# Research Papers — 2026-07 · bundle page 53 of 222 > Full-text markdown of papers #2601–2650 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/53 [← Previous bundle](https://aipod.dev/papers/archive/2026-07/page/52.md) [Next bundle →](https://aipod.dev/papers/archive/2026-07/page/54.md) ### Hiding in Plain Sight: An Effective Physical Adversarial Patch Attack against Visual-Infrared Fused Face Detection Authors: Qiucheng Yu, Tao Ni, Yihe Zhou, Jiayimei Wang, Qingchuan Zhao Published: 2026-07-25 Categories: cs.CR, cs.CV arXiv: https://arxiv.org/abs/2607.23292v1 Deep learning-based visual-infrared fused face detection models are increasingly deployed across a wide range of applications, yet they remain susceptible to adversarial patch attacks. Most prior attacks target either the visual or the infrared image alone in the digital domain, which renders them ineffective against fused models in the physical world. Moreover, many of these methods are readily noticeable, as their patch patterns deviate substantially from those seen in the real world. In this paper, we introduce VIPatch (Visual-Infrared Patch), a novel physical adversarial patch attack that produces inconspicuous, realistic, and natural-looking patches for facial images. Specifically, VIPatch crafts a gradient-color mask together with a band-aid sticker across both the visual and infrared images, and jointly optimizes these two elements; the resulting digital patches further guide the fabrication of their physical counterparts. Experimental results show that VIPatch achieves competitive attack success rates (over 90%) in both the digital and physical domains, while keeping the patches unobtrusive to human observers. --- ### RareLens: Towards End-to-End Rare Disease Care via Aligning Divergent Large Language Model Reasoning Authors: Xi Chen, Hongru Zhou, Shiyu Feng, Hanyu Zhou, Huahui Yi, Rongsheng Wang, Tiancheng He, Kun Wang, Pingping Liu, Qiankun Li, Sicheng Lin, Huiying Ou et al. Published: 2026-07-25 Categories: cs.AI arXiv: https://arxiv.org/abs/2607.23290v2 Rare diseases represent one of the most challenging settings for clinical decision-making, where heterogeneous presentations, sparse evidence and limited expertise create persistent uncertainty throughout the care pathway. Although artificial intelligence could help, existing systems largely address isolated tasks, particularly diagnosis, and usually rely on downstream investigations rather than information available at initial presentation. Here we show that clinical AI performance under uncertainty can be improved not by scaling a single model, but by exploiting the diversity of multiple imperfect reasoning systems. Across heterogeneous large language models, we identify divergent reasoning trajectories with complementary error patterns and develop RareLens, which learns to reconcile these perspectives into actionable decisions across four stages of rare disease care: risk screening, diagnosis, treatment planning and prognosis prediction. Built on RarelensBench, a real-world dataset of 157,525 cases spanning all 33 Orphanet categories and more than 7,000 conditions, RareLens outperformed every frontier model tested, including GPT-5, DeepSeek-R1, Claude-3.7-Sonnet and Gemini-2.5-Pro, across all stages. It achieved an area under the curve of 0.917 for screening and top-1 accuracies of 65.5% and 89.8% for diagnosis and treatment. In an external evaluation involving 1,287 cases and 23 physicians, autonomous RareLens and physicians assisted by RareLens both outperformed unaided physicians, while demonstrating that effective human-AI collaboration requires more than simply providing model outputs. These findings establish divergent model reasoning as an exploitable source of information and suggest a general strategy for building AI systems that operate reliably under high clinical uncertainty. --- ### RareLens: Towards End-to-End Rare Disease Care via Aligning Divergent Large Language Model Reasoning Authors: Xi Chen, Hongru Zhou, Shiyu Feng, Hanyu Zhou, Huahui Yi, Rongsheng Wang, Tiancheng He, Kun Wang, Pingping Liu, Qiankun Li, Sicheng Lin, Huiying Ou et al. Published: 2026-07-25 Categories: cs.AI arXiv: https://arxiv.org/abs/2607.23290v1 Rare diseases collectively affect an estimated 3.5% to 5.9% of the population, yet more than 70% of patients are misdiagnosed and many endure years of evaluation before a diagnosis is reached, because early presentations are nonspecific and relevant expertise is scarce and unevenly distributed. Artificial intelligence could provide support, but existing systems address isolated stages of care, overwhelmingly diagnosis. They typically depend on the results of downstream investigations, and they treat the variability between models as noise to be eliminated. Here we present RareLens, a system that supports clinical decision-making across the entire rare disease trajectory by exploiting this variability. When heterogeneous large language models evaluate the same case, they generate divergent but complementary reasoning, which RareLens aligns and calibrates into a single convergent, actionable decision at each stage. Four coordinated modules perform primary-visit risk screening, diagnosis, treatment planning and prognosis. Developed and evaluated on RareBench, a real-world dataset of 157,525 cases spanning all 33 Orphanet categories and more than 7,000 conditions, RareLens outperformed every frontier model tested, including GPT-5, DeepSeek-R1, Claude-3.7-Sonnet and Gemini-2.5-Pro, at each stage. It achieved an area under the curve of 0.917 for screening and top-1 accuracies of 65.5% and 89.8% for diagnosis and treatment. In an external study spanning 1,287 cases and 23 physicians, autonomous RareLens and physicians assisted by RareLens both substantially outperformed unaided physicians. These findings indicate that aligning divergent model reasoning, rather than scaling a single model, offers a generalizable strategy for high-uncertainty clinical decision-making. --- ### Continuous surrogates versus threshold Boolean networks for modeling Arabidopsis ISR gene regulation Authors: Gonzalo A. Ruz Published: 2026-07-25 Categories: q-bio.MN, cs.LG, cs.NE arXiv: https://arxiv.org/abs/2607.23289v1 Gene regulatory network modeling often requires balancing predictive accuracy and mechanistic interpretability. In this work, we compare continuous surrogate models and a discrete mechanistic model on the same \textit{Arabidopsis thaliana} induced systemic resistance (ISR) dataset, using both the raw continuous gene-expression measurements and their sign-binarized representation. The study considers eight defense-related genes measured over nine time points and evaluates two continuous predictors, Random Forest (RF) regression and a Multi-Layer Perceptron (MLP), against a threshold Boolean network (TBN). The models are assessed using rolling-origin one-step prediction, recursive multi-step rollout, and interpretability analysis. RF achieved the best average one-step numerical performance in the continuous domain, with an MAE of 1.910 and an RMSE of 2.836, compared with 2.089 and 3.106 for the MLP. In the binary domain, the TBN obtained the best average one-step qualitative performance, with a binary accuracy of 0.550 and a Hamming distance of 3.600, compared with 0.500 and 4.000 for RF, and 0.495 and 4.040 for the MLP. In recursive rollout, the TBN exactly reproduced the observed binarized trajectory, while the MLP also showed near-perfect fidelity, with a trajectory binary accuracy of 0.986, and RF accumulated substantially larger deviation, with a trajectory binary accuracy of 0.708. These results highlight that local numerical accuracy and global qualitative dynamical fidelity are not necessarily aligned, and suggest that continuous surrogates and threshold Boolean networks should be viewed as complementary tools for modeling biological regulation. --- ### Approximate reservoir computing with a semiconductor laser for reducing energy consumption Authors: Tatsuki Ito, Kazutaka Kanno, Satoshi Kawakami, Atsushi Uchida Published: 2026-07-25 Categories: physics.optics, cs.LG, nlin.CD arXiv: https://arxiv.org/abs/2607.23288v1 Photonic reservoir computing is a promising physical machine-learning technique for predicting time-series data. The quantization of the response signal from the reservoir is required for the implementation of photonic reservoir computing, and the number of quantization bits and sampling frequency need to be optimized to achieve high performance and low energy consumption. However, few studies have been reported to investigate the effect of bit quantization and sampling frequency. In this study, we introduce a concept of approximate reservoir computing with a semiconductor laser by quantizing the amplitude of node states in the reservoir and output weights. We evaluate the performance of a chaotic time-series prediction task and energy consumption per sample. We achieve significant reduction of energy consumption by optimizing the number of quantization bits, the sampling frequency, and the injection current of the semiconductor laser, while maintaining the prediction performance. --- ### Learning Asymptotics with Convergence-Rate Guarantees using Linear Least Squares Authors: Christos N. Efrem Published: 2026-07-25 Categories: stat.ML, cs.LG, math.CO, math.NA arXiv: https://arxiv.org/abs/2607.23287v2 We introduce a new research area that is called Asymptotics Learning Theory (ALT) and combines optimization with asymptotic analysis. In particular, ALT provides a unified approach for computing unknown constants/parameters in proven asymptotic expansions using optimization theory. In this paper, we focus on a general asymptotic form which includes a broad class of asymptotics. Furthermore, we study two powerful numerical methods, namely, sliding Linear Least Squares (sLLSQ) and sliding Tikhonov Linear Least Squares (sT-LLSQ). For these techniques we rigorously prove asymptotic estimates that lead to sufficient conditions for convergence (to the correct values of unknown parameters) and convergence-rate guarantees. Despite their strengths, both methods have also limitations, e.g., slow convergence---or even, counterintuitively, divergence---in some cases. Moreover, we present fundamental applications in analytic combinatorics, a beautiful field of mathematics that deals with asymptotic enumeration of discrete structures using complex analysis. The proposed techniques complement existing approaches, such as the ratio method and its variants. Numerical examples also verify the theoretical results. Finally, we discuss interesting research directions in ALT. --- ### Learning Asymptotics with Convergence-Rate Guarantees using Linear Least Squares Authors: Christos N. Efrem Published: 2026-07-25 Categories: stat.ML, cs.LG, math.CO, math.NA arXiv: https://arxiv.org/abs/2607.23287v1 We introduce a new research area that is called Asymptotics Learning Theory (ALT) and combines optimization with asymptotic analysis. In particular, ALT provides a unified approach for computing unknown constants/parameters in proven asymptotic expansions using optimization theory. In this paper, we focus on a general asymptotic form which includes a broad class of asymptotics. Furthermore, we study two powerful numerical methods, namely, sliding Linear Least Squares (sLLSQ) and sliding Tikhonov Linear Least Squares (sT-LLSQ). For these techniques we rigorously prove asymptotic estimates that lead to sufficient conditions for convergence (to the correct values of unknown parameters) and convergence-rate guarantees. Despite their strengths, both methods have also limitations, e.g., slow convergence---or even, counterintuitively, divergence---in some cases. Moreover, we present fundamental applications in analytic combinatorics, a beautiful field of mathematics that deals with asymptotic enumeration of discrete structures using complex analysis. The proposed techniques complement existing approaches, such as the ratio method and its variants. Numerical examples also verify the theoretical results. Finally, we discuss interesting research directions in ALT. --- ### Adaptive Graph-of-Islands Evolution for Automatic Feature Engineering with LLMs Authors: Sha Li, Naren Ramakrishnan Published: 2026-07-25 Categories: cs.AI, cs.LG arXiv: https://arxiv.org/abs/2607.23286v2 Automatic feature engineering (AutoFE) for tabular data requires discovering informative transformations from a large program space. Existing approaches suffer from three limitations: classical methods rely on fixed operator libraries with limited expressivity, LLM-based methods generate proposals from static prompts without retaining search experience, and evolutionary methods use fixed migration policies that ignore task-specific cross-family transfer utility. We introduce TOPOFE, a framework that formulates AutoFE as graph-structured multi-island evolutionary program search. The transformation space is partitioned into semantically coherent families, each explored by an island through LLM-guided mutation and crossover. Each island maintains a Prompt Adaptation Memory that accumulates accept/reject feedback to steer proposals toward productive regions without parameter updates. To coordinate global exploration, TOPOFE dynamically learns a directed topology graph whose edge weights encode transfer utility between transformation families. Cross-island transfer is triggered by adaptive saturation detection and performed through LLM-mediated hybrid synthesis, enabling discovery of compositional feature programs that cannot emerge from isolated local search. Experiments on 29 tabular datasets show that TOPOFE consistently outperforms most state-of-the-art AutoFE methods on classification and regression tasks. Beyond predictive performance, TOPOFE produces feature sets with lower redundancy and higher representational coverage, while the learned topology graph acquires meaningful task-specific transfer structure correlated with downstream gains. The discovered feature programs transfer reliably across diverse predictors and LLM backbones, demonstrating that improvements arise from TOPOFE's structured search and adaptive coordination rather than backbone-specific generation capability. --- ### TopoFE: topology-aware LLM-guided Automated Feature Engineering Authors: Sha Li, Naren Ramakrishnan Published: 2026-07-25 Categories: cs.AI, cs.LG arXiv: https://arxiv.org/abs/2607.23286v1 Automatic feature engineering (AutoFE) for tabular learning can be naturally formulated as a program synthesis problem, where the objective is to discover predictive feature transformations from an exponentially large search space. Recent advances in large language models (LLMs) have expanded the expressiveness of AutoFE by enabling feature program generation beyond predefined operator libraries. However, existing LLM-based approaches remain fundamentally limited by stateless generation and homogeneous search: feature proposals are produced from static prompts without accumulating search experience, while single-population exploration quickly converges to dominant transformation patterns and rarely discovers complementary feature compositions across transformation families. We propose TOPOFE, a topology-aware multi-island evolutionary framework for LLM-guided feature engineering. TOPOFE combines family-specialized exploration, adaptive prompt memory, and topology-guided knowledge transfer to efficiently discover diverse and compositional feature programs. Experiments on 29 public tabular datasets demonstrate consistent improvements over state-of-the-art AutoFE methods across classification and regression tasks. Beyond predictive performance, TOPOFE discovers more diverse and transferable feature programs that generalize across multiple downstream predictors and LLM backbones. --- ### Photonic reservoir computing with complex networks Authors: Sion Park, Kohei Watabe, Satoshi Sunada, Tomoki Yamagami, Atsushi Uchida Published: 2026-07-25 Categories: physics.optics, cs.LG, nlin.CD arXiv: https://arxiv.org/abs/2607.23285v1 Photonic reservoir computing has attracted increasing attention as a fast and low-cost approach for time-series prediction. Photonic reservoir computing utilizes the high speed, broad bandwidth, and spatial parallelism of light. However, the effect of the internal connection structure (network topology) on the computing performance has not been investigated for large-scale photonic reservoirs. In this study, we experimentally and numerically demonstrate photonic reservoir computing using a spatial light modulator to systematically evaluate the relationship between the network topology and the performance of reservoir computing. We introduce complex network structures such as small-world and scale-free network topologies of the internal nodes in the reservoir. We perform the memory capacity measurement and the one-step-ahead prediction task of the chaotic time series to compare the performance. We found that the small-world network exhibits the maximum memory capacity and the best prediction performance. Our numerical calculations reveal that the performance of the time-series prediction can be optimized by changing the rewiring probability of the network and the leak rate of the reservoir. We also implement photonic human brain network as a reservoir, which is designed by the connectomes of human brain activities. We found that the network topology strongly affects the performance of reservoir computing, and the small-world network structure outperforms the other configurations. --- ### StageGuard: Physiologically Constrained Sleep Staging Authors: Juntang Wang, Yihan Wang, Hao Wu, Jiayu Gao, Shixin Xu, Dongmian Zou Published: 2026-07-25 Categories: cs.LG, q-bio.QM arXiv: https://arxiv.org/abs/2607.23284v1 Automated sleep staging is increasingly used in large-scale studies to derive sleep-architecture endpoints: total sleep time, REM latency, sleep efficiency, and bout-duration statistics. Deep learning models achieve epoch-level accuracy approaching inter-rater agreement, yet often produce hypnograms that violate physiological invariants, such as rare transitions (e.g., direct Wake -> REM) or excessively fragmented sequences. Such violations can bias downstream sleep metrics, regardless of overall accuracy. We propose StageGuard, a plug-and-play, backbone-agnostic structured-inference framework that wraps any neural sleep-staging backbone with physiology-informed priors. StageGuard combines (1) a differentiable soft transition penalty that discourages physiologically rare transitions during training, and (2) a semi-Markov constrained decoder with a duration-augmented state space that jointly enforces transition penalties and minimum bout durations at inference. Unlike hard-prohibition methods, it admits rare transitions when emission evidence is overwhelming, leaving informative pathological events recoverable rather than blocked. StageGuard constrains staging outputs to satisfy known physiological priors rather than modeling sleep generatively. We quantify the validity gap using transition-violation rate (TVR) and fragmentation index (FI) and demonstrate that, across six backbones and four datasets, StageGuard reduces TVR to physiologically plausible levels and lowers FI by 56-62%, while maintaining or slightly improving classification accuracy. Crucially, improved constraint satisfaction translates into 59-79% lower error on derived sleep-architecture statistics not directly optimized by the method, and recovers the direction and effect size of expert-defined subgroup differences (OSA severity, age) more faithfully than the unconstrained baseline. --- ### Co-Evolving Graph and Text Memory for Training-Free Multi-Hop Question Answering Authors: Hieu Man, Thien Huu Nguyen Published: 2026-07-25 Categories: cs.CL, cs.MA arXiv: https://arxiv.org/abs/2607.23278v1 Multi-hop question answering requires coordinating relational and textual evidence across reasoning steps, a combination neither a text corpus nor a knowledge graph can supply alone. Prior work often emphasizes only part of this loop: graph-augmented RAG retrieves from a pre-built or query-updated graph, KGQA systems search within topic-centered subgraphs, and memory-augmented agents maintain evolving memories without continuously reconciling graph memory with textual context. We propose Co-E, a training-free system built around synchronized bidirectional graph-text working memory. A synchronization cycle consolidates textual memory, extracts relational triples into graph memory, and injects graph facts back into the generation context. Because both memories are maintained, they shape subsequent retrieval and generation. Evaluated on six multi-hop QA benchmarks, Co-E improves over comparable training-free open-backbone baselines and is competitive with larger or trained systems. --- ### Statistically Supported LLM Ingredient and Recipe Data Collection in Computational Nutrition Authors: James Izzard, Hassan Eshkiki, Fabio Caraffini Published: 2026-07-25 Categories: cs.IR, cs.AI, cs.ET arXiv: https://arxiv.org/abs/2607.23273v1 Computational nutrition needs precise ingredient data, but current databases are incomplete, inconsistent, and built for human reference rather than automated reasoning. LLMs could help fill these gaps, but single-pass outputs are unreliable and can introduce silent errors into downstream computation. We present a quality-controlled LLM pipeline for ingredient data acquisition that combines robust statistical estimation, domain-specific invariant checks, and a web-fetch fallback. An illustrative Heap's Law fit to 233 recipes suggests that unique-ingredient growth is sub-linear and front-loaded: the projected ratio of unique ingredients to recipes falls from 1.74 at 100 recipes to 0.19 at 5,000. For each ingredient attribute, repeated LLM queries are treated as samples from a model-induced answer distribution, and we apply robust point estimators and normalised confidence scores across numerical, Boolean, multiple-choice, open categorical, and optional integer types. An invariant guard layer enforces nutritional and logical self-consistency within each ingredient record. Minor numeric inconsistencies are reconciled via a linear program that minimises worst-case percentage deviation while preserving semantic zeros, and major violations are escalated to web-evidence-grounded repair, then human review only if that fails. On a curated 30-ingredient reference set, the pipeline achieves 98.4% exact match on nutrient flags and cuts median absolute percentage error on nutrient ratios from 31.9% for the median-aggregated baseline to 10.1%, a reduction of 21.8 percentage points, at an API cost of about $1 per ingredient. This frames LLM-assisted database construction as a controlled data-engineering workflow that makes uncertainty operational rather than discarding it. --- ### What CLIP Knows but Cannot Say: Recovering Negation from Frozen Intermediate Features Authors: Chen-Yi Lu, Yueh-Shao Chen, Somali Chaterji Published: 2026-07-25 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2607.23271v1 Contrastive vision-language models such as CLIP map semantically opposite phrases (e.g., "a dog" vs. "not a dog") to nearly identical embeddings, rendering them insensitive to negation. We attribute this failure to a phenomenon we call Representational Collapse: by tracking compositional divergence and visual alignment across the CLIP text encoder, we show that middle layers build compositional syntax, but the final layers collapse this structure as visual alignment rises, producing a syntax-blind final representation. To recover the lost negation signal without altering pretrained weights, we propose PeakPatch, a lightweight post-hoc correction system that intercepts the encoder at its compositional peak. An Embedding Correction Network (ECN) uses cross-attention to extract a negation-specific signal from the peak layer, anchored to a stable baseline, and predicts a deviation vector that re-injects the lost syntax into the final-layer embedding space. A complementary Score Correction Network (SCN) predicts bounded scalar score offsets for discriminative tasks. Both modules are trained jointly end-to-end while all CLIP parameters remain frozen, adding only 5.2M parameters (3.5% of the backbone) and preserving the standard cosine similarity interface. On NegBench, PeakPatch achieves 74.3% on COCO MCQ (+35.1 over CLIP, +17.8 over the best encoder fine-tuning method) and 65.5% on VOC MCQ, while outperforming all fine-tuning baselines on fully out-of-distribution negation retrieval despite training only 3.5% of the parameters. The corrected embeddings also transfer to text-to-image generation (+18.4 negation score) and generalize across ViT-B/32, ViT-L/14, and SigLIP backbones. Project URL: https://stevencylu.github.io/PeakPatch/. --- ### WaveZip: Wavelet-Driven Space-Time Decoupling for Video Token Condensation Authors: Yuhui Zeng, Wang Chen, Jinfa Huang, Tianyu Xie, Yongdong Luo, Jiayi Ji, Xiawu Zheng, jiebo Luo Published: 2026-07-25 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.23265v2 Existing Large Vision-Language Models (LVLMs) struggle with long-form video understanding due to the quadratic computational cost of visual tokens. While recent efficient methods attempt to compress tokens via hard pruning or uniform merging, they operate strictly in the spatial feature domain, where robust structural context and discriminative semantic details are inherently entangled. In this work, we propose WaveZip, a joint signal-frequency-domain framework for efficient video inference. Driven by the insight that temporal redundancy resides in low-pass approximation scales while spatial saliency strongly correlates with high-frequency components, WaveZip leverages Discrete Wavelet Transforms (DWT) to disentangle these signals. Temporally, it employs 1D DWT to analyze query-frame relevance, and the resulting high-frequency coefficients are further gated by inter-frame differences, with both signals jointly driving the dynamic allocation of a precise frame-level token budget. Spatially, a 2D DWT decomposes features into low-frequency approximations and high-frequency detail components, where the high-frequency coefficients are modulated within query-salient regions to regulate spatial reconstruction. Importantly, WaveZip requires no task-specific training and can be seamlessly integrated into off-the-shelf LVLMs to boost inference efficiency. Extensive experiments on long video understanding benchmarks demonstrate that WaveZip retains 99.6% of the full performance under an extreme 10x compression ratio, consistently outperforming state-of-the-art methods. --- ### X-Stage: Modeling Post-Issue Backpressure in GPU Communication--Computation Fusion Authors: Jianwen Xian, Zhiyuan Xu, Yuchen Li, Ziliang Lai, Kang He, Zhen Huang, Aichen Feng, Jinyan Chen, Yilin Zhang, Qinqin Chen, Julien Lai, Chengru Song Published: 2026-07-25 Categories: cs.DC, cs.AI arXiv: https://arxiv.org/abs/2607.23264v2 Fine-grained, device-initiated communication allows fused GPU kernels to issue remote stores directly from their compute pipelines, a pattern increasingly used in expert parallelism (EP), tensor parallelism (TP), and Ulysses-style sequence parallelism (UP). Existing designs reason about where communication is issued and when remote data becomes ready, but lack a quantitative model of the sender-side interval after a remote store is accepted and before it becomes visible at the destination. This interval determines whether communication remains decoupled from computation or backpressures it. We identify X-Stage, a software-visible post-issue stage with finite decoupling. Downstream pressure can dissipate while the issuer resumes useful work, whereas sustained injection consumes X-Stage headroom and eventually stalls the compute pipeline. We characterize this behavior and build a calibrated model that predicts whether remote-store arrivals accumulate backpressure or recover during intervening computation. Guided by the model, we reshape bursty arrivals when they would exhaust X-Stage headroom and exploit natural compute windows when headroom can recover concurrently. Evaluation across representative EP, TP, and UP workloads shows up to 1.62x fused-kernel, 1.75x end-to-end, and 1.43x sender-visible speedup, respectively. Microbenchmarks further validate the model's predictions of backlog accumulation, recovery, and sender-side backpressure. --- ### X-Stage: An Overlooked Pipeline Stage for Communication-Computation Overlap in DiT Inference Authors: Jianwen Xian, Zhiyuan Xu, Yuchen Li, Ziliang Lai, Kang He, Zhen Huang, Aichen Feng, Jinyan Chen, Yilin Zhang, Qinqin Chen, Chengru Song Published: 2026-07-25 Categories: cs.DC, cs.AI arXiv: https://arxiv.org/abs/2607.23264v1 Fine-grained, device-initiated communication lets persistent GPU kernels in distributed diffusion transformer (DiT) inference issue remote stores and overlap data movement with Tensor Core computation. Existing systems schedule when communication is issued and when received data becomes consumable, but omit post-issue progress before remote-visible completion, making sender backpressure hard to predict. We identify X-Stage, a software-visible post-issue pipeline stage. Measurements on an eight-GPU node with a recent NVIDIA architecture show that short remote-store bursts drain as the issuer resumes work, whereas sustained injection exhausts finite outstanding capacity and delays later issues. A lightweight Burst-Gap model parameterized by backpressure-free issue time, effective drain rate, and outstanding capacity predicts issue overhead, recovery between bursts, and the onset of backpressure. Guided by the model, we redesign two communication-computation fused kernels. For DeepGEMM MegaMoE, interleaving Linear-1 and Linear-2 work across expert waves places computation between concentrated remote-store bursts, yielding a 1.18x geometric-mean and 1.62x maximum kernel speedup over the Expert-Wave baseline across 84 configurations. For Ulysses sequence-parallel attention, tile-granular fusion of the post-attention All-to-All with FlashAttention lets an output-tile owner issue remote stores and resume computation without a dedicated communication warp or streaming multiprocessor. FlashAttention-3 and FlashAttention-4 reach maximum sender-visible speedups of 1.43x and 1.42x over serial execution, and at long sequences their steady-state times approach those of FlashAttention alone. These results establish post-issue progress as a measurable scheduling lever for shaping bursts, avoiding backpressure, and hiding sender-side overhead. --- ### SeekJudge: A Practical Reward Framework for Reinforcement Learning in Computer-Use Agents Authors: Yang Wan, Zhenhao Zhang, Jierui Wang, Xiangxin Zhou, Linchao Zhu Published: 2026-07-25 Categories: cs.AI arXiv: https://arxiv.org/abs/2607.23263v2 Deciding whether a trajectory actually fulfills its instruction determines how we measure computer-use agents on long-horizon graphical-user-interface tasks and how we train them with reinforcement learning (RL). This judgment has long relied on rule-based evaluation. Rule-based evaluation often disagrees with human intention and becomes outdated when an app updates or its online content drifts. Existing model-based judges attempt to address these problems, but their judging accuracy remains limited. We propose the \textbf{SeekJudge} framework, whose four agents, a Condense, a Ground, a Seek and an Analyze agent, reach a verdict through a Seek--Analyze loop over the trajectory. To train a specialized model on densely labeled trajectories, we propose a seed-driven distillation pipeline that expands a few human-labeled seed trajectories into $10$K such trajectories. To evaluate step-level judgments, we build CUAStepBench, a human-annotated benchmark that pairs trajectory verdicts with dense step labels. Beyond accuracy, SeekJudge costs far less than a closed-source large model. We further propose rollout overlap, a reward-server design that reduces the overhead of a reward model in RL training. To our knowledge, SeekJudge is the first model-based reward to match or surpass native rule-based reward in online RL, measured by downstream success rate on held-out RL test goals. In offline judging, SeekJudge-9B also exceeds rule-based evaluation by $9.5$ F1 on AgentRewardBench. --- ### SeekJudge: A Practical Reward Framework for Reinforcement Learning in Computer-Use Agents Authors: Yang Wan, Zhenhao Zhang, Jierui Wang, Linchao Zhu Published: 2026-07-25 Categories: cs.AI arXiv: https://arxiv.org/abs/2607.23263v1 Deciding whether a trajectory actually fulfills its instruction governs how we measure computer-use agents on long-horizon graphical-user-interface tasks and how we train them with reinforcement learning. This judgment has long relied on rule-based evaluation, which struggles to align with human intention and goes stale when an app updates or its online content drifts. Existing model-based judges attempt to address these problems but still leave a performance gap to the rule-based evaluation. We propose the \textbf{SeekJudge} framework, in which four role-specialized agents, a Condense, a Ground, a Seek and an Analyze agent, reach a verdict through a Seek--Analyze loop over the trajectory. A seed-calibrated distillation pipeline trains one specialized $9$B model to serve as the shared backbone for all four agents. Measured by downstream success rate on held-out RL test goals, SeekJudge is the first practical model-based reward to match or surpass native rule-based supervision in online RL. Beyond accuracy, SeekJudge provides step-level judgments, runs far cheaper than a closed-source large model, and keeps a small per-call context that scales to much longer trajectories. We further contribute a general architectural improvement to the reward server that speeds up judging in RL. Together these make model-based reward a practical drop-in for rule-based supervision in CUA reinforcement learning. --- ### CAPT: A Multi-task Continuous Autoregressive Transformer enabling Cross-dataset and Cross-species Transfer for Calcium Population Dynamics Authors: Xinhong Xu, Yimeng Zhang, Yuanlong Zhang Published: 2026-07-25 Categories: cs.AI, cs.LG arXiv: https://arxiv.org/abs/2607.23258v2 Large-scale calcium imaging has created an opportunity to build foundation-style models for neural population dynamics, but a central question remains unresolved: \textbf{whether a model pretrained on one collection of recordings can generalize to new datasets, experimental paradigms, and even species.} Existing approaches are often designed for specific tasks and evaluated on a single dataset, making it unclear whether their learned representations are reusable for new calcium trace datasets. To tackle this gap, we present \textbf{CAPT}, a \textbf{C}ontinuous \textbf{A}utoregressive \textbf{P}opulation \textbf{T}ransformer for calcium population dynamics. CAPT models continuous calcium traces directly through a continuous patch tokenization strategy and is trained autoregressively, enabling end-to-end pretraining and adaptation to diverse downstream tasks. We first pretrain CAPT on a large-scale mouse calcium imaging dataset and evaluate its transferability across independent mouse, larval zebrafish, and \textit{C. elegans} datasets collected by different laboratories. In these transfer settings, the pretrained backbone is frozen and only adaptation modules are updated. Across neural population forecasting and behavior decoding tasks, CAPT consistently outperforms specialized and general-purpose baselines. Alongside predictive performance, multimodal analyses using NeuroPAL annotations in \textit{C. elegans} datasets show that CAPT embeddings form a shared functional space across datasets and capture anatomical cell-identity-related structure. These results suggest that the continuous autoregressive modeling opens up possibilities for a simple route towards general-purpose neural foundation models for calcium imaging, which can generalize across datasets, experimental paradigms, and species. Code is available at https://github.com/TSuXinH/CAPT. --- ### FedTaste: Topology-Aware Structural Transfer for Multimodal Federated Learning with Missing Modalities Authors: Haochen Liang, Jie Zhang, Hideya Ochiai Published: 2026-07-25 Categories: cs.MM, cs.LG arXiv: https://arxiv.org/abs/2607.23245v1 Multimodal Federated Learning is often challenged by arbitrary modality missingness and Non-IID data distributions, which lead to severe representation drift and hinder effective collaboration across clients. Existing methods typically rely on generative imputation, external auxiliary data, or isolated unimodal training to bridge modality gaps, often incurring substantial communication and computational costs as well as potential privacy risks. To address these limitations, we propose FedTaste, a parameter-efficient framework for topology-aware structural transfer in Multimodal Federated Learning with missing modalities. Instead of aligning fragile first-order features, FedTaste focuses on more stable group-level semantic relations. Specifically, FedTaste leverages frozen foundation models to extract a joint multimodal topology from full-modality clients, which is then consolidated by the server into a global structural blueprint. To adapt clients with missing modalities, we introduce Modality-Adaptive Structural Prompts together with spectral consistency regularization, enabling lightweight branch-specific adaptation that aligns local partial representations with the shared blueprint. In this way, FedTaste avoids explicit modality imputation while preserving shared semantic structure across clients. Extensive experiments demonstrate that FedTaste consistently achieves superior performance across multiple datasets and challenging Non-IID settings, while substantially reducing communication overhead compared with existing methods. --- ### Characterisation of Density-based FM generation methods in the context of Information Fusion Authors: Yanhao Huang, Christian Wagner Published: 2026-07-25 Categories: cs.AI arXiv: https://arxiv.org/abs/2607.23243v1 Fuzzy Integral (FI) based aggregation provides a powerful mechanism for nuanced aggregation, for example, in ensemble approaches or decision-level fusion more generally. The main challenge of this approach is the appropriate parametrization of the Fuzzy Measure (FM), which captures the worths of the individual components--and their combinations--which are being fused. Here, widely used approaches including the Sugeno-$λ$ and Decomposable FMs, parametrize the FM by extrapolating from the densities, i.e. the weights associated with individual sources, while respecting the FM's monotonicity constraint. This paper articulates that this information is, in general, insufficient to uniquely identify a discrete FM; but shows how an interval-valued FM can indeed be determined uniquely. We proceed to show how the incorporation of additional information beyond the above, such as the choice of a specific FI and a dataset, then allows for obtaining even more specific interval-valued FMs. In practice, establishing the quality of an empirically determined FM is not trivial. To help address this, we show how the likelihood with which a resulting interval FM encompasses the `ideal', i.e. the commonly intangible, best, or ground-truth numeric FM, can be determined, producing a confidence interval at a given confidence level. Finally, based on a series of experiments, we demonstrate empirically that the Choquet FI output based on this FM can also be regarded as the confidence interval for the `ideal' information fusion result, providing a novel means to characterize FI fusion outcomes a priori and charting a pathway for future research. --- ### IndicTalk: A Large-Scale Persona-Based Multilingual Conversational Corpus for Indic Languages Authors: Sahil Deepak Gawande, Mayank Singh Published: 2026-07-25 Categories: cs.CL, cs.LG arXiv: https://arxiv.org/abs/2607.23242v1 Large Language Models (LLMs) have transformed conversational AI, yet high-quality multilingual code-mixed dialogue resources remain scarce, particularly for Indic languages where speakers naturally alternate between English and their native language in both native-script and Romanized forms. We present IndicTalk, one of the largest multilingual Indic code-mixed conversational corpora, comprising over 13,28,604 event-grounded multi-turn conversations across 18 language varieties covering 9 Indic languages. The corpus is generated through a fully automated pipeline that combines real-world news grounding, persona-conditioned dialogue generation using multilingual LLMs, and automatic quality validation. Extensive linguistic, automatic, and human evaluations demonstrate that IndicTalk produces fluent, coherent, and naturally code-mixed conversations across both script variants. We will release IndicTalk to support the development and evaluation of multilingual conversational AI for underrepresented Indic languages. The dataset is available at: https://huggingface.co/datasets/LingoIITGN/IndicTalk . --- ### SARATR-X-v2: Scale-Aware Structural Pre-Training for SAR Foundation Models Authors: Weijie Li, Yafei Song, Yongxiang Liu, Bowen Peng, Jie Zhou, Jingyuan Xia, Wei Yang, Tianpeng Liu, Zhen Liu, Li Liu Published: 2026-07-25 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.23238v4 Masked image modeling has become a dominant paradigm for SAR pre-training, yet the design of the reconstruction target remains fundamentally unsettled. This article argues that a SAR pre-training target should satisfy two conditions to produce transferable representations: (i) physics-grounded stability, i.e., approximate invariance of the target operator to multiplicative speckle inherent in coherent imaging; and (ii) semantic scale compatibility, i.e., coverage of the heterogeneous spatial scales that downstream tasks demand. These two conditions are individually achievable but jointly difficult: physics-grounded stability favors fixed operators, while semantic scale compatibility favors data-driven composition. To this end, SARATR-X-v2 reconciles both within a single design. The target is constructed through fixed structural extractors spanning six receptive fields, from blind-spot local aggregation to directional log-ratio region contrast, and fused via learnable weights into one unified supervision signal for masked reconstruction. On twelve SAR benchmarks across classification, detection, and segmentation, SARATR-X-v2 achieves state-of-the-art transfer performance. Under synthetic speckle variation, the proposed target reduces perturbation drift in the learned supervision by nearly two orders of magnitude relative to pixel-space supervision. Taken together, these results support physics-grounded stability and semantic scale compatibility as a principled framework for pre-training target design under coherent imaging, and suggest that effective SAR pre-training is not about reconstructing more signal, but about reconstructing the right structural target. --- ### SARATR-X-v2: Scale-Aware Structural Pre-Training for SAR Foundation Models Authors: Weijie Li, Yafei Song, Yongxiang Liu, Bowen Peng, Jie Zhou, Jingyuan Xia, Wei Yang, Tianpeng Liu, Zhen Liu, Li Liu Published: 2026-07-25 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.23238v1 Masked image modeling has become a dominant paradigm for SAR pre-training, yet the design of the reconstruction target remains fundamentally unsettled. This article argues that a SAR pre-training target should satisfy two conditions to produce transferable representations: (i) physics-grounded stability, i.e., approximate invariance of the target operator to multiplicative speckle inherent in coherent imaging; and (ii) semantic scale compatibility, i.e., coverage of the heterogeneous spatial scales that downstream tasks demand. These two conditions are individually achievable but jointly difficult: physics-grounded stability favors fixed operators, while semantic scale compatibility favors data-driven composition. To this end, SARATR-X-v2 reconciles both within a single design. The target is constructed through fixed structural extractors spanning six receptive fields, from blind-spot local aggregation to directional log-ratio region contrast, and fused via learnable weights into one unified supervision signal for masked reconstruction. On twelve SAR benchmarks across classification, detection, and segmentation, SARATR-X-v2 achieves state-of-the-art transfer performance. Under synthetic speckle variation, the proposed target reduces perturbation drift in the learned representation by nearly two orders of magnitude relative to pixel-space supervision. Taken together, these results establish physics-grounded stability and semantic scale compatibility as a principled framework for pre-training target design under coherent imaging, and suggest that effective SAR pre-training is not about reconstructing more signal, but about reconstructing the right structural target. --- ### Context-Aware Concept Distillation for Trustworthy Flood Prediction Authors: Eli Levinkopf, Efrat Morin, Claudia V. Goldman Published: 2026-07-25 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2607.23237v1 Effective flood risk management relies on accurate forecasting, yet the "black box" nature of stateof-the-art Deep Learning models creates a barrier to trust and accountability in high-stakes public safety decisions. While existing Explainable AI (XAI) methods offer local attributions, they fail to provide the verifiable, operationally meaningful causal narratives required by disaster response authorities. To address this societal challenge, we propose Context-Aware Concept Distillation (CACD), a framework developed in collaboration with domain experts to distill opaque LSTMs into interpretable, hydrology-aware surrogate models. We introduce an unsupervised pipeline to discover a "Hydrological Language" and a Residual Hypernetwork that dynamically modulates these concepts based on static basin characteristics. Evaluated on 5,203 basins globally, our model achieves high fidelity (Median NSE 0.70), significantly outperforming black-box baselines (e.g., Multi Layer Perceptrons) on unseen future data. By demonstrating that human-interpretable concepts are sufficient to reconstruct flood dynamics, this work balances AI accuracy with the transparency required for responsible environmental decision-making. --- ### FedSLIM: Privacy-Preserving Federated MDL-Based Descriptive Pattern Mining Across Data Silos Authors: Samar Samir Khalil, Noha S. Tawfik, Marco Spruit Published: 2026-07-25 Categories: stat.ML, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2607.23236v1 Federated learning has achieved considerable success for predictive modelling, yet federated descriptive analytics remains largely unexplored. Existing federated pattern mining approaches are predominantly support-based and do not optimise a principled global objective such as Minimum Description Length (MDL). We introduce FedSLIM, the first federated MDL-based framework for descriptive pattern mining. Building on the SLIM principle, FedSLIM enables collaborative optimisation of compact pattern models across distributed databases without sharing raw transactions. We propose two complementary variants that balance privacy, communication, and optimisation fidelity under different deployment assumptions. To evaluate federated MDL mining, we introduce fidelity and discovery-oriented metrics that quantify agreement with a centralised baseline and assess recovery of globally informative patterns. Experiments on multiple real-world datasets under IID and non-IID partitioning show that both variants preserve high-quality compression structure while requiring orders of magnitude less search than the centralised baseline. We further reveal a local-global discovery gap in distributed MDL mining, where globally compressive patterns may be undiscoverable through isolated local optimisation. Both variants recover globally informative patterns absent from all standalone local models, demonstrating the benefits of federated optimisation beyond independent local mining. These results establish federated MDL mining as a practical foundation for privacy-preserving descriptive analytics across distributed data silos. --- ### A Reconstruction-Based Framework for Caption Evaluation Beyond Reference Captions Authors: Zhijiang Tang, Jiaxin Qi, Kaihua Tang, Yuhua Zheng, Jianqiang Huang Published: 2026-07-25 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.23235v1 Image captioning is a primary task in vision--language research, yet assessing how faithfully a caption preserves image semantics without relying on reference captions remains unsettled. Prevailing evaluations rely on human-annotated references, whose content reflects annotator intent and captioning proficiency. In this paper, we study a reconstruction-based principle for caption evaluation: a caption is as good as its capacity to enable reconstruction of the original image. However, because captioning inherently compresses visual information, it is impossible to recover all details, and pixel-wise comparison between reconstructed and source images is neither feasible nor meaningful. Through our in-depth analysis of the nature of captions, whose fundamental purpose is to transmit the semantic content of an image, we propose a revised principle: a caption is as good as its capacity to enable a reconstruction that is semantically equivalent to the original. To assess semantic equivalence, we test whether the reconstruction matches the original image across a suite of downstream vision--language tasks, yielding a reference-free, task-conditioned caption score. We characterize component-dependent limitations and introduce the lower-cost Captioning Turing Test Dataset (CTTD) surrogate. --- ### From Score Learning to Discretized Sampling: An End-to-End Generalization Analysis of Diffusion Models Authors: Jinshu Huang, Yiming Jiang, Chunlin Wu Published: 2026-07-25 Categories: cs.LG, math.ST arXiv: https://arxiv.org/abs/2607.23226v1 Despite the empirical success of score-based diffusion models, a complete theoretical understanding of how finite-sample learning, network parameterization, and numerical discretization jointly dictate generative quality remains underdeveloped. Existing sampling analyses often evaluate the generative performance conditional on an oracle score or a pre-specified error threshold. In this work, we establish a unified convergence and generalization framework for score-based diffusion models parameterized by practical ResNet-type architectures. We analyze the generalization and convergence properties from the practical finite-sample, discrete-time learning problem of the score function to the ideal continuous-time, population-level objective. Based on the generalization result of the learning problem of score function, we analyze the sampling process induced by the learned score function and provide an end-to-end total variation distance estimate for the generated terminal distribution. This estimate explicitly decomposes the overall generative error into four interpretable components: the truncation error of the forward process, the reverse-time discretization error, the generalization error incorporating both finite data and forward-time discretization, and the training optimization gap. Our results quantitatively characterize how the training sample size, temporal discretization grids, and optimization accuracy jointly control the final fidelity of samples generated by diffusion models. --- ### ParasGB: A Graph Benchmark Suite for Parasitic Estimation on AMS Circuits Authors: Jiajun Zou, Jiawei Liu, Ao Liu, Junnong Tian, Yibin Zhang, Chengjie Liu, Yuxi Wang, Shan Shen, Wenhua Gu, Jun Yang, Wenjian Yu Published: 2026-07-25 Categories: cs.LG arXiv: https://arxiv.org/abs/2607.23225v1 As chip manufacturing processes advance to deep submicron nodes, parasitic interconnect effects increasingly dominate the performance of analog and mixed-signal (AMS) circuits and often lead to costly layout iterations. This makes early-stage estimation of parasitic capacitance and resistance important for parasitic-aware design exploration before full physical implementation. However, progress on GNN-based parasitic modeling has been hindered by the lack of public, high-fidelity RC benchmarks that support reproducible evaluation. To address this gap, we introduce ParasGB, the first open-source benchmark suite for pre-layout parasitic parameter prediction on circuit graphs. ParasGB provides large-scale, heterogeneous RC networks extracted with commercial EDA tools from tape-out-proven designs, together with a unified evaluation protocol covering node-level ground capacitance, edge-level resistance, and edge-level coupling capacitance. Within this framework, we benchmark diverse GNN architectures using a standardized training pipeline and expose challenges such as extreme label imbalance, long-tailed parasitic distributions, and strong structural heterogeneity. By establishing a physically grounded and standardized benchmark for early-stage parasitic prediction, ParasGB provides an open platform for reproducible research on circuit graph learning and parasitic-aware model development. All datasets, preprocessing scripts, and configurations are publicly available in our code repository https://github.com/ShenShan123/ParasGB.git. --- ### BoneAgeTW2: Automated Skeletal Maturation Assessment via the Tanner-Whitehouse 2 Method, Deep Learning, and Clinical Report Generation with Distribution Curves Authors: Juan Manuel Castillo Pinto Published: 2026-07-25 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2607.23224v1 We present BoneAgeTW2, the first fully open-source system to automate the complete Tanner-Whitehouse 2 (TW2) clinical protocol for skeletal maturity assessment end-to-end. The system employs YOLOv8 for precise detection and localization of the 20 TW2 hand bones from radiographic images, and an EfficientNet-B3 backbone with 20 independent classification heads to assign maturation stages (A-I) to each bone simultaneously. From these predictions, the system automatically generates clinical PDF reports including interactive Gaussian distribution curves for all 20 bones, enabling direct comparison with population norms. The model is trained on the public RSNA Pediatric Bone Age Challenge dataset (12,611 hand radiographs) using a pseudo-labeling strategy to derive per-bone stage labels from global bone age annotations. The full codebase is publicly available at https://github.com/jmmana/BoneAgeTW2. --- ### An Ontology for Machine Learning Interatomic Potentials Authors: Daniel Hernández, Jong Hyun Jung, Yuji Ikeda, Yongliang Ou, Pranav Kumar, Tom Schächtel, Wenchuan Liu, Xin Li, Xi Zhang, Xiang Xu, Lifang Zhu, Fritz Körmann et al. Published: 2026-07-25 Categories: cs.AI, cs.DB, cs.DL arXiv: https://arxiv.org/abs/2607.23219v1 Machine learning interatomic potentials (MLIPs) approximate quantum-mechanical energies and forces---conventionally computed by density functional theory (DFT) or wave-function methods---at a fraction of the cost. The field encompasses a growing ecosystem of algorithms, training datasets, hyperparameters, and target materials, yet the metadata needed to systematically compare, reproduce, and build upon MLIP studies remains scattered across papers, scripts, and ad-hoc file formats. We present the MLIPs ontology, an OWL 2 DL ontology that captures the concepts needed to describe MLIP methods, their hyperparameters, training datasets with DFT provenance, and published benchmarks. The ontology is organized into three modules---Method, Training Data, and Benchmark---and connects existing ontologies in materials science (MDO, CMSO/ASMO) and machine learning (ML-Schema), complementing dataset-side schemas such as Croissant. It declares 27 formal axioms enforcing data completeness and consistency, including property chains that link trained models to their methods and training data. We demonstrate the ontology through a running example based on Moment Tensor Potentials and evaluate it through competency-question execution on a 20-paper seeded knowledge graph, OWL reasoning, and comparison with existing ontologies. --- ### Counterfactual Motion Reliability Learning for Robust UAV Tracking Authors: Yuehai Chen Published: 2026-07-25 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.23209v1 Infrared unmanned aerial vehicle (UAV) tracking is challenging because the target is often small, low-contrast, and easily confused with thermal distractors or cluttered backgrounds. Recent Transformer-based trackers have achieved promising performance by learning strong appearance representations, but their responses can still be dominated by background structures when the target appearance is weak or ambiguous. A natural solution is to introduce temporal motion cues. However, in infrared UAV tracking, motion cues are not always reliable: camera jitter, dynamic backgrounds, sensor noise, and target disappearance may produce temporal variations that are stronger than the true target motion. Therefore, the key challenge is not simply how to use motion, but how to distinguish target-consistent motion from background-induced pseudo motion. To this end, we propose CMRTrack, a counterfactual motion reliability learning framework for robust infrared UAV tracking. CMRTrack first extracts temporal evidence from adjacent search regions using a lightweight motion evidence encoder. During training, a counterfactual target-erased history branch is introduced to construct hard motion references, encouraging the motion encoder to learn reliable target-consistent motion rather than arbitrary temporal changes. The learned motion evidence is then incorporated into a one-stream tracking framework through motion-guided token modulation and reliability-aware score fusion, enabling adaptive feature enhancement and response refinement. Extensive experiments on Anti-UAV410 demonstrate that CMRTrack consistently outperforms representative state-of-the-art trackers and significantly improves the OSTrack baseline, with ablation studies and qualitative analysis verifying the effectiveness of the proposed counterfactual motion reliability learning. --- ### Low-Latency Turn-Taking via Context-Aware Preface Generation in a Real-World Dialogue Robot Authors: Yuki Okafuji, Koji Inoue, Yoshiki Ohira Published: 2026-07-25 Categories: cs.RO, cs.CL, cs.SD arXiv: https://arxiv.org/abs/2607.23204v1 Large language model (LLM)-based dialogue systems suffer response delays because generation begins only after final speech recognition. While fixed fillers are a workaround, they become unnatural over time. We propose a two-stage incremental framework that decouples prefatory-response preparation from speech onset. Once user intent becomes predictable, an intent readiness detector triggers LLM-based generation of a short prefatory response. Concurrently, a voice activity projection (VAP) model determines when to deliver it. Through a field experiment with a route-guidance robot in a shopping mall, we evaluated three conditions: no-filler, fixed-filler, and contextual-preface. Both fixed-filler and contextual-preface significantly reduced initial response latency relative to no-filler. Relative to fixed-filler, contextual-preface had significantly longer initial response latency but a significantly shorter initial-to-main gap. Exploratory ratings showed no significant differences. These results indicate a timing trade-off. --- ### Variance-Preserving Orthogonal Selection (VPOS): Greedy Feature Selection via Orthogonal Deflation in PCA Loading Space Authors: Baran Koseoglu, Berrin Yanikoglu Published: 2026-07-25 Categories: cs.LG arXiv: https://arxiv.org/abs/2607.23198v3 We present Variance-Preserving Orthogonal Selection (VPOS), an unsupervised feature-selection method that performs sequential orthogonal deflation in the variance-weighted principal component analysis (PCA) loading space $\mathbf{V}_d\mathbfΛ_d^{1/2}$. After each feature is selected, its loading direction is projected out of all remaining candidates, so subsequent selections cover complementary directions of the rank-$d$ covariance approximation while returning original variables. We establish rank-reduction guarantees and a determinant-growth interpretation, and distinguish VPOS from greedy selection on raw data, unweighted eigenvector pivoting, Principal Feature Analysis (PFA), and Principal Variable Selection (PVS). Experiments enforce $k\leq d$, tune method-specific parameters on validation observations, and evaluate on unseen outer folds. Across seven labelled benchmarks, VPOS improves held-out normalised reconstruction error over matched PCA without deflation on every dataset, with reductions of 1--78%. It obtains the lowest mean reconstruction error on Wine, Breast Cancer, and MNIST and is within 1.7% of the lowest error on CIFAR-10 and HighDim. On CIFAR-10, VPOS is approximately 24$\times$ faster than the closely related PVS baseline while incurring a 1.7% reconstruction gap. These results establish VPOS as an efficient covariance-coverage method, particularly when correlated high-dimensional data must be represented by a small set of identifiable original variables. --- ### Domain-Prior-Regularized Graph Modeling for Anomaly Detection in Cyber-Physical Systems Authors: Youngseok Hwang, Joonsung Kwon, Geonwoo Lee, Hyunwoo Park Published: 2026-07-25 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2607.23197v1 Anomaly detection on multivariate sensor time series is critical for industrial monitoring of cyber-physical systems (CPS), where even subtle deviations from normal behavior can indicate process disruption. Recent graph-based approaches have made significant progress, but they often struggle in small-scale physical systems with scarce labeled anomalies and limited normal data. In such settings, graph-based models tend to capture spurious correlations and produce unstable sensor topologies. We propose DPR-GM (Domain-Prior-Regularized Graph Modeling), a forecasting-based framework that incorporates system design knowledge into graph construction. DPR-GM leverages a large language model (LLM) to extract directed physical couplings between sensor pairs from system documentation, which are encoded as a binary domain adjacency matrix serving as a structural gate over sensor relations. This gate is then modulated by Pearson correlations estimated from normal training data. The anomaly score is further weighted by sensor-level reliability derived from the coefficient of variation. All graph and weighting components are fixed prior to training and add no learnable parameters. On the SKAB benchmark, DPR-GM outperforms graph-based, statistical, and deep learning baselines across F1, AUROC, and AUPRC, showing that domain-structured graph priors are a practical alternative to fully learned topologies in data-scarce CPS. --- ### Out-of-Length Scene Text Recognition: A Two-Axis Diagnosis and a Training-Free Fix Authors: Zobeir Raisi, John Zelek Published: 2026-07-25 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.23194v1 Scene Text Recognition (STR) models are trained almost exclusively on word crops of at most 25 characters, yet real deployments (signage, product labels, dense captions) require reading much longer text. This paper diagnoses that failure and then closes it. The diagnosis separates out-of-length failure into two simultaneously extrapolating axes (the encoder's width axis and the decoder's time axis) and shows that encoder width, not decoder length, is the dominant failure mode. Representation-side fixes bring only partial relief: training-free rotary rescalings recover at most 2-4 points of character error rate (CER), and a weighted fine-tuning recipe recovers 6-8 points while improving standard-benchmark accuracy, yet word accuracy on the Long Text Benchmark (LTB) stays near zero, because the residual gap lies in the decoding mechanism rather than the representation. We then close that gap at inference time, on an unmodified word-level checkpoint: the long image is sliced into overlapping crops at the model's training width, each decoded independently and in-distribution, and the reads stitched by geometry-anchored edit-distance alignment. This procedure reaches 42.79-43.05% bucket-average word accuracy on LTB across two base checkpoints, matching the published state of the art (41.57%) and beating it by 11-12 points on the hardest bucket, at wall-clock parity with plain decoding; applied unchanged to the public PARSeq checkpoint it reaches 47.11%. Once chunking is applied fine-tuning no longer helps: the decoding-side fix alone matches purpose-built architectures. We release the diagnosis harness and implementation. --- ### OmniScope: Modality-Decoupled Token Compression for Omnimodal Large Language Models Authors: Jinsen Su, Yongdong Luo, Yuexiao Ma, Yibo Hu, Meiguang Jin, Xiawu Zheng Published: 2026-07-25 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.23193v3 Existing token compression methods for omnimodal large language models typically rely on one modality to determine what to retain in the other. We show that this assumption often breaks down: for the same query, audio and video relevance often peaks at different moments. This cross-modal salience mismatch makes unidirectional guidance prone to discarding answer-critical cues under aggressive compression. We propose OmniScope, a training-free token compression framework that uses the query as a shared semantic anchor while estimating relevance separately for audio and video. OmniScope allocates modality-specific token budgets, prunes visual tokens with an anchor-delta strategy that preserves both global context and temporal changes, and merges audio tokens within each second to reduce redundancy while maintaining temporal continuity. Across four audio-video benchmarks and two Qwen2.5-Omni model scales, OmniScope achieves the best average accuracy across all compression settings. At 25% overall token retention, it delivers up to 3.53x prefill speedup and more than 15% GPU memory reduction, with only a 0.35-point drop in average accuracy. These results suggest a simple design principle for OmniLLM inference: share the query across modalities, but not the salience estimates. The code is available at https://github.com/MAC-AutoML/OmniScope. --- ### Data-Driven Diffusion Processes on Differential Forms via the Projected Ambient Connection Laplacian Authors: Alvaro Almeida Gomez, Jorge Duque Franco Published: 2026-07-25 Categories: math.NA, cs.LG arXiv: https://arxiv.org/abs/2607.23192v1 We develop a data-driven approximation of the projected ambient connection Laplacian acting on differential forms over smooth Riemannian manifolds sampled by point clouds. The proposed construction extends the classical framework of diffusion maps and Vector Diffusion Maps from scalar functions and tangent vector fields to differential forms of arbitrary degree. Our approach is based on a novel representation of differential forms as alternating differential arrays obtained through an extension of the classical musical isomorphism. This representation enables the construction of a matrix-valued diffusion operator that approximates the projected ambient connection Laplacian directly from point cloud data without requiring a mesh or simplicial complex. The proposed discretization admits the asymptotically optimal kernel bandwidth scaling inherited from diffusion maps, leading to sharper convergence guarantees than previous data-driven approximations of the Hodge Laplacian. Building upon this operator, we derive a fully data-driven explicit Euler scheme for the heat equation on differential forms and validate the proposed methodology through numerical experiments on the unit sphere. The experiments confirm the predicted decay of the analytical solution and demonstrate the effectiveness of the proposed discretization. The proposed framework provides a natural generalization of Vector Diffusion Maps to differential forms of arbitrary degree and establishes a practical foundation for the numerical approximation of geometric partial differential equations directly from point cloud data. --- ### Wrong Design Intent Is Worse Than Never Conditioning: A Derangement-Control Diagnosis of Header Conditioning in CAD Program Completion Authors: Yang Xiao Published: 2026-07-25 Categories: cs.LG, cs.GR arXiv: https://arxiv.org/abs/2607.23191v4 Fine-tuned code LLMs are routinely conditioned on a design-intent specification, but the correctness axis of such a signal -- a wrong intent rather than an absent one -- has not been tested, and the benefit of conditioning is usually scored with the same detector that defines the signal. We study CADCON, a five-feature design-intent header prepended to CadQuery-style programs during LoRA fine-tuning of Qwen2.5-Coder-1.5B, scoring adherence with executable geometric assertions that share no code with the header-defining extractor. On a pre-registered sample of 400 deduplicated held-out programs stratified over eleven intent profiles, at 40% prefix and three seeds, a semantically wrong header degrades adherence below the never-header-trained baseline on 3/3 seeds under both tokenizations at the program level, and on 3/3 token and 2/3 text seeds at the 298 distinct model inputs they present. Wrong-header executability is not depressed relative to that baseline. A derangement control, retrained so every program receives another program's header -- holding the header marginal fixed while destroying its correlation with the program -- saw the same programs, indices and wrong headers. Its correct-to-wrong change is -0.006/+0.016/-0.003 against 0.124/0.241/0.230 for the standard model, and the interaction is significant on 3/3 seeds (p <= 5.9e-7), so the model's sensitivity to whether the header is right or wrong requires the learned mapping. The control sits below the baseline by the same margin under a correct as under a wrong header, so we claim that sensitivity and not the below-baseline level. On features the true intent lacks, the standard model realizes a feature far more often when the wrong header names it; the control does not. Ground truth itself scores only 0.567 here, the scale on which arm levels should be read. Wrong design intent is not inert: it actively misdirects generation. --- ### Wrong Design Intent Can Be Worse Than None: A Derangement-Control Diagnosis of Header Conditioning in CAD Program Completion Authors: Yang Xiao Published: 2026-07-25 Categories: cs.LG arXiv: https://arxiv.org/abs/2607.23191v2 Fine-tuned code LLMs are often conditioned on a design-intent header to steer parametric CAD generation, but whether the model reads that header's content has been tested neither under execution-level scoring nor with a causal control. We study CADCON, a five-feature design-intent header prepended to CadQuery-style sketch-extrude programs during LoRA fine-tuning of Qwen2.5-Coder-1.5B, re-scored by executable geometric assertions on the produced B-rep solid. Across three seeds and a pre-registered {0%, 40%}-prefix $\times$ {correct, wrong, masked}-header matrix -- with inference duplicate-aware over the 38 unique-program clusters a submission-stage audit found held out -- we report: (i) in conditional completion (40% prefix), a semantically wrong header degrades adherence below the no-header baseline (0.43 $\to$ 0.30/0.21 text/token) on the intents the model can render unconditioned, namely polygonal and thin. The drop is significant on 3/3 token-header seeds but 1/3 text-header seeds, so the pre-registered combined rule, which required the text side, does not pass; (ii) a derangement control -- retrained with shuffled headers: identical marginal, greatly reduced content correlation -- stays competent yet shows no detectable correct-to-wrong drop while M does (text headers; interaction significant on 3/3 seeds, p $\le$ 0.024, one seed below the frozen non-zero-count guardrail), so the harm requires the learned header$\to$program mapping, not the header marginal; (iii) requiring a generated program to execute removes almost all of a correct header's apparent benefit (token: +0.22 ungated regex $\to$ +0.03 gated regex $\to$ +0.02 gated geometry), so the deflation is execution blindness, not detector disagreement; (iv) the harm is regime-specific: at 0% prefix the baseline generates no valid CAD at all. Wrong intent is not noise: where it is detectable, it actively misdirects generation. --- ### Fashion-3DLR: A Controllable 3D Garment Generation Using Pairwise Fashion Elements for Intelligent Design Authors: Shenghao Yang, Hongtao Zhang, Yuhan Yi, Zhihao Tang, Zihao Cui, Lian Wen, Han Yan, Yuan Gao, Mingbo Zhao Published: 2026-07-25 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2607.23189v1 AI-generated content (AIGC) has made significant progress, with 2D generative models becoming ready-to-use tools for the digital fashion industry. However, 3D garment generation remains in its nascent stage, where in the realm of fashion, the semantic information of diverse design elements exhibits intricate coupling relationships in 3D representations, posing substantial challenges for generating diverse 3D garments. In this work, to handle the above problem, We introduce Fashion-3DLR, a novel 3D garment generation framework that utilizes diverse design elements to create high-quality, versatile 3D garment assets. Specifically, to bridge the semantic gaps between different fashion elements, we propose a Garment Feature Fusion Diffusion Transformer (GFF-DiT) module to integrate 2D fashion design elements, e.g., sketch and texture, into latent space. Within the latent space, we then employ a rectified flow transformer to generate geometry latents, which can be decoded into various 3D garment representations, including 3D Gaussians and meshes. Furthermore, we integrate Fashion-3DLR into downstream tasks, achieving the 3D Gaussian Splatting (3DGS)-driven cloth physical simulation and mesh-based virtual try-on. Experimental results indicate that Fashion-3DLR surpass the previous state-of-the-art methods, which verify that the proposed work can generate well-structured, non-watertight garments capable of physical simulation and virtual try-on, underscoring its potential as a versatile 3D garment design tool. --- ### A Scale-adaptive Vision Model Links C. elegans Neuronal Morphology to Behavior for Neurotoxicity Assessment Authors: Haochao Ying, Shenchong Lv, Yutao Sun, Zijian Tu, Xufeng Jin, Yuyang Xu, Yizhe Wang, Wei Yang, Xiaomin Yue, Jian Wu, Peilin Yu Published: 2026-07-25 Categories: cs.CE, cs.CV arXiv: https://arxiv.org/abs/2607.23183v1 Neurological disorders are a leading cause of global disability and are increasingly linked to environmental chemical exposures. Yet neurotoxicity assessment still relies on hand-scored morphological readouts that are subjective and poorly predictive of behavioral outcomes. Caenorhabditis elegans provides a genetically tractable, 3R-compliant alternative, but quantifying neuronal phenotypes from confocal microscopy at scale remains computationally challenging: existing vision foundation models, trained on natural or radiological images, cannot resolve the sparse signals and multi-scale lesions of neuronal imaging. Here, we introduce a dedicated self-supervised vision model for C. elegans dopaminergic neurons, together with CeNeuMorph, a multi-grained confocal benchmark of 27,117 annotated images. Specifically, moving beyond standard Masked Autoencoders, we propose a scale-adaptive masked image modeling strategy that jointly learns representations across resolutions and patch sizes under a fixed token budget. By decoupling structural semantic learning from rigid grid constraints, the model effectively resolves the full spectrum of neurodegenerative lesions - ranging from fine dendritic beading to gross soma shrinkage - within a tractable computational framework. Finally, our model surpasses both generalist and biomedical foundation models across classification, segmentation and detection tasks. Fusing visual features with morphological descriptors enables prediction of dopamine-dependent behavioral deficits ($R^2=0.498$). Screening 180 agrochemicals, we identify the benzimidazole moiety as a previously unrecognized determinant of dopaminergic neurotoxicity. Together, the work demonstrates how scale-adaptive self-supervised learning can connect morphology to function for a scalable alternative to mammalian in vivo models for neurotoxicity assessment and drug discovery. --- ### Beyond ICA: Identifiability by Symmetry Breaking Authors: Pengzhou Wu Published: 2026-07-25 Categories: stat.ML, cs.LG, math.ST arXiv: https://arxiv.org/abs/2607.23182v1 We prove the identifiability of deep generative models (DGMs) with piecewise-affine (PWA) decoders and Gaussian mixture model (GMM) priors, in a purely unsupervised setting. We introduce three algebraic contrast principles for symmetry breaking: domain contrast, which trivializes the mixture symmetry group; mechanism contrast, which ensures every decoder branch is witnessed by a unique boundary; and interaction contrast, which forbids parameter conspiracies between latent components and decoder branches. Together they exploit the interplay between the discrete combinatorics of the PWA map and the continuous symmetry structure of the latent GMM. Continuity is replaced by algebraic symmetry conditions; injectivity is decoupled from structural identification and required only for pointwise inversion. Our results form a hierarchy: from law identifiability (LID; latent distribution up to a global affine map) through map identifiability (MID; decoder up to the same map) to posterior and pointwise identifiability. The ICA-form ambiguity emerges under conditions on diagonal component covariances. Assumptions are only on the data-generating process, not on learning methods, except for the interaction contrast. To our knowledge this is the first to make algebraic symmetry-breaking the engine of nonlinear identifiability, the first to admit discontinuous decoders, and the first to handle fully non-injective decoders, where every observation admits multiple latent codes. --- ### Towards Dual-Brain Minimal Sufficient Representation for Vision-Language Navigation Authors: Yihao Wu, Chenyi Xu, Liqi Yan, Chenhuan Cai, Geyong Min, Bin Lin, Fangli Guan, Jianhui Zhang, Pan Li Published: 2026-07-25 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.23181v1 Vision-and-Language Navigation in continuous environments (VLN-CE) requires an agent to ground language in egocentric observations and plan in unseen scenes. Although recent multimodal large models and world-model-based methods have improved navigation, they often preserve excessive task-irrelevant detail, weakening generalization and increasing computational burden. We propose BrainNav, a navigation framework grounded in the Principle of Minimal Sufficiency. BrainNav consists of three components: a Logical Anchor Model that implements instruction-aware selective perception to suppress environmental noise, a Minimalist Constraint Alignment module that serves as a compact cross-modal bottleneck, efficiently synchronizing discrete linguistic intent with continuous latent dynamics while filtering out redundant information, and a Compression World Model that predicts action-conditioned states within a condensed, low-rank latent space. These modules align semantic intent with spatial perception, enhancing the agent's robustness and efficiency in complex tasks. Experiments show that BrainNav improves over prior SOTA by 2.0 % / 1.0 in SR/SPL on R2R-CE val-unseen and 0.94 % / 0.78 on RxR-CE val-unseen. These results indicate that minimally sufficient world representations provide an effective foundation for robust VLN. --- ### Beyond a Global Norm: Personalizing Toxicity Sensitivity in Language Models Without Retraining Authors: Rares A. C. Diaconescu, Iulia Slanina, Alina Florea, Andrei B. Trache, Miruna E. Coroi, Anne Arzberger, Jie Yang, Enrico Liscio Published: 2026-07-25 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2607.23175v1 Reducing toxicity is often framed as a global alignment problem, yet perceptions of harmful language are subjective and context-dependent. We present the first comparative evaluation of training-free methods for aligning language generation to user-specific toxicity sensitivities across three inference-time intervention stages: pre-decoding (prompt conditioning and rewriting), in-decoding (token, logit, and representation steering), and post-decoding (candidate re-ranking). Evaluated against toxicity sensitivity targets derived from the PRISM dataset, all methods reduce alignment error by 28-47%. However, the results reveal a fundamental trade-off between alignment effectiveness, personalization, and general language quality, showing how toxicity sensitivity alignment is an inherently multi-objective problem. --- ### Adaptive Multi-Scale Forecasting and Gate-Localized Conformal Prediction for Multivariate Nonstationary Time Series Authors: Ziling Ma, Junshu Jiang, Ángel López-Oriona, Ying Sun, Hernando Ombao Published: 2026-07-25 Categories: stat.ML, cs.LG, stat.AP, stat.ME arXiv: https://arxiv.org/abs/2607.23165v1 We propose ABF-T-GLCP, a model-agnostic framework for forecasting and uncertainty quantification in nonstationary multivariate time series. The central idea is to learn an adaptive predictive state representation for point forecasting and reuse it for conformal calibration. The forecasting module combines horizon-specific temporal experts through a learned gate and refines predictions using sparse predictive transfer across related series. The uncertainty module, Gate-Localized Conformal Prediction (GLCP), uses the learned gate state, together with temporal recency, to select locally relevant calibration residuals, thereby coupling uncertainty calibration to the predictive regimes used by the forecasting model. This shared representation allows point forecasts and prediction intervals to adapt consistently under evolving temporal dynamics while retaining the model-agnostic nature of conformal prediction and yielding approximate local coverage under mild stability conditions. Experiments on a large-scale high-frequency commodity forecasting benchmark show consistent gains in point forecasting accuracy and substantially narrower prediction intervals with empirical coverage close to the nominal level. Additional results indicate that the framework extends beyond the motivating financial application. --- ### CachedSearch: Training-Free Cached Exploration for Test-Time Search in Video Diffusion Authors: Shreshth Saini, Neil Birkbeck, Yilin Wang, Balu Adsumilli, Alan C. Bovik Published: 2026-07-25 Categories: cs.AI, cs.CV, cs.MM arXiv: https://arxiv.org/abs/2607.23159v2 Test-time search lets small video diffusion models rival larger ones, but costs 2-10x more. All candidates are fully denoised, although most are discarded. Training-free caching makes each rollout 2-3x faster at near-lossless quality. Composition is safe only if lossy caching preserves verifier rankings. We present the first study of whether caching corrupts candidate ranking in video test-time search. On Wan2.1-T2V-1.3B with an adaptive caching wrapper (~2x per-candidate speedup), ImageReward scores seed-matched cached and full rollouts. Median per-prompt Spearman rank correlation is 0.905, with 72% top-1 agreement on the VBench suite. VBench-2.0 replicates this result on a harder suite. Recomputing the cached winner at full compute retains 90-94% of the full-search gain. Errors cluster among near-tied candidates, making corruption self-limiting. This finding leads to CachedSearch. It explores every candidate with aggressive caching, then re-generates only the winner at full compute. At N=8, it captures 94.7% of best-of-N's gain at 63% of the cost. Capture rises with width. At matched budget, it searches twice as wide for 38% more gain. The result holds from 1.3B-14B across six models and four families: Wan, LTX, CogVideoX, and Hunyuan. Wan2.1-14B matches the 1.3B model's fidelity. Mid-trajectory pruning multiplies the exploration saving to 3.11x at 88.6% capture. Ports to other model families require recalibrating a single parameter, showing that fidelity tracks architecture rather than parameter count. CachedSearch is training-free, verifier-agnostic, and orthogonal to the search algorithm, making it a plug-in multiplier for test-time scaling. --- ### Anytime-Valid Confirmation of Covariate Balance for Prespecified Corrections Authors: Seungjin Choi Published: 2026-07-25 Categories: stat.ME, stat.ML arXiv: https://arxiv.org/abs/2607.23157v1 Many covariate-shift adaptation methods construct a correction $w(x)$, but users must still determine whether the corrected distribution is sufficiently balanced for the target stream. We study anytime-valid confirmation of prespecified corrections from sequential unlabeled target inputs. Our primary contribution is a procedure for confirming covariate balance. For a prespecified class of balancing functions and tolerances, time-uniform confidence sequences permit continuous monitoring and data-dependent stopping once all plausible target moments lie within their tolerance bands. If the correction is out of tolerance for at least one function, the probability of ever incorrectly confirming balance is at most the prescribed level. Upon stopping, the procedure yields a certificate local to the chosen functions and tolerances, yet providing an absolute downstream-adequacy statement that ordinary shift diagnostics generally do not. With finite source data, contracted bands preserve this guarantee while accounting for uncertainty in weighted source moments, whereas expanded bands support only compatibility diagnostics. As complementary information, we study a source-calibrated likelihood-ratio e-process whose KL-drift identity characterizes correction directions relative to the source. Under the source-reference distribution, the probability of ever crossing its evidence threshold is controlled, but crossing does not confirm balance. We also give an exponential-tilt test for departures beyond an acceptable correction region and deploy balance-confirmed corrections in weighted conformal prediction. Experiments illustrate false-confirmation control, locality to the balancing-function class, KL-drift diagnostics, acceptable-region monitoring, finite-source effects, and downstream conformal coverage under covariate shift. --- ### In-Context Learning as Implicit Policy Gradient Authors: Masahiro Kaneko, Timothy Baldwin Published: 2026-07-25 Categories: cs.LG, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2607.23153v1 Recent work has shown that large language models (LLMs) can iteratively improve their outputs by incorporating generated samples and their corresponding evaluation scores as in-context examples. Despite these empirical findings, the theoretical foundations underlying this phenomenon remain poorly understood. In this paper, we show that score-conditioned In-Context Learning (ICL) admits a structural correspondence to policy gradient optimization. We first provide a constructive proof that self-attention mechanisms can implement reward-weighted aggregation analogous to the REINFORCE algorithm under specific weight matrix configurations, and discuss the relationship between this construction and the behavior of pretrained transformers. The correspondence is directional in hidden-state space and holds exactly only under the stated simplifying conditions; we quantify its strength empirically. Within our simplified hidden-state model, we furthermore derive an exact upper bound on the distribution shift induced by a bounded attention update, yielding a trust-region-like analogy to KL-constrained policy optimization. We validate our theory through extensive experiments across multiple LLMs, demonstrating that LLMs effectively utilize score information to shift output distributions toward high-scoring exemplars, and that attention weights exhibit a strong correlation with example scores. ---