# Research Papers — 2026-09 · bundle page 6 of 266 > Full-text markdown of papers #251–300 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/6 [← Previous bundle](https://aipod.dev/papers/archive/2026-09/page/5.md) [Next bundle →](https://aipod.dev/papers/archive/2026-09/page/7.md) ### Why Do Conventional World Models Fail to Learn Cellular Automata? Authors: Shaoyang Guo, Ziming Liu Published: 2026-09-30 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.39604v1 Although conventional world models - auto-regressive or diffusion models based on transformers or convolutional networks - may learn surface statistics of world dynamics, can they learn the exact world dynamics from its observed history? Leveraging cellular automata as a simple testbed, we find the answer to be no in many cases. Conventional architectures predict most pixels correctly yet rarely complete a rollout: a CNN predicts 96.3% of cells but completes 18.9% of rollouts; a joint diffusion model completes none. We trace the gap to three failure modes of these world models - namely, they fail to exactly capture spatial locality, temporal locality or temporal stability. Simple changes repair each: (1) for spatial locality, two-dimensional rotary positions lift a transformer from 39.1% to 100% on the Game of Life; (2) for temporal locality, handing each token its cell's previous-frame neighbourhood lifts the same transformer from 25.8% to 99.9% on unseen rules; (3) for temporal stability, causal freezing lifts the same diffusion weights from 42.2% to 99.9%. None of the three changes touches the architectural backbone; each only modifies the information flow within it. We also compare joint and ordered sampling on billiards and, in an exploratory study, on a simulated Burgers equation. --- ### GroundingPI: A Grounding Foundation Model towards Physical Intelligence with Visual Primitives Authors: Qize Yu, Lianrui Fan, Boyu Chen, Jiaqi Liang, Xini Ding, Yue Chen, Zetian Song, Yuran Wang, Yi Zou, Kaixuan Wang, Tianxing Chen, Wenxuan Song et al. Published: 2026-09-30 Categories: cs.CV, cs.AI, cs.LG, cs.RO arXiv: https://arxiv.org/abs/2609.39601v1 Precise grounding matters. It specifies which object is the target and where that object is, even in clutter and for tiny objects, and it has to be fast enough for closed-loop control. Yet vision-language-action (VLA) and world-action models (WAMs) take perception from general-purpose vision-language and video-generation backbones, which still fail in these settings. We introduce GroundingPI, a 4B grounding foundation model that generates points and boxes as quantized coordinates in a shared vocabulary. Training combines multimodal and spatial pretraining, supervised fine-tuning, and reinforcement learning with GRPO, using supervision from public datasets and dedicated data engines. Against 44 baselines across 34 grounding benchmarks spanning 11 perceptual capabilities, GroundingPI establishes a new state of the art, averaging 73.68%, above the larger GPT-6 Astra (71.54%). As a downstream visual backbone, GroundingPI improves performance on robotic manipulation and autonomous driving. On RoboTwin 2.0, it outperforms every mainstream backbone we evaluate in all four out-of-distribution settings, by up to 24.8% relative to the strongest backbone. On RoboCasa-GR1, GroundingPI trained with 50% of the demonstrations outperforms those baselines trained with 75%. On nuScenes, used as the visual backbone, GroundingPI attains an average open-loop L2 error of 0.296 m. We systematically analyze GroundingPI's pretraining in scale and data composition. Downstream autonomous driving and robotic manipulation improve as the pretraining is scaled. Analyzing the data recipe across these 11 perceptual capabilities shows dense grounding's substantial benefits for both, and OCR's potential as a catalyst for perceptual learning. These results support grounding as a perceptual foundation, and dedicated perceptual pretraining as a promising direction for foundation models of physical intelligence. --- ### GroundAnything: Reconciling Parallel Decoding with Precise Visual Grounding at Flash Speed Authors: Qize Yu, Lianrui Fan, Bowen Ping, Xini Ding, Zetian Song, Junbo Niu, Kaixuan Wang, Tianxing Chen, Yue Chen, Minghua He, Yuran Wang, Jie Huang et al. Published: 2026-09-30 Categories: cs.CV, cs.AI, cs.LG, cs.RO arXiv: https://arxiv.org/abs/2609.39600v1 Autoregressive (AR) grounding models serialize spatial predictions, introducing sequential latency and imposing a causal order on output tokens. We view grounding as visual evidence extraction: objects, locations, and spatial relations are jointly constrained by the image and query, yet their dependencies do not imply an intrinsic left-to-right generation order. This distinction makes bidirectional diffusion a natural fit, allowing spatial hypotheses to emerge in parallel and be jointly refined through iterative denoising. We introduce GroundAnything, a 4B-parameter grounding foundation model that reconciles fast parallel decoding with precise localization through blockwise denoising. Training combines grounding pretraining from public datasets and dedicated data engines, direct AR-to-diffusion conversion with joint AR and diffusion objectives, supervised fine-tuning, and GRPO-based reinforcement post-training. Across 30 grounding benchmarks, our autoregressive variant, GroundAnything-VLM, establishes a new overall state of the art among similarly sized models at 72.42%, remaining competitive with GPT-6 Astra (71.35%). With entropy-guided decoding, GroundAnything also surpasses the prior state of the art at this scale, averaging 61.75% versus 53.32% for the fast MTP-based LocateAnything model. We further explore decoding strategies, showing that an optional self-speculative mode achieves a $4.51\times$ speedup over the AR counterpart with a 0.74 percentage-point drop in COCO F1mIoU. Infrastructure experiments show that progressive inference optimizations translate parallel decoding into practical speedups. These support efficient visual grounding in latency-sensitive real-world systems. --- ### Text-to-3D Policy: Fine-Grained Language-Behavior Alignment for Unseen Specification Generalization Authors: Xinhao Yang, Wenhao Wu, Ning Lv, Yanshen Ding, Zhenhong Sun, Daoyi Dong, Chunlin Chen, Zhi Wang Published: 2026-09-30 Categories: cs.RO, cs.AI arXiv: https://arxiv.org/abs/2609.39599v1 3D visuomotor policies provide a strong foundation for spatially precise manipulation, yet current text-to-3D policies struggle to follow unseen fine-grained behavioral specifications beyond those covered by demonstrations. We study this challenge as unseen specification generalization, where language specifies behaviorally significant variations, such as target position, displacement, or articulated state, that are absent from policy training. We find that pretrained language representations and conventional global behavior-language alignment capture coarse task semantics but often blur nearby specifications that require distinct behaviors. We introduce T3DP, a Text-to-3D Policy framework for fine-grained language-behavior alignment. Rather than compressing each instruction and demonstration into a single global embedding, T3DP preserves their local structures and establishes bidirectional token-level correspondence between linguistic elements and behavioral segments. This directly grounds subtle linguistic variations in the behavior components they affect, preventing closely related specifications from collapsing in the representation space. The resulting specification-sensitive language representation conditions a point-cloud-based 3D diffusion policy, enabling more precise control over unseen behavioral specifications without modifying the underlying policy architecture. Across Meta-World, ManiSkill, and RoboTwin, T3DP improves average held-out-specification success over global language-behavior alignment by +11.0-14.2 points, with gains on all 15 task families; on real-robot tasks, it further raises average success from 47.5% to 65.0% (+17.5 points). Representation and action-probe analyses show that fine-grained alignment better preserves specification geometry and action-relevant variation, linking local behavior grounding to downstream control. --- ### Convergence of Practical Muon with Finite Newton-Schulz Iterations and Nesterov Momentum Authors: Hanyng Peng, Hui Wang, Yue Yu Published: 2026-09-30 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.39595v1 Practical Muon maintains momentum and performs a small, fixed number of Newton--Schulz iterations separately for each parameter matrix, often with a Nesterov correction. We analyze these layer-wise finite-step updates jointly on a coupled nonconvex objective, rather than replacing them by exact polar factors or one global orthogonalization. Under gradient-dependent $(\mathcal L_0,\mathcal L_1,q)$-smoothness and conditionally unbiased stochastic gradients with bounded layer-wise variance, we establish an $\mathcal O(T^{-1/4})$ bound on the expected average Frobenius gradient norm. The analysis retains the Nesterov recursion and requires neither bounded stochastic gradients, symmetric noise, nor a uniform positive lower bound on the nonzero output singular values. Its constants contain no explicit matrix-dimension or rank factors when the number of blocks and problem constants are fixed. The proof follows a descent inequality and a decomposition of the momentum tracking error into initialization, noise, and drift. For the original five-step quintic, we verify the required scalar-map bounds analytically; the result also allows step-dependent coefficients satisfying the same bounds. A complementary nuclear-norm result quantifies rank dependence under a stronger spectral condition. The vanishing rate uses coupled learning-rate and momentum schedules, including the standard single-coefficient Nesterov rule. --- ### Structural Limits of the Information-Theoretic Uncertainty Decomposition Authors: Jakob Lønborg Christensen, Christian F. Baumgartner, Morten Rieger Hannemose, Anders Bjorholm Dahl, Vedrana Andersen Dahl Published: 2026-09-30 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.39591v1 Uncertainty estimation in machine learning typically decomposes uncertainty into aleatoric uncertainty (AU) and epistemic uncertainty (EU) using the standard information-theoretic framework. However, in practice, two critical issues arise: entanglement (AU and EU are highly correlated) and epistemic collapse (EU magnitude shrinks with increasing model capacity). We analyze this framework on a functional level and discover that significant portions of the assumed AU, EU range are infeasible in finite settings, and cannot be attained with any class probabilities. We characterize how this infeasible region scales with the number of classes and Monte Carlo samples $N$ (e.g., from ensembles with $N$ members), revealing it is bounded by $\text{AU} \leq \log(2)/N$. Crucially, the infeasible region's boundary helps explain epistemic collapse: when model confidence is high, $\text{AU} > \text{EU}$ is guaranteed by this fundamental structural limitation. Our findings show that increasing ensemble size mitigates epistemic collapse by reducing the infeasible area. Lastly, we caution against interpreting AU and EU as independent quantities in low AU regimes, since we show they are coupled when $\text{AU} \leq \log(2)/N$. --- ### SPOON: Towards Coherent Compositional 3D Scene Generation from Uncalibrated Multi-view Images Authors: Guibiao Liao, Mochu Xiang, Heng Li, Ken Deng, Zijie Wang, Guanbin Li, Ping Tan, Shenghua Gao, Yizhou Yu Published: 2026-09-30 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.39590v1 Compositional 3D scene generation aims to recover complete 3D object shapes and their spatial arrangement from visual observations. Recent image-conditioned 3D generators provide strong priors for producing high-quality object geometry, making the generation of complex scenes increasingly practical. A central challenge is therefore to spatially organize these generated assets into a globally coherent scene while remaining consistent with multi-view observations. Existing approaches either entangle scene layout with object generation or separately estimate spatial placement from view-specific observations, where pose hypotheses may remain ambiguous and inconsistent across views, often resulting in an incoherent object-camera soup. We introduce SPOON, a framework that reformulates multi-view compositional 3D generation as scene-level, geometry-grounded pose reasoning. Rather than treating view-specific object pose hypotheses independently, SPOON coordinates them using reconstruction-derived multi-view geometry through a Guide-Route-Reconcile paradigm. This progressively organizes object poses and camera configurations into a coherent scene-level spatial arrangement. Extensive experiments on ARSG-110K and MIDI-3D-Front demonstrate consistent improvements in object placement and scene composition across varying numbers of input views. On ARSG-110K, SPOON reduces scene-level and object-level Chamfer distances by 12.7% and 17.7%, respectively, compared with a strong baseline. --- ### KilometerVision: A New Frontier for Large-Scale Spatial Intelligence in VLMs Authors: Aravindh Mahendran, Michael King, Matthew Koichi Grimes, Antoine Yang, Tyler Zhu, Joseph Heyward, Tengda Han, Shiry Ginosar, Chen Sun, Dima Damen, Simon Osindero, Noah Snavely et al. Published: 2026-09-30 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2609.39588v1 We push the frontier of large-scale spatial intelligence in Vision-Language Models (VLMs) and introduce the first benchmark that probes geographical layout understanding from real-world videos, spanning up to 1km distances. Inspired by the cognitive science literature, we evaluate models against the hierarchical stages of human spatial awareness: anchoring via landmarks, connecting them through routes, and integrating these into global mental maps. Extensive experiments reveal a fundamental divergence in how current AI models process spatial information. Instead of utilising true path integration or forming geometric survey knowledge, we find that VLMs rely almost entirely on 2D visual recognition and text-matching to bypass complex spatial reasoning. The benchmark is publicly available at https://perception-test-challenge.github.io/kilometervision.html. --- ### A Generalizable and Explainable Framework for Synthetic Video Detection Using First-Digit Gradient Statistics Authors: Sidharth Shanu, Gautam Kumar, Tej Singh Published: 2026-09-30 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.39585v1 AI video generators have not only become harder to detect but are used to generate a diverse set of scenarios from landscapes to street views to animal videos. This creates a problem where CNN-based detectors are effective but offer no insight into their inner workings, while forensics-based detectors are often pretrained for a set scenario or become too complex to derive meaningful insights. We present a novel approach to AI video detection using Sobel gradient values analysed with the first-digit law. Using linear discriminant analysis, we visualise the discriminatory signal, while a multi-layer perceptron is used for classification. The detection method has no generator- or scenespecific features, and the model has no knowledge of container formats, codec, bitrate, or compression artefacts. The model is trained and tested on GenBuster-200K, GenBusterBench, GenVA, FaceForensics++ C23, and CelebDF. We also show how zero-shot detection fails even though the feature set carries a discriminatory signal. --- ### Cybersecurity in Edge Computing: A Trust-Aware Federated Hybrid Intrusion Detection Framework Authors: Zawad Yalmie Sazid, Robert Abbas Published: 2026-09-30 Categories: cs.CR, cs.LG arXiv: https://arxiv.org/abs/2609.39584v1 Edge computing has emerged as a critical computing paradigm in modern distributed systems by migrating data processing closer to end users and Internet of Things (IoT) devices. While this paradigm decentralizes processes, minimizes latency, and reduces backhaul bandwidth congestion, it exponentially enlarges the cyberattack surface. Heterogeneous, resource-constrained edge devices deployed across unmanaged administrative domains present highly vulnerable targets. To address these vulnerabilities without compromising global data privacy regulations, this paper proposes a novel Trust-Aware Federated Hybrid Intrusion Detection Framework (TA-FHIDF). The proposed framework integrates an Autoencoder, a 1D Convolutional Neural Network (1D-CNN), and a Bidirectional Long Short-Term Memory (BiLSTM) model into a unified, localized deep learning engine capable of autonomous spatial and temporal feature extraction. Model training is performed collaboratively via federated learning, ensuring raw network telemetry remains isolated at local gateways. Furthermore, to defend against adversarial model poisoning attacks, we introduce a robust server-side trust-aware aggregation mechanism that evaluates client reliability using a cosine similarity metric before global model integration. Empirical evaluations across multi-vector benchmark datasets (UNSW-NB15, CICIDS2017, and Edge-IIoTset) demonstrate the framework's superior detection accuracy, rapid convergence, and high Byzantine fault tolerance under adversarial attack scenarios. --- ### MADGRAV: a multilevel anomaly-detection pipeline for gravitational-wave searches applied to LIGO data Authors: Gianluca Inguglia, Huw Haigh, Ulyana Dupletsa, Alessandro Longo Published: 2026-09-30 Categories: gr-qc, astro-ph.IM, cs.LG arXiv: https://arxiv.org/abs/2609.39583v1 We present the results of \textbf{MADGRAV}, a deep-learning-based search for high-mass compact binary coalescences, applied to the data collected by the LIGO interferometers during the third observing run and during the first and second part of the fourth observing run. The \textbf{MADGRAV} pipeline consists of a series of sequential convolutional neural networks that perform anomaly detection, glitch classification, coherence testing, and signal ranking. Data from the Hanford and Livingston LIGO detectors are studied (both individually and in coherence) by way of 1 second Q-transform windows. Of the candidates that survive every stage of the pipeline, 48 reach the significance threshold, and we report 47 gravitational wave detections characterised by a false alarm rate below $1\,{\rm yr}^{-1}$ with a probability of astrophysical origin $p_{\rm astro}>0.9$. Of the 47 detections, 44 are shared with the minimally modelled coherent WaveBurst search. The observed total source-frame masses, extracted from official gravitational wave transient catalogues, are in the $14-236 M_{\odot}$ range with a median of $69 M_{\odot}$, and a median SNR of 16. We note that the recovered fraction of confident detections rises with mass: for LIGO detectors network SNR $>10$ the pipeline recovers $8.1\%$ of confident catalog events below $30 M_{\odot}$, $39.8\%$ between $30$ and $100 M_{\odot}$, and $53.3\%$ above $100 M_{\odot}$, corresponding to $33.3\%$, $45.5\%$ and $53.3\%$ of the events detected by coherent WaveBurst in the same bins. These results suggest that anomaly detection pipelines can serve as an independent detection channel complementary to matched filtering in the high-mass high-SNR regime. --- ### Robust Transfer Learning for Paper ECG Recognition Authors: Yinghao Xie, Zhenbang Dai, Haojun Wang, Jinyu Cai, Fabio Bonassi, Hongwu Chen, Johan Sundström, Jiawei Li, Antônio H. Ribeiro Published: 2026-09-30 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2609.39581v1 Paper ECG recognition is challenging because real-world ECG images vary in layout, physical artifacts, and label availability. We introduce RobECG-CL, a rank-aware contrastive learning framework for robust paper ECG representation learning. Starting from standard 12-lead ECG recordings, we construct progressively degraded paper ECG views with heterogeneous layouts and train the model to balance same-recording invariance with degradation-aware ordering. Across synthetic stress tests on CODE-II and EchoNext, RobECG-CL improves robustness under severe degradation and few-shot transfer, outperforming contrastive learning baselines and surpassing the waveform-based foundation model, ECG-FM, in the 1% labeled setting. On 312 samples of hospital data with 37 labels, RobECG-CL achieves the best macro AUROC. --- ### AVERT-VLN: Abstention-aware Visual Error Recovery and Training for Vision-and-Language Navigation Authors: Minrui Liu, Jingke Wang, Yuehao Huang, Hao Su, Jiajun Lv, Yukai Ma, Yong Liu Published: 2026-09-30 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.39579v1 Deploying vision-and-language navigation (VLN) agents in unseen environments remains challenging because unfamiliar layouts and visual conditions can cause execution to go off track. Rather than relying on continuous human supervision, a practical strategy is to selectively request corrective guidance, recover the ongoing task, and reuse corrective interactions to improve subsequent navigation. We propose Abstention-aware Visual Error Recovery and Training for Vision-and-Language Navigation (AVERT-VLN), a closed-loop framework that uses a plug-in vision-language Monitor for online human-assisted recovery and offline preference learning. The Monitor operates separately from navigation decision generation and assesses instruction-execution consistency from the instruction, visual history, and current observation. To train the Monitor for deviation recognition, we construct LOSTNAV DATASET with 20K counterfactual risk trajectories and rule-based deviation labels. The Monitor is first fine-tuned on 40K normal trajectories to assess instruction progress and then jointly fine-tuned on normal and risk trajectories to recognize semantic deviations. At runtime, Asynchronous Sidecar Monitoring evaluates execution alongside the navigation model. When the controller accepts a LOST verdict, it suspends autonomous execution and requests human guidance for recovery. For offline policy improvement, Trajectory-Anchored Preference Learning converts deviation-associated failures into decision-level preference pairs under shared decision contexts, restricting supervision to the decisions targeted for correction. Under human-assisted evaluation, the full AVERT-VLN system achieves success rates of 76.2% and 66.3% on the val-unseen splits of R2R-CE and RxR-CE, respectively. The same monitoring and human-assisted recovery interface also improves success rates across the three evaluated navigation architectures. --- ### Thinking Outside the Box: Can Language Models Rely on External Guidance Selectively? Authors: Minghan Wang, Boyuan Wang, Jinhang Zuo, Yuxin Tao, Fang kong Published: 2026-09-30 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.39578v1 Agent harnesses often improve language models with human-designed workflows, but as models grow more capable, unreliable guidance can increasingly constrain their execution. We call the ability to benefit from useful guidance while overriding unreliable guidance thinking outside the box. We introduce Box$^2$-Bench, which holds the model and task fixed while varying workflow reliability to isolate how models regulate their reliance on guidance. On Box$^2$-Bench, frontier models often benefit from reliable guidance but remain vulnerable when it is misleading or becomes unreliable. To test whether this capability can be learned, we train two open-weight models using bad workflows, reserving good workflows for evaluation. We explore two complementary training strategies: counterfactual supervised fine-tuning improves robustness, while outcome-based reinforcement learning can shift the balance toward greater use of helpful workflows. We further find that this behavior extends beyond workflows to other forms of external information, improving peer correction and robustness to corrupted memory. Together, our results identify selective reliance on fallible external information as a dimension of agent reliability not captured by task performance alone. --- ### ECHO-G: Embodied Co-speech Humanoid mOtion Generation Authors: Yizhao Li, Pusen Gao, Ming Wang, Shaojie Shen, Shuo Yang, Hao Xu Published: 2026-09-30 Categories: cs.RO, cs.AI arXiv: https://arxiv.org/abs/2609.39575v1 Generating full-body co-speech motion for humanoid robots requires coordinating speech prosody, linguistic content, and embodiment-specific motion. To this end, we present ECHO-G, a framework jointly conditioned on speech audio and timed transcripts. Its Speech-Grounded Diffusion Transformer (SGDiT) combines frame-aligned acoustic features with token-level linguistic context, preserving their distinct granularities. Trained with rectified flow matching, it models one-to-many utterance-motion relationships directly in robot space. To support training and evaluation, we introduce a BEAT2-derived audio-text-robot dataset and a benchmark covering co-speech characteristics, robot-motion quality, and runtime efficiency. Comparative evaluation supports direct robot-space generation over the evaluated human-motion generation and retargeting pipelines, while modality ablations highlight the benefits of joint audio-text conditioning. We further demonstrate deployment on a physical humanoid robot. A complementary video-rating study also favors joint conditioning over the alternatives. The dataset and training, inference, and evaluation code are available through our project page. --- ### Steering Fields: Adaptive Vector Fields for Safe Image Generation and Beyond Authors: Simone Facchiano, Jan Eric Lenssen, Bernt Schiele, Wolfgang Stammer, Fabio Galasso, Jonas Fischer Published: 2026-09-30 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2609.39573v1 As state-of-the-art text-to-image flow models achieve near-photorealistic quality, controlling their outputs, e.g., suppressing harmful content while promoting benign alternatives, has become a central challenge. The current steering paradigm consists of adding a global steering vector to selected activations. While functional, a fixed and example-agnostic vector applied uniformly along the entire trajectory cannot adapt to the changing state of the generation and often causes unintended global changes. We introduce Steering Fields, a generalization of steering vectors that adaptively re-estimates the steering direction at each step of the generative process. Steering Fields operate on the noisy states of flow models, expose a continuous trade-off between steering strength and content preservation, and are compositional, enabling the simultaneous induction and inhibition of concepts, setting a new state of the art on safety steering benchmarks. Despite using no explicit spatial masks or object priors, the trajectory-adaptive estimation naturally preserves local structure, in a manner reminiscent of image editing. In fact, Steering Fields can serve as a structure-preserving image-editing technique that achieves state-of-the-art semantic fidelity (CLIP, VQAScore), while remaining model-agnostic and inversion-free. --- ### Compact Language, Complex Model Shifts: How and Where Ambiguity and Underspecification Affect LLMs Authors: Michaela Regneri, Nina Scheller, Sören Laue Published: 2026-09-30 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.39572v1 We analyze how lexical ambiguity and underspecification affect language model training. We create artificial homonyms and artificial hypernyms as pseudowords and analyze the generative performance of language models as they are trained with increasing amounts of these ambiguous or underspecified pseudoword types. We further analyze whether the models disambiguate ambiguous or underspecified statements and provide a first mechanistic account of how ambiguity and disambiguation are represented internally. Our main results show that both ambiguity and underspecification increase model performance in ways that scale with their influence on the language's type-token ratio. However, the accuracy of generating sequences containing ambiguous words or their synonyms decreases compared to other texts. We also show that internal representations of pseudowords reflect disambiguation of pseudo-homonyms, but underspecification of pseudo-hypernyms is maintained during the generative process. --- ### Self-Spec Verifiable Code Generation Authors: Jiaru Qian, Yihong Dong, Yongmin Li, Hao Zhu, Bin Gu, Ge Li Published: 2026-09-30 Categories: cs.SE, cs.AI arXiv: https://arxiv.org/abs/2609.39568v1 Large language models (LLMs) may generate unreliable code on corner cases missed by testing, while formal verification can provide machine-checkable guarantees. Recently, researchers have proposed several benchmarks to evaluate the capabilities of LLMs in generating formally verifiable code, where LLMs need to formulate formal specifications, generate the corresponding code, and verify its correctness. However, existing benchmarks have two key limitations: (I) They primarily evaluate specification and code generation stage-wise, with code generation typically conditioned on an oracle specification. This setup overlooks whether strong stage-wise performance translates into end-to-end success. (II)They mainly focus on a single proof-oriented language and mathematically structured tasks, offering limited coverage of tasks common in software development. In this paper, we introduce VeriCodeBench, a benchmark for self-spec verifiable code generation, where the LLM relies solely on its own generated specification and code throughout the entire process. VeriCodeBench contains 400 language-native problems across C, Java, Rust, and Python, covering practical concerns in software development. We evaluate specification coverage, code validity, and joint problem-level success. We further introduce CodeNova to enhance the capabilities of LLMs in self-spec verifiable code generation. CodeNova makes requirements explicit through constraint-guided specification and uses verifier feedback to guide targeted implementation repairs. Experimental results reveal that self-generated specifications remain a major bottleneck, while providing more sophisticated specifications may not necessarily lead to higher verification success rates. CodeNova substantially improves performance across all evaluation metrics, enabling Claude Sonnet 5 to achieve the strongest results under the self-spec protocol. --- ### Invariant Shape Analysis of Surfaces with Spherical Topology Authors: T. Shaska, M. -R. Siadat Published: 2026-09-30 Categories: cs.CV, math.AG arXiv: https://arxiv.org/abs/2609.39567v1 Spherical harmonic descriptors of closed 3D shapes depend on the parameterization, the pose and the scale of the surface, and the standard rotation-invariant reductions, the power spectrum and the bispectrum, discard the relative orientation of the harmonic bands and cannot distinguish a shape from its mirror image. We construct a descriptor that removes all three dependencies exactly and loses nothing else: a conformal parameterization normalized by its conformal barycenter, followed by polynomial invariants of the rotation group. Identifying each harmonic band with a binary form turns the rotation quotient into classical invariant theory and makes reflections visible as the sign of an invariant, so chirality is recorded. The descriptor is complete for the truncated expansion, stable in the orbit distance, and comes with numerical diagnostics. Benchmarks confirm the guarantees, and on bilateral anatomical structures the descriptor separates mirror-image pairs from asymmetric pairs, which parity-blind descriptors cannot. --- ### From Given to Gathered Evidence: Agentic Learning for Longitudinal Medical Reasoning Authors: Minye Shao, Chaohui Yu, Yixuan Wu, Fan Wang, Ling Shao, Yang Long Published: 2026-09-30 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.39566v1 Foundation models can serve as clinical agents through tool-use harnesses. However, conventional medical benchmarks assess reasoning over preselected evidence rather than the ability to seek it across clinical records and longitudinal imaging. We propose CASE: a series of role-specific Clinical Agents for Seeking Evidence, together with a tool-use harness and an agentic post-training framework for compact vision-language policy models. We further introduce a longitudinal multimodal benchmark built on UK Biobank, comprising 50,401 clinical questions derived from real-world ICD-10-coded diagnoses of 4,739 participants. Each question links to a patient-specific environment containing clinical context and multi-sequence MRI from baseline and follow-up visits, where agents autonomously select which visits, organs, modalities, slices, and specialist tools to inspect and compare. Supervised fine-tuning transfers evidence-seeking workflows from 14,734 frontier-model interaction trajectories, followed by agentic reinforcement learning on the learner's own environment interactions. Privileged on-policy self-distillation and rubric-based LLM feedback refine evidence-to-conclusion reasoning without prescribing tool sequences. Experiments show that CASE moves beyond question-answer imitation toward transferable investigation policies, strengthening evidence-grounded longitudinal reasoning. Under matched evaluation conditions, our Qwen3-VL-8B based agent achieves over 16% and 10% relative improvements in answer accuracy over GPT-5.4 and Claude Opus 4.8. Code will be available at https://github.com/VinyehShaw/CASE. --- ### A2Z GameSpec-Bench: How Faithfully Can Coding Agents Generate Games from Game Design Specifications? Authors: Seonho Lee, Wonryeol Jeong, Alberto Cereser, Inha Kang, Hyeonjong Kim, Seungmin Kwak, Dongmin Park Published: 2026-09-30 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.39564v1 Delegating complete application development to coding agents requires preserving the intended design rather than simply producing plausible outputs through naive prompting. Game development provides a demanding testbed, as long-form Game Design Documents (GDDs) describe requirements that must work together across game logic, visual rendering, and player interactions. However, existing game-development benchmarks typically use compact specifications and provide limited support for evaluating interdependent requirements across these aspects in long-form GDDs. We introduce A2Z GameSpec-Bench, a benchmark of 100 long-form GDDs for evaluating end-to-end game development by agents. We measure faithfulness by checking whether the game satisfies the GDD requirements and preserves the relationships among them. Each GDD is turned into a dependency-aware contract that contains rules, constraints, and prerequisite relations. Following game-development practices, we combine source-code inspection with agent-generated test policies for scenario-based replay and adaptive playtesting. The contract remains fixed across agents and revision rounds, while judgments and evidence linked to the same requirements support consistent comparison and failure detection. Our evaluations show that current agents struggle to jointly satisfy interdependent requirements across code implementation and actual play. Requirement-specific feedback improves GDD Fidelity by 10.9% relative to self-revision after two rounds. A2Z GameSpec-Bench assesses end-to-end specification-following ability beyond implementation judgments and provides targeted feedback to support more faithful game development. Code and datasets are available at https://a2z-gamespec-bench.github.io. --- ### RESUME: Recurrent State Updates from Motion and Residual Signals for Efficient Video Language Modeling Authors: Can Zhang, Xiaotian Han, Junyuan Shang, Yuchen Ding, Zhenyu Zhang, Shuohuan Wang, Dianhai Yu, Ruirui Li Published: 2026-09-30 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.39563v1 Existing video language models encode sampled RGB frames independently, so a long video must either exhaust the token budget or drop the changes between sampled frames. Codec-aware front-ends read the motion vectors and residuals that encoding produced, but in their deployed form each predictive frame is still tokenized on its own: the tokens are a function of the current primitives, not of a carried reference. We argue that a more natural function is of both---the current primitives and a carried reference. A clip and its time reversal share the same frames and differ only in the order of changes---an axis that symmetric pooling discards by construction, and that is non-empty in the frozen vision features VideoLMs use---and the codec recurrence already composes those changes in order against a reference state. We introduce RESUME, a stateful codec representation: an anchor I-frame initializes a compact latent state, each subsequent predictive frame is consumed as an update to that state, and a shared readout exposes VideoLM-compatible tokens from the accumulated state. Codec prediction is thereby kept at the representation level and handed to the language model as a trajectory, not as a set of independent token groups. At the same per-predictive-frame token budget as prior codec-aware methods, a predictive frame enters the language model as a readout of what the front-end already knows, not as an encoding of the current primitives alone. Across ten benchmarks, the gains concentrate on temporal reasoning: on all three temporal benchmarks RESUME improves over both the RGB-frame baseline LLaVA-Video-7B (by 2.8, 5.1, and 3.9 points on TempCompass, TOMATO, and MVBench) and the codec-based baseline CoPE-7B, while staying competitive on general and long-form QA. Frozen-transition tests further show anchor dependence, order sensitivity, and useful rollout behavior beyond the training horizon. --- ### Candidate Retention for Abductive Learning Authors: Hao-Yuan He, Yu Liu, Ming Li Published: 2026-09-30 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2609.39561v1 Abductive learning combines neural perception with symbolic reasoning, using explanations generated by abduction to supervise the perception model. Multiple valid explanations of the same symbolic target can assign conflicting labels to the same inputs. Common policies select a single candidate as a pseudo-label, which may reinforce mistaken assignments, or weight all candidates, which may spread supervision across competing labels. These risks motivate selecting a retained subset to balance supervision sharpness and model-mass coverage. To guide this choice, we bound the coordinate-level supervision error using retained uncertainty, discarded model mass, and model mismatch. For a fixed model and training pair, only the first two terms depend on the retained set. We propose Abductive Candidate Retention (ACR), which uses these terms to guide greedy additions, accepting a candidate when its recovered mass exceeds the increase in retained uncertainty. Experiments show that ACR improves concept accuracy over single-candidate baselines and A3BL in most evaluated aggregated mod-addition settings. Objective ablations support the joint use of uncertainty and posterior mass. --- ### Self-Repulsive Sampling for Diffusion Language Models Authors: Michael Helcig, Martin Jaggi Published: 2026-09-30 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.39560v1 Sampling several responses and voting over their answers can improve a language model's accuracy, but repeated answers limit the benefit of additional samples. Raising temperature increases diversity at a potential cost to per-sample accuracy. We introduce Self-Repulsion (SR), a sampler for masked diffusion language models that uses peer commitments to diversify the pool. At each penalized denoising step, each path lowers a token's logit according to how many peers have committed that token at the same position. Paths share a batched forward pass and then commit in sequence, so later paths observe choices made earlier in the same step. This coupling requires no training or additional forward or backward pass and can produce distinct paths even at temperature zero. When all paths commit a position together from identical logits, the update exactly maximizes total logit minus a convex duplication cost. On LLaDA-8B-Instruct with ten paths and 128 denoising steps, deterministic SR reaches 80.38% plurality accuracy on GSM8K, compared with 70.17% for the unpenalized greedy decoder. At temperature 0.6 and matched model-evaluation budgets, the count penalty improves over self-consistency by 2.06 percentage points in blocks of 32 and 14.50 under pure diffusion. Experiments on GSM8K, MATH and TruthfulQA show that voting gains arise mainly from higher coverage of correct answers, with gains that vary by benchmark and decoding regime. --- ### Divide and Collapse: MAPF-Collapse via Exact Decomposition into Independent Sub-Instances Authors: Oren Salzman Published: 2026-09-30 Categories: cs.AI, cs.RO arXiv: https://arxiv.org/abs/2609.39559v1 In this work we study the problem of MAPFC, a post-optimization step for Multi-Agent Path Finding (MAPF) plans where we are given a feasible plan produced by a modern MAPF solver and are tasked with removing avoidable moves while preserving feasibility. This NP-hard problem naturally arises when using learning-based state-of-the-art (SOTA) solvers which construct plans that contain redundant moves that can be removed. Recently, Tang et al. presented Judgelight, which uses Integer Linear Programming (ILP) to solve MAPFC. Importantly, the ILP is constructed over all agents jointly, so its cost is governed by the full instance rather than by the small coupled residue that actually requires joint reasoning. Our key insight, motivating this work, is that MAPFC instances naturally decompose into independent sub-problems, most of which involve a single agent and can be solved without any inter-agent reasoning. To this end, we first identify which agents need to coordinate their motion and partition the instance into sub-problems accordingly. For the cases where no coordination is required, we introduce an extremely lightweight solver that is $\approx\!1{,}900\times$ faster than Judgelight. For cases where coordination is required, Judgelight can be used but we introduce an alternative CBS-like solver which is more efficient on easier problems. The resulting framework is exact, uses no commercial ILP solver, and matches Judgelight's quality while running substantially faster on the coordination-light majority of instances; on the coordination-heavy instances we propose a regime-aware hybrid planner that falls back to Judgelight. Over all benchmarks tested, this planner achieves a median $10.5\times$ per-instance speedup over Judgelight. --- ### General Performance Guarantee for Human Torque Estimation-Based Task-Agnostic Assistive Exoskeleton Control Authors: Duy Hoang, Bastien Berret, Olivier Bruneau, Laurent Fribourg Published: 2026-09-30 Categories: cs.RO, cs.LG, eess.SY arXiv: https://arxiv.org/abs/2609.39558v1 Accurate human torque estimation is crucial for enabling task-agnostic control in robotic exoskeleton systems. However, estimation errors may cause mismatches between the robot assistance and the human intention, degrading controllability and task performance. In this paper, we address this issue by formally defining matched assistance as scenarios in which the robot positively contributes to human movement. Based on this definition, we develop a theoretical framework to design the robot's desired interaction torque that guarantees a lower bound on the matched assistance probability. Importantly, the proposed guarantee holds over the entire torque distribution, including unseen data beyond the training tasks. This provides our method with strong reliability and generalization, both of which are critical for effective exoskeleton control. The proposed strategy is implemented on the ABLE upper-limb exoskeleton and evaluated in a multi-task setup. Experimental results validate the theoretical guarantees and demonstrate that the proposed strategy achieves effective general performance across several tasks, guaranteeing movement smoothness while reducing human physical effort. --- ### EffGS: Efficient and High-Fidelity Gaussian Splatting Authors: Changbai Li, Shuo Yang, Yichen Yang, Shuwei Shao, Huobin Tan Published: 2026-09-30 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.39553v1 3D Gaussian Splatting (3DGS) enables real-time novel view synthesis, but existing general-purpose acceleration methods suffer severe rendering quality degradation when extended to more complex, large-scale scenes. To address this issue, we propose EffGS, a more general acceleration framework that improves training and rendering efficiency while maintaining reconstruction quality comparable to or better than vanilla 3DGS across bounded and large-scale scenes. EffGS combines frequency-aware guidance, localized density control, and adaptive primitive scale modulation. First, an importance scoring mechanism combines pixel-wise reconstruction errors with a difference-of-Gaussians mask scheduled over training to provide stage-dependent spatial guidance. Second, localized densification and pruning restricts density modifications to Gaussians with valid projected footprints in the sampled views. Third, learnable per-Gaussian scale modulation adjusts effective primitive extent during optimization while retaining the Compact Box rasterization rule. Extensive experiments on bounded and large-scale scene datasets demonstrate a favorable balance between reconstruction quality, training time, and primitive count. Component ablations and matched-primitive-budget comparisons further support the effectiveness of the framework. --- ### Ghost in the Encoder: Decodable Artist Identity Representations in Lyrics-to-Song Generation Authors: Arhan Vohra, Choenden Kyirong, Laura Ibáñez-Martínez, Martín Rocamora Published: 2026-09-30 Categories: cs.SD, cs.LG arXiv: https://arxiv.org/abs/2609.39552v1 Text-to-song generation models can be prompted to imitate specific artists or regurgitate entire songs from their training data. Although these phenomena have been documented behaviorally on small datasets, little is known about the internal representations that may give rise to them. Prior interpretability work on generative audio has focused on locating semantic concepts such as genre or time signature within model activations. In this work, we show that a trained model can be probed for linearly decodable representations of artist identity from song lyrics alone, without any additional identifiers. Through a controlled case study of ACE-Step 1.5 spanning 2,000 songs across 100 artists, we demonstrate that the artist associated with a given set of lyrics can be identified within the model's internal activations, and that this conditioning signal propagates from the lyric encoder to the diffusion backbone during inference. These findings indicate that lyrics constitute an artist-level conditioning channel not addressed by prompt-side replication safeguards. More broadly, our work highlights how latent-space analysis can be used to audit what generative music models have implicitly learned from their training data. --- ### RankEvolve: A Reliable Multi-Agent Auto-Research Harness for Evolving Ranking Models Authors: Zheng Chen, Linfeng Liu, Hong Li, Hong Yan Published: 2026-09-30 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.39551v1 Auto-research agents, LLM systems that propose, implement, train, and evaluate model changes across iterations, promise to automate applied ML's experimental loop. Over long horizons, execution accuracy is a binding constraint: a change can silently leak held-out data, omit normalization, disconnect a gradient, or leave a train/eval flag unwired, invalidating expensive runs and compounding error across iterations. We present RankEvolve, an auto-research framework for evolving generative ranking models. An Executable Operating Protocol (EOP) declares phases, gates, branches, and loops, and the runtime enforces the compiled state machine. A meta-meta-harness composes complete black-box coding-agent products, including Claude Code and Codex, as execution-graph nodes that review and repair one another's work. In a budget-matched evaluation, heterogeneous composition raises all-oracle execution accuracy from the best single-product baseline of 45.8 percent to 62.5 percent (paired +16.7 points, 95 percent CI (6.6, 26.7)) while achieving a 10.4 percent silent critical-defect rate. An implemented knowledge layer carries findings, including negative results, across iterations. In a twelve-iteration deployment on the open-source HSTU recommender, RankEvolve reported NDCG@10 of 0.2192 on MovieLens-20M LARGE (+4.48 percent over the published anchor) and 0.1948 on BASE (+2.80 percent). ExecML-HSTU, seeded by incidents from that deployment, provides the oracle benchmark for the execution-accuracy evaluation. A pre-specified LitGPT transfer split replicates the heterogeneous-composition effect beyond recommendation (+12.5 points, 95 percent CI (3.0, 22.0)), and a paired ablation isolates per-step from full-protocol instruction injection. These results characterize when runtime-controlled composition of coding-agent products improves execution accuracy. --- ### Hyperbolic Prototype Routing for Rehearsal-Free Class-Incremental Learning Authors: HongWei Zhao, Rui Liu, Yong Chen Published: 2026-09-30 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.39550v1 Class-Incremental Learning (CIL) aims to continually learn new classes while preserving prior knowledge. Parameter-efficient fine-tuning with pre-trained models enables CIL with minimal parameter updates, but existing approaches still suffer from catastrophic forgetting caused by cumulative interference and suboptimal module-sample matching at inference. We propose Hyperbolic Prototype Routing (HyPro), a rehearsal-free framework for continual learning. HyPro allocates a dedicated LoRA-Expert module to each incremental task for isolated representation learning, then projects routing features onto a Poincare ball and performs geodesic nearest-prototype matching for reliable task-level discrimination. Extensive experiments on standard CIL and Few-Shot CIL benchmarks show that HyPro consistently improves average and final-stage accuracy over strong baselines. --- ### Speculative Safety Honeypot: Toward Proactive Defense Against Multi-turn Agent Attacks Authors: Zezhong Wang, Xueyang Tang, Rui Lian, Yang Lou, Heqing Huang Published: 2026-09-30 Categories: cs.CR, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2609.39549v1 As Large Language Model (LLM) agents are increasingly deployed in complex environments, multi-turn interaction attacks have become a significant security challenge. Existing detection methods typically rely on historical context. However, this retrospective logic struggles to identify deep malicious intents that are split across turns to hide future risks. Inspired by speculative decoding, we propose the Speculative Safety Honeypot (SSH) framework. SSH uses a multi-agent simulation system composed of small LLMs to build an action-level speculate-and-verify workflow. In the speculation stage, SSH predicts future behaviors of the target agent and asynchronously builds a trajectory tree to expose potential risks in advance. In the verification stage, the system uses the target agent's real actions to calibrate and prune the trajectory tree, effectively reducing false positives. As a plug-and-playable component, SSH provides existing detectors with rich decision redundancy beyond the current interaction slice. By judging risk based on the evolution of the entire trajectory tree rather than a single point in time, the system reduces the reliance on the absolute precision of individual detection components. This improves the defense resilience and the warning lead-time of agent systems against complex temporal attacks. --- ### Learning Normal Diffusion Dynamics for Backdoor Defense in Text-to-Image Models Authors: Junjian Li, Xiaolong Liu, Peng Sun, Liantao Wu, Linghan Chen, Yudong Gao, Honglong Chen Published: 2026-09-30 Categories: cs.CV, cs.AI, cs.CR arXiv: https://arxiv.org/abs/2609.39548v1 Backdoor attacks pose a serious threat to the secure deployment of text-to-image (T2I) diffusion models. Existing defenses typically detect backdoors from specific abnormal patterns in internal representations, which may limit their generalizability with the emergence of increasingly diverse attack mechanisms. In this paper, we study backdoor defense of T2I diffusion models from a transition-dynamics perspective. We observe that benign diffusion trajectories exhibit structured and timestep-dependent transition patterns from cross-attention, latent and noise spaces, whereas backdoor attacks tend to induce deviations from such normal evolution. Motivated by these observations, we propose Normal Diffusion Dynamics Learning (NDDL), a novel backdoor defense framework that learns the normal transition dynamics of diffusion trajectories utilizing only benign samples. NDDL constructs compact multi-space trajectory representations and trains a timestep-conditioned dynamics model to predict the diffusion evolution. In the inference phase, deviations between the observed and predicted transitions are exploited to quantify dynamics inconsistency for backdoor detection. NDDL further enables trigger localization without any prior knowledge of the embedded backdoor by performing substitution with low-semantic words. Extensive experiments for diverse backdoor attacks demonstrate the effectiveness and generalizability of our proposed NDDL. --- ### Learning Reliable GUI Agents under Imperfect Priors Authors: Bo Han, Qianyi Wang, Shuai Liu, Xiong Zifan, Changqiao Wu, Yuanfa Li, Pengzhi Gao, Wei Liu, Jian Luan, Heng Qu, Yunpeng Song, Zhongmin Cai Published: 2026-09-30 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.39547v1 GUI agents built on large language and vision-language models still struggle on unseen applications and complex multi-step tasks, as completing real GUI tasks depends on app-specific, temporally volatile operational knowledge that is scarce in pretraining corpora. Retrieval-augmented execution offers a natural remedy but faces two coupled bottlenecks: knowledge at scale is hard to acquire, and self-collected priors inevitably drift from the live environment due to version updates, promotions, ads, A/B tests, and personalization. We therefore argue that GUI agents should not pursue perfect knowledge but learn to act correctly under imperfect priors, and propose our framework that couples knowledge acquisition with noise-robust utilization: a structured exploration strategy traverses interactive elements, builds a UI state-transition graph, and synthesizes (task, trajectory) pairs via a VLM without human annotation; a noise-aware training strategy, grounded in a taxonomy of real GUI drift patterns, injects five types of realistic errors into self-explored trajectories to teach the agent to assess prior reliability before acting. Experiments on physical devices and online emulator benchmarks show that our method discovers more unique screens, covers more benchmark tasks, and more effectively rejects erroneous priors while leveraging correct ones, with accuracy gains that transfer across datasets. --- ### Growing an Agent/Prover Interface: Evolutionary Tool Design for Cost-Efficient Theorem Proving in Rocq and Lean Authors: Jules Viennot, Guillaume Baudart, Marc Lelarge Published: 2026-09-30 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.39544v1 Recent achievements in AI-assisted mathematics require intensive interaction of agents with proof assistants to generate machine-checked proof certificates. Agents interact with proof assistants such as Rocq or Lean through an interface that controls what the agent receives from the prover and the cost of these interactions. Today, these interfaces are adapted from tools designed for humans and not optimized for agents. We propose an evolutionary method where a frontier model incrementally proposes new features and only keeps the ones that improve the overall performance of smaller models. We demonstrate the effectiveness of our method by growing, on a curated set of mathematical problems, \rme, a new MCP server for the Rocq prover. On the held-out \texttt{test} split of miniF2F-Rocq, an agent equipped with \rme outperforms both the baseline that only exposes the Rocq compiler and an established MCP server, across four models from two families, in success rate, cost per solve, and time per solve. Although evolved for Rocq, the resulting server transfers to Lean, improving cost and time per solve on a subset of PutnamBench. We release \rme and its port to Lean. --- ### Comparative study of adapting pre-trained models for driving behavior video captioning Authors: Sayak Mallick, Philipp Geiger, Augustin Kelava Published: 2026-09-30 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2609.39542v1 This report examines and compares some of the many fine tuning and prompting methods existing, applying them within the domain of autonomous driving. The idea is to compare these methods by adapting a Large Language Model (LLM) on a video dataset. LLM's have become extremely good at achieving a good understanding of different forms of data and this study aims to induce a low dimensional understanding of driving situations into our primary test model SpaceTimeGPT. Experiments on BDD-X (Berkeley DeepDrive eXplanation) dataset demonstrate good performance of the full fine tuning framework on some automatic metrics, and in some metrics, it even surpasses the baseline. We also try Low-Rank Adaptation (LoRA) and prompt engineering on VideoLLaVA model and discuss its limitations. --- ### A Reusable Semantic Web Framework for Evidence-Grounded Fundamental Rights Impact Assessments under the EU AI Act Authors: Faith Olopade, Delaram Golpayegani, David Lewis Published: 2026-09-30 Categories: cs.CY, cs.AI arXiv: https://arxiv.org/abs/2609.39537v1 The EU AI Act (Art. 27) requires deployers of high-risk AI systems to conduct Fundamental Rights Impact Assessments (FRIAs) before deployment, yet the evidence needed for credible assessments is fragmented across incompatible incident repositories, risk vocabularies, and legal texts. We present a reusable Semantic Web-based framework that consolidates this evidence for two high-risk public sector categories: employment and worker management (Annex III(4)) and access to essential public services (Annex III(5)(a)). A curated 150-record corpus is annotated along four axes using keyword, LLM, and hybrid methods and serialised as a SPARQL-queryable knowledge graph of 1,351 RDF triples. Five FRIA demonstration scenarios surface 103 records (68.7% coverage). Evaluation against a 69-record gold standard reveals that LLM-assisted classification of the employment domain achieves only $κ= 0.045$, a cautionary result for automated fairness-related evidence retrieval in this domain. All artefacts are released openly to support adoption by regulators, national authorities, and SMEs. --- ### CATCH: A Controllable Analysis Testbed for Reward Hacking in Coding RL Authors: Shouli Wang, Yanfeng Jia, Zhihao Ou, Zitao Su, Ruize He, Haotong Xie, Hao Peng, Juanzi Li, Xiaozhi Wang Published: 2026-09-30 Categories: cs.CL, cs.LG arXiv: https://arxiv.org/abs/2609.39533v1 During reinforcement learning with verifiable rewards (RLVR), large language models (LLMs) can exploit loopholes in their environments to obtain high rewards without improving the intended capabilities, i.e., reward hacking. Despite its risks to training efficiency and safety, monitoring and mitigating reward hacking during training remain challenging, which is limited by a lack of testbeds that reproduce hacking and reliably identify it. We introduce CATCH, a controllable testbed for studying reward hacking in coding RL. CATCH deliberately exposes environmental loopholes and provides execution-based gold labels by comparing success under a vulnerable evaluator with task correctness under an independent audit. It also can control the model's initial hacking tendency through supervised fine-tuning data mixtures and the difficulty of earning rewards through reward designing, enabling systematic comparisons of hacking dynamics and interventions. Experiments show that CATCH can produce diverse RL training trajectories with clear reward hacking, and analyses demonstrate that both initial models and reward difficulties shape the emergence of reward hacking. We further evaluate the effectiveness of different reward hacking detection and mitigation methods. A key finding is that a chain-of-thought monitor initially suppresses hacking, but this protection erodes as the policy model learn to mislead the monitor with code comments. This highlights the need to evaluate hacking mitigations throughout training with CATCH. The source code and resources are publicly released at https://github.com/THUAIS-Lab/CATCH. --- ### WEIRDO: WEak resIdual Regularized DOob's h-transform diffusion alignment Authors: Denis Suchkov Published: 2026-09-30 Categories: math.ST, cs.LG arXiv: https://arxiv.org/abs/2609.39531v1 We study the problem of estimating the guidance that steers the distribution learned by a diffusion generative model toward a tilted target $q_0 \propto w\,p_0$ at inference time. Relying on the stochastic optimal control approach, we observe that the exact drift correction is the gradient of the logarithm of Doob's $h$-function, and we study the problem of estimating it from a sample. In the present paper, we assume that the score of the pretrained model is available, that the tilting weight is bounded and positive, and that the reference distribution has a bounded support, no smoothness of the weight is required. Introducing a penalized least-squares risk in which the penalty is the residual of the space-time harmonicity equation satisfied by the $h$-function, measured in a dual Sobolev norm, we derive high-probability bounds on the squared error of the resulting guidance estimate. Since the penalty vanishes at the target, the estimator is free of regularization bias, and in favourable scenarios its rate of convergence is faster than the minimax rate of estimating first-order derivatives of a smooth regression function. Assuming that $w$ is bounded and positive with $\mathbb{E}_{p_0}(w^{-\mathrm{s}}) < \infty$ for some $\mathrm{s} \in (0,\infty)$, and that the reference data are compactly supported, we prove that the guidance is estimable in squared $L^2$ at rate $\varepsilon_n^{\mathrm{s}/(\mathrm{s}+4)}$, where $\varepsilon_n = n^{-2(β-1)/(2(β-1)+d)}.$ We also transfer the obtained bounds to the total variation distance between the marginals of the estimated and the exactly guided samplers, and illustrate the performance of the suggested approach with numerical experiments. --- ### Lens Flare Removal and Reconstruction Authors: Tarun Yenamandra, Jonathon Luiten, Daniel Cremers, Nathan Matsuda Published: 2026-09-30 Categories: cs.CV, cs.GR arXiv: https://arxiv.org/abs/2609.39527v1 The presence of lens flares in images can significantly reduce the quality of downstream application results for tasks such as 3D scene reconstruction. This is because lens flares are a property of the camera imaging system, and not a part of the underlying scene being modeled. There are previous methods that tackle the removal of small flares focused around a light source. However, existing methods struggle with large flares, such as those that fill the entire image. In this work, we compile a novel dataset for large-flare removal, combining publicly available real-world data with a procedural generation pipeline. We fine-tune a diffusion-based model on our dataset to remove complex, large lens flares. On the other hand, lens flares remain effective artistic tools, widely used in the media. While there are ways to simulate 2D flares, representing and reconstructing lens flares consistently across multiple views has not yet been explored. To achieve this, we introduce a flare representation model that leverages the symmetry of lens flares about the camera's principal point. We propose a computational pipeline to jointly optimize this flare model and a Gaussian splatting model (3DGS). This enables the decomposition of a 3D scene into lens flares and the scene itself, using our flare-removal model. Because the reconstructed flare is explicit and re-renderable, it can be edited and transferred to novel images and new 3D scenes. We evaluate removal on an established benchmark and a new one for large reflective flares, quantify the flare/scene decomposition directly, and show that the pipeline is robust to errors in automatic light-source localization. --- ### Mitigating Representation Gaps in Amortized Bayesian Inference with Auxiliary Supervision Authors: Hans Olischläger, Svenja Jedhoff, Šimon Kucharský, Aayush Mishra, Stefan T. Radev, Paul Bürkner Published: 2026-09-30 Categories: stat.ML, cs.LG, stat.CO, stat.ME arXiv: https://arxiv.org/abs/2609.39525v1 Casting Bayesian inference as a neural network optimization problem targeting an amortized posterior is attractive, as it extends to otherwise intractable statistical models and offers near instantaneous inference for new datasets after prepaying the training cost. Although theory guarantees faithfulness under ideal convergence, practical amortized inference still requires iterating over architectures and optimization choices and ultimately ``satisficing'' under finite simulation, compute, and time budgets. Even the best-performing solution may thus retain avoidable representation gaps that typically require problem-specific fixes. Here, we propose a generic alternative which improves training dynamics with auxiliary guidance losses applied to internal representations. Specifically, we show how such guidance leads to faster convergence when training data is abundant and to better performance when it is scarce. We formalize representation gaps as getting stuck in a local optimum at the information bottleneck between the parts of the network tasked with feature learning and those tasked with conditional distribution learning, and offer a generic diagnostic to separate summary failures from inference failures. Finally, we demonstrate that auxiliary supervision improves convergence speed and accuracy on a range of challenging real-world inference problems. --- ### CIDER-FM: Foundation Models for Causal Inference from Diverse Experimental Regimes Authors: Yuche Gao, Arik Reuter, Siyuan Guo, Anish Dhir, Bernhard Schölkopf, Adrian Weller Published: 2026-09-30 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.39523v1 Causal foundation models (CFMs) amortise causal inference over priors of synthetic structural causal models (SCMs), predicting the effect of an experiment on a specific variable. However, observational data alone may leave multiple causal models compatible with available evidence, while experimental data with interventions on exactly the variable of interest might be unavailable. This work studies CFMs as a method to combine finite observational and surrogate-interventional datasets in order to predict a target conditional interventional distribution (CID) more accurately than with observational data alone. We first formalise the conceptual benefits of surrogate experiments. Building on this analysis, we introduce \textsc{Foundation Models for Causal Inference from Diverse Experimental Regimes} (\emph{CIDER-FM}), a causal foundation model that uses an intervention-aware representation and hierarchical three-axis attention to exchange information across variables, samples, and experimental regimes. We evaluate CIDER-FM against a wide range of baselines across diverse synthetic graph and mechanism families, as well as on both simulated and real-world data from Causal Chambers. Our results demonstrate strong CID prediction performance and show that incorporating experimental context can improve predictions over observational data alone. --- ### Referential Uncertainty in Human--AI Collaboration Authors: Christian Poelitz, Finale Doshi-Velez, Siân Lindley Published: 2026-09-30 Categories: cs.AI, cs.HC arXiv: https://arxiv.org/abs/2609.39518v1 Effective human-AI collaboration requires partners to establish references through interaction, which becomes fragile when descriptions are ambiguous, similar referents compete, or partners see different things. We study referential uncertainty - uncertainty over which candidate object a description refers to - in a collaborative puzzle task where a human Helper instructs an AI Worker to place pieces. The Worker must identify and communicate its uncertainty, and the Helper must recognize and act on it. We show that a separately elicited belief distribution over candidate pieces is better calibrated (ECE 0.15) and better discriminates correct from incorrect placements (AUROC 0.65) than raw action-token probabilities, which are severely overconfident (0.97 mean confidence, ECE 0.44). Across three frontier vision-language models (GPT-4.1, GPT-5, GPT-5.5), this elicited uncertainty rises predictably with instruction vagueness, but not with competing referents in context, even when those increase errors. The models seldom externalize it, asking for clarification on only 3.5-16.7% of turns. In a controlled human study (N=210), participants given only the Worker's default message accept 78% of wrong placements and cannot tell right from wrong (AUC 0.50). Precise descriptions and, especially, well-targeted hedges cut wrong-move acceptance to 36% while largely preserving correct-move acceptance, compensating for missing shared awareness such as not seeing the Worker's action. But this benefit depends on targeting: a deployable hedge derived from the model's own belief entropy inherits that signal's weakness and can do more harm than good. Externalized uncertainty helps a human partner only when it is accurately targeted. --- ### Spike-driven Vision-Language-Action Model Authors: Shuai Wang, Malu Zhang, Mingquan Liu, Weihui Dai, Dehao Zhang, Jieyuan Zhang, Yimeng Shan, Zijian Zhou, Yang Yang Published: 2026-09-30 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.39514v1 Vision-language-action (VLA) models bridge multimodal understanding and robotic control, advancing the dominant paradigm for embodied intelligence. However, most existing models rely on large Transformers, whose latency and energy costs hinder deployment on resource-constrained platforms. Through sparse event-driven computation, spiking neural networks offer a promising paradigm for high-performance and energy-efficient computing. Here, we propose the first Spike-driven VLA framework enabling end-to-end direct training for robotic manipulation, which mainly comprises three core components. First, we develop spiking visual and instruction encoders for multimodal perception, encoding visual observations and language instructions into sparse, reliable spike representations for subsequent cross-modal fusion. Then, we introduce Multi-Winner Spike Fusion for instruction-guided scene understanding, using bidirectional top-$k$ winner-take-all spike routing to suppress background interference and yield fused memory. Finally, we propose a Spike Action Chunking Transformer that incorporates spiking cross-attention over the fused memory and the current robot state, enabling efficient end-to-end generation of continuous action chunks for robotic control. Extensive experiments on LIBERO and Meta-World demonstrate that Spike-driven VLA achieves competitive performance with fewer parameters and lower estimated inference energy than conventional VLA models. This work establishes a foundational framework for neuromorphic VLA modeling, paving the way for future advances in resource-efficient embodied intelligence. --- ### Can Domain Generalization be Guaranteed in Small-Sample Learning? Authors: Hong Zheng Published: 2026-09-30 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.39512v1 The small-sample learning problem remains a fundamental challenge in machine learning because limited training data lead to unstable model estimation and generalization. Structural Risk Minimization (SRM) has long been regarded as a principled solution under the classical i.i.d. assumption. However, domain generalization (DG) violates this assumption, leaving the theoretical role of SRM in DG largely unexplored. To bridge this gap, we establish the first theoretical guarantees for SRM in DG under mild assumptions. Specifically, based on the concept of stability, we derive learning consistency and generalization error bounds and prove that these bounds become tight when the hypotheses satisfy the stability condition. Building upon this, under a specific hypothesis space assumption, we establish stability, learning, and generalization bounds for SRM. We further discuss the applicability of these bounds to deep learning. This work establishes theoretical foundations for SRM under distribution shifts and sheds light on the design of robust DG algorithms in small-sample scenarios. --- ### PartiCam: Camera Controlled Video Generation with Reward Guidance Authors: Amine Ouasfi, Runjia Li, Junlin Han, Eric Marchand, Philip H. S. Torr, Adnane Boukhayma Published: 2026-09-30 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2609.39504v1 We present PartiCam, a training-free Particle filtering rooted method for improved Camera controlled video generation. Generating videos that follow a precisely specified camera trajectory remains challenging for large video diffusion models. Training-free approaches are backbone-agnostic and avoid the need to construct large camera-annotated datasets by steering pretrained models toward the desired camera motion at test time. This enables the generation of camera-controlled video data that can subsequently be used to train camera-conditioned video diffusion models. Existing sampling-based guidance approaches often suffer from unstable trajectories: they either explore too broadly and fail to respect the target camera motion or collapse early and lose visual diversity over time. We introduce a global-local refinement framework for diffusion reward guidance, enabling accurate and consistent camera control during video generation. Our method builds on Sequential Monte-Carlo (SMC) guidance, but introduces a local refinement stage based on particle filtered resampling. Experiments show large improvements in camera trajectory adherence, reduced drift, and better visual quality, without requiring model retraining. --- ### The Geometry of Randomized Smoothing on Feasible Sets Authors: Syed Izhan Khilji, Alireza Furutanpey, Schahram Dustdar Published: 2026-09-30 Categories: cs.LG, stat.ML arXiv: https://arxiv.org/abs/2609.39497v1 Randomized smoothing certifies the probability of a fixed output event as the center of Gaussian noise moves. Feasibility or confidence filtering reports label probabilities only among retained proposals, producing a ratio. Its numerator is a fixed Gaussian event mass, while its denominator is the probability of retention and can change with the center. Substituting this ratio into the ordinary smoothing formula can therefore certify a ball that contains a decision boundary. We separate the problem into a geometric question and a certification question. Geometry determines when conditioning preserves Gaussian comparisons. Convex retained sets preserve the full comparison, while general sets require geometric control of the retained law as the center moves. Without such control, conditional probabilities imply no positive universal radius. Joint retention-and-label probabilities always yield a valid certificate for the same filtered predictor. A uniform covariance bound transfers divergence certificates to the retained law and can yield larger radii even when the Gaussian event comparison fails. Both methods admit finite-sample bounds. For a learned image classifier with a training-selected nonconvex filter, conditional Rényi bounds certify more images than joint-mass bounds without additional model evaluations. A released confidence filter exhibits verified label changes inside radii obtained by conditional substitution. An application of adaptive Gaussian composition covers causal finite-horizon executions with history-dependent center shifts under a pathwise energy bound. --- ### When the Right Answer Is Missing: An Arithmetic-Dependent Rejection Bottleneck in Jev Authors: Jike Zhong, Ming Li, Yuxiang Lai Published: 2026-09-30 Categories: cs.LG, cs.CL arXiv: https://arxiv.org/abs/2609.39496v1 Typed decision models such as Jev offer an efficient alternative to generative LLMs in decision-making workflows by selecting directly from predefined options. When candidate sets contain no valid answer, TypeSafe recommends including an "other" or "none-of-the-above" option to enable rejection. In this report, however, we identify an arithmetic-dependent rejection bottleneck: Jev reliably selects correct numerical answers when available but frequently accepts incorrect alternatives when they are absent despite an explicit rejection option. On paired arithmetic problems, answer-present accuracy reaches 99%, while correct rejection falls to 7%. Moreover, this gap persists across numerical magnitudes, operation depths, contextual formulations, and rejection labels, and extends to scenarios such as time calculation and capacity rounding. Yet native Boolean verification achieves 99% exact-match accuracy on the same answer-absent arithmetic cases, showing that categorical rejection can fail even when the model successfully verifies candidate correctness. Finally, we show that a simple decision threshold selected on separate development problems raises arithmetic rejection accuracy from 7% to 79% while retaining 97% answer-present accuracy, substantially mitigating the failure without retraining or additional inference. --- ### Disentangling Self-Distillation: Measuring and Modeling Acquisition and Retention Authors: Luis Zuin, Alexis Huet, Dario Rossi, Zied Ben Houidi Published: 2026-09-30 Categories: cs.AI, cs.LG arXiv: https://arxiv.org/abs/2609.39494v1 Self-distillation with privileged context adapts a language model from demonstrations by letting the model, once conditioned on a reference response, teach its context-free copy token by token. Our taxonomy reveals existing methods differ along three entangled axes: (i) the rollout source (student or teacher), (ii) the teacher coupling (frozen, or an exponential moving average of the student at some coupling rate) and (iii) the KL direction (reverse or forward), yet these axes are usually studied in fixed combinations and have led to conflicting conclusions. We formalize a unifying framework to encompass all self-distillation methods vs classic supervised fine-tuning: we train every combination of the three axes, on Qwen2.5-7B and Ministral-3-3B across ordinary and contradictory tasks, totaling 1,200 adaptation runs, to systematically investigate the impact of the above axes. We propose a controlled model of the same objective to explain the resulting acquisition-retention trade-offs. We find that (i) the rollout source matters mostly where the task contradicts the pretrained behavior: there teacher rollouts raise acquisition well above what student rollouts achieve, with almost no change in retention; (ii) the teacher coupling changes acquisition most, on every task: acquisition rises with the coupling rate, then falls past a task-specific rate; (iii) switching the KL direction costs retention in one model but not the other so which axis to tune first depends on the model. The controlled model reproduces the three trends. --- ### Front-to-Back: Benchmarking Vision-Language Models for Asymmetric Cross-View Vehicle Re-Identification Authors: Moseli Mots'oehli, Thulani Babeli Published: 2026-09-30 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.39492v1 Matching the same vehicle across front and rear cameras is difficult because the cameras do not share a view and the vehicle's appearance changes substantially. We introduce Front2Back-ReID, a benchmark of 500 manually verified vehicle handovers from 20 recording sequences in South Africa. Each example asks a model to match a vehicle highlighted in a front-camera image to the same vehicle among at least three candidates in a later rear-camera image. We evaluate seven zero-shot vision-language models, four image-retrieval baselines, and 25 human participants. Models are tested using full front RGB images, cropped target vehicles, and binary silhouettes. The strongest VLM achieved 76.6 percent Rank-1 accuracy on target crops, compared with 74.0 percent for the frozen SigLIP2 baseline; this difference was not statistically clear. Human participants achieved 94.0 percent accuracy with full images and 92.2 percent with target crops. Under our evaluation setup, enabling reasoning improved accuracy across all three input conditions for every model evaluated in both modes. We also found that VLMs generally performed worse on full scenes than on target crops. These results show that general-purpose VLMs do not yet consistently outperform strong visual retrieval for front-to-rear vehicle matching, while humans remain substantially more reliable. --- ### OmniReasoning: Pushing the Limits of Audio-Visual Joint Reasoning Authors: Junming Lin, Yuxuan Wang, Zhenxin Lei, Yuxin Liu, Ruixun Liu, Yinsong Yan, Ling Wang, Minghao Han, Yunfei Chu, Shun Lei, Xueyao Zhang, Qize Yang et al. Published: 2026-09-30 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2609.39490v1 Recent advances have enabled unified omni-modal models in understanding audio, vision, and language. However, existing benchmarks, training data, and learning methods largely treat the modalities independently, leaving the capability of audio-visual joint reasoning poorly evaluated and insufficiently elicited. We address this gap with a benchmark, data engine, and learning method. First, we introduce OmniReasoningBench, a benchmark where both audio and visual evidence are indispensable. It comprises 1,150 multiple-choice and open-ended questions across two tasks, reasoning over video and reasoning beyond video. Second, we develop a data engine OmniQA. It automatically constructs evidence-grounded QA pairs that explicitly necessitate audio-visual joint reasoning, together with time-stamped clue chains that guide the annotation of thinking process. Besides our benchmark, this engine produces training data OmniReasoning-SFT-112K and OmniReasoning-RL-19K. Finally, we propose an on-policy self-distillation method Modality-Factored Self-Distillation (MFSD). It evaluates each sampled response under modality-specific clue contexts, disentangling the contributions of individual clues and their cross-modal interactions for token-level credit assignment. With our training data and learning method, our model OmniReasoning-30B-A3B achieves 50.0% on OmniVideoBench and 42.5% on OmniReasoningBench, improving the base model Qwen3-Omni-30B-A3B-Thinking by 12.8 and 9.3 percentage points, respectively. Moreover, it delivers substantial gains on general and long-video benchmarks, including Video-MME-v2. We hope our work offers a solid step for facilitating future research in omni-modal joint reasoning. ---