# Research Papers — 2026-07 · bundle page 32 of 222 > Full-text markdown of papers #1551–1600 of 11075 published in 2026-07, newest first. AI-readable mirrors of the full 11075-paper 2026-07 archive (50 papers per file). Canonical page: https://aipod.dev/papers/archive/2026-07/page/32 [← Previous bundle](https://aipod.dev/papers/archive/2026-07/page/31.md) [Next bundle →](https://aipod.dev/papers/archive/2026-07/page/33.md) ### WildShadowRemover: In-the-Wild Video Shadow Removal via Detail-Preserving Video Diffusion Models Authors: Jiamin Xu, Cong Wang, Zheng Dong, Chi Wang, Renshu Gu, Weiwei Xu, Gang Xu Published: 2026-07-28 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.26203v2 Video shadow removal in the wild remains challenging due to complex illumination, diverse shadow appearances, and limited training data. Despite its importance to numerous vision and graphics applications, it remains largely unexplored in unconstrained real-world scenarios. To address this gap, we present WildShadowRemover, a framework that adapts a pretrained video diffusion model for robust video shadow removal via LoRA fine-tuning. To preserve fine image details while retaining the model's powerful generative prior, we augment the frozen VAE decoder with a detail injection module and introduce a shadow-mask-guided frequency-decomposed modulation module to selectively restore high-frequency textures while suppressing shadow artifacts. Monocular depth priors from Depth Anything 3 further provide geometry-aware guidance under challenging lighting conditions. We also construct WildShadow, a large-scale paired video shadow removal dataset and benchmark, covering diverse synthetic scenes. Extensive experiments demonstrate that our method outperforms existing approaches in shadow removal quality and temporal consistency, producing temporally coherent shadow-free videos with superior visual quality and strong generalization across challenging in-the-wild scenarios. --- ### WildShadowRemover: In-the-Wild Video Shadow Removal via Detail-Preserving Video Diffusion Models Authors: Jiamin Xu, Cong Wang, Zheng Dong, Chi Wang, Renshu Gu, Weiwei Xu, Gang Xu Published: 2026-07-28 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.26203v1 Video shadow removal in the wild remains challenging due to complex illumination, diverse shadow appearances, and limited training data. Despite its importance to numerous vision and graphics applications, it remains largely unexplored in unconstrained real-world scenarios. To address this gap, we present WildShadowRemover, a framework that adapts a pretrained video diffusion model for robust video shadow removal via LoRA fine-tuning. To preserve fine image details while retaining the model's powerful generative prior, we augment the frozen VAE decoder with a detail injection module and introduce a shadow-mask-guided frequency-decomposed modulation module to selectively restore high-frequency textures while suppressing shadow artifacts. Monocular depth priors from Depth Anything 3 further provide geometry-aware guidance under challenging lighting conditions. We also construct WildShadow, a large-scale paired video shadow removal dataset and benchmark, covering diverse synthetic scenes. Extensive experiments demonstrate that our method outperforms existing approaches in shadow removal quality and temporal consistency, producing temporally coherent shadow-free videos with superior visual quality and strong generalization across challenging in-the-wild scenarios. --- ### Randomizing the Number of Centers in k-means++ Authors: Vaclav Rozhon Published: 2026-07-28 Categories: cs.DS, cs.LG, stat.ML arXiv: https://arxiv.org/abs/2607.26202v1 The $k$-means++ algorithm is a standard and widely used seeding method for $k$-means clustering, but for a fixed number $k$ of centers its worst-case expected approximation ratio is $Θ(\log k)$. We consider the same algorithm when an adversary first fixes the dataset and some $K$; the number of centers $k$ is then chosen uniformly from $\{K,\ldots,2K-1\}$. We prove that $k$-means++ is an $O(1)$-approximation with constant probability in this budget-smoothed setup. --- ### (EC)2: Event-Centric Explainability for Cybersecurity Through Multi-Agent LLM Investigations Authors: Neta Kirmayer, David Tayouri, Andrés Murillo, Motoyoshi Sekiya, Asaf Shabtai, Rami Puzis Published: 2026-07-28 Categories: cs.CR, cs.AI arXiv: https://arxiv.org/abs/2607.26201v1 Security operations centers rely on anomaly detection systems to flag suspicious events. Feature-level explanations for anomaly detectors offer limited value for operational investigations. To effectively handle alerts, analysts need to know contextual relationships and need actionable understanding of the entities involved. This paper introduces an event-centric detector-agnostic approach for explaining cybersecurity alerts in small- to medium-sized enterprise networks. We present (EC)2, a multi-agent framework that performs structured, hypothesis-driven investigation to provide explanations grounded in verifiable evidence. Evaluation results show that the proposed framework improves post-detection analysis by generating operationally meaningful explanations, which also enhance event classification accuracy. --- ### Choosing Where and How to Moderate: End-to-End Trade-offs in Filter Placement and Response Rewriting Authors: Mengya Hu, Susie Park, Suzana Ilic, Qiong Wei, Sandeep Atluri, Myra Deng, Tucker Fross, Curt Tigges Published: 2026-07-28 Categories: cs.CL arXiv: https://arxiv.org/abs/2607.26200v1 Content-moderation classifiers are usually evaluated in isolation, but deployment requires choosing where to intervene and what follows a flag. We evaluate these choices using two end-to-end customer-outcome metrics rather than component accuracy: Usefulness, the fraction of turns with a shown, non-harmful, relevant response, and Harmful Exposure, the fraction with a shown harmful response. Latency and error rates are diagnostics. We compare Input only, Response only, and Input + response hard blocking on a human-labelled product benchmark and public ToxicChat evaluation. At the evaluated operating points, Response only achieves the highest filter-only Usefulness in both settings, while Input + response achieves lower Harmful Exposure. Replacing Response only blocking with Response + rewrite recovers most blocked traffic and yields the same observed Harmful Exposure count as Response only blocking for the selected configuration; this equality is not an equivalence result. Probe routing substantially reduces conditional route-and-generation time relative to LLM routing at comparable measured outcomes. A focused output review shows how rewrites balance filter passage with usefulness by generalizing triggering language while retaining benign intent and safe redirection; some sensitive-domain outputs nevertheless omit potentially safety-relevant support information. These results support comparing moderation configurations under deployment-specific safety and latency constraints rather than applying a universal placement rule. Code and public artifacts are available at https://github.com/microsoft/mod-frontier --- ### Rad-JEPA 3D: Radiology Joint-Embedding Predictive Model for 3D Computed Tomography Authors: Quoc-Huy Trinh, Minh-Van Nguyen, Ulas Bagci Published: 2026-07-28 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.26196v1 Self-supervised pretraining is central to 3D medical image analysis, where unlabeled CT volumes are abundant but expert annotations are scarce. Yet existing volumetric encoders often fail to preserve the coarse spatial and geometric structure that downstream reasoning depends on, limiting their performance on organ disentanglement, abnormality detection, and spatial understanding when paired with language models. We introduce Rad-JEPA 3D, a joint-embedding predictive framework that learns volumetric CT representations by predicting the latent features of a complete scan from a masked view. At its core is a hybrid H-Mamba encoder that fuses a Mamba state-space branch, which models inter-slice continuity through sequential scanning, with a grouped-query attention branch, which captures cross-plane spatial context, combined through a lightweight per-token router. To improve the quality of intermediate representations, we further propose Hidden States Orthogonal Regularization (HSOR), which aligns student-teacher hidden states and reduces feature redundancy throughout the encoder. This layer-wise regularization produces more consistent and discriminative volumetric representations, leading to improved performance on organ recognition and spatial reasoning tasks. Pretrained on approximately 120,000 CT scans, Rad-JEPA 3D attains state-of-the-art results despite its compact size: with only 4.0B total parameters, it achieves competitive results with state-of-the-art on closed-ended VQA and the best average spatial-reasoning score on the Spatial-Med benchmark. Ablation studies confirm that the hybrid block and HSOR contribute complementary gains, and that the induced spatial structure can substitute for raw language-model scale on volumetric reasoning tasks. --- ### Dynamic Parameterization Is Not Dynamic Inference Authors: Zongfei Li, Yuan-yih Shang, Guozhong Luo Published: 2026-07-28 Categories: cs.LG arXiv: https://arxiv.org/abs/2607.26192v1 Input-dependent controller coefficients are often treated as evidence of dynamic inference or computational savings. This interpretation conflates three properties: coefficient variation, dependence of a frozen model on how coefficients are assigned to inputs, and conditional execution. We focus on the second property and formulate a general principle of frozen-controller auditing. We provide one concrete implementation, Frozen-Controller Auditing (FCA), which caches the complete coefficient tensor along an unperturbed trajectory, disables the controller, and replays the frozen model with cross-input reassignment, token shuffling, and static profiles estimated from an independent calibration set. Because the coefficients are cached before any intervention, performance changes under replay measure assignment dependence without feedback from recomputing the controller on perturbed hidden states. Across seven independently trained 76M FeatureGate Transformers and three 504M models, static layerwise profiles retain 98.70% and 99.43% of the Correct-to-GlobalMean performance gap, respectively. Layer identity explains 87% to 96% of the coefficient variance. FeatureGate nevertheless executes every Transformer block, and its measured inference is 30.8% slower than Dense. On the public MUDDPythia-1.4B checkpoint, cross-input reassignment and token shuffling increase NLL by 1.9067 and 2.9637, respectively. These penalties show that the model depends strongly on content-conditioned cross-layer assignment. MUDDPythia also executes every Transformer block. The results show that dynamic parameterization alone does not establish dynamic inference and that functional dynamics do not establish computational savings. Claims about dynamic models should separately report coefficient variation, functional dependence of the frozen model, and actual execution. --- ### Position: Evaluation Scores Are Perishable Knowledge Claims Authors: Sankalp Gilda, Shlok Gilda Published: 2026-07-28 Categories: cs.AI, cs.CL, cs.LG, cs.SE arXiv: https://arxiv.org/abs/2607.26191v1 Evaluation methodologies for language models increasingly combine multiple signals, from automated metrics and LLM-as-judge ratings to human assessments and benchmark suite results. When these signals are aggregated via averaging, evaluation confidence can then substantially exceed the reliability of the weakest signal: a phenomenon we call trust inflation in evaluation. We argue that evaluation scores should be treated as epistemic claims with three properties: formality (human evaluation provides stronger evidence than an automated metric), scope (a benchmark result applies to the tested distribution, not universally), and validity windows (benchmark results expire as contamination accumulates and distributions shift). Several converging research traditions (chain-of-thought analysis, possibilistic logic, and algebraic theory) establish weakest-link aggregation as the conservative endpoint of a parameterized operator family controlled by a single pessimism parameter. Drawing on those traditions, and on concrete lessons from building an evaluation harness for agentic AI, we propose that evaluation results carry explicit metadata (formality tier, scope declaration, and expiration date) to make their epistemic status transparent. We illustrate the cost of mean aggregation on the public HELM leaderboard: across 54 frontier models on ten scenarios, the top-five models ranked by mean score and by weakest-link are completely disjoint. --- ### GoGoTB: Agentic RTL Verification with Specification-Grounded Coverage Closure Authors: Xin Xin, Jincheng Lou, Junhui Li, Jinglin Yan, Panda Xiao, Di Wu, Haixiao Li, Weicong Lu, Weijian Fan, Xinyu Qu, Yuxiang Zhao, Min Yu et al. Published: 2026-07-28 Categories: cs.AI arXiv: https://arxiv.org/abs/2607.26181v1 Functional verification dominates integrated circuit (IC) front-end engineering effort, and a single missed bug that escapes to silicon can trigger a costly respin. Recent large language models (LLMs) offer new opportunities to automate this process, yet existing LLM-based approaches generate each component through independent single-turn calls with no shared context, leaving interface mismatches undetected and reported coverage disconnected from specification requirements. To address these challenges, we present GoGoTB, an agentic framework that achieves end-to-end verification closure through three subsystems: an agentic execution control layer, an evolvable knowledge system, and specification-grounded coverage closure. The execution control layer separates deterministic enforcement from LLM reasoning at every tool and stage boundary. The knowledge system dispatches methodology and design-specific expertise on demand. The coverage framework anchors every bin to a named specification behavior so that each residual gap has a diagnosable root cause and a targeted remedy. Tested on 8 register transfer level (RTL) designs without any human intervention, GoGoTB achieves 100\% environment generation success and averages 98.4\% line, 97.2\% branch, 97.0\% toggle, and 83.2\% functional coverage. No prior work successfully generates a complete verification environment or achieves meaningful coverage on the same benchmarks. --- ### Cognitive Convergence: Deep Similarities Between Large Language Models and Human Cognition Authors: Chandra Sripada, Richard Lewis Published: 2026-07-28 Categories: q-bio.NC, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2607.26179v1 LLMs are widely regarded as alien intelligences, systems whose cognitive operations are fundamentally unlike our own. Apparent similarities to human cognition are therefore often seen as the result of anthropomorphic projection. We argue that this framing is mistaken. LLMs clearly differ from humans in important respects, including their physical substrate, learning history, and the environments with which they interact. These differences make it all the more striking that contemporary LLM-based systems converge with human cognition on a number of principles of cognitive organization with longstanding support in cognitive science. We identify structural correspondences across five dimensions: inferential organization, computational architecture, representational structure, prediction-driven learning, and reinforcement-learning-like mechanisms supporting goal-directed action. These correspondences support a broader model of intelligent cognition in which core principles long used to explain human intelligence also characterize contemporary LLM-based systems. --- ### DuplexGen: Adaptive Synthesis of Human-AI Turn-Taking Dialogues Authors: Takyoung Kim, Kang-wook Kim, Sang Hoon Woo, Julia Hirschberg, Gunhee Kim, Dilek Hakkani-Tür Published: 2026-07-28 Categories: cs.CL arXiv: https://arxiv.org/abs/2607.26178v2 Turn-taking is a central component of full-duplex interaction. Which turn-taking behaviors are appropriate varies with the scenario, yet current models apply a single norm regardless of context. This limitation originates in their training data: human-human speech corpora capture natural timing phenomena but provide little role grounding or scenario-specific norms, while heuristic or prompted synthesis methods inject turn-taking behaviors without basing them on human preferences. We introduce DuplexGen, a framework for generating dialogues with scenario-adaptive turn-taking by calibrating LLM predictions against a small set of slot-level human preference annotations. In six cooperative and competitive tasks, human turn-taking preferences differ systematically, and DuplexGen aligns substantially more closely with those preferences than uncalibrated prompting or training solely on generic human-human data; a full-duplex model trained on DuplexGen-generated data exhibits distinctive, human-preferred turn-taking behaviors. These results show that human calibration, not corpus scale or prompt design alone, is what allows turn-taking synthesis to be scenario-specific. --- ### DuplexGen: Adaptive Synthesis of Human-AI Turn-Taking Dialogues Authors: Takyoung Kim, Kang-wook Kim, Sang Hoon Woo, Julia Hirschberg, Gunhee Kim, Dilek Hakkani-Tür Published: 2026-07-28 Categories: cs.CL arXiv: https://arxiv.org/abs/2607.26178v1 Turn-taking is a central component of full-duplex interaction. Which turn-taking behaviors are appropriate varies with the scenario, yet current models apply a single norm regardless of context. This limitation originates in their training data: human-human speech corpora capture natural timing phenomena but provide little role grounding or scenario-specific norms, while heuristic or prompted synthesis methods inject turn-taking behaviors without basing them on human preferences. We introduce DuplexGen, a framework for generating dialogues with scenario-adaptive turn-taking by calibrating LLM predictions against a small set of slot-level human preference annotations. In six cooperative and competitive tasks, human turn-taking preferences differ systematically, and DuplexGen aligns substantially more closely with those preferences than uncalibrated prompting or training solely on generic human-human data; a full-duplex model trained on DuplexGen-generated data exhibits distinctive, human-preferred turn-taking behaviors. These results show that human calibration, not corpus scale or prompt design alone, is what allows turn-taking synthesis to be scenario-specific. --- ### Shared SFT Lessons Across Alignment, Model Organisms, and Toy Models Authors: Anton de la Fuente, Arthur Conmy Published: 2026-07-28 Categories: cs.LG arXiv: https://arxiv.org/abs/2607.26173v1 Alignment training, model organisms, and toy models are usually treated as separate research areas. But projects in all three frequently use supervised fine-tuning (SFT) to pursue the same underlying goals. When projects share a goal, we should test whether lessons learned from one area transfer to the other areas. We study three such transfers, each taking a lesson developed in one SFT setting and testing it in another. First, we port a lesson about behavior generalization from alignment training into toy models. Training on the reason for a behavior, as in Teaching Claude Why, can make the behavior generalize better than training on examples of the behavior alone. Second, we port a lesson about capability preservation from model organisms into the Model-Spec Midtraining alignment setting. SFT on outputs written by a model other than the student (off-model outputs) can damage capabilities when trained on. Mixing in benign on-model (and on-policy) data into our training can prevent most of this damage while still embedding the target behavior. Third, we port a lesson about robustness from model organisms into the same alignment setting. We find that follow-up benign SFT can erase the alignment behavior while preserving capabilities, showing that capability preservation alone does not ensure robustness to subsequent training. Our work illustrates how porting SFT lessons between different research fields can uplift them all, suggesting more researchers should borrow techniques from outside their own areas. --- ### A Picture Says Thousands of Words - Harnessing Dermal Exposure Data from Images through Hybrid Deep Learning for Enhanced Safety Assessment Authors: Hua Qian, Manisha Kotha, Tuan Tran, Jennifer Shin, Haining Zheng Published: 2026-07-28 Categories: cs.CV, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2607.26170v1 This study developed a hybrid computer vision method to quantify exposed skin from images for dermal exposure assessment. Using 170 indoor-painting images, Mask R-CNN first identified human subjects and removed background interference; a color-based algorithm then segmented exposed skin. The resulting exposed-skin-to-body pixel ratios showed approximately 80% agreement with human estimates. The approach demonstrates a scalable way to extract semi-quantitative exposure information from images, with future extensions to body-part recognition, PPE detection, and video-based exposure analysis. --- ### DVPSFormer: Efficient Online Depth-aware Video Panoptic Segmentation for Autonomous Driving Authors: Yung-Hsu Yang, Luigi Piccinelli, Siyuan Li, Mattia Segu, Lei Ke, Martin Danelljan, Yuqian Fu, Zuria Bauer, Fisher Yu, Hermann Blum, Marc Pollefeys Published: 2026-07-28 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.26165v1 Safe autonomous navigation requires a holistic understanding of dynamic environments, necessitating the simultaneous estimation of metric depth, semantic segmentation, and instance trajectories. While depth-aware video panoptic segmentation (DVPS) unifies these tasks, existing approaches often rely on computationally expensive, multi-stage pipelines or offline tracking, rendering them unsuitable for real-time decision-making. To address this, we propose DVPSFormer, a unified online architecture designed for efficient 4D scene understanding. Central to our approach is explicit scene discretization (ESD), a novel mechanism that leverages segmentation queries to represent foreground and background regions, enabling a discrete-to-continuous (D2C) depth head to decode metric depth in a single pass. This tightly couples semantic and geometric learning while significantly reducing latency. Furthermore, we propose an online majority voting (OMV) mechanism that exploits temporal consistency to refine classification during instance tracking. DVPSFormer establishes a new state-of-the-art on the Cityscapes-DVPS and SemKITTI-DVPS benchmarks, offering a streamlined solution for online robotic perception. Code and models are available at https://royyang0714.github.io/DVPSFormer. --- ### Data Fusion and Contrastive Alignment for Unconstrained IR Molecular Structure Elucidation Authors: Ethan J. Mick, Campbell A. Sweet, Matthias J. Young, Derek T. Anderson Published: 2026-07-28 Categories: cs.LG arXiv: https://arxiv.org/abs/2607.26164v1 Automated molecular structure elucidation from infrared (IR) spectroscopy data has seen significant advancements in recent years, but its broad applicability is limited by a reliance on pre-determined chemical formulas provided as auxiliary model inputs. This limits model predictions to isomer identification rather than full molecular structure prediction. Although transformer models have been shown to identify molecular isomers with high accuracy, their reliability for unconstrained structure elucidation is comparatively low and poorly understood. In this work, we propose and evaluate key modifications to the traditional encoder-decoder transformer. To better address the vast chemical space of the unconstrained problem, we implement a novel Mixture-of-Experts (MoE) decoder module that utilizes non-additive aggregation via linear-order statistics and the Choquet integral. We further modify the transformer to utilize these non-additive operators when aggregating spectral representations as well. Together with an auxiliary contrastive alignment loss term, these enhancements improve Top-K prediction accuracy by over 10 percentage points compared to baseline IR-only models. Through sub-structure fragment analysis of molecular predictions, we further confirm that infrared spectra encode the vast majority of relevant chemical information, implying that the higher performance of isomer-ranking models is largely due to underrepresented or overlapping absorption bands for molecules in the explored chemical space. Ultimately, by demonstrating the efficacy of automated molecular structure elucidation from measured IR spectra, this work serves to significantly broaden the utility of AI in analytical chemistry. --- ### GuideSkill: Evolving Executable LLM Agent Skills for Guideline-Grounded Clinical Reasoning Authors: Lang Cao, Yuhao Shen, Tianyang Luo, Simo Du, Hao Peng, Yue Guo Published: 2026-07-28 Categories: cs.AI arXiv: https://arxiv.org/abs/2607.26160v1 Clinical practice guidelines (CPGs) encode diagnostic criteria, but LLM systems typically retrieve guideline text or absorb it through training rather than execute its rules. We introduce GuideSkill, an external reasoning layer that compiles disease-specific criteria into executable functions returning ordinal diagnostic-support scores. GuideSkill-Zero is initialized from guidelines, while GuideSkill-Evo uses case--diagnosis pairs to refine covered skills and add missing diagnoses. At inference, an LLM proposes a differential diagnosis, grounds the features required by each matched skill, and fuses its ranking with the executed skill scores. Across four benchmarks and four backbones, GuideSkill-Zero improves macro-average accuracy over guideline RAG by 13.45% on average. GuideSkill-Evo achieves the highest macro-average for every backbone, improves over direct inference by 18.49% relatively, and increases gold-label skill coverage from 56.5% to 99.5%. On Qwen3.5-9B, it also exceeds the strongest parameter-update baseline by 11.16% without updating the backbone. Expert evaluation further indicates that GuideSkill produces clinically sound and broadly acceptable skills, suggesting that its initialized and evolved rules are reliable and practically meaningful. These results support executable skills as a model-agnostic mechanism for combining guideline-derived procedures with case-derived diagnostic patterns. --- ### When benchmark inferences do not compose: Projectibility in AI evaluation Authors: Brett Reynolds Published: 2026-07-28 Categories: cs.AI, cs.CY, cs.LG arXiv: https://arxiv.org/abs/2607.26159v2 An AI benchmark result rarely reaches a consequential claim in one step. Evaluators generalize it to further cases, interpret it as evidence of capability, extrapolate it to new tasks, transport it to another system or site, and combine it with assumptions about human review and downstream consequences. Validity-centred approaches require evidence for each claim. This paper makes explicit and operationalizes a problem those approaches leave to the analyst: warranted links don't automatically make a warranted chain. The target of one study may not be the source of the next; system, population, outcome, or conditions may change at the interface; and shared data or model lineage may make apparently independent support dependent. Projectibility concerns whether a bounded extension from observed to unobserved cases is warranted. Goodman supplies the problem of rival extensions; argument-based validity supplies an architecture for testing them. The contribution is an interface audit for distributed AI evidence: typed source and target descriptions, and a procedure separating endpoints that never meet from endpoints that meet while warrant fails to cross. A legal-research case shows how benchmark evidence and a deployment study can each be sound while remaining parallel. A known-truth demonstration shows why aggregate stability can erase distinctions a later projection requires. The resulting projectibility audit diagnoses unsupported joins in benchmark-to-use arguments. --- ### When benchmark inferences do not compose: Projectibility in AI evaluation Authors: Brett Reynolds Published: 2026-07-28 Categories: cs.AI, cs.CY, cs.LG arXiv: https://arxiv.org/abs/2607.26159v1 An AI benchmark result rarely reaches a consequential claim in one step. Evaluators generalize it to further cases, interpret it as evidence of capability, extrapolate it to new tasks, transport it to another system or site, and combine it with assumptions about human review and downstream consequences. Validity-centred approaches require evidence for each claim. This paper identifies a further epistemic problem: warranted links don't automatically make a warranted chain. The target of one study may not be the source of the next; system, population, outcome, or conditions may change at the interface; and shared data or model lineage may make apparently independent support dependent. Projectibility concerns whether a bounded extension from observed to unobserved cases is warranted. Goodman supplies the problem of rival extensions; argument-based validity supplies an architecture for testing them. The paper's distinctive claim is a non-composition principle: support for adjacent projections warrants their composition only when endpoints and assumptions align and dependence and uncertainty are carried through. A legal-research case shows how benchmark evidence and a deployment study can each be sound while remaining parallel. A reanalysis and simulation show why aggregate stability can erase distinctions a later projection requires. The resulting projectibility audit diagnoses unsupported joins in benchmark-to-use arguments. --- ### ClinLens: Towards Long-Horizon LLM Agents for Longitudinal Multimodal Clinical Data Science Authors: Yuan Zhu, Ethan B. Liu, Frank Nie, Wei Fan, Haibo Pu, Jindong Han Published: 2026-07-28 Categories: cs.AI arXiv: https://arxiv.org/abs/2607.26155v2 Automating clinical data science requires agents to translate patient-centric objectives into executable workflows grounded in longitudinal multimodal evidence. Existing benchmarks provide limited coverage of two essential capabilities: grounding analyses in the correct patient episode and harmonizing multimodal observations. We introduce ClinLens, a patient-centric benchmark for evaluating large language model (LLM) agents on long-horizon clinical data science workflows. ClinLens comprises 126 executable tasks spanning five data modalities, four analytical scopes, and five analytical capabilities. The tasks assess agents' ability to maintain a coherent patient-specific analytical context across interdependent stages of evidence grounding, multimodal integration, and analytical pipeline generation and execution. We evaluate general-purpose, coding, and biomedical agents, with task runs averaging 36.93 steps and 8.06 minutes. The strongest evaluated agent achieves an execution success rate of 86.4% and a final-answer accuracy of 54.0%, but a strict pass rate of only 29.4%. Failure analysis identifies patient-episode grounding as the most frequently recorded failure category under the evaluation procedure. By jointly evaluating analytical outcomes and the validity of the supporting clinical context, ClinLens provides a testbed for advancing reliable agents for longitudinal clinical data science. --- ### ClinLens: Towards Long-Horizon Coding Agents for Longitudinal Multimodal Clinical Data Science Authors: Yuan Zhu, Ethan B. Liu, Frank Nie, Jindong Han Published: 2026-07-28 Categories: cs.AI arXiv: https://arxiv.org/abs/2607.26155v1 Clinical data-science agents must transform heterogeneous longitudinal records into auditable analyses, yet existing benchmarks largely isolate medical question answering, structured-table reasoning, or generic scientific repositories. We introduce CLINLENS, a benchmark of 200 executable tasks over five linked MIMIC resources spanning structured electronic health records, notes, electrocardiograms, chest radiographs, and echocardiograms. A 4 x 5 taxonomy crosses four patient-time scopes with five analysis capabilities. Program-first reverse synthesis pairs each bounded semi-raw package with an evaluator-private reference workflow and checks required artifacts, cohort and temporal semantics, and the final answer. On a fixed 126-task suite, the strongest of 24 standardized model-scaffold configurations achieves 56.3% scope-macro STRICTPASS despite 100% EXECSUCCESS. For reference, a separately configured coding agent solves 83 of 126 tasks, while five biomedical systems adapted to GPT-4o-mini reach at most 2.9% scope-macro STRICTPASS. These results expose a substantial gap between runnable submissions and correct clinical analyses. --- ### Towards Trustworthy Embodied Intelligence: A Systems Framework and Graded Trustworthiness Levels Authors: Xinyu Yang, Tianxing Chen, Honghao Su, Minxuan Wang, Chenze Yu, Zhangzheng Tu, Yue Chen, Yuxiao Huo, Lingfeng Zhang, Yan Huang, Yan Qin, Shaolong Zhu et al. Published: 2026-07-28 Categories: cs.RO, cs.AI, cs.CY arXiv: https://arxiv.org/abs/2607.26121v1 Embodied intelligence integrates learned perception and decision making with real-time computation, control, and physical interaction. Because failures can cause immediate physical or operational harm, task completion alone does not establish trustworthiness. We define trustworthy embodied intelligence as the sustained capacity to execute specified tasks reliably under environmental and system variation while maintaining risk within acceptable bounds. We term this objective sustained safe success. Its supporting mechanisms are organized into four interdependent layers. The model layer generates task-competent action proposals with calibrated uncertainty and explicit safety preferences. The system layer realizes authorized actions dependably through integrated sensing, computation, control, hardware safeguards, fault containment, and fallback. The evidence layer substantiates bounded claims through evaluation, verification, validation, traceability, and structured assurance arguments. The deployment layer maintains claim validity through runtime monitoring, authority management, intervention, incident response, and controlled updates. Because assumptions and failures propagate across these layers, neither model capability, isolated safeguards, nor benchmark performance alone can establish end-to-end trustworthiness. Drawing on embodied AI, robotics, control, dependable computing, distributed systems, and autonomous driving, we further propose a non-normative hierarchy of trustworthiness levels. This hierarchy grades the strength of bounded deployment claims across task capability, safety, system assurance, operational governance, and supporting evidence, providing a basis for bounded deployment, comparative evaluation, research prioritization, and future standardization. --- ### Even More Deception: Objective Misalignment in Mixed-Motive LLM Multi-Agent Systems Authors: Marylou Fauchard, Florian Carichon, Margarida Carvalho, Golnoosh Farnadi Published: 2026-07-28 Categories: cs.AI arXiv: https://arxiv.org/abs/2607.26120v1 Large Language Models (LLMs)-powered multi-agent systems are increasingly deployed in mixed-motive environments, where agents operate under asymmetric information and strategic deception due to conflicting or hidden objectives. In these settings, misalignment with collective goals becomes a central concern. We propose a novel framework for evaluating objective misalignment using the social deduction game Werewolf, modifying the objective of a single agent while preserving its assigned role. Across LLMs from four different model families and sizes, four player roles, and three objective formulations, we introduce a dual analysis of the agents' internal reasoning and their public cheap-talk behavior (i.e costless, non-binding communication that does not directly affect the agents' utilities), complemented by an analysis of game outcomes. Our results show that objective misalignment undermines outcomes in inherently adversarial environments, an effect exacerbated by asymmetric information and specialized roles. While compromised agents consistently develop distinct objective-dependent reasoning strategies, these adaptations remain largely invisible in their public behavior. More broadly, our findings suggest that even subtle objective misalignment can profoundly affect collective decision-making, highlighting the need for effective mitigation strategies for LLM-based multi-agent systems. --- ### Probing the Origins of Reasoning Performance: Representational Quality for Mathematical Problem-Solving in RL vs. SFT Fine-Tuned Models Authors: Antyabha Rahman, Akshaj Gurugubelli, Omar Ankit, Kevin Zhu, Aishwarya Balwani Published: 2026-07-28 Categories: cs.AI, cs.CL arXiv: https://arxiv.org/abs/2607.26119v1 Large reasoning models trained via reinforcement learning (RL) have been increasingly shown to outperform their supervised fine-tuned (SFT) counterparts on mathematical reasoning tasks; Yet the mechanistic basis for this advantage remains unclear. We therefore ask, what internal representational differences enable RL models' superior performance? Our work presents two converging lines of evidence: First, linear probes trained on layer-wise hidden states reveal that RL models tend to achieve higher accuracy in predicting answer correctness compared to SFT models, indicating more linearly separable and structured representations. Second, mean ablation studies show that RL models develop a hierarchical architecture where deeper layers become progressively more critical, whereas SFT models distribute importance uniformly across layers. Together, these findings demonstrate that RL training fundamentally restructures how models represent and process reasoning problems. Finally, we analyze token-count variability under repeated sampling across problems to assess adaptive compute allocation. While we observe higher variability in some RL-tuned models than in their SFT counterparts, we see strong consistency in others, suggesting that token allocation may depend more on the overall training pipeline than on RL versus SFT alone. We believe this token-allocation variability reveals the spread of plausible on-policy reasoning, highlighting which models exhibit stable policies versus those that are under-determined, potentially non-identifiable solution behaviour. --- ### Try Again, Don't Look Back: Blind Resampling Outperforms Self-Repair in Small Code Models Authors: Yuvraj Verma Published: 2026-07-28 Categories: cs.SE, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2607.26117v1 Self-repair - returning a failed program to the model together with its test output and asking for a correction - is a standard component of code agents, and is almost always evaluated against a baseline that does not retry at all. We argue that this comparison confounds the value of the feedback with the value of the extra attempt. Using a placebo-controlled design on MBPP+ at three model scales (1.5B, 3B, 7B), we compare four matched-budget retry conditions: blind resampling, a content-free failure notice, genuine execution feedback, and feedback augmented with verbal self-reflection. Blind resampling is the strongest condition below 7B, and remains statistically tied with the best condition at 7B, while consuming 2.5-5.5x fewer tokens; conditioning on the model's own failed attempt costs 6.1 points at 1.5B (p=0.006), and the informational content of execution feedback adds nothing measurable over the placebo. We attribute this to anchoring: when shown its previous attempt, a model reproduces a near-identical program in 33-68% of retries, against 2-14% under blind resampling. Two further experiments delimit the effect. Retrieved solutions to other tasks change nothing (bounded to +/-3.5 points), which localizes the harm to self-conditioning rather than context length; and reflection, the only condition that measurably weakens the anchor, remains dominated on cost. Replication rules out two competing explanations: the penalty is unchanged at full precision, and it reproduces on an independent model family. Across six configurations spanning two families and two precisions, its magnitude is predicted by baseline quality alone (r=0.96) - the cost of anchoring is the cost of committing to a bad first attempt. --- ### GPT-Red: Automated Red Teaming via Self-Play at Scale Authors: Eric Wallace, Christopher A. Choquette-Choo, Nikhil Kandpal, Sam Toyer, Dylan Hunn, Stephanie Lin, Yuxin Wen, Xiangyu Qi, Christopher Wolff, Zizhao Wang, Milad Nasr, Sicheng Zhu et al. Published: 2026-07-28 Categories: cs.CR, cs.AI, cs.CL, cs.LG arXiv: https://arxiv.org/abs/2607.26115v1 We introduce \textbf{GPT-Red}, an automated red-teaming agent that is trained to discover novel prompt injection attacks against frontier LLMs. The goal of this model is to evaluate and improve the robustness of our production systems. To this end, we use it to adversarially train GPT-5.6, our most robust model to prompt injections to date. To create GPT-Red, we design a scalable self-play algorithm where the model is tasked with attacking a diverse population of simultaneously-trained defender agents. We train the model on realistic red-teaming environments using compute on the same scale as some of our largest RL post-training runs, making it the single-largest LLM safety training run ever documented. GPT-Red excels at red-teaming: it reliably breaks our past models up to GPT-5.5, it finds more successful attacks than human red-teamers, and it generalizes to held-out environments, defender models, and harnesses. In the future, we expect that as we improve the robustness of each new GPT model, it will in turn will provide better learning signal for \textit{even stronger} red-teamer agents, thus unlocking a self-improvement flywheel. --- ### Two2Four: Generative Quadruped Puppeteering from Human Motion Authors: Fatemeh Zargarbashi, Zehong Qiu, Dhruv Agrawal, Stelian Coros, Robert W. Sumner, Martin Guay, Jakob Buhmann Published: 2026-07-28 Categories: cs.GR, cs.LG arXiv: https://arxiv.org/abs/2607.26108v1 Realistic animal motion for virtual production is typically obtained either through motion capture of highly trained performers who accurately mimic animal behavior, or by retargeting ordinary human motion using complex control setups. Both approaches are challenging and often fail to fully reproduce the nuances of natural animal motion, motivating data-driven alternatives. We present an automatic human-to-quadruped puppeteering framework that produces plausible and controllable quadruped motions from ordinary human motion data. Our approach employs a two-stage generative diffusion model trained purely on quadruped motion data. By introducing a structured conditioning and inpainting strategy, our method supports a wide range of actions, including walking, running, jumping, sitting, and lying. Furthermore, we enable fine-grained intuitive control of the quadruped motion such as head movement control and individual limb puppeteering. Experimental results demonstrate improved motion realism and controllability compared to existing retargeting approaches, highlighting the effectiveness of our framework as a tool for animation and virtual production applications. --- ### TraceCLIP: Recovering Local Semantics from Patch-to-CLS Contributions Authors: Xinran Liu, Shouqian Shi, Yutong Chen, Ge Wang, Xin-Wei Yao, Sheng Zhong Published: 2026-07-28 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2607.26107v1 Dense vision-language understanding, including object localization, region recognition, and open-vocabulary semantic segmentation, requires associating language concepts with spatially grounded visual regions. CLIP provides a strong foundation for these tasks by learning a shared image-text embedding space from large-scale contrastive pre-training. However, its image-level objective aligns text with a CLS-derived global representation, leaving local vision-language correspondence only indirectly constrained. Existing methods either introduce additional supervision, external models, or task-specific adaptation, while training-free approaches mainly recover dense responses from existing patch features without examining where local semantics become most accessible within CLIP. We introduce TraceCLIP, a training-free framework that recovers latent patch-level semantic evidence by isolating the patch-specific terms written into the CLS attention output. TraceCLIP further converts contribution-derived semantic responses into a semantic-geodesic topology gate that calibrates final-layer patch affinity for dense feature reconstruction. Diagnostic experiments show that these contribution features exhibit strong local semantic discrimination and text-conditioned spatial alignment. On eight zero-shot semantic segmentation benchmarks, TraceCLIP achieves gains of 1.3 to 4.5 points in average mIoU over the strongest prior training-free methods across both backbones and background settings, without additional training, external vision foundation models, or region-level supervision. More broadly, these findings suggest that spatially localized semantics may remain accessible within the internal construction of globally aligned representations. --- ### ScalablePromptus: Scalable and High-Fidelity Prompt-Based Video Streaming Authors: Zehao Cao, Bowei Xu, Xun Cao, Zhan Ma, Hao Chen Published: 2026-07-28 Categories: eess.IV, cs.CV, cs.MM arXiv: https://arxiv.org/abs/2607.26106v1 Prompt-based video streaming transmits compact semantic prompts instead of pixel-level content for generative reconstruction, enabling ultra-low-bitrate communication. However, the state-of-the-art Promptus framework is vulnerable to network fluctuation, where partially received prompts lead to catastrophic quality collapse. We propose ScalablePromptus, which enhances Promptus with semantic and color-aware prompt inversion, spherical linear interpolation for intermediate frames, and--most critically--a dropout training strategy that produces rank-ordered prompt representations. This allows the receiver to reconstruct meaningful video from arbitrarily truncated prompts without any adaptation. Under stable networks, ScalablePromptus achieves modest quality gains. Under lossy conditions, it reduces the performance degradation caused by truncation by 82%-95% compared to the baseline, making prompt-based streaming robust enough for real-world deployment. --- ### Weight and Height Estimation from a Single Human Image Captured in the Wild Authors: Hira Yaseen, Arif Mahmood, Waqas Sultani Published: 2026-07-28 Categories: cs.CV, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2607.26104v1 A person's physical characteristics such as weight and height are important indicators of his physical and mental health, daily life routines and finances. Body Mass Index (BMI) is a well known measure that encodes the characteristics of both the weight and the height. BMI has been used as a self-monitoring tool, and it has long-term implications on one's life. For example, it may help predicting the risk of various diseases and estimating longevity. Automatic BMI estimation using a single person image in the wild is a challenging task due to wide variations in human pose, camera geometry, personal appearance and distracting backgrounds. In this paper, we explore the performance of deep neural networks using single and multi-task learning by employing different modalities including RGB, depth-maps, pose-affinity maps, and edge-maps to predict BMI, weight, and height from daily life images available on social networking websites. Currently, no full body image dataset for BMI estimation is publicly available, therefore we propose a new dataset consisting of 6105 images with ground truth labels of height, weight and BMI. Our proposed dataset is collected in the wild containing images from various ethnicity and distributed over varying age groups and gender. It consists of frontal, back, full and half body, side poses, mirror selfies with varying backgrounds and scale variations and may contain artifacts hiding partial or full face. Extensive experimentation is performed using full body, half body and face images only using different CNN backbones including VGG, Densenet and ResNet. Our experimental results demonstrate that full body images have produced better results than the other half body and facial images in the wild. --- ### A Reference-Free Score for Detecting Silent Reasoning Failures in Large Language Models Authors: Vivek Shukla, Varun Shukla, Atul, Divya Mishra, Mehul Kumar Das Published: 2026-07-28 Categories: cs.CR, cs.AI arXiv: https://arxiv.org/abs/2607.26102v1 Mathematical chain of thought (CoT) evaluation is commonly reduced to whether the final answer matches a reference. This conflates producing a correct conclusion with producing a valid derivation an invalid chain can accidentally reach the right answer, while a valid calculation can be followed by a transcription error. We call this mismatch the reasoning answer consistency gap. This framework paper introduces the Reasoning Answer Faithfulness Score (RAFS), a reference free, instance level diagnostic of whether an emitted mathematical trace is locally credible, supports its answer, and is stable under resampling and targeted counterfactual interventions. RAFS combines step validity, reasoning to answer entailment and counterfactual sensitivity, answer consensus, and conditional reasoning stability. It evaluates transcript level agreement, not a models private computation and not factual correctness outside the tested mathematical setting. We retain a preregistered, results blind confirmatory study on GSM8K and MATH, with hypotheses, admissibility rules, calibration, and tests fixed before confirmatory outcomes are inspected. A separate feasibility pilot is specified to verify end to end execution and estimate interven tion coverage before that freeze numerical pilot claims are re ported only when trace level artifacts are available. We formalize four reasoning answer outcomes, justify the non compensatory aggregator, instantiate semantic trace distance, quantify compute and abstention tradeoffs, and define verifier independence and power analyses. RAFS is intended to complement mathematical answer accuracy with an auditable warning signal for silent reasoning failures and answer extraction errors --- ### Lilith: Backdoor Generalization under Training-Inference Trigger Shift Authors: Zhou Feng, Jiahao Chen, Chunyi Zhou, Yuan Su, Tianyu Du, Yuwen Pu, Jianhai Chen, Jinbao Li, Shouling Ji Published: 2026-07-28 Categories: cs.CR, cs.LG arXiv: https://arxiv.org/abs/2607.26099v1 Machine-learning services increasingly rely on public data, third-party providers, and outsourced training, creating opportunities for data-poisoning attacks that implant persistent malicious behavior while preserving benign utility. However, existing backdoor studies largely evaluate exact trigger reuse, training-exposed trigger diversity, or variations along predefined transformation axes. They therefore leave a critical blind spot: whether a backdoor learned from one training-time trigger can generalize to an inference-time trigger family absent from victim training. We formulate this problem as backdoor generalization under training--inference trigger shift and introduce Lilith, a black-box anchor-to-family framework. Using only disjoint surrogate resources, Lilith first induces a compact target-side vulnerability with a single training anchor, then constructs a bounded inference-only family that preserves the anchor-induced representation geometry. We characterize this mechanism through anchor clearance and family reach, deriving sufficient conditions for family-wise target preservation under local regularity and bounded surrogate--victim discrepancy. Experiments across datasets, architectures, poisoning rates, and defenses show that Lilith achieves high family-wise attack success with limited utility degradation and a small trigger generalization gap. Additional analyses show that family activation depends on representation alignment rather than the proposal mechanism, exposing a broader threat overlooked by exact-trigger evaluation. --- ### Knowledge-guided Disentanglement with Atomic Actions for Action Recognition Authors: Tianci Wu, Siqi Cao, Guangming Zhu, Jiang Lu, Siyuan Wang, Longfei Zhang, Jincai Huang, Jun Sheng, Liang Zhang Published: 2026-07-28 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.26097v2 Action recognition in complex scenes often involves multiple concurrent fine-grained actions, making it challenging to model internal action structures. Most existing methods rely on holistic representations, which are insufficient for capturing subtle interactions and fine-grained semantics. While recent prompt-based approaches introduce disentanglement, they lack explicit semantic guidance, and methods based solely on visual or structured cues remain coarse-grained. In this paper, we propose Knowledge-guided Disentanglement with Atomic Actions (KDA), which leverages fine-grained semantic knowledge to enhance action representations and enable more precise disentanglement. Specifically, we use Large Language Models (LLMs) to decompose action labels into atomic actions, providing explicit spatial-temporal semantics. A Knowledge Injection Module (KIM) first integrates atomic action knowledge into video features. Based on this enhanced representation, a Knowledge Disentanglement Module (KDM) further disentangles atomic action knowledge to produce more precise semantic guidance for action disentanglement. A Knowledge Disentanglement Loss (KD Loss) is introduced to encourage clearer disentanglement of knowledge components within KDM. Extensive experiments demonstrate that KDA improves feature discriminability and achieves state-of-the-art performance on multi-label action recognition benchmarks. Moreover, KIM and KDM can be readily integrated into other methods, demonstrating strong generality. --- ### Meta-Learned Reward Shaping for Reinforcement Learning from Human Feedback Authors: Yunpeng Chu Published: 2026-07-28 Categories: cs.LG, cs.CL arXiv: https://arxiv.org/abs/2607.26094v1 Reinforcement Learning from Human Feedback (RLHF) is the standard approach for aligning large language models with human preferences, but its quality is limited by static, task-agnostic reward models. This mismatch leads to sparse learning signals and suboptimal alignment. We introduce MeRLa (Meta-Learned Reward Shaping), a principled framework that meta-learns a task-aware shaping function $Φ(x,y;φ)$ across auxiliary tasks before RLHF training. The learned shaping produces a composite reward that preserves policy optimality while providing task-specific learning signals. Our meta-objective combines task discrimination, entropy regularization, and potential-based conservation for stable convergence. We provide theoretical guarantees for policy invariance, analyze representation drift sensitivity, and formally address incentive misalignment from entropy maximization. Experiments on LLaMA-3-8B across four benchmarks show consistent improvements over PPO, DPO, GRPO, and DAPO, achieving a 90.8% length-controlled win rate on AlpacaEval 2.0 and a score of 9.14 on MT-Bench, with 41% less training instability. MeRLa retains its benefits when combined with process-based and rubric-based enhanced rewards. --- ### Pass the Baton: Trajectory-Relayed On-Policy Distillation Authors: Haolei Xu, Xiaowen Xu, Haiwen Hong, Zixuan Ni, Hongxing Li, Yiwen Qiu, Weiming Lu, Yongliang Shen Published: 2026-07-28 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2607.26057v1 On-policy distillation (OPD) grounds token-level supervision in the student's own trajectory, yet suffers from prefix failure: once the student commits to a wrong reasoning direction, all subsequent generation builds on this deviation, producing misdirected continuations that elicit unreliable supervision and waste compute. We identify a teacher-student continuation asymmetry on failed prefixes, where the teacher tends to redirect while the student continues along the original direction, and convert it into a label-free handoff trigger in Relay On-Policy Distillation (Relay-OPD). During training, Relay-OPD constructs relay trajectories by letting the teacher briefly take over at detected trigger points to produce a teacher leg, after which the student resumes and is optimized on the resulting trajectory. A limited relay budget concentrates intervention on critical early positions while limiting departure from the student policy. With a Qwen3-4B-Instruct-2507 teacher and Qwen3-0.6B/1.7B-Non-Thinking students on eight mathematical reasoning benchmarks, Relay-OPD achieves the best or second-best results on every benchmark, outperforming standard OPD by +5.73% and the strongest baseline FastOPD by +1.49% on average for 1.7B, with consistent gains at 0.6B. Training trajectory length is reduced by over 50%. --- ### $π\mathbf{R}^2$: Reactive Real-time Flow Policies Authors: Sungjae Park, Shubham Tulsiani Published: 2026-07-28 Categories: cs.RO, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2607.26055v1 Generalist manipulation policies increasingly take the form of action-chunking flow policies built on large pretrained backbones. Such chunks run open-loop, so the policy cannot react to sensory input arriving mid-execution, sacrificing \emph{reactivity}. Replanning more often would restore it, but the perception-to-action pipeline (a large backbone plus multiple denoising steps) is too slow: this \emph{latency} forbids frequent replanning and leaves committed actions stale, making such policies ill-suited for dynamic, closed-loop control. We present $π\mathbf{R}^2$, which makes these policies reactive and real-time while retaining large backbones, expressive multi-modal policies, and multi-action prediction. Built on the per-position noise schedule of diffusion forcing, $π\mathbf{R}^2$ contributes two ideas. First, it splits conditioning into a fast channel (proprioception, fresh every tick) and an asynchronously updated slow channel (vision-language features), so the policy reacts to proprioception within a chunk while tolerating stale vision. Second, a latency-adaptive flow schedule treats in-flight actions as inpainting conditioning and emits actions in one denoising step per call, letting one trained model adapt to varying hardware latency. Requiring minimal modification to existing architectures, $π\mathbf{R}^2$ can be finetuned from a pretrained policy: applied to GR00T-N1.7 on a real xArm6+XHand platform, it replans closed-loop roughly $4\times$ faster than the base policy (~$25$Hz on an A5000 GPU), acting on a fresh observation every $40$ms. Across simulation and real-world manipulation tasks, $π\mathbf{R}^2$ improves the success rate by up to $23\%$ in simulation and $30\%$ in the real world over the strongest baseline. Project page: https://pi-r2-flow.github.io/ --- ### Spend Experts Where You Are Unsure: Confidence-Adaptive Routing for Mixture-of-Experts LoRA Authors: Tom Saliencro, Rohan Desai, Priya Nair, Maya Lindqvist, Daniel Whitmore Published: 2026-07-28 Categories: cs.LG arXiv: https://arxiv.org/abs/2607.26052v2 Mixture-of-Experts (MoE) variants of Low-Rank Adaptation (LoRA) route every token to a fixed number of experts $k$. Tokens differ in how uncertain the model is about them, so a single k over-spends on easy tokens and under-serves hard ones. We observe that the router's output distribution is already a per-token uncertainty signal: peaked mass indicates confidence, while a flat distribution indicates ambiguity. We introduce CARE (Confidence-Adaptive Routing of Experts), which admits experts in a nucleus fashion. Experts are activated in decreasing router weight until their cumulative mass reaches a threshold, with a small extension when the admitted experts disagree. A budget thermostat calibrates the threshold so that the average number of active experts matches any target. CARE is a drop-in, single-forward-pass rule with no extra parameters. Across eight commonsense benchmarks on LLaMA-3.1-8B and Qwen2.5-7B, as well as math, code, and knowledge tasks, CARE improves over fixed top-k MoE-LoRA at matched compute and matches the fixed-k=4 baseline while activating fewer experts. The same confidence and disagreement signals also improve out-of-distribution detection over MSP, entropy, and multi-pass proxies. We support the design with nucleus fidelity, budget optimality, and an epistemic reading of disagreement, and we release code. --- ### Re-thinking Mammography Transfer Learning: The Dataset-Informed Transfer Learning (DITL) Framework for Breast Cancer Screening and Lesion Diagnosis Authors: Adarsh Bhandary Panambur, Siming Bayer, Andreas Maier Published: 2026-07-28 Categories: cs.LG arXiv: https://arxiv.org/abs/2607.26043v1 Enhancing classification performance in mammography remains a persistent challenge across both small curated datasets and large-scale clinical cohorts. Conventional transfer learning approaches often neglect dataset-specific characteristics, while recent neighborhood-informed methods have been restricted to narrow tasks with rigid formulations, limiting their scalability to population-level datasets. To address these challenges, we propose the Dataset-Informed Transfer Learning (DITL) framework, which integrates dataset-derived difficulty signals with neighborhood-based triplet supervision in a unified objective. DITL introduces two adaptive components: (i) Adaptive Difficulty-Weighted Cross-Entropy (A-DWCE), which assigns per-sample weights based on k-nearest neighbor label purity in a self-supervised feature space, and (ii) Adaptive Neighborhood Representation Triplet (A-NR-Triplet), which enforces intra-class compactness and inter-class separation using a learnable margin. Unlike focal loss, DITL requires no hyperparameter tuning, removes heuristic weighting and fixed margins, and incurs negligible computational overhead, yielding a robust and scalable optimization strategy. On the large-scale VinDR-Mammo dataset, DITL achieves state-of-the-art performance for whole-image breast density classification, with significant improvements across accuracy, F1-score, and AUC (p < 0.0001). Beyond large cohorts, DITL also delivers consistent, statistically significant gains on small ROI datasets (p < 0.0001). By bridging small-scale lesion analysis with large-scale density estimation, DITL establishes a clinically relevant, scalable, and generalizable framework for mammography classification, spanning the full breast cancer screening-to-diagnosis spectrum. --- ### VetClaw: An Edge-Cloud Multimodal Agentic System for Veterinary Disease Screening Authors: Syed Mhamudul Hasan, Anas AlSobeh, Hussein Zangoti, Abdur R. Shahid Published: 2026-07-28 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2607.26042v1 We present VetClaw, an edge-cloud multimodal agentic system for early veterinary disease screening. VetClaw uses a camera module as an edge sensing device and sends captured images, together with optional symptom descriptions, to a server-hosted vision-language model for zero-shot disease classification. The system separates agent interaction from workflow orchestration: OpenClaw provides scheduling, tool access, user interaction, and notification services on the edge device, while LangGraph manages the stateful screening workflow, including input validation, image transmission, model invocation, safety checks, conditional routing, failure handling, and structured logging. This design moves beyond static image classification by enabling the system to collect visual evidence, invoke external models, apply deterministic safety rules, and generate diagnostic-support alerts. Results show that image-only VLM prediction remains limited, whereas symptom-guided and multimodal inputs improve zero-shot classification performance. Thus, VetClaw transforms a static prediction model into a coordinated, safety-aware system that can use tools, manage workflows, handle failures, and escalate uncertain cases. --- ### Desktop-Delta Bench: Do Computer-Use Models Understand Desktop GUI Transitions? Authors: Abhishek Pillai, Samir Kumar Nayak, Yuan Chen Published: 2026-07-28 Categories: cs.AI, cs.CV arXiv: https://arxiv.org/abs/2607.26041v2 Computer-use agents (CUAs) increasingly act through desktop GUIs to complete long-horizon tasks. Current benchmarks primarily measure end-task success or single-frame grounding. Neither isolates whether a model can reconstruct the causal, task-relevant transition produced by an action- crucial for rejecting stale observations, verifying progress, and recovering from failure. This is difficult because inference, remote input, app rendering, and screenshot capture are asynchronous: the next observation may be delayed, occluded, transient, or unrelated, then misread as progress and carried into subsequent planning. We introduce Desktop-Delta Bench (DDB), an offline step-level benchmark with 2,013 human-verified instances from novel, multi-app Linux trajectories across ~15 applications and 50 task domains. DDB trajectories targets 3 failure dimensions- state verification, source tracking, and context-aware control- through 2 complementary tasks: 463 3-frame temporal-ordering instances, including 105 with a cross-trajectory decoy, and 1,550 before-after pairs labeled from 5 actions + its payload. We evaluate 8 closed and open-source model families across 32 ordering and 16 single-action settings, observing consistent gaps. Ordering remains unsaturated: best non-decoy and decoy exact-match rates are 65.1% and 65.7%. Task context improves decoy identification by 6.9 percentage points but reduces non-decoy exact match by 2.2 points; error analysis reveals systematic copying of the presented A-B-C order. Single-action results show that inferring the action family is harder than locating it: click F1 is 0.96 vs, 0.76 for drag, while recognized drags are generally localized well. DDB, thus, complements end-to-end benchmarks by filling the missing diagnostic layer between GUI grounding and final task success, enabling targeted improvements to desktop CUA verification, reliability, and recovery. --- ### Reinformed Dreamer: An Asymmetric World Model Efficiently Trained through Latent Guidance Authors: Gaspard Lambrechts, Adrien Bolland, Daniel Ebi, Damien Ernst Published: 2026-07-28 Categories: cs.LG, stat.ML arXiv: https://arxiv.org/abs/2607.26040v1 Much like humans benefit from guidance while learning, reinforcement learning algorithms may benefit from additional supervision beyond rewards. Leveraging additional information during training to learn better representations and behaviors has been the focus of asymmetric reinforcement learning. This learning paradigm has proven effective under partial observability when additional state information is available, but also under full observability when more refined state information is available. Focusing on model-based reinforcement learning, we study the effect of asymmetric learning on observation representations and on privileged information representations. First, we identify a limitation in the privileged information representations learned by an asymmetric model-based algorithm known as the Informed Dreamer. Then, we propose a novel asymmetric representation learning objective using latent guidance, resulting in a new algorithm called the Reinformed Dreamer. Experiments across several benchmarks show a more consistent improvement over Dreamer than previous asymmetric approaches. --- ### Collaborative System Failure Prognostics via Federated Longitudinal-Survival Modeling Authors: Fan Yang, Madelyn Weller, Dimuthu Fernando, Hila Livneh, Yuxin Wen Published: 2026-07-28 Categories: cs.LG arXiv: https://arxiv.org/abs/2607.26038v1 Time-to-event modeling provides a systematic framework for estimating time-dependent failure risk, reliability, and remaining useful life (RUL) from longitudinal condition monitoring data. However, applying these models to distributed prognostics remains challenging because sensor trajectories and failure-time records are often stored across organizations or operational sites and cannot be centrally pooled due to privacy or proprietary constraints. Moreover, the classical Cox proportional hazards model relies on a nonseparable partial likelihood involving global risk sets, making direct optimization difficult under standard federated learning protocols. This paper presents a federated longitudinal-survival modeling framework for collaborative system failure prognostics. The proposed framework combines longitudinal sensor representation learning with a client-separable discrete-time hazard objective, enabling multiple clients to collaboratively train a prognostic model without sharing raw sensor measurements or individual failure records. Time-dependent representations extracted from multivariate sensor histories are used to estimate interval-specific failure hazards, reliability curves, and system RUL. Experiments on the four C-MAPSS turbofan engine degradation subsets under simulated decentralized settings demonstrate that the proposed framework consistently improves prognostic performance over isolated local training while maintaining performance comparable to centralized training across heterogeneous operating conditions and failure modes. These results demonstrate the potential of federated longitudinal-survival modeling for collaborative, data-aware condition monitoring and system failure prognostics. --- ### Wonder: Video World Model Done Better Authors: Jiacong Xu, Hanwen Jiang, Zhixin Shu, Kalyan Sunkavalli, Vishal M. Patel, Yiqun Mei Published: 2026-07-28 Categories: cs.CV, cs.GR arXiv: https://arxiv.org/abs/2607.26037v1 We present Wonder, a general-purpose video world model for real-time, camera-controllable world exploration. Given an image or a conditional video, Wonder constructs a playable world where users can navigate interactively by moving the camera, discovering unseen regions, and revisiting previously observed areas in real time and over a long-term horizon. Achieving this capability requires a system-level co-design of control method, memory mechanism, and training strategy. We introduce a novel camera conditioning with a dense coordinate field whose renderings provide spatially aligned motion and orientation cues, allowing the model to interpret camera motion directly as visual evidence. To support fast and precise memory retrieval over a growing generation context, we propose an efficient sparse attention-based memory mechanism, enabling the model to selectively attend to a small set of relevant context tokens at inference time, regardless of actual context length. We further develop several techniques to rectify the self-forcing-style distillation pipeline, improving the student model's ability to respect control signals, as well as maintaining diverse generation modes and long-term memory from the teacher. Together, these components enable Wonder to synthesize diverse, minute-scale videos at 16 FPS while preserving coherent geometry, appearance, and dynamics across long rollouts. Beyond image-to-video generation, Wonder naturally supports video-conditioned generation, allowing existing dynamic scenes to be re-shot in real time. --- ### Falling Behind Drives Unsafe Development in an Idealised AI Race Experiment Authors: Elias Fernández Domingos, The Anh Han Published: 2026-07-28 Categories: cs.AI, cs.CY, cs.GT, econ.GN arXiv: https://arxiv.org/abs/2607.26034v1 Technological races create tension between speed and safety: actors may gain by moving faster than competitors, even when risky development is harmful. This is prominent in debates about artificial intelligence (AI), where competitive pressure is often argued to incentivise riskier, less safety-conscious development. We study this using a framed behavioural experiment based on an idealised AI race, in which paired participants repeatedly chose between Safe and Unsafe development under an uncertain time horizon. Unsafe development gave faster progress and higher immediate payoffs but accumulated private risk up to a treatment-specific maximum of 10\%, 60\%, or 90\%; the race's competitive structure was held constant, and only this maximum risk varied. Neither the pre-registered comparison between risk levels nor the role of elicited risk preferences was supported by the data. Instead, exploratory analyses motivated by the task's repeated structure show that Unsafe behaviour is shaped less by risk preferences than by the evolving strategic state of the race: participants are more likely to choose Unsafe after their opponent does so, being ahead reduces Unsafe play while falling behind increases it, and first-round choices predict later behaviour. To interpret these effects we introduce a reduced evolutionary model with four strategies -- Always Safe, Always Unsafe, Conditionally Safe, and Conditionally Antisocial Safe -- which reproduces the treatment effect and shows how conditional Unsafe behaviour can be favoured by competitive race dynamics. Together, the experiment and model show that unsafe development can emerge from early behavioural momentum, opponent behaviour, and fear of falling behind, rather than from risk preferences alone, suggesting policy should focus on reducing competitive pressure and promoting cooperation in AI development rather than only individual risk. --- ### CHARM: A Multimodal Graph Foundation Model with Hierarchical Context Modeling for Zero-Shot Transfer Authors: Ankang Yang, Jitao Zhao, Di Jin, Yuxiao Huang, Dongxiao He Published: 2026-07-28 Categories: cs.AI arXiv: https://arxiv.org/abs/2607.26023v1 Graph foundation models (GFMs) have emerged as a promising paradigm for transferring knowledge across graph domains and tasks. Real-world graphs associate nodes with text, images, and other modalities, making multimodal graphs essential for representing complex entities and relations. Moreover, collecting labels and adapting models for every new graph domain is costly and often infeasible, motivating zero-shot transfer. Unfortunately, zero-shot transfer on multimodal graphs remains underexplored. Existing GNN-based graph foundation models typically require downstream adaptation, whereas LLM-based graph methods mainly address unimodal graphs or tasks within a single domain. This setting presents two key challenges. First, models must generalize knowledge from individual modalities while capturing transferable cross-modal relations. Second, without target-domain fine-tuning, node representations remain entangled with domain-specific structures and modality-specific characteristics, obscuring shared concepts in unseen domains. To address these challenges, we propose CHARM, a multimodal graph foundation model with hierarchical context modeling for zero-shot transfer. CHARM replaces isolated raw nodes with hierarchical graph contexts that capture multimodal semantics and cross-modal relations. These contexts map domain-specific node patterns to shared high-level concepts, reducing reliance on target-domain supervision or adaptation. A modality-aware graph context encoder integrates multimodal information with graph structure and converts the resulting representations into graph tokens for a large language model . Experiments show consistent improvements on zero-shot multimodal graph tasks. --- ### UniMem: Complementary Episodic-to-Parametric Memory for Boundary-Agnostic Task Streams Authors: Siyu Xia, Chenheng Zhang, Yanting Wu, Haoxuan Li, Jiajun Chai, Xiaohan Wang, Guojun Yin, Wei Lin, Zhouchen Lin, Haifeng Zhang, Jun Wang Published: 2026-07-28 Categories: cs.CL arXiv: https://arxiv.org/abs/2607.26017v1 Memory is essential for LLM agents to accumulate task experience and reuse task-specific execution strategies. However, real-world deployment over boundary-agnostic and evolving task streams exposes a fundamental stability-plasticity dilemma. External retrieval-based memory can rapidly absorb new evidence, but it often fails to internalize recurring execution patterns and incurs inference-time retrieval overhead. Parametric memory enables stable and efficient execution once learned, but typically relies on explicit task boundaries and fixed parameter budgets. Inspired by the human brain, which balances plasticity and stability through complementary episodic storage and gradual consolidation, we propose UniMem, a self-routing framework for autonomous memory management. UniMem uses learnable routing tokens as memory controllers, enabling adaptive coordination between complementary memory pathways: novel or sparse tasks are retained in an episodic buffer for retrieval-augmented execution, while recurring and reliable patterns are consolidated into expandable parametric memory. By decoupling task identification from task execution with routing tokens and parametric memory blocks, UniMem expands memory on demand without task labels during deployment or uncontrolled parameter growth. Experiments on long-horizon streaming task sequences show that UniMem consistently outperforms baselines while maintaining execution fidelity, achieving an average gain of 4.0 EM points across three backbone models. --- ### MDTransformer: A Hardware-Software Co-Design of Mode-Division Photonic Transformer Accelerator with Inverse-Designed Coherent Crossbar Authors: Solomon Micheal Serunjogi, Rachmad Vidya Wicaksana Putra, Ayat Taha, Muhammad Shafique, Mahmoud Rasras Published: 2026-07-28 Categories: cs.AR, cs.AI, cs.DC arXiv: https://arxiv.org/abs/2607.26016v1 Recently, photonic transformer accelerators (PTAs) have successfully achieved significant speedup and energy efficiency improvements over electronic accelerators for expediting Transformer inference. However, state-of-the-art rely on expensive multi-wavelength light generation and large dot-product units due to active phase-shifter components, thus making their approach inefficient and impractical. To address this, we propose MDTransformer, a novel hardware-software co-design of PTA based on mode-division optical dataflow and operations. Specifically, MDTransformer performs complex matrix operations using spatial-mode interference, that leverages the inverse-designed multi-mode couplers, crossings, and Mach-Zehnder IQ modulators into a compact mode-division photonic tensor core (MPTC), capable of executing matrix multiplications in the optical domain. Its each guided mode (i.e., TE0-TE3) acts as an independent computational lane, enabling four-fold parallelism-per-waveguide without spectral filtering or free-spectral-range limitations. Moreover, its coherent detection and IQ modulation jointly encode amplitude and phase, realizing complex-valued arithmetic for full-range operations in transformers. MDTransformer offers analog multiplication with sub-4-bit effective precision and inter-modal crosstalk below -30 dB. Its inverse-designed approach also offers scalable and full compatibility with single-laser continuous-wave operation at 1550 nm. Experimental results show that MDTransformer achieves 40.4% area reduction, 63.6% power saving, 40.6% energy saving, and comparable latency over the state-of-the-art PTA across different workloads (i.e., DeiT-Tiny/Small/Base and BERT-Base/Large). These results show that MDTransformer offers a practical solution for high-performance and energy-efficient transformer-based systems. --- ### Instruction-Tuned Models Locally Reuse Human Syntax More Than Humans Do Authors: Zandi Eberstadt Published: 2026-07-28 Categories: cs.CL arXiv: https://arxiv.org/abs/2607.26015v1 Syntactic convergence (the tendency of speakers to adapt in language towards the grammatical profiles of their interlocutors) is a well-documented feature of human dialogue widely considered to operate below conscious awareness. Whether large language models exhibit analogous syntactic convergence toward human users relative to human baselines and across a broad range of syntactic constructions remains an open question. Using substitution-paradigm data in which model generations replace one speaker's turns in pre-existing human dialogues, this study measures turn-adjacent reuse of context-free grammar (CFG) rules across sixteen open-weight Llama and Gemma models (1B-70B, pretrained and instruction-tuned) at 1,901 matched positions per model. Every model showed greater CFG-rule overlap with the preceding human turn than with a sampled unrelated human prime, and in every model this actual-versus-random difference was larger for lower-frequency rules. Each instruction-tuned model also showed greater natural-output overlap with the actual prime than the human response it replaced, and all eight matched architecture pairs exhibited greater actual-prime overlap after instruction tuning. However, relative to pretrained variants, instruction-tuned outputs overlapped more with unrelated primes, showed a smaller actual-versus-random increment, and had lower conditional rule-reuse odds once target rule-set size was held constant. In exploratory analyses, each model exhibited greater mean lexical and semantic similarity to the preceding turn than the matched human responses did. Instruction-tuned models additionally produced responses with greater mean semantic similarity than their pretrained counterparts in all eight architecture pairs, whereas the lexical similarity results were more heterogeneous. --- ### Pictura: Perspective-View Self-Play at Scale for Driving Authors: Yuan Yin, Elias Ramzi, Marc Lafon, Valentin Charraut, Victor Bares, Yihong Xu, Éloi Zablocki, Alexandre Boulch, Thibault Buhet, Andrei Bursuc, Matthieu Cord Published: 2026-07-28 Categories: cs.CV, cs.AI, cs.RO arXiv: https://arxiv.org/abs/2607.26005v1 Self-play in simulation produces robust driving policies at scale. Demonstrations of such behavior have been made using privileged vectorized observations such as exact poses and velocities, even for occluded agents. This assumes that perception is solved and introduces a representation gap with the partial observation of a deployed agent driving from the perspective view of egocentric cameras. A common fix, distilling the privileged policy into a camera-input student, leaves the student imitating decisions its own view cannot justify. Instead, we establish perspective-view self-play as a practical training regime. We introduce Pictura, a GPU-accelerated multi-agent driving simulator that renders each agent's egocentric view at every step, mitigating the representation gap at its source. Pictura sustains up to 500K agent-steps/s (2M images/s) on a single H100. Using Pictura, we train Alberti by self-play with plain PPO. It is the first large-scale driving self-play policy trained directly from perspective images, without privileged observations. Training spans 50B agent steps for ~35M km of driving. It approaches the driving performance of its privileged vectorized counterpart, and transfers zero-shot to Waymo Open Motion Dataset layouts re-rendered in Pictura, where it outperforms privileged vectorized agents. Project page: https://valeoai.github.io/Pictura/ --- ### Parallel Decoding Distillation for Fast Image and Video Generation Authors: Neta Shaul, Chao Liu, Arash Vahdat, Julius Berner Published: 2026-07-28 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2607.26004v1 Generation in video diffusion or flow models is computationally expensive due to the slow and iterative sampling process. Current state-of-the-art (SOTA) acceleration methods heavily rely on variational score distillation (VSD) and adversarial losses to distill diffusion models into few-step generators. Albeit achieving high-quality video generation, these training losses are notoriously hard to optimize and suffer from mode collapse, leading to loss of video diversity and lack of motion. In this paper, we introduce Parallel Decoding Distillation (PDD), a simplified and scalable trajectory-based distillation method for fast inference of diffusion and flow matching models. Our architecture and training procedure are compatible with any pre-trained model and support sampling with a varying number of function evaluations (NFE). PDD accelerates generation by predicting multiple denoising steps per network evaluation. Conceptually, it learns a representation of the mean velocity without regressing its derivative using JVPs or finite-difference approximations. Our method achieves SOTA performance with 4-8 NFE on LTX-2.3 Text-to-Video/Audio, Wan 14B Text-to-Video, and Qwen-Image Text-to-Image. Moreover, PDD presents a significant improvement in generated video diversity. ---