, evaluated on ViT architectures spanning supervised, self-supervised, contrastive, hierarchical, and multimodal families, as well as on the ground-truth-controlled FunnyBirds benchmark. The largest gains arise in modern Vision Language Models (VLMs), with +27-29% localization and up to 3.4x faithfulness scores. Beyond benchmarks, ResLRP localizes Sparse Autoencoder (SAE) features in input space, and our residual amplification measure serves as an architecture-level diagnostic predicting where attribution degrades. --- ### Kernel-Based Metrics Learning for Uncertain Opponent Vehicle Trajectory Prediction in Autonomous Racing Authors: Hojin Lee, Youngim Nam, Sanghun Lee, Cheolhyeon Kwon Published: 2026-09-15 Categories: cs.RO, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2609.17147v1 Autonomous racing confronts significant challenges in safely overtaking Opponent Vehicles (OVs) that exhibit uncertain trajectories, stemming from unknown driving policies. To address these challenges, this study proposes heterogeneous kernel metrics for Deep Kernel Learning (DKL), designed to robustly capture the diverse driving policies of OVs, and carry out precise trajectory predictions along with the associated uncertainties. A key virtue of the proposed kernel metrics lies in their ability to align similar driving policies and disjoin dissimilar ones in an unsupervised manner, given the observed interactions between the Ego Vehicle (EV) and OVs. The efficacy of the proposed method is substantiated through experimental studies on a 1/10th scale racecar platform, demonstrating improved prediction accuracy and thereby safely overtaking against OVs. Furthermore, our method is computationally efficient for onboard computing units, affirming its viability in fast-paced racing environments. The video and source code can be found at https://github.com/HMCL-UNIST/OpponentPredictionWithKMDKL.git. --- ### Continual Learning for Traversability Prediction with Uncertainty-Aware Adaptation Authors: Hojin Lee, Yunho Lee, Daniel A Duecker, Cheolhyeon Kwon Published: 2026-09-15 Categories: cs.RO, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2609.17141v1 Traversability prediction is a critical component of autonomous navigation in unstructured environments, where complex and uncertain robot-terrain interactions pose significant challenges such as traction loss and dynamic instability. Despite recent progress in learning-based traversability prediction, these methods often fail to adapt to novel terrains. Even when adaptation is achieved, retaining experience from previously trained environments remains a challenge, a problem known as catastrophic forgetting. To address this challenge, we propose a continual learning framework for traversability prediction that incrementally adapts to new terrains using a generative experience recall model. A key virtue of the proposed framework is two folds: i) retain prior experience without storing past data; and ii) incorporate the uncertainty of the generated samples from the recall model, enabling uncertainty-aware adaptation. Real-world experiments with a skid-steering robot validate the effectiveness of the proposed framework, demonstrating its ability to adapt across a series of diverse environments while mitigating catastrophic forgetting. --- ### From Foundation Embeddings to Cropland Maps: Label Efficiency, Temporal Transferability and Independent Human Validation Authors: Mohammad Ammar Mughees, Giovanni Montefoschi, Zhongxin Chen, Maria Antonia Brovelli Published: 2026-09-15 Categories: cs.CV, cs.LG, eess.IV arXiv: https://arxiv.org/abs/2609.17138v1 Geospatial foundation models provide reusable representations of satellite imagery that support downstream mapping with limited task-specific modelling. We evaluate whether annual AlphaEarth embeddings support binary cultivated-versus-non-cultivated mapping in Maine, USA, using 192 spatially separated patches and labels derived from the USDA Cropland Data Layer (CDL). Without fine-tuning the foundation model, a lightweight classifier reaches 93.7% overall accuracy and 90.8% balanced accuracy on held-out patches. Logistic regression is within 0.3 percentage points of a gradient-boosted ensemble, while a nearest-class-centroid rule, which uses class centroids but fits no parameters, reaches 90.2%. A balanced sample of 60,000 labelled pixels is within 1.3 percentage points of the full pool of 8.6 million pixels; because pixels are spatially autocorrelated, this result concerns pixel-sample efficiency rather than 60,000 independent annotation sites. In a same-region transfer experiment, classifiers trained in one year remain accurate across 2018 to 2023. Against a blind, two-interpreter consensus at 385 randomly sampled points in one contiguous 2023 block, the AlphaEarth-plus-random-forest map agrees at 95.3% ($κ=0.82$), compared with 91.7% for the CDL ($κ=0.72$; exact two-sided McNemar $p=0.0161$). This local result is consistent with partial smoothing of CDL label noise, but it does not establish statewide correction of the reference product. On the same points, the difference from a fine-tuned TerraMind segmentation model is not statistically significant (95.3% versus 93.5%; $p=0.14$), and the experiment is not a controlled comparison of computational cost. These results support frozen geospatial embeddings as a low-compute candidate for regional cropland mapping, subject to the limits of a single-state study, a 30 m-derived training reference, and a one-block human validation. --- ### Event-based Selective Attention for Multi-resolution Fast Region of Interest (ROI) Detection Authors: Luca Peres, Giulia D'Angelo, Chiara Bartolozzi, Oliver Rhodes Published: 2026-09-15 Categories: cs.CV, cs.NE arXiv: https://arxiv.org/abs/2609.17134v1 Neuromorphic vision systems operate under strict constraints on bandwidth, memory, and energy, particularly at the edge, motivating early mechanisms for data reduction and selective processing. In this work, we investigate a multi-scale training-free, saliency-based, bottom-up visual attention model that operates directly on low-resolution event-based input and selects Regions of Interest (ROI) from the visual scene. The model is evaluated across multiple downscaling factors applied to the incoming event stream, with input resolutions reduced by up to 256x relative to full resolution. Performance is assessed on the Prophesee Automotive dataset, the largest publicly available event-based dataset, demonstrating robust ROI selection across different scales on a real-world use-case. The proposed approach is capable of detecting ROIs belonging to multiple object classes, including various vehicle types, pedestrians, traffic lights, and traffic signs, with accuracy up to 70.8%, while operating at millisecond temporal resolution, 16x finer than the temporal resolution provided by the dataset ground truth. These results highlight the potential of combining early event downscaling with saliency-based attention as an effective front-end for efficient edge neuromorphic vision systems. --- ### Predicting Human Disagreement for Calibrated Dynamic Facial Expression Recognition Authors: Yiming Wang, Frederick W. B. Li, Jingyun Wang Published: 2026-09-15 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.17130v1 Dynamic facial expression recognition (DFER) benchmarks such as DFEW provide multiple annotator votes per clip, yet most models collapse them to a majority label and cannot represent human disagreement at inference time. We propose a disagreement-aware DFER framework that trains directly on the raw annotator count vector using a Dirichlet-Multinomial likelihood. Unlike mean-only soft-label objectives, the proposed likelihood provides scale-sensitive supervision for the Dirichlet concentration while preserving the predictive mean. A separate ambiguity head predicts annotation entropy for unseen clips, and a monotone Chow-style reject rule combines predicted ambiguity, vacuity, temporal instability, and input quality for selective prediction. On DFEW, the method preserves recognition accuracy while reducing ECE by 30% and AURC by 15%, and predicted ambiguity reaches a Spearman correlation of 0.52 with the annotation entropy of test clips. The calibration and selective-prediction gains transfer to FERV39k and remain under identity- and movie-disjoint DFEW splits. --- ### FirmCORe: A Benchmark for Structured Reasoning about Inter-Firm Collaboration Opportunities Authors: Tian Du, Tiantong Wu, Yafei Wang, Mengyu Liu, Xingyan Chen, Mu Wang Published: 2026-09-15 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.17128v1 Comprehensive structured data on inter-firm relationships is often scarce or inaccessible because many relationships are privately negotiated, selectively disclosed, and fragmented across proprietary databases. This scarcity hinders the discovery of collaboration opportunities, particularly for startups and small and medium-sized enterprises. Firm profiles are readily available, but collaboration potential cannot be inferred from business similarity alone, since similar firms may be competitors, whereas dissimilar firms may offer complementary products, technologies, channels, capabilities, or capital. We present FirmCORe (Inter-Firm Collaboration Opportunity Reasoning), a human-annotated benchmark for pairwise reasoning over weakly structured firm profiles, comprising 2,805 labeled firm pairs. Given two firm profiles, a model must determine whether the available evidence supports a collaboration opportunity and, for positive pairs, jointly predict its strength, primary collaboration type, and role direction. FirmCORe also provides parallel Chinese- and English-language evaluation sets containing identical instances and gold labels, enabling controlled analysis of input-language sensitivity. Experiments with representative locally deployed and hosted large language models (LLMs) show that the strongest model achieves a macro-F1 score of 74.51 for opportunity detection but only 61.57% exact match across all four output fields. Language effects vary across models, and high cross-language agreement can mask errors shared across languages. These results indicate that current LLMs are substantially more reliable at detecting broad collaboration opportunities than at identifying their specific types and role directions. --- ### AI for Science with GPT-6 Astra: Thermal Design and Electrothermal Analysis of 2D CFET Authors: Min-Hui Kim, Khushi Sharma, Sarah Zhang, Ye Wang Published: 2026-09-15 Categories: cond-mat.mtrl-sci, cs.AI, cs.AR arXiv: https://arxiv.org/abs/2609.17123v1 Thermal optimization of 2D CFET inverters requires testing structural proposals against their electrical costs. We examine these research tasks using an AI agent workflow within a supplied electrothermal model. At 12 nm, Astra selects a redistributed source-interconnect geometry, while a coordinating agent proposes a substrate-directed heat-removal path. The combined design reduces peak temperature rise by 1.67 K at fixed metal volume and 20 μW. A subsequent metal-resistance sensitivity gives about 0.6-K inverter cooling alongside a 2% nFET on-current loss. Effective contact-length scaling further shows that lower temperature can accompany higher thermal resistance when current falls. Reproduction identifies agreeing implementations and retains a 104.95-K failure for diagnosis. These results show that an AI scientist workflow can propose thermal structures, test them under common constraints, and quantify their electrical cost. --- ### An Empirical Study of Counterfactual Self-Explanations in LLMs Authors: Giannis Kalyvas, Giorgos Filandrianos, Orfeas Menis Mastromichalakis, Vassilis Lyberatos, Giorgos Stamou Published: 2026-09-15 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.17119v1 Large language models can easily generate explanations for their own outputs, but such self-explanations are not necessarily faithful to the model's behavior. We study this issue through counterfactual self-explanations, where a model minimally edits an input so that its own prediction changes. Across sentiment analysis and natural language inference, we evaluate ten instruction-tuned models from the LLaMA-3 and Qwen-2.5 families, measuring faithfulness, minimality, and alignment with human-annotated rationales. Our results show that model scale is the strongest determinant of explanation quality: larger models are substantially more likely to generate counterfactuals that flip their own predictions and target decision-relevant evidence. In contrast, the rationale-guided condition produces edit-minimal counterfactuals that are also more human-aligned. However, it does not consistently improve faithfulness. Overall, counterfactual self-explanations can provide useful behavioral evidence about model decisions, but their reliability depends strongly on model capacity and should be empirically validated rather than assumed. --- ### Intrinsic Robot Rewarding: Reusing VLA Representations for Autonomous Evaluation and Policy Improvement Authors: Tobias Schaffer, Mohab Elkhayat, Daniela Nicklas, Mustafa Almohamad, Elham Al-Fuqara Published: 2026-09-15 Categories: cs.RO, cs.LG arXiv: https://arxiv.org/abs/2609.17115v1 Vision-language-action (VLA) systems already bring together two valuable resources for robot learning: rich visual representations and demonstrations of successful task execution. Intrinsic Robot Rewarding (IRR) proposes to use these resources for a second, complementary purpose: evaluating the robot's own outcomes and providing feedback for policy improvement. Successful demonstration endpoints define task-specific references, and the policy's frozen visual encoder provides the feature space in which new outcomes are assessed. The core reward mechanism adds a reference bank and a scoring operation to the existing pipeline, without requiring a separate learned evaluator or an additional perception backbone. Our position is that this reuse offers a promising route to lower integration effort, efficient reward computation, and reduced recurring human outcome scoring. Building on established research in visual rewards and learning from experience, IRR brings these ideas into the robot's existing perception and demonstration pipeline. An operational COMAU Racer 3 demonstrator is available at technology readiness level 4 (TRL 4). This laboratory foundation supports the next research step: connecting internal outcome evaluation to physical policy improvement. We present the reward formulation, central research questions, and an evaluation methodology linking reward reliability to task success and supervision effort. The intended contribution is a reusable approach to learn and improve from the data and experience already available in industrial robot systems. --- ### Not Another Text Benchmark: Putting the "Visual" Back in Visual Question Answering for Large Video Models Authors: Rwiddhi Chakraborty, Yinong, Wang, Cheng Zhang, Fan Bai, Zhuoran You, Michael Kampffmeyer, Yong Jae Lee, Fernando De la Torre, Robert Jenssen Published: 2026-09-15 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.17112v1 Large video models have exhibited impressive performance on a wide range of visual question answering tasks, owing to the rise of powerful, pretrained text and vision encoders. The usefulness of such models have also been demonstrated on a wide range of benchmarks, with an important caveat - the dominant approach in these benchmarks evaluates multiple choice reasoning via text options. This is a natural way to test text-based reasoning in these models, and has led to significant insights regarding model behavior in the community. In this work, we ask a different question - what happens when the evaluation modality is visual, rather than text? We introduce three new vision-centric evaluation benchmarks in temporal frame retrieval, video future prediction, and causal memory distortion, all designed around evaluating visual understanding capabilities in large video models. Our approach complements the existing approaches to evaluate video understanding in frontier models. We show that current frontier models exhibit significant weakness when attempting to reason through visual queries, rather than text. We conclude with an extended analysis section that provides pointers for future improvements in visual understanding for large video models. --- ### Finding Common Mistakes In Modelling With Mathematical Formalisms Using LLMs Authors: Lilian Killich, Marko Schmellenkamp, Fabian Vehlken, Thomas Zeume Published: 2026-09-15 Categories: cs.CY, cs.AI, cs.LO arXiv: https://arxiv.org/abs/2609.17111v1 Modelling with mathematical formalisms like logical formulas, mathematical equations, or regular expressions is an important yet challenging task for students of computer science and other STEM disciplines. Identifying common mistakes occurring in this context is an important step towards helping struggling students by providing targeted high-quality feedback, e.g. in interactive learning systems. We present a tool-supported workflow that allows to (1) identify candidates for common mistakes that explain many student mistakes in large educational data sets, (2) cluster candidates according to similarities, and (3) visualize resulting clusters for instructors and CS education researchers. The visualization is designed to help researchers to identify common modelling mistakes. The candidates for common mistakes are represented by bug fixing transformations that translate incorrect formalizations into correct formalizations; they are generated by an LLM and validated algorithmically. We show that this approach works well by reproducing common mistakes in propositional logic modelling that were identified by hand in the literature; showing that, unlike other algorithmic approaches, the LLM-based approach is suitable for very large sets of data; and applying it to multiple other formalisms to showcase it generalizes beyond propositional logic. --- ### Shared-Prefix KV Reuse Across Standard LoRA Adapters: Quality and Serving Tradeoffs Authors: Dushyant Rajput Published: 2026-09-15 Categories: cs.AI, cs.CL arXiv: https://arxiv.org/abs/2609.17109v1 A common small-model deployment runs one shared backbone with several LoRA specialists that answer over the same context. Serving them naively re-prefills that shared context once per specialist. We study a narrow, practical question: for already-trained standard LoRA adapters -- not adapters retrained for cache compatibility -- how much task quality is preserved if the backbone's prefill KV cache is computed once and reused across specialists, and what does that buy in serving cost? On a Qwen3-1.7B backbone with two adapters (extractive QA on HotpotQA, arithmetic reasoning on GSM8K), we sweep the boundary at which the specialist takes over from the reused base cache and measure paired quality differences and serving cost. Full-prefix reuse had the lowest prefill cost and a small quality difference on held-out GSM8K (Delta = -4.6 EM at a 160-token budget; -3.0 at 320 tokens; -0.8 under a second training seed -- all favoring native, only the first excluding zero, and the magnitude not consistent). Partial recomputation provided no demonstrated advantage. Neither quality equivalence nor a general boundary-selection rule is established. We also report a closed-form ridge KV translator that did not beat direct reuse, and specialist-dependence contrasts whose intervals all include zero. The measured serving benefit is warm-cache time-to-first-token, which grows with context (~16x at 8K); two-branch peak memory was only 12% lower and, on inspection, the prefix was never physically shared across branches -- this implementation reuses KV values but copies their storage, so shared-cache memory savings are not achieved. --- ### Symbolic Separation: Grounding Deep Agents in Knowledge Graphs for Trustworthy Operational Data Analytics Authors: Baibek Davletiyarov, Junaid Ahmed Khan, Andrea Bartolini Published: 2026-09-15 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.17107v1 Generative AI promises natural language access to the massive numerical telemetry of data centers and Industry 4.0 installations, yet text-to-query and tool-using agents stay unreliable: even frontier models answer little more than half of real-world database questions, and far fewer of the multi-step, operational ones, because the LLM must compose how heterogeneous sources relate and hallucinates the relations, not just the fields. We propose symbolic separation: a deep agent reasons freely but may act on data only through an ontology-constrained Virtual Knowledge Graph with deterministic pre-execution validation. Unlike a tool API's interface contract, this domain-semantic contract turns a complex question into one validated graph traversal instead of LLM-inferred joins. Instantiated as the Neurosymbolic Deep Analyst and evaluated on 49.9 TB of superconputer telemetry against a rigid workflow and a non-symbolic ablation, it raises end-to-end task success from 43% to 86%, prevents silent data-integrity errors that no syntactic check catches, and cuts token cost by 2.4x, letting a smaller on-premise model outperform a larger one. --- ### High-Fidelity Digital Twin Data Models by Randomized Dynamic Mode Decomposition and Deep Learning with Applications in Fluid Dynamics Authors: Diana A. Bistrian Published: 2026-09-15 Categories: cs.LG, math.NA arXiv: https://arxiv.org/abs/2609.17101v1 The purpose of this paper is the identification of high-fidelity digital twin data models from numerical code outputs by non-intrusive techniques (i.e., not requiring Galerkin projection of the governing equations onto the reduced modes basis). In this paper the author defines the concept of the digital twin data model (DTM) as a model of reduced complexity that has the main feature of mirroring the original process behavior. The significant advantage of a DTM is to reproduce the dynamics with high accuracy and reduced costs in CPU time and hardware for settings difficult to explore because of the complexity of the dynamics over time. This paper introduces a new framework for creating efficient digital twin data models by combining two state-of-the-art tools: randomized dynamic mode decomposition and deep learning artificial intelligence. It is shown that the outputs are consistent with the original source data with the advantage of reduced complexity. The DTMs are investigated in the numerical simulation of three shock wave phenomena with increasing complexity. The author performs a thorough assessment of the performance of the new digital twin data models in terms of numerical accuracy and computational efficiency. --- ### Semi-Supervised Learning-Based Genetic Biomarkers Dataset for Multiple-Stage Hepatocellular Carcinoma Prediction Authors: Ahmed Ammar Kubba, Manar Abu Talib, Jibran Sualeh Muhammad, Ali Bou Nassif, Abdalla Sayed Mohamed, Darko Castven, Jens U. Marquardt Published: 2026-09-15 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.17100v1 Liver cancer is a complex disease responsible for a high number of deaths across the globe each year, making automated solutions for liver cancer classification urgent. The most common form of liver cancer is hepatocellular carcinoma (HCC), accounting for over 90% of liver cancer cases. There is a distinct lack of publicly available HCC datasets utilizing genomic data, which is necessary for training artificial intelligence (AI) models for automated HCC classification. This study proposes constructing a multi-stage HCC dataset using XGBoost and Semi-Supervised learning on three separate datasets of genomic biomarkers, utilizing their existing labels in the Semi-Supervised learning process to label the proposed dataset. The proposed dataset consists of 770 patient samples in total, categorized into five classes that represent normal tissue alongside different stages of HCC. Each sample in the dataset consists of 11,150 different gene expression levels. The XGBoost model demonstrated a final classification accuracy of 96.5% during the Semi-Supervised learning process. ---