# Research Papers — 2026-09 · bundle page 218 of 266 > Full-text markdown of papers #10851–10900 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/218 [← Previous bundle](https://aipod.dev/papers/archive/2026-09/page/217.md) [Next bundle →](https://aipod.dev/papers/archive/2026-09/page/219.md) ### DF26: We Cannot Tell Fake From Real Anymore Authors: Severyn Shykula, Andrii Yermakov, Ivan Samarskyi, Dmytro Mishkin, Jan Cech, Anastasiia Mishchuk Published: 2026-09-07 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.07369v1 We introduce DF26, a novel benchmark for detecting AI-generated videos containing fully synthetic clips produced by recent text-to-video and image-to-video models. The videos capture single-person public-speaking scenarios, spanning direct-to-camera recordings, official statements, and studio interviews - 271 real and 2,420 synthetic videos generated by seven modern video models. The study on DF26 shows that human performance in detecting AI-generated videos, as well as state-of-the-art deepfake detectors, is close to random chance. Our results highlight the limitations of current evaluation protocols and motivate the need for benchmarks that explicitly measure robustness to modern generative model distribution shifts. --- ### BlueprintAgent: Constraint-Triggered Targeted Revisits for Simulation-Ready Generation from Scanned Structural Blueprints Authors: Zhouyuan Xu, Chen Yang, Linhao Wang, Jiansheng Fan, Chen Wang Published: 2026-09-07 Categories: cs.CL, cs.AI, cs.CV arXiv: https://arxiv.org/abs/2609.07362v1 Converting in-service reinforced-concrete (RC) building blueprints into simulation-ready models---structured frame representations that support deterministic FEM export and qualified-engineer review---underpins safety assessment and seismic retrofit, but the process remains manual. Direct prompting of a multimodal large language model (MLLM) over a scanned sheet is unreliable: outputs often violate engineering constraints on beam--column support, span count, or 3D continuity. We present BlueprintAgent (BPA), a constraint-triggered multimodal agent for simulation-ready frame extraction from scanned blueprints. BPA treats the MLLM as the primary reader and decision maker, with OCR and computer vision supplying localized evidence. Its central mechanism realizes engineering constraints as callable validators whose entity-level conflict reports trigger targeted MLLM revisits over the local region---an inference-time control distinct from fixed pipelines and free-form self-reflection. We evaluate BPA on 300 real scanned blueprint sheets from 20 anonymized RC frame projects, against five baselines and six ablations. BPA reaches a macro-averaged Beam F1 of 0.994, against 0.301 for single-MLLM zero-shot and 0.820 for a fixed pipeline; removing MLLM-led axis adjudication collapses Beam and Column F1 on complex multi-sheet projects. For dense technical drawings, engineering constraints are best deployed as triggers for entity-level targeted revisits rather than as post-hoc output filters. --- ### Human-like moral judgments conceal divergent motive attributions in large language models Authors: Xiaoyan Wu, Jean-Claude Dreher Published: 2026-09-07 Categories: cs.AI, cs.CY arXiv: https://arxiv.org/abs/2609.07353v1 Large language models (LLMs) are used to simulate human participants in psychological research. We asked whether LLMs that reproduce human evaluations of a whistleblower's moral character also reproduce the motive attributions that accompany them. Five LLMs and two human samples (N = 125 and N = 742) evaluated a physician who either remained silent about fraudulent billing or reported it to a hospital, regulator, or newspaper. Models reproduced the human ranking of the physician's moral character but portrayed whistleblowers as more helpful, less self-interested, and less hostile. In four of five models, competitive motives were less strongly associated with moral-character judgments. Model ratings changed little when prompts reproduced the narratives and demographic profiles of both human samples, although this comparison cannot isolate a perspective effect. Thus, agreement in average ratings can conceal differences in attributed motives, relationships among judgments, and sensitivity to context. Validating LLMs as simulated participants therefore requires testing psychologically informative response patterns, not average agreement alone. --- ### D3ARC: Time-Critical Distributed Disaster Detection for Asynchronous Cooperative Multi-Robot Systems Authors: Nikolaos Koursioumpas, Lina Magoula, Nancy Alonistioti, Ramin Khalili Published: 2026-09-07 Categories: cs.RO, cs.AI, cs.CV, cs.MA arXiv: https://arxiv.org/abs/2609.07350v1 Climate change is increasing the severity and unpredictability of natural disasters. In time-critical crises such as wildfires, traditional monitoring practices remain limited by coverage, cost, and personnel risk, paving the way for autonomous and adaptive monitoring solutions. Within this context, this paper introduces D3ARC, an asynchronous distributed hierarchical framework for time-aware and reliable wildfire detection. D3ARC integrates multiple robotic agents that cooperate under uncertainty through distributed perception, shared situational awareness and coordinated actions. A remote controller asynchronously decides upon each robot's motion, while each robotic agent senses the environment and decides where and how to execute the wildfire detection. All robotic operations require time, and as time progresses, wildfires continue to spread, reducing the opportunity for early intervention. As such, all agents share a common objective: to detect a wildfire with a certain performance threshold as fast as possible and within a time limit. D3ARC integrates mechanisms for safe navigation, coverage efficiency, cooperation and reliability. It introduces a forward-looking capability that allows agents to anticipate the future by evaluating candidate strategies before execution. The framework is evaluated through realistic robotics simulations, ablation studies, and baseline comparisons, achieving an overall mission success up to 94% with 89.4% detection confidence. --- ### Inferring Urban Mobility Interactions from Aggregated Dynamics Authors: Yi Wang, Jing Li, Jinliang Deng, Zhenghong Wang, Yizhi Zhang, Fan Zhang, Ivor W. Tsang, Yu Liu Published: 2026-09-07 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.07349v1 Real-time urban governance depends not only on knowing where people are, but on how they move between places, directional flows that could be conventionally resolved by tracking individuals through space, i.e., expensive to sustain and built on traces that are highly unique and readily re-identifiable. Here we show that this directional structure need not be observed to be known: aggregated counts which cities already collect retain enough information to reconstruct the temporal evolution of origin-destination (OD) matrix. Using an uncertainty-aware physics-informed framework, we infer future OD flows from area-level counts alone across twelve mobility datasets from cities in the United States and China, reaching accuracy comparable to models that take historical OD matrices as input. Probabilistic modeling corrects the systematic underestimation of sparse, high-value corridors and yields calibrated predictions consistent with observed flows. Architectures that respect the generation-before-assignment logic of transport planning recover interactions more faithfully, indicating that location-level spatial heterogeneity should be preserved before pairwise interactions are reconstructed. Because inference requires only aggregated observations after training, recovering interactions this way reduces reliance on continuous individual-level tracking, pointing toward a more deployable and less exposure-heavy basis for real-time urban intelligence. --- ### From Explicit References to Scene Manifolds: Distributional Fidelity and Realism for Radiance Field Quality Assessment Authors: Saeed Mahmoudpour, Gi-Mun Um, Hyon-Gon Choo, Peter Schelkens Published: 2026-09-07 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.07346v1 Radiance field representations such as 3D Gaussian Splatting (3DGS) enable high-quality novel view synthesis but can introduce complex, view-dependent artifacts from reconstruction, rendering, and compression. Reliable perceptual quality assessment (QA) is thus essential for evaluating rendered views and guiding the design of perceptually faithful scene representations. Existing full-reference QA metrics require an aligned reference image, while recent cross-reference metrics relax this requirement by comparing a test view with non-aligned references. However, under wide-baseline radiance field settings, selecting a reliable nearby reference can be difficult, particularly when evaluating views along arbitrary trajectories and poses. We propose SCODA, a lightweight scene-conditioned objective QA method that shifts QA from explicit image-to-image comparison to scene-manifold modeling. High-quality observations of each scene are represented as a multivariate Gaussian distribution in deep feature space, producing a semantic fidelity score that measures deviation from the scene distribution. A weakly-supervised distortion-aware patch discriminator provides a complementary realism signal, and both cues are combined through an unsupervised bounded fusion strategy. Experiments on multiple benchmarks show strong agreement with human judgments and robust generalization across GS- and NeRF-generated views and trajectories. Code is publicly available at https://gitlab.com/saeedmp/scoda. --- ### Monadic Second-Order Logic in HOL: Deep and Shallow with Automated Faithfulness (Extended Preprint) Authors: Christoph Benzmueller, Daniel Kirchner Published: 2026-09-07 Categories: cs.LO, cs.AI, math.LO arXiv: https://arxiv.org/abs/2609.07345v3 In Isabelle/HOL, we apply the deep-and-shallow embedding methodology of our prior work to monadic second-order logic (MSO). Three embeddings are developed side by side: a deep embedding (an inductive datatype with an explicit satisfaction relation); a maximal-shallow embedding that translates the connectives and quantifiers directly into HOL, carrying the interpretation and both assignments explicitly; and a minimal-shallow embedding -- a locale that fixes those parameters, collapsing the formula type to bool. The enabling new ingredient is a two-sorted substitution apparatus -- capture-avoiding substitution, renaming, and a substitution lemma per namespace -- in which each binder is transparent for the other; faithfulness of all three embeddings is mechanised and largely automated. Our central contribution is a fully mechanised two-sorted downward Loewenheim-Skolem theorem: the minimal embedding recovers deep validity relative to the (countable) assignment ranges, and this range-relative reading is shown to coincide with the general (Henkin-style) reading of MSO, whereas the standard reading validates strictly more formulas, witnessed by comprehension. Both readings are nonetheless recovered from the minimal embedding, differing only in the admitted interpretations: all of them for the general reading, only the elementary substructures of the full model for the standard. We exercise the embeddings on classical MSO landmarks: the Boolean-closure and graph schemata hold under the full second-order domain yet fail in the minimal embedding, making the dichotomy concrete, while reachability and 2-colorability are refuted throughout. --- ### Staying on the Attack Path: Structured State for Long-Horizon Automated Penetration Testing Authors: Weizhe Wang, Yitong Zhang, Yao Zhang, Xiaoqiang Di, Zhigang Li, Bin Wu, Guangquan Xu Published: 2026-09-07 Categories: cs.CR, cs.AI arXiv: https://arxiv.org/abs/2609.07344v2 Large language model (LLM) based agents are increasingly applied to cybersecurity tasks such as vulnerability discovery and automated penetration testing. On long-horizon security tasks, however, such agents remain limited by context forgetting and intent drift: early critical facts and causal reasoning chains are lost over extended interactions, and the agent falls into aimless, repetitive exploration. This paper proposes Intentest, an intent-graph-guided automated penetration testing agent that externalizes long-horizon state from the LLM's context window onto a persistent fact-intent directed acyclic graph (DAG), thereby substantially reducing invalid transitions. We evaluate Intentest on automated penetration testing of web applications, a representative long-tail task in cybersecurity. In the DAG, verified network states are stored as immutable fact nodes, and exploration directions are constrained as intent edges bounded by predecessor facts. The system adopts a three-layer architecture, in which the fact-intent mapping layer maintains the global state, the task scheduling and allocation layer ensures execution stability through two-phase degradation recovery and multi-dimensional adaptive load balancing, and the intent retrieval and prediction layer provides tactical priors through a top-down five-stage filtering algorithm. On a benchmark of real CTF challenges covering more than ten vulnerability types across three difficulty levels, Intentest achieves an overall success rate of 88.2% and a success rate of 75.0% on hard tasks, improving over the baseline by approximately 44 and 50 percentage points. Ablation experiments further show that the intent retrieval and prediction reduce the average number of rounds on successful medium and hard tasks by about 33% and 48%, respectively, without changing the set of solvable tasks. --- ### CRISP: Corneal Confocal Microscopy Real-Time Image Stitching Pipeline Authors: Qincheng Qiao, Puli Zhang, Jian Zhou, Xinguo Hou Published: 2026-09-07 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.07336v2 Morphology of the sub-basal nerve plexus (SNP) reflects peripheral nerve health, and corneal confocal microscopy (CCM) provides an important means for in vivo, real-time, non-invasive observation of the SNP. However, mainstream CCM devices offer a limited field of view per frame, whereas the SNP is spatially non-uniform; discrete image sampling is therefore sensitive to sampling location and frame selection, which limits the reproducibility and clinical adoption of CCM as a quantitative assessment tool. Wide-field stitching can reconstruct larger SNP mosaics by integrating sequentially acquired CCM images, but existing methods largely rely on offline post-processing, additional hardware, or specific acquisition protocols, and lack open-source real-time solutions for conventional CCM video streams. This paper presents CRISP (Corneal confocal microscopy Real-time Image Stitching Pipeline), an open-source real-time SNP wide-field stitching framework for conventional CCM examination video streams. CRISP excludes defocused and discontinuous segments via focus-aware gating, propagates poses through local pairwise registration, and maintains non-redundant spatial coverage with a sparse anchor map; when local temporal continuity is interrupted, the system completes relocalization and subgraph merging through global appearance retrieval followed by geometric verification. The framework prioritizes low-latency coverage feedback during examination while outputting accepted frames, poses, and anchor information to initialize offline fine stitching. To our knowledge, CRISP is the first open-source real-time SNP wide-field stitching framework released for conventional CCM video streams. By lowering the barrier to adoption and reproduction of wide-field stitching, CRISP may help move SNP wide-field imaging from a research tool into routine clinical examination workflows. --- ### AAS-RAIL: Improving Information Extraction for Asset Administration Shells through Retrieval-Augmented In-Context Learning Authors: Janek Groß, Jens Heidrich Published: 2026-09-07 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.07334v1 The Asset Administration Shell (AAS) is a cornerstone of Industry 4.0 and the Digital Product Passport, providing standardized digital representations of industrial assets. While manufacturers already maintain extensive technical product documentation, generating AAS instances from existing product datasheets remains a labor-intensive task because technical information is extracted from heterogeneous document structures and often involves company-specific terminology and conventions. In this work, we present AAS-RAIL, a retrieval-augmented information extraction (IE) approach that automatically generates Asset Administration Shells from PDF product datasheets using large language models (LLMs). Instead of relying on a fixed set of few-shot examples, the proposed retrieval-augmented in-context learning (RAIL) approach retrieves LLM-generated extraction helpers from similar Asset Administration Shells to provide instance-specific in-context learning (ICL). This enables the model to adapt its extraction behavior to company-specific naming conventions and formatting styles without fine-tuning. Our core contribution is the dynamic selection of company-specific AAS examples for each datasheet, replacing static prompting with an extraction pipeline that adapts to instances and combines semantic retrieval and structured information extraction. The proposed approach is evaluated on a collection of industrial product datasheets using a selection of open- and closed-weight LLMs. Experimental results show that RAIL consistently improves extraction quality over conventional few-shot prompting, yielding relative improvements of 30.4-52.4%. These results demonstrate that our approach provides an effective improvement for company-specific AAS generation. --- ### PV-WM: A Heterogeneous Micro-Macro World Model for Articulated Pedestrian-Vehicle Co-Rollout Authors: Haozhuang Chi, Jingsong Liang, Ziying Song, Lei Yang, Shihao Li, Haoruo Zhang, Chen Lv Published: 2026-09-07 Categories: cs.RO, cs.AI, cs.MA arXiv: https://arxiv.org/abs/2609.07328v1 Local pedestrian-vehicle forecasting spans heterogeneous physical scales: pedestrians combine root locomotion with articulated motion, whereas vehicles are rigid bodies described by kinematic state and oriented extent. Existing road-agent forecasters typically omit pedestrian articulation, while pose forecasters leave vehicle futures outside the learned rollout. We introduce PV-WM, a history-only world model over structured post-perception tracks. It recurrently advances pedestrian root motion, 15-joint articulation, and learned vehicle states within a synchronized heterogeneous state. The generated pedestrian and vehicle chunks supply the next recurrent boundary; vehicle boxes are reconstructed from predicted center and heading with observed extent, and P-V geometry is recomputed after every transition. Relative to a matched one-shot complete-state predictor, recurrent execution reduces Root ADE by 12.7% and MPJPE by 14.8%. Feedback interventions show that later predictions depend on the content, temporal order, and pedestrian identity of generated articulation. Across 824 aligned Waymo contexts, with 797 providing valid future vehicle support, PV-WM reduces Root ADE by 5.2%, MPJPE by 7.6%, P-V distance error by 11.9%, and oriented-box closest-approach error by 5.8% relative to a validation-selected Modular Specialist. The single-network model uses 57.1% fewer parameters, 96.5% lower average FLOPs per local scene, and 25.5% lower measured p95 latency. PV-WM unifies this heterogeneous future state while preserving type-specific pedestrian and vehicle dynamics. --- ### PLATOS: A Power and Latency-Aware Task-Oriented Scheduling Strategy for Healthcare IoT in Fog Computing Authors: Mohammed Alaa Ala'anzy, Zulfiqar Ahmad, Zhanar Mukash Published: 2026-09-07 Categories: cs.DC, cs.AI arXiv: https://arxiv.org/abs/2609.07318v1 Healthcare Internet of Things (HIoT) technology is revolutionising the healthcare industry by enabling real-time data collection and analysis for personalised patient care. However, the rapid expansion of HIoT technology introduces challenges such as increased latency and higher energy consumption in fog computing environments, particularly when managing battery-operated devices. To address these issues, this work proposes a novel scheduling strategy that optimises both power consumption and latency through task-oriented scheduling for HIoT tasks. The proposed strategy, named PLATOS (Power and Latency Aware Task Oriented Scheduling), is implemented in four sequential phases. In the first phase, HIoT tasks are categorised into three groups: priority-oriented, storage-oriented, and computational-oriented. The second phase focuses on latency optimisation by identifying the fog computing resources that yield the lowest execution delay for each task category. In the third phase, power optimisation is achieved by selecting the resources that minimise energy consumption. Finally, in the decision-making phase, high-performance fog resources are allocated to high-priority tasks while the remaining tasks are scheduled based on a mapped list derived from the latency and power optimisation phases. Simulation experiments conducted in iFogSim2 demonstrate that PLATOS reduces energy consumption by 18.72% and latency by 8.65% when compared to the state-of-the-art. These improvements enhance the efficiency and responsiveness of HIoT systems and contribute to more effective patient care and proactive healthcare service delivery. --- ### DGCPath: Distribution-Aware Generative Contrastive Framework for Self-supervised Path Representation Learning -- Extended Version Authors: Sean Bin Yang, Hao Miao, Zongyi Xu, Jilin Hu, Xiangmeng Wang, Hua Lu, Bin Yang, Christian S. Jensen Published: 2026-09-07 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.07316v2 Due to the proliferation of vehicle trajectory data enabled by advanced sensing technologies, path representation learning has become a pivotal task in intelligent transportation systems. Although existing self-supervised approaches have achieved promising performance, their dependence on deterministic contrastive learning paradigms and handcrafted view augmentation strategies inherently restricts their cross-scenario generalization capabilities. To address these limitations, we present DGCPath, an innovative Distribution-aware Generative Contrastive learning framework for Path representation. This framework establishes a synergistic connection between generative modeling and distributional contrastive learning, enabling the acquisition of robust and transferable feature embeddings. Specifically, our framework incorporates: (1) a diffusion-based view generator that autonomously produces semantically coherent yet diverse trajectory views from Gaussian noise; (2) a variational contrastive mechanism that enforces latent feature alignment at the distribution level, transcending conventional instance-wise consistency; and (3) a novel generative cross-supervision module that reinforces view-level consistency through cross-view reconstruction learning. Comprehensive evaluations on three real-world trajectory datasets demonstrate that DGCPath outperforms state-of-the-art baselines on two distinct downstream tasks, validating its enhanced generalization capability and representation effectiveness. --- ### Content-Based Addressing for Long Context Authors: Mahesh Godavarti Published: 2026-09-07 Categories: cs.CL, cs.LG arXiv: https://arxiv.org/abs/2609.07314v2 Rotary position embedding (RoPE) uses each token's integer position to determine the rotation applied inside attention. This works well for local token order, but increasing context length creates a positional train-test mismatch: RoPE produces relative rotations at offsets not seen during training. Methods that rescale, interpolate, randomize, or bias positions specify how attention handles those offsets, but still derive positional information from a growing token counter. We instead divide a token stream into units, retain ordinary RoPE positions within each unit, and assign every completed unit an address computed from its content. Adding units then applies the same learned map to new content rather than extending a positional range or an identifier table. We prove that this construction preserves local RoPE exactly, leaves the attention comparison between two fixed tokens unchanged when other units are inserted or reordered, and does not create new relative rotations merely because more units are added. In a character-level Tiny Shakespeare diagnostic, all-token validation perplexity remains approximately constant from contexts of 256 to 4096 characters. A second diagnostic shows that content-based addressing can retrieve and use information from multiple serialized facts. These are controlled shallow experiments, not scale benchmarks, but they support a direct prescription: use position to address locally and content to address across units. --- ### Weakly supervised neural network: segmentation of complex structures in X-ray microCT Authors: Daniele Rusconi, Michela Ascolese, Stephanie Fest-Santini, Alberto Bravin, Maurizio Santini Published: 2026-09-07 Categories: cs.AI, cs.CV arXiv: https://arxiv.org/abs/2609.07313v1 Segmentation of complex structures in X-ray tomographic data is a fundamental task in biomedical research, but it often requires large amounts of precisely annotated data, making fully supervised approaches costly and difficult to scale. In this study, weakly supervised deep learning is investigated as a strategy to reduce annotation effort while maintaining accurate segmentation. A two-dimensional convolutional neural network based on the nnU-Net framework was adapted to a weak supervision setting using sparse dot-based annotations, complemented by a limited number of fully segmented images. The approach was evaluated on high-resolution microCT slices of rat kidneys, targeting the segmentation of renal glomeruli, which are small, low-contrast anatomical structures. Results indicate that weak supervision provides a meaningful learning signal, enabling reliable localization of glomeruli even in the absence of dense labels. Incorporating a small set of high-quality annotations substantially improves segmentation performance, approaching that of a fully supervised model. These findings highlight the potential of weakly supervised learning as an annotation-efficient strategy for the analysis of complex structures in X-ray tomographic data, and suggest that alternative loss formulations tailored to sparse annotations may further enhance performance. --- ### Robust Decentralized Personalized Federated Learning via Prediction-Constrained Neighborhood Collaboration Authors: Xiao Ma, Hong Shen, Hui Tian, Wenqi Lyu, Wei Ke Published: 2026-09-07 Categories: cs.LG, cs.CR, cs.DC arXiv: https://arxiv.org/abs/2609.07312v1 This paper proposes a robust decentralized personalized federated learning method R-DPFL, that enables clients to reduce the impact of Byzantine attacks via robust neighborhood direction estimation and history-based update trend prediction, rather than purely aggregating client models as in the existing work. In R-DPFL, each client first computes the current-round model update by aggregating the received neighborhood update vectors. It then predicts what this update should be based on its historical values and local model changes. Finally, R-DPFL computes the difference between these two quantities, adaptively clips this difference, and adds it to the local update. We prove convergence of the learning process through rigorous analysis and show that honest clients maintain stable personalized descent dynamics under Byzantine neighbor perturbations without requiring consensus among neighboring models. Extensive experiments on CIFAR-10 demonstrate that RDPFL consistently outperforms state-of-the-art decentralized and personalized federated learning baselines under heterogeneous and adversarial settings. --- ### LANTERN: Language Model Assessment on Noisy and Transformed Tasks for Understanding Error and Robustness Nuances Authors: Vamsi Krishna Kodavali, Rituraj Singh Published: 2026-09-07 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2609.07309v1 Robustness evaluation of large language models (LLMs) remains a critical challenge, particularly in assessing their sensitivity to perturbations in input data. In this work, we systematically evaluate LLM robustness across multiple dimensions, including word error rate, character repetition and duplication, modifications in choices, and variability in instruction following. To facilitate this evaluation, we construct a synthetic and augmented dataset encompassing a diverse set of LLM benchmarks, specifically targeting multiple-choice question (MCQ) datasets and instruction-following tasks. We conduct extensive experiments on LLMs of varying scales-small, medium, and large-as well as across base and instruction-tuned variants. Our analysis quantifies the variability in model responses under perturbed conditions and highlights discrepancies relative to baseline models. The findings provide insights into the stability of LLMs across different evaluation scenarios contributing to the development of more robust and reliable language models as well as robust evaluation methodologies. --- ### SPARROW: Scalable Taxonomy Induction via Structure-Preserving Partitioning and Constraint-Guided Merging Authors: Yirui Zhang, Yixuan Tang, Yandong Sun, Mong-Li Lee, Anthony Kum Hoe Tung Published: 2026-09-07 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.07307v1 Taxonomy induction aims to organize concept sets into coherent hierarchical structures. Recent LLM-based methods can induce taxonomies directly from flat term lists, avoiding the need for corpora, but degrade sharply as concept sets scale up. We argue that this degradation stems not only from context length limitations, but also from structural failures in hierarchical reasoning. To address this, we adopt a divide-and-merge paradigm that partitions concepts into smaller subsets, induces local taxonomies, and merges them into a global hierarchy. However, we identify two structural failure modes inherent to this paradigm: Structural Fragmentation, where partitioning weakens local hierarchical signals, and Parent Displacement, where locally plausible relations are misplaced in the global hierarchy. To address both, we propose SPARROW, a scalable taxonomy induction framework that combines structure-preserving spectral partitioning to retain hierarchical connectivity within each block, and constraint-guided incremental fusion that treats block-level relations as structural constraints rather than ground truth for global placement. Experiments on large-scale benchmarks show that SPARROW consistently achieves the strongest global structural quality across backbones. The code is available at https://github.com/rebeccazyr/SPARROW. --- ### RouteRelay: Event-Triggered Cross-Layer Route Reuse for Efficient Dynamic Sparse Attention Authors: Bin Li, Sisi Liu, Chenyang Hu, Chaoyang Zhang, Wei Li, Hui Song Published: 2026-09-07 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.07306v1 Dynamic sparse attention reduces long-context prefill cost by routing each query chunk to a small set of key chunks at every Transformer layer. The sparse attention kernel avoids most token interactions, but the router still rebuilds a chunk--chunk score matrix layer after layer, even when the selected routes change little. We introduce RouteRelay, a router-agnostic method that reuses only route metadata across depth while continuing to compute attention with the current layer's queries, keys, and values. Anchor layers perform full routing. Intermediate layers rescore the previous top-$k$ route and a compact sentinel set of near-miss and randomly probed chunks. A query row is rerouted only when a sentinel challenges its weakest selected chunk. We give a top-$k$ stability condition, a probabilistic bound on missed challengers, and a row-selective GPU execution design. In a reproducible empirical evaluation, RouteRelay retains at least 99.99% route recall while rerouting 25.0%, 55.4%, and 78.2% of rows under low, moderate, and high cross-layer drift, respectively. Across routing scales, RouteRelay retains 100.0% recall while evaluating 38.4--51.6% of full-routing score pairs as the key-chunk count grows from 128 to 1024. Its unfused CPU execution remains slower than dense matrix multiplication, exposing row compaction and ledger updates as the main kernel-engineering targets. --- ### Marginal Fidelity Does Not Establish User Simulation in Demographic Synthetic Survey Panels: Response Contracts, Support Collapse and Conditioning Failure Authors: Alexander Doudkin Published: 2026-09-07 Categories: cs.CL, cs.HC arXiv: https://arxiv.org/abs/2609.07305v1 Demographic synthetic survey panels are often validated by matching aggregate answers to published surveys. We test what that certificate establishes across six multiselect batteries from four survey organisations in three countries. The headline analysis is restricted to three instruments whose synthetic cohort and human target share the stated population frame; three other batteries remain sensitivity analyses. The response contract dominates measured fidelity. In the aligned instruments, committed sets leave 66 of 128 model-battery option slots empty in panels of up to 500 respondents, versus 0 of 128 under per-option probability elicitation. Across eight uncapped model-instrument comparisons, probabilities reduce option-marginal MAE by 4.53 to 7.30 points. The capped instrument reverses on two models until the vectors are projected onto its stated maximum. These are measurement effects: human targets are realised check-all responses, whereas the vectors are latent inclusion propensities. Published marginal agreement also fails to discriminate respondent simulation from direct population estimation. On nine aligned model-battery pairs, a no-persona population-prevalence query averages 6.27 MAE versus 12.39 for committed panels and wins all nine comparisons. Constraint-aware probability vectors average 5.34 and beat the query on four of nine, so the baseline challenges the validation criterion rather than proving direct estimation uniformly best. On three unpublished demographic cells, neither approach beats reciting the national distribution. Population-marginal agreement is therefore evidence about an elicitation contract and an estimand obtainable without simulated respondents, not evidence of individual simulation. --- ### Long-Horizon Language Model Reinforcement Learning via Progressive Point Matching Authors: Preston Fu, Kevin Frans, Oleh Rybkin, Sergey Levine, Aviral Kumar Published: 2026-09-07 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.07303v1 Current paradigms for training language models via reinforcement learning rely heavily on sparse outcome rewards. However, as we pursue tasks that require longer and more complicated trajectories, such strategies result in slow learning. Prior work has attempted to address this problem by rewarding partial progress; however, naive formulations are often biased and converge to suboptimal policies. We show that a simple and unbiased dense reward formulation, which we term progressive point matching, scales exponentially more efficiently to long-horizon tasks by rewarding partial progress on a segment level, both theoretically and empirically via synthetic environments. We then show how progressive point matching can be practically instantiated using a single reference trajectory per task. On extremely hard math reasoning problems, sparse outcome rewards cannot make any progress, whereas segment-level rewards enable improvements at larger test-time token budgets when measured by success rate or pass@k. --- ### PCFlow: Physics-Conditioned Flow Matching for GPR B-Scan Image Synthesis Authors: Zhijie Shen, Chenchen Fu, Xuanhao Chang, Hongtao Bai, Lili He Published: 2026-09-07 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.07300v1 Ground-penetrating radar (GPR) B-scan image synthesis is important for data augmentation, algorithm validation, and simulation acceleration, yet generating radargrams with both visual realism and physical consistency remains challenging. Existing learning-based generative models often emphasize visual appearance but provide limited control over response geometry. In this paper, we propose PCFlow, a physics-conditioned flow matching framework for fast GPR B-scan image synthesis. The core of PCFlow is a Maxwell-informed dense physical condition field constructed from the parameterized physical model used for electromagnetic simulation, including material properties, target geometry, propagation cues, and response-domain priors. This condition field provides an interpretable interface between physical scene parameters and radar response geometry, and guides conditional flow matching in the VAE latent space toward physically feasible generation paths. We evaluate PCFlow on a gprMax-based buried-pipeline dataset with both in-distribution and out-of-distribution test cases. Experimental results show that PCFlow generates images with more accurate response geometry and high visual fidelity, demonstrating its effectiveness for controllable and physically faithful radar image synthesis. --- ### World Models Under Asynchronous Sensor Observations Authors: Akash Anand, Abhay Anand, Yash Vishe Published: 2026-09-07 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.07299v1 Learned world models typically assume that observations arrive synchronously, an abstraction inherited from simulators that return a complete state vector at each environment step. Physical sensing instead operates at heterogeneous rates, leaving most observation channels stale at any given instant. Interpolating stale channels introduces measurements that were never observed, while downsampling to the slowest sensor discards valid measurements. A natural alternative is to zero-order-hold the most recent reading and provide the known sampling schedule to the model through two features, staleness and time-to-refresh. We test this prediction using transformer world models across three regimes of increasing causal coupling: open-loop rollouts in continuous-control locomotion, closed-loop model-predictive planning in which each learned model serves as the planner dynamics, and a linear latched-actuator system in which refresh events apply a zero-order-held command to the plant. Our findings show that the effectiveness of time-to-refresh depends on the causal role of the sampling schedule, specifically when refresh events affect the system rather than merely report its state. These results establish when sampling schedules provide useful information for predictive world models operating under asynchronous physical observations. --- ### KODAMA: Multimodal Digital Twin Reconstruction for Urban RF Propagation Modelling Authors: Maximiliano Wardle, A. Ryo Koblitz Published: 2026-09-07 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.07298v2 3D reconstruction typically strives for geometric fidelity or visual plausibility. Radio frequency digital twins (RFDT) are instead judged by whether communication channels behave in them as they do in the real world. RFDTs promise site-specific channel prediction but current practice forces a choice between coarse automated scenes and hand-built, measurement-calibrated models that take weeks to construct per-site. We present KODAMA, an automated pipeline that reconstructs ray tracing-ready RFDTs at city scale from off-the-shelf geospatial data alone: aerial imagery, LiDAR, and photogrammetry yield terrain and watertight building meshes, while exposure-weighted multi-view fusion of street-level imagery recovers façade relief, electromagnetic materials, and clutter---all without site visits or calibration. Across three sites spanning 3.6 to 28 GHz, KODAMA's uncalibrated predictions achieve single-digit RMSE, reducing point-to-point error by up to 5.35 dB over automated baselines and coming within 0.22 dB of a measurement-calibrated, hand-built RFDT. --- ### Probing the Structure and Dynamics of LLM Value Expression through Value Conflicts Authors: Kaicheng Zhang, Jingyi Xiao, Renjun Hu, Xiaoling Liu, Yunshi Lan, Xuan Zhou Published: 2026-09-07 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.07296v1 Ethical evaluation of Large Language Models (LLMs) often characterizes model values as static and monolithic. In contrast, we argue that LLM value expression is better understood as a structured yet dynamic phenomenon. To investigate this, we introduce Conflict-driven Value Probing, a controlled framework that places LLMs in value conflicts and implements four types of interventions that perturb these conflicts to probe LLM value expression. Applying this framework to ten LLMs, we identify three recurring patterns. (1) Expression duality: models shift from broad idealistic orientations in abstract assessment toward more pragmatic priorities in concrete conflicts. (2) Functional steerability: models readily reconfigure their expressed value profiles toward task-defined value objectives. (3) Bounded plasticity: such reconfiguration is not without constraints, i.e. pressure induces a security- and goal-oriented priority shift while negative framing distinguishes protected values from those more amenable to redirection. Together, these findings characterize both the structure and dynamics of LLM value expression: context flexibly reconfigures expressed priorities, yet within behavioral boundaries. This behavioral account provides a foundation for understanding controllability, alignment, and safety in LLMs. Code and data are available at https://github.com/ZeroGen-Lab/CFProbe. --- ### Constitutive State-Space Modeling of Path-Dependent Plasticity: A Resolution-Consistent and Parallelizable Computational Framework Authors: Rui Barreira, Taylan Soydan, Francesco Scipione, Miguel A. Bessa, Dirk Mohr Published: 2026-09-07 Categories: cs.LG, math.NA, physics.comp-ph arXiv: https://arxiv.org/abs/2609.07294v1 Data-driven constitutive models for path-dependent plasticity are commonly formulated using nonlinear recurrent neural networks, whose sequential state evolution limits parallel training and whose predictions may depend on the discretization of the applied strain path. We introduce a Constitutive State Space (CSS) model that reformulates structured state-space dynamics as an incremental constitutive operator. The strain increment is decomposed into magnitude and direction: the loading direction drives the latent state-space system, while the increment magnitude enters the zero-order-hold discretization of its continuous-time linear recurrence. This mechanics-tailored construction guarantees stationarity under zero increments, strongly reduces sensitivity to strain-path resolution, and retains the parallel-scan structure of S5 for efficient training on long constitutive histories. The CSS and Minimal State Cell (MSC) architectures are compared for four multiaxial path-dependent material models including isotropic J2 plasticity, pressure-sensitive foam plasticity, and combined isotropic-kinematic hardening. CSS matches or exceeds the prediction accuracy of the MSC, including one order of magnitude lower validation losses for the plastically incompressible materials. Importantly, CSS maintains low errors across large changes in strain-path discretization, whereas the MSC error increases substantially when evaluated at coarser resolutions than used for training. CSS trains substantially faster and requires fewer strain-stress pairs to attain comparable or better accuracy. Analysis of the learned state further reveals latent structure consistent with the dimensionality of the underlying physical constitutive models. These results establish mechanics-tailored structured state-space dynamics as a computational framework for efficient and discretization-robust data-driven constitutive modeling. --- ### Quality Metrics for LLM-Generated Asset Administration Shells: A Perturbation-Based Evaluation Approach Authors: Janek Groß, Elena Zentgraf, Jens Heidrich Published: 2026-09-07 Categories: cs.SE, cs.AI arXiv: https://arxiv.org/abs/2609.07290v1 The rapid digital transformation of manufacturing, often referred to as Industry 4.0, relies on seamless interoperability between physical and software assets. A central enabler is the Asset Administration Shell (AAS), a standardized digital representation of such assets. Recent advances in large language models (LLMs) enable the generation of AAS submodels from unstructured sources such as product datasheets but raise challenges for quality assurance. In particular, unexpected errors, the lack of ground truth references, and the absence of standardized quality metrics hinder reliable adoption. In this work, we evaluate quality metrics for AI-generated AAS using a perturbation-based evaluation framework. By systematically degrading AAS generation along multiple dimensions, we assess how well different metrics reflect quality changes. Based on a dataset of 200 products from multiple manufacturers, we generate 6,400 AAS instances using GPT-4o-mini, Qwen3, and DeepSeek-R1. Our results show that metrics based on exact matching of property names and similarity-based soft matching of property values, in particular value-based recall and name-based F1 score, provide the most reliable indicators of quality degradation. Furthermore, we quantify the impact of different perturbation types and analyze differences across model families and product segments. These findings support the selection of suitable metrics, the tuning of LLM-based pipelines, and the integration of AI-generated AAS into industrial applications. --- ### TAD: Token-Adaptive Contrastive Decoding with Confidence-Guided Gating for Hallucination Mitigation in Large Audio-Language Models Authors: Heyu Chang, Nianwen Si, Hao Zhang, Wenlin Zhang, Dan Qu Published: 2026-09-07 Categories: cs.SD, cs.CL, cs.CV arXiv: https://arxiv.org/abs/2609.07286v1 Large audio-language models (LALMs) can hallucinate audio objects, answering "yes" to absent sound events, thus undermining reliability in audio question answering. We propose Token-Adaptive Decoding (TAD), a training-free strategy for hallucination mitigation that grounds the initial yes/no decision by contrasting logits under real audio with a matched silent reference. TAD introduces a token-adaptive, confidence-guided gate that is decision-critical at the first decoding step and class-conditional on affirmative tokens, using the audio-silent margin to avoid overcorrection when evidence is weak or already sufficient. Experiments on AudioCaps-Hallucination show that, relative to Audio-Aware Decoding (AAD), a contrastive baseline with fixed contrast strength, TAD improves F1 for Qwen2 by 0.059 to 0.117 across Popular, Adversarial, and Random splits, and for Gemma by 0.025 to 0.064, while on Clotho-AQA it raises F1 from 0.810 to 0.816 on Qwen2 and remains comparable to AAD on Gemma. --- ### Separating Stream Stability from Long-Term Recall in Language Models Authors: Peipei Cao, Xin Zhang, Jie Tang, Xiao Li, Siying Li, Qing Pei Published: 2026-09-07 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.07282v1 Methods for streaming language models are often discussed alongside long-context and memory systems, although they solve different problems. An attention sink can stabilize autoregressive generation over an indefinitely long stream while the model remains unable to use content that has left its recent-token cache. We argue that this distinction should be explicit in system claims and evaluation. We introduce three horizons: the stability horizon, over which predictive behavior remains well behaved; the access horizon, over which past content can still causally affect the output; and the utility horizon, over which a task retains acceptable performance. We show constructively that the stability horizon can be infinite while the access and utility horizons are finite. We then propose ThreeH, an evaluation contract that measures all three horizons under a common state and compute budget. Applying the framework to attention-sink streaming clarifies its strength, constant-memory, stable generation, without treating anchor tokens as semantic memory. The framework exposes roles for cache policies, recurrent state, retrieval, and external memory. Experiments on 128K-token streams, delayed binding recall, and delayed decisions show that attention sinks preserve local modeling but not content beyond the active cache; recurrent and retrieval state extend the semantic horizon. --- ### Matryoshka Hash Representations for Model-Aware Compact Semantic Retrieval Authors: Peichun Hua, Yunming Xiao Published: 2026-09-07 Categories: cs.IR, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2609.07276v1 Retrieval-augmented generation (RAG) depends on dense retrieval: each document is stored as a learned vector, and a query is answered by finding its nearest neighbors in that vector space. Keeping one full-precision vector per document is the dominant index cost at corpus scale, so retrieval systems replace each vector with a short code of a few bytes---a step called quantization. Standard quantizers such as product quantization (PQ) pick the code that reconstructs the original vector most closely. A single code is even more useful if it serves several byte budgets at once: when its short prefixes are each directly searchable, a deployment can set its efficiency--quality operating point without re-encoding the corpus. But training all prefixes under one objective makes the early bits a compromise across budgets---short codes improve while the full-width code degrades. Quantization to low-bit representation, such as binary codes, further sharpens the conflict. We introduce Matryoshka Hash Representations (MHR), a two-stage procedure that separates full-width training from prefix organization. MHR first learns a longer binary code, then freezes the model and trains additional zero-initialized residual code adaptors for directly searchable prefixes. Documents are stored at one bit per coordinate, while queries keep continuous logits like PQ to attain sufficient expressivity. We implement the search process with FAISS FastScan. Trained on MS MARCO and zero-shot transferred to seven BEIR datasets, MHR reaches .5561 NDCG@10 and .6535 Recall@100 at 32 bytes, surpassing the best baseline of the same budget. The advantage is more pronounced in lower budgets. The same code also strengthens two common pipelines: shortlisting candidates for full-precision reranking, and pruning a low-storage graph index such as LEANN. --- ### LightSplat: Real-Time High-Fidelity 3D Gaussian SLAM with Loop Closure Authors: Junze Bao, Ye Gao, Yiming Huang, Xiaolong Yu, Chen Dong, Qing Gao, Wei Wang, Jinhu Lü Published: 2026-09-07 Categories: cs.RO, cs.CV arXiv: https://arxiv.org/abs/2609.07274v1 SLAM systems based on 3D Gaussian Splatting (3DGS) have recently demonstrated promising reconstruction accuracy for dense 3D scene representations. However, current 3DGS systems struggle to meet the strict demands of real-world deployments due to severe limitations in operational performance and map adaptability. To this end, we propose LightSplat, a hybrid-representation RGB-D SLAM framework. It synergizes local sparse features for robust and fast tracking with a dual-thread backend that progressively constructs dense Gaussian submaps. Crucially, we enable online loop closure through feature-accelerated 3DGS registration, refining overall map consistency through pose graph optimization. Ultimately, LightSplat achieves the online reconstruction of high-fidelity Gaussian map. Extensive experiments on multiple datasets and real-world robotic platform demonstrate that our method achieves near state-of-the-art reconstruction quality and the capability to accommodate practical camera motions, maintaining an average framerate of 8 FPS. Overall, LightSplat provides an efficient and robust foundation for deploying high-fidelity 3DGS in real-world environments. --- ### Dense Structural Compression of Transformers via Gauge-Correct Channel Removal Authors: Jed A. Duersch, Naïm Es-Sebbani, Nathanaël Haas, Zied Bouraoui Published: 2026-09-07 Categories: cs.LG, stat.ML arXiv: https://arxiv.org/abs/2609.07264v1 Inference energy per token drives the cost and carbon footprint of deployed transformers. It is dominated by dense matrix products that incur fused multiply-accumulate (FMA) operations and memory traffic. To reduce these computations while retaining dense tensors for high GPU throughput, we develop a methodology from first principles to adapt structural complexity during training to maximize inference utility per unit compute. Channel penalties drive entire tensor slices to zero to enable physical removal while preserving density and the network function. The natural approach, penalizing the norm of operator components acting through each channel, is provably destabilized by gauge freedom. We resolve this pathology with GaugeLasso: additive symmetric group-lasso penalties that recover a monotone function of product-norms when the network converges to gauge balance. Our equilibrium analysis enables per-channel calibration to correctly suppress slices that under-perform in inference utility per unit compute. Under adaptive pressure, the network reorganizes into depth-dependent structural profiles that can be far smaller than the architecture required to learn the task. On polynomial long division over $\mathbb{F}_{31}$, compute compresses from 148 to 255 times with perfect accuracy. On character-level language modeling, compressed models outperform the hand-designed baseline at equal FMA. On masked autoencoding, a compression trial exposes which axes were over-provisioned and which saturated, guiding a better second design. Compaction also accelerates training monotonically as the model progresses. Post-hoc pruning with the same utility ranking cannot reach these structures, showing that sustained pressure is central to discovery of efficient models. Retraining a discovered architecture recovers baseline quality on our statistical tasks, but fails on our exact algorithmic task. --- ### Query-Aware Token Budgeting for Efficient Late-Interaction Visual Document Retrieval Authors: PS Rishi, Rajeev Ranjan Dwivedi, Vinod K Kurmi Published: 2026-09-07 Categories: cs.IR, cs.AI arXiv: https://arxiv.org/abs/2609.07262v1 Late-interaction visual document retrievers preserve fine-grained page evidence by storing many token embeddings per page, but the resulting storage and query-time interaction costs make large-scale deployment expensive. Pooling document tokens before indexing offers a natural remedy, yet static pooling must decide which visual evidence to preserve before the query is known. We study an alternative: a heavily compressed hot-path index generates candidates, after which query-aware token budgeting operates on the original token sets of the shortlisted pages. We formulate this stage-two selection as a budgeted MaxSim coverage problem, show that a clipped version is monotone submodular, and compare coverage-only, cluster-guided, token-wise, and marginal-gain policies. On ten ViDoRe tasks with ColModernVBERT, direct static pooling reduces macro normalized discounted cumulative gain at rank five from 0.6309 without compression to 0.4738 at a thirty-two-fold pool factor. Under the same candidate-generation regime and a pool-factor-eight-equivalent reranking budget, token top-k recovers 93.93 percent of the full-token score, while greedy marginal-gain selection recovers 98.39 percent. Held-out and leave-one-dataset-out evaluations yield positive greedy improvements over token top-k on every dataset. The latency analysis reveals two useful operating points: token top-k for interactive retrieval and the naive greedy implementation as a quality upper envelope. Together, these results show that late-interaction visual retrieval benefits from query-aware allocation rather than query-agnostic pooling alone. --- ### MV-STRIDE: Enabling MLLMs to Master Multi-View Spatial Reasoning via Hierarchical Capability Modeling Authors: Jin Xu, Xiaojian Huang, Zhuodong Luo, Zhihong Zhang, Xin Liu, Jiansheng Wei, Xinzhi Wang, Jie Zhao, Xuejin Chen Published: 2026-09-07 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2609.07258v1 Despite the rapid progress of Multimodal Large Language Models (MLLMs) in 2D vision-language tasks, robust multi-view spatial reasoning remains a fundamental bottleneck due to the lack of structured 3D cognitive pathways in existing datasets. To address this, we introduce MV-STRIDE, a Multi-View hierarchical SpaTial Reasoning dataset with Interdependent and DEcomposed capabilitiEs. Moving beyond flat data structures, MV-STRIDE explicitly models the dependency relationships between foundational perception, scene understanding, and complex contextual reasoning, providing a coherent learning pathway aligned with human spatial cognition. We develop a systematic QA generation pipeline leveraging diverse 3D scene sources that enforces cross-view dependency constraints to prevent single-view solvability, generating multi-level spatial reasoning tasks supported by cognitively grounded chain-of-thought supervision for complex inference. Extensive evaluations demonstrate that our multi-stage training framework based on our hierarchical dataset achieves state-of-the-art performance across multiple spatial reasoning benchmarks, notably the multi-view oriented MMSI-Bench. Our approach enables MLLMs to maintain robust, 3D-consistent spatial reasoning across diverse viewpoints. The code and dataset are available at https://co1dspring.github.io/MV-STRIDE/. --- ### SkillAlign: Aligning Skill Interfaces for LLM-based Agents Authors: Shuo Ren, Xiaomian Kang, Jiajun Zhang Published: 2026-09-07 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.07255v1 Language-model agents increasingly rely on skills: reusable procedural knowledge for reasoning, tool use, and interaction. Existing work studies how skills are acquired, retrieved, compressed, or composed, but often assumes that once a skill is selected, its interface to the agent is fixed. We argue that this overlooks a key source of skill utility: the same skill can help, distract, or mislead depending on how it is exposed. We propose SkillAlign, a provider-agnostic framework that represents candidate skills as multi-view procedural cards and renders them through alternative exposure interfaces, including full instructions, hints, compressed summaries, workflows, or no exposure. This enables counterfactual evaluation where the task, agent, and candidate skills are fixed while only the exposure interface varies. Across ALFWorld and SkillsBench, we show that exposure form substantially affects task success and rendered context cost, and that compact top-k exposure can outperform full-library injection. We further conduct a replay-based policy-learning analysis on ALFWorld, showing that adaptive exposure contains learnable signal but remains far from oracle selection. Our results suggest that skill-augmented agents should optimize not only which skills to use, but also how those skills are presented. --- ### Mathematical Programming in Machine Learning and Artificial Intelligence: A Unified Taxonomy of Models and Applications Authors: Chaosheng Dong Published: 2026-09-07 Categories: math.OC, cs.AI arXiv: https://arxiv.org/abs/2609.07254v1 Mathematical programming provides a common language for many decisions embedded in modern machine-learning (ML) and artificial-intelligence (AI) systems: selecting retrieval context, routing tokens, allocating inference compute, fitting structured predictors, protecting against distribution shift, and balancing competing objectives. However, the relevant literature is fragmented across optimization, information retrieval, recommendation, natural-language processing, computer vision, and learning theory. This paper organizes various applications under common mathematical programming paradigms: linear, quadratic, binary and mixed-integer, conic, bilevel, multi-objective, inverse, distributionally robust, submodular, and min--max optimization. We normalize the models with a mostly unified notation and, for every application, identify inputs, decision variables, a principal formulation, structural properties, solution strategies, and limitations. Across paradigms, we compare tractability, relaxation quality, decomposition, approximation guarantees, and scalability bottlenecks. The paper shows that mathematical programming is most useful not as a claim that all learning is LP or MIP, but as a disciplined interface between predictions and constrained decisions. --- ### Adaptive Conformal Inference Under Delayed Feedback: Coverage Guarantees and a Delay-to-Memory Diagnostic Authors: Lama El Halabi, Adam Brandt Published: 2026-09-07 Categories: stat.ME, math.ST, stat.ML arXiv: https://arxiv.org/abs/2609.07251v1 Adaptive Conformal Inference (ACI) extends conformal prediction to non-exchangeable settings by adjusting the nominal miscoverage level online in response to recent coverage errors. When forecasts are issued with horizon $τ$, however, the outcome needed to evaluate a prediction is observed only $τ$ steps later, so these adaptive updates must rely on delayed feedback. We study this setting and ask how the effect of delay depends on the persistence of the residual process. First, we show that the $τ$-delayed ACI recursion can be decomposed into $τ$ interleaved ACI-like sequences. This representation yields a finite-sample bound on long-run empirical coverage with explicit dependence on $τ$. We also derive an approximate marginal coverage bound that relates coverage deviation to changes in the underlying environment across the forecast horizon and to the adaptation rate $γ$. We then introduce the delay-to-memory ratio $r=τ/L$, where $L$ is the time scale over which the temporal signal driving non-exchangeability decays. Simulation results show that the usefulness of this ratio depends on the form of temporal dependence: it strongly organizes performance under AR(1) dependence, is less predictive of overall performance under GARCH(1,1) and Markov switching, but more clearly characterizes when scale normalization remains useful in those settings. Abrupt mean- and variance-shift experiments further show that the preferred adaptation rate depends on the residual dynamics. Overall, the results show that the effect of forecast delay is best understood relative to the time scale over which past residual information remains relevant. --- ### Elastic Horizon: Discovering the Effective Interaction Frontier in Agentic Reinforcement Learning Authors: Gangyi Zhang, Junjie Meng, Letian Zhang, Wei Wu, Yang Zheng, Dong Wang, Yang Liu, Guanjun Jiang, Chongming Gao Published: 2026-09-07 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.07247v1 Scaling the interaction horizon-the maximum number of environment interactions per episode-improves LLM agents on long-horizon tasks, and curriculum-based methods that progressively expand the horizon outperform fixed-horizon alternatives. However, existing schedules are open-loop: they monotonically increase the horizon until a manually specified maximum, with no mechanism to detect when further expansion stops helping. We propose the effective interaction frontier hypothesis: a dynamic boundary beyond which additional interactions yield diminishing returns while cost grows linearly. We then introduce Elastic Horizon, a closed-loop controller that tracks this boundary via the 90th percentile of successful trajectory lengths. On AppWorld and BFCL, fixed-horizon sweeps reveal clear saturation plateaus; Elastic Horizon stabilizes the horizon inside the saturation band from both under- and over-capacity initializations, attains the best success rates across 7B and 14B backbones, and saves up to 25% of per-step trajectory tokens. Our work shifts the paradigm from how to scale interaction horizons to when to stop scaling. --- ### Towards a Resilience-Theoretic Foundation for Adversarial Robustness in Industrial Control System Anomaly Detection Authors: Branka Stojanović, Andreas Flatscher, Michael Somma Published: 2026-09-07 Categories: cs.CR, cs.AI arXiv: https://arxiv.org/abs/2609.07244v1 Anomaly-based intrusion detection systems in industrial control systems (ICS) and operational technology (OT) environments are increasingly required to meet formal resilience criteria: absorbed adversarial disturbances, graceful degradation under sustained attack, and certified system-level guarantees. Existing resilience frameworks for cyber-physical systems define absorb-recover-adapt trajectories at the architectural level but do not treat machine learning anomaly detectors as first-class components, leaving a gap between component-level robustness evaluation and system-level resilience certification. In this paper, we establish that adversarial robustness in ICS anomaly detection is a specific instantiation of system resilience, and formalise this connection by mapping four resilience constructs, i.e. disturbance class, absorption capacity, recovery trajectory, and degradation function, onto the adversarial machine learning setting. We derive a compositional resilience bound for heterogeneous ICS detection networks, showing that the binding constraint on system-level resilience is the coupling-adjusted absorption capacity of each node along the attack path, not the per-node capacity -- so the binding node need not be the weakest one. Empirical validation on the BATADAL water distribution system benchmark demonstrates that the resulting metrics surface operationally significant phenomena invisible to standard benchmarks: the absorption-degradation divergence under adversarial training, and the paradox that hardening the binding node in isolation reduces system-level resilience. Implications for ICS architecture design and certification standards are discussed. --- ### Kolmogorov--Arnold stability for discontinuous functions Authors: Sviatoslav V. Dzhenzher Published: 2026-09-07 Categories: cs.LG, math.FA arXiv: https://arxiv.org/abs/2609.07240v1 Here we investigate the stability of the Kolmogorov--Arnold representation theorem (KART) under adversarial reparameterisations of the hidden layer for multivariate discontinuous and unbounded functions. Our results provide a rigorous mathematical foundation for the structural robustness of modern deep learning architectures, such as Kolmogorov--Arnold Networks (KANs), under adversarial configurations. --- ### CEDAR: Error-Bounded Residual Routing for Efficient Long-Context Attention Authors: Siyu Li, Dong Wang, Jie Zhou, Wei Li, Yang Xu, Sijie Song Published: 2026-09-07 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.07237v1 Post-hoc sparse attention accelerates long-context prefill by routing each query to a small set of token-level interactions. Hard selection, however, assigns zero probability to every omitted chunk: a routing miss cannot be recovered, and a fixed expansion budget spends the same work on easy and ambiguous queries. We introduce Coarse-to-fine Error-aware Dynamic Attention Routing (CEDAR), a coarse-to-fine method that keeps the language model frozen while preserving global coverage. Each semantic chunk contributes a cheap key--value summary to a residual attention path; chunks with high estimated approximation error are then expanded to exact token attention. Exact and summarized contributions are combined in a single softmax normalization, so refinement replaces, rather than duplicates, coarse evidence. We derive an output-error bound governed by within-chunk key/value dispersion and use it to allocate a variable refinement budget. A controlled clustered-attention study shows that residual summaries reduce reconstruction error by more than 98% relative to hard dropping at equal exact-chunk budgets. Experiments on long-context benchmarks demonstrate that CEDAR recovers most of the quality lost by hard sparse routing while maintaining approximately $3\times$ kernel speedup at 128K context. --- ### Parallelism Strategy Chaining for Fast Training Convergence Authors: Minchul Kang, Changyong Shin, Younghun Go, Hyunho Lee, Jinwoo Jeong, Chuck Yoo, Gyeongsik Yang Published: 2026-09-07 Categories: cs.LG, cs.AI, cs.DC arXiv: https://arxiv.org/abs/2609.07236v1 Selecting a parallelism strategy - the configuration of data, tensor, and pipeline parallelism degrees together with micro- and global-batch sizes - largely determines the training efficiency of large language models. State-of-the-art methods search for a parallelism strategy offline and select the single strategy that minimizes per-iteration time. But we find that they neglect the target validation perplexity and time-to-perplexity (TTP). In particular, our analysis reveals that the best strategy yielding the fastest perplexity improvement changes multiple times during training. As a result, state-of-the-art methods are 1.8-11.4x slower in TTP than the strategy sequence that selects the best strategy at each iteration. This paper proposes CONA, a new training method that introduces online strategy chaining. Instead of a single strategy selected offline, CONA ranks candidate strategies during training using a surrogate metric built from compute throughput and gradient statistics, and switches the current strategy to a new strategy with a higher metric. In our evaluation with GPT-3 1.3B, BERT-Large, and Llama-3.2-1B, CONA reaches the target validation perplexity 1.4-9.6x faster than state-of-the-art methods. Moreover, CONA closely tracks the perplexity achieved by the sequence that selects the best strategy at each iteration, within 2.6%. --- ### Robust Decentralized Federated Distillation via Multi-Modality Knowledge Collaboration Authors: Xiao Ma, Hong Shen, Hui Tian, Wei Ke, Wenqi Lyu Published: 2026-09-07 Categories: cs.LG, cs.DC arXiv: https://arxiv.org/abs/2609.07230v1 This paper propose a robust decentralized federated distillation method that enables clients with heterogeneous models to collaborate through predictions on shared unlabeled public data. In the proposed method, each client first evaluates the received predictions in three modalities of class prediction, boundary decision, and prediction correlation. It then filters unreliable clients, assigns reliability-based weights to the retained clients, and constructs a teacher for each type of knowledge. Finally, the corresponding distillation gradients are validated using a supervised gradient computed from private data. Conflicting prediction and boundary gradients are removed, and conflicting relation gradients are suppressed before the final model update. We prove the convergence of the proposed method by showing stable local optimization for honest clients under Byzantine distillation. Particularly, we show that our method ensures a bounded Byzantine influence on both distillation gradients and individual client private gradients after cross-modality fusion, thereby enabling stable local optimization for honest clienunder Byzantine distillation. Extensive experiments on CIFAR-10 and CIFAR-100 demonstrate that the proposed method improves the prediction accuracy of heterogeneous models of clients under non-IID data and Byzantine attacks. As the booming demands of federated learning in decentralized environments such as edge computing and mission-oriented UAV collaborations, our method has a great potential for adoption of DFL in unreliable real-world scenarios where clients are exposed to receiver-specific Byzantine messages of malicious predictions. --- ### Proximity-CLIP: Text-Guided Semantic Proximity Learning for Zero-Shot Anomaly Detection Authors: Manwen Yang, Leqian Ding, Yu Guo, Fei Wang Published: 2026-09-07 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.07229v1 Vision-language models offer a promising approach for zero-shot anomaly detection (ZSAD). However, due to object-centric bias, normal and anomalous text prototypes exhibit a high semantic overlap. While enforcing strict orthogonality between them improves discriminability, mapping highly contiguous visual inputs onto drastically orthogonal prototypes introduces a geometric dilemma, disrupting the pre-trained structural continuity. To address this problem, we propose Proximity-CLIP, a framework that visually calibrates the semantic margin to guide visual adaptation. First, we introduce a visually-calibrated semantic proximity learning mechanism that uses a bounded dynamic regularization to learn an appropriate semantic margin, ensuring discriminative separation while preserving structural alignment. Second, we design an Anomaly Query Module (AQM) driven by these text priors. Using the calibrated anomalous prototype as a semantic query, the AQM actively retrieves localized defect cues from contextual visual patches, mitigating the dilution of subtle anomalies during global pooling. Extensive experiments demonstrate that Proximity-CLIP outperforms current state-of-the-art methods across multiple ZSAD benchmarks with minimal architectural modifications. --- ### Iterative Audio Separation with Mixture Consistency via MIMO Model Extension Authors: Yukara Ikemiya, WeiHsiang Liao, Yuki Mitsufuji Published: 2026-09-07 Categories: cs.SD, cs.LG arXiv: https://arxiv.org/abs/2609.07226v1 This paper proposes a general framework for stable and effective iterative audio separation with mixture consistency by extending source separation models to a multi-input multi-output (MIMO) configuration. In the field of audio separation, mixture consistency is an essential property for many applications that require accurate phase and timbral information of target sources. While iterative approaches such as diffusion models achieve perceptually superior results in speech enhancement or user-guided target source separation tasks, most existing methods focus on single-step separation with a single-input single-output (SISO) or single-input multi-output (SIMO) configuration through architectural improvements, since mixture-consistent audio separation is generally regarded as a regression problem that admits a unique solution. By extending these architectures to a MIMO configuration, we introduce iterative prediction without compromising the architectural advantages or the characteristics of mixture consistency. We conduct a comprehensive ablation study of combining the framework with discriminators and extending it to a generative model. Experimental results demonstrate significant performance improvements when applying the proposed framework to state-of-the-art separation models. --- ### Distance-Aware Attention and Wall-Distance Expert Routing for Transformer-Based 3D Flow Prediction Authors: Sanghyeon Kim, Sunwoong Yang, Namwoo Kang Published: 2026-09-07 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.07222v1 Transformer surrogates for 3D flow prediction compress an industrial mesh into a small set of tokens from which every prediction point reads. Two operations follow: the retrieval step in which a point gathers information from the compressed representation, and the feed-forward layer that transforms what it retrieved. In current backbones both are blind to where the point sits in the flow. We condition both on wall-related physical signals. Distance-aware cross-attention (DA-CA) reshapes each volume query by its wall distance before retrieval, so that a point deep in the boundary layer draws different geometric information than one in the outer flow. Surface-volume mixture-of-experts (SVMoE) replaces the shared feed-forward layer with a small set of experts, routed by wall distance for volume points and by local geometry for surface points. Neither mechanism is tied to one architecture, so we apply both unchanged to AB-UPT and Transolver-3. On DrivAerML with 50 training cases, DA-CA reduces the volume pressure error by 10.1%, and DA-CA and SVMoE together reduce it by 12.5%; DA-CA improves the near-wall region at some cost in the far region, which SVMoE recovers, and the volume experts settle into near-wall, transition, and free-stream bands without routing supervision. Retrained on 300 cases, the conditioning improves every field quantity, reducing volume pressure and velocity errors by 33.1% and 18.6% on AB-UPT and by 21.4% and 21.3% on Transolver-3. Under Leave-One-Body-Out evaluation on DrivAerNet++, it reduces the volume pressure error on unseen body types by up to 14.2%. --- ### The Price of Consistency: Exploiting Visual Anchors for Multimodal Jailbreaking in Video Generation Authors: Peng Li, Qianqian Xu, Yangbangyan Jiang, Zhipeng Yu, Qingming Huang Published: 2026-09-07 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.07216v1 The rapid evolution of video generation has shifted the paradigm from pure text-driven to multi-conditional controllable generation, with reference images now widely adopted as conditional inputs to achieve superior spatiotemporal consistency. While these reference images serve as powerful visual anchors that significantly enhance controllability, their impact on safety remains largely unexplored. In this work, we reveal the visual anchoring effect: by enforcing consistency, the mechanism prevents the generated content from drifting away from the original harmful intent, thereby eliminating the model's natural safety escape route from harmful to benign content. Consequently, visual anchors inherently increase the safety risk---this is the price of consistency. Building on this insight, we propose Decoupling Intent via Visual Anchors (DIVA), a training-free multimodal jailbreak framework for video generation that exploits this vulnerability. DIVA decouples harmful intent into a static visual anchor image and a dynamic motion text prompt, and employs dual-criteria selection to balance attack stealthiness with semantic preservation. Extensive experiments across various leading commercial platforms and mainstream open-source video generation models demonstrate that DIVA achieves a substantially higher Attack Success Rate than existing text-only methods. To facilitate future research, we additionally contribute TI2VSafetyBench, the first safety benchmark for multi-conditional video generation. --- ### Unraveling the Real Working Mechanism and Inherent Flaws of GAE: A Method for Interpreting Transformer Processes from an Economic Perspective Authors: Yongjin Cui, Xiaohui Fan Published: 2026-09-07 Categories: cs.AI, cs.CV arXiv: https://arxiv.org/abs/2609.07213v1 We observe a phenomenon that current algorithmic research in the field of explainable artificial intelligence primarily pursues better performance on several proxy metrics. On the one hand, these proxy metrics themselves are more or less flawed and cannot properly measure the quality of methods. On the other hand, metric-oriented research approaches often lead to the neglect of the rationality and interpretability of the methods themselves. Explainable artificial intelligence is abbreviated as XAI. The metric-driven research paradigm has resulted in a lack of interpretability of the relevant XAI methods themselves. Accordingly, there is a need for interpretability research on XAI methods, which can be playfully referred to as XXAI. This paper is one of our works on XXAI. This paper takes Generic Attention-model Explainability (GAE), a widely influential model interpretation method , or rather, XAI method that represents an important technical route, as the research object, and explores the real working mechanism and flaws of this method as well as the technical route it represents. Based on the conclusions of this study, it may be necessary to re-examine or verify GAE-related methods and their domain applications. We argue that GAE is an interpretation method that focuses on the attention process. After pointing out the working mechanism and flaws of GAE, we propose Cumulative Asset Holdings (CAH), a more reasonable Transformer interpretation method integrating both process-based and feature-based ideas from an economic zero-sum games perspective. In addition, it is worth noting that our method is applicable to models with special tokens, where existing methods may suffer from limitations. The model simplification research method and the analysis of additive operations adopted in this study may provide inspiration for other research works in XAI. --- ### REFINE: Trajectory Representation Learning via Closed-Loop Transcription -- Extended Version Authors: Sean Bin Yang, Ying Sun, Jilin Hu, Zongyi Xu, Kristian Torp, Hua Lu, Bin Yang, Christian S. Jensen Published: 2026-09-07 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2609.07206v1 Trajectory representation learning underpins a wide range of trajectory analytics tasks; however, most existing self-supervised approaches, whether discriminative or generative, adopt an open-loop paradigm, relying on fixed data augmentations or random masking without feedback, which limits their ability to generalize and scale. We propose REFINE, a simple yet effective Representation lEarning Framework vIa closed-loop traNscription rEfinement for trajectory data. Drawing upon feedback control theory, REFINE tightly couples road-network-aware generative reconstruction with feedback-driven contrastive learning, enabling the model to capture fine-grained local movement semantics and global spatio-temporal dependencies without manually designed augmentation views. We further provide a control-theoretic analysis that establishes convergence guarantees for the proposed closed-loop optimization. Extensive experiments on four real-world datasets demonstrate that REFINE consistently outperforms state-of-the-art methods across multiple downstream tasks while remaining computationally efficient and scalable. This paper is an extended version of REFINE: Trajectory Representation Learning via Closed-Loop Transcription, to appear in KDD 2026. --- ### Agentic Algorithm Engineering: Improving Shared-Memory Exact Minimum Cuts Authors: David A. Bader, Adil Chhabra, Ernestine Großmann, Monika Henzinger, Alexander Noe, Christian Schulz Published: 2026-09-07 Categories: cs.AI, cs.DC, cs.DS arXiv: https://arxiv.org/abs/2609.07204v1 The minimum cut problem for an undirected edge-weighted graph asks us to divide its set of nodes into two blocks while minimizing the weighted sum of the cut edges. Over the last years, we engineered a range of fast algorithms for this problem. Our fastest exact algorithm uses an inexact algorithm to obtain a better bound for the problem, reductions that depend on this bound, improved data structures and parallel contraction routines. It is available in the open-source package VieCut and, on real-world instances, outperformed the previously fastest solvers by a factor of up to 2.5 sequentially and up to 12.9 when run in parallel. We improve this algorithm using agentic algorithm engineering (AAE), a methodology that we introduce here, in which autonomous large language model agents run the algorithm engineering cycle on an existing code base: they form hypotheses about where running time is lost, implement them, benchmark the result on a fixed instance set and keep or discard the change. Even though we had already tuned our algorithm by hand extensively, the agent finds significant optimizations, in particular on the DIMACS core instances: factors of 1.28 (sequential) and 1.63 (32 threads) on real-world k-cores, and 6.26 and 127 on the DIMACS core instances. ---