跳到正文
研踪PaperTrail

论文库

检索计算机领域论文,按研究方向、作者或关键词筛选。

部分采集异常
正在读取账号与研究记录…

全站更新概况

+0新增论文
+0其中 AI 相关
+0已有论文修订
+91新增 AI 导读

arXiv 最近完成扫描:2026-11-22 21:02 · 全站最近入库:2026-11-21 12:19各信源进度 →

统计口径与数据新鲜度

统计 2026-11-21 23:54 — 2026-11-22 23:54(北京时间) · 点击数字查看对应简报;AI 相关包含在新增论文内。

最近完成检查:2026-11-22 21:02最近发现新增版本:2026-11-21 12:19库内论文最新更新时间:2026-11-20 02:59

每分钟刷新已保存的数据。空检查不代表采集异常;新收录可能包含历史论文。AI 导读按新增版本计数。

29,158 已收录论文24,181 AI 相关8,992 有修订112 HF 社区发现

浏览论文库

时间范围跟随排序:版本更新、本站收录或首次发表。最新收录可能包含历史论文。部分信源异常,查看状态 →

找到 23,345 篇论文 · 按版本更新时间排序上轮扫描起点 2026-11-22 21:00(北京时间)
新论文 · v1arXiv:2611.25629

NIMIS: A Structure-Agnostic, Chemically Informed Slot-Attention Framework for Demixing 1D $^1$H-NMR Mixture Spectra

Anna Marija Sumrova, Jonathan M. Goodman

原文摘要

The products of chemical reactions are usually complex mixtures that must be separated into their components before interpretable $^1$H-NMR spectra can be obtained. Direct analysis of the spectra of crude mixtures would accelerate reaction analysis and facilitate the exploration of new chemical space. We present NIMIS, a structure-agnostic, attention-based deep learning framework that groups the peaks of a 1D $^1$H-NMR mixture spectrum according to the component from which they originate. J-coupling information is introduced directly into the attention mechanism of a multi-head Transformer encoder, and the resulting contextualised peak embeddings are separated into components by slot attention. NIMIS was trained on 158,005 synthetic mixtures constructed from experimental spectra in the NMRexp database, covering molecules of up to 20 heavy atoms composed of C, H, N, O, S, F, Cl, Br and I. On held-out molecules it reaches a peak-to-component assignment accuracy of 82.28% for noise-free input and 77.34% with 10% integral noise, with a mean NIMIS score of 0.83 for correct assignments. NIMIS outperforms a conventional Transformer and simple baselines, and its performance carries over to a scaffold-disjoint evaluation. By grouping peaks into component-level sets, NIMIS provides a foundation for more robust mixture analysis methods, including methods capable of handling overlapping signals.

cs.LG

新论文 · v1arXiv:2611.25628

RelGeo: Pair-Conditioned Relational Calibration for Frozen Diffusion Correspondence

Zixuan Dai, Zhiliang Zhang, Daosong Hu, Tang Zheng, Xinkai Liu, Tianyu Lei, Wei Zhan

原文摘要

Frozen diffusion descriptors retrieve strong semantic correspondence candidates, yet cosine nearest-neighbor matching can select incorrect locations despite valid high-ranking alternatives. Across DIFT-SD2.1 and DiTF-FLUX on SPair-71k, AP-10K, and SOCO (CC, SOC, Cross-SOC), 41-69% of Top-1 errors retain a PCK@0.10-valid match among the original Top-5 spatial peaks, revealing a persistent candidate-selection bottleneck. We find that descriptor rankings and current-pair relational evidence can favor different candidates: reciprocal diffusion Query/Key associations and pretrained RoMa transport exhibit distinct, partly complementary preferences. To incorporate this evidence into matching decisions, we introduce RelGeo, a Pair-Conditioned Relational Calibration framework expressing heterogeneous relations as shared source-target coordinates. Spectral Relational Coordinates (SRC) encode reciprocal diffusion affinities, while Warp Identity Coordinates (WIC) encode reliability-weighted bidirectional transport endpoints. Their inner products augment frozen descriptor similarity while preserving full-domain cosine nearest-neighbor matching without additional semantic-correspondence training. RelGeo improves all ten evaluations by an average of 5.51 PCK@0.10 points; fixed-candidate reranking and pair-misalignment controls corroborate improved selection and dependence on current-pair relational alignment.

cs.CV

新论文 · v1arXiv:2611.25627

Uncontrolled source titles: When Nature, Science, and PNAS meet Scopus

Weishu Liu, Ruifeng Zhang, Haifeng Wang

原文摘要

Bibliographic databases such as Scopus and Web of Science play a critical role in literature discovery, systematic reviews, bibliometric analyses, and research evaluation, yet their reliability depends on effective control of the quality of the bibliographic data. This study examines the problem of uncontrolled source titles in Scopus by comparing five possibly adopted journal-level retrieval strategies across three flagship journals-Nature, Science, and PNAS. Substantial discrepancies across retrieval results indicate persistent fragmentation in the source title field. A comparative analysis using the Web of Science Core Collection shows that no comparable degree of uncontrolled source-title variation is observed for the same journals. The study further explores possible mechanisms, including inheritance of different title forms from (different) upstream providers, automated parsing and OCR-related artifacts, and insufficient cross-validation. Based on these findings, the study proposes recommendations at both the user and database-provider levels to improve source-title normalization and strengthen source-level metadata governance, with implications for the accuracy and reproducibility of journal-level retrieval and evaluative uses.

cs.DL · cs.IR

新论文 · v1arXiv:2611.25626

Physical Laws without Physics Loss? Probing Learned Dynamics in Video World Models through Forced Response

Yutao Li, Chen Wang, Wangmeng Zuo, Hui Li

原文摘要

Do video world models learn physical laws without a physics loss, or merely imitate motion? Overall motion and appearance scores cannot settle it. We investigate this through nonlinear forced response: an excitation shakes the support of a suspended object. We introduce dual visual-numeric conditioning and adapt six image-to-video backbones to predict this motion using only the flow-matching loss. Wan predicts motion accurately and generalizes to unseen excitations, objects, poses, and scenes. Counterfactual tests suggest capabilities beyond replaying training videos. But what laws have they learned? We design three tiers of tests using Taylor expansions of the dynamics and geometry. All six reproduce the linear response, including resonance; five capture higher-order dynamics, including nonlinear softening. Remarkably, softening is barely present in training: only 0.74% of the swing's squared displacement. Extensions to a double pendulum and a rocking block show sustained dynamics are learned more faithfully than brief impacts. Yet learned dynamics may fail to appear when state estimation or propagation goes wrong. We turn these tests and findings into strategies to improve video training and generation.

cs.CV · cs.LG

版本更新 · v2arXiv:2603.02825

A simple Path-based LP Relaxation for Directed Steiner Tree

Kanstantsin Pashkovich, Marta Pozzi, Laura Sanità

原文摘要

We study the Directed Steiner Tree (DST) problem in layered graphs through a simple path-based linear programming relaxation. This relaxation achieves an integrality gap of O(l log k), where k is the number of terminals and l is the number of layers, which matches the best known bounds for DST previously obtained via lift-and-project hierarchies. Our formulation bypasses hierarchy machinery, offering a more transparent route to the state-of-the-art bound, and it can be exploited to provide an alternative simpler proof that O(l) rounds of the Sherali-Adams hierarchy suffice for reducing the integrality gap on layered instances of DST.

cs.DS

新论文 · v1arXiv:2611.25624

Deviations from the NTK Regime in CNNs under Joint Depth-Width Scaling

Maria Matveev, Pascal Esser, Gitta Kutyniok, Mariia Seleznova

原文摘要

The concentration and training-time constancy of the neural tangent kernel (NTK) depend on the interplay between network width, depth, and initialization. While joint depth-width scaling has been studied in fully-connected, residual, and recurrent networks, its effects in convolutional neural networks (CNNs) remain less understood. We investigate these effects numerically in CNNs, examining NTK dispersion at initialization and kernel rotation during training. Let depth $L$ be the number of convolutional layers, and width $C$ the number of channels per layer. We find that NTK dispersion increases with $L/C$ in the critical and chaotic initialization regimes, but remains small in the ordered regime. At critical initialization, dispersion curves approximately collapse as a function of $L/C$, suggesting a nontrivial joint scaling regime. During training, kernel rotation generally increases with $L/C$, even in the ordered regime where initial dispersion is small. Together, these results suggest that $L/C$ may govern NTK deviations from the infinite-width limit in certain regimes, and that NTK concentration at initialization and stability during training exhibit different scaling behaviors.

cs.LG

版本更新 · v3arXiv:2609.19513

QVAC Genesis III: A Large-Scale, High-Quality Open Synthetic STEM Corpus for Efficient Language Model Pre-Training

Davide Vitabile, N. Ranjan, Akshay Nambiar, Kamal K. Gupta, Amril Nazir

原文摘要

High-quality pre-training data is a critical bottleneck for educational and STEM-specific language models targeting edge AI and on-device deployment where token budgets are tightly constrained. While major organizations train ever-larger models on private corpora, the open ecosystem lacks STEM-focused synthetic datasets that deliver high per-token learning value efficiently for small models. To address this gap, we introduce QVAC Genesis III, a 191.43B-token, STEM-focused multi-domain synthetic corpus covering 19 domains across several difficulty levels and different educational styles. QVAC Genesis III is built via a dual generation strategy that performs targeted teacher distillation using a weak edge-scale student model as signal: the student's failures are converted into corrective explanations, while its successes are expanded into contrastive option-level reasoning over all answer choices. We further introduce an LLM-as-a-parser evaluation protocol that extracts final answers from free-form outputs and tracks both accuracy and answer validity. To validate the effectiveness of our QVAC Genesis III data, we conduct controlled from-scratch ablations with 1.7B-parameter models, showing that models trained with QVAC Genesis III consistently outperform both models trained with the open-source synthetic corpus Cosmopedia-v2 and the publicly released Cosmo-1B model across ARC, GPQA Diamond, and MMLU STEM benchmarks, achieving up to +28.57% on ARC-E and +21.35% on ARC-C, while reaching a Valid Answer Rate of up to 99.45%.

cs.AI · cs.LG

版本更新 · v2arXiv:2605.10850

Verification Mirage: Mapping and Repairing the Reliability Boundary of Self-Verification in Medical VQA

Ruinan Jin, Beidi Zhao, Myeongkyun Kang, Jiaqi Liu, Behzad Bozorgtabar, Qiong Zhang, Xiaoxiao Li

原文摘要

Self-verification is increasingly used to improve the reliability of vision-language models (VLMs) and provide feedback for iterative self-correction, yet it is useful only when checking an answer provides information that discriminates correct from incorrect predictions rather than simply re-solving the same visual problem. We introduce VeriMap, an auditing framework for mapping this reliability boundary in medical visual question answering, where many answers hinge on fine-grained but auditable visual distinctions. Across six VLMs, five datasets, and seven task types, VeriMap reveals a verification mirage: vanilla verifiers accept 71.0% of incorrect candidates on average, and all six models are more likely to accept their own errors than errors from other models. When reused as feedback, these false acceptances can persist across self-correction rounds. We find that stronger access to the correct image or candidate-hidden re-solving is insufficient. Instead, verification becomes more selective when the candidate is reduced to a falsifiable visual distinction and the resulting observation governs the verdict. Guided by this finding, we introduce MirageCheck, which separates test construction, candidate-hidden visual measurement, and evidence-to-verdict binding through CONTRAST, MEASURE, and BIND. Across models and datasets, MirageCheck reduces mean false acceptance by 14.6 percentage relative to the strongest candidate-independent baseline while improving balanced accuracy, and substantially reduces false acceptance during iterative correction. Together, our results show that reliable visual self-verification requires turning acceptance to discriminative evidence from the current image rather than another judgment of the same problem.

cs.CV

新论文 · v1arXiv:2611.25620

CaT-Movie: Balancing Global Planning and Local Rendering for Long Multi-Shot Video Generation

Jinpeng Lu, Haoyuan Shi, Yuanshen Guan, Jiyuan Xia, Bin Wu, Yinda Chen, Zhiwei Xiong

原文摘要

Current multi-shot video generators render only limited durations per pass, requiring long movies to be stitched from local segments across which narrative progression and visual identity drift. To synthesize long-horizon movies with Coherence across Time, we present CaT-Movie, a multi-shot generation system consisting of a global story planner and a local window renderer. The planner expands user requests via a large language model to preserve, infer, and complete story details, while deterministic code maintains the resulting storyboard across shots to compile self-contained shot prompts and index reference frames from generated history. To execute this guidance, the renderer applies reference-routed sparse self-attention and shot-wise cross-attention to route identities from the indexed frames while isolating per-shot textual conditions. Given the scarcity of high-quality open-source multi-shot data, we construct a dataset of over 350K film shots annotated with semantic state transitions, entity grounding, and multi-granularity requests. We further build CaT-MBench, a benchmark on held-out films with the same annotations. It gives every baseline its native format and counts cross-shot consistency only when a cut opens a new camera setup, preventing repeated framing from passing as consistency. On CaT-MBench, CaT-Movie attains the highest average score and the most consistent characters among all methods. The project page is available at https://jinplu.github.io/CaT-Movie/.

cs.CV

版本更新 · v6arXiv:2606.06840

Characterize Then Distill: Mechanistic Reasoning in Large Output Spaces

Debjyoti Saha Roy

原文摘要

Reasoning-trained language models can pick a document's few labels from thousands of candidates, zero-shot. We ask which components make these choices, and whether a student can be taught where to look as well as what to write, so its decisions trace to a few components and what they read. We study clinical coding of MIMIC-IV discharge summaries with all 5,651 candidate codes in the prompt. Each choice is one token: a region of the code list, a code within it, or the departure from a similar code named earlier. We measure each attention head's causal effect there by removing it and transplanting it into another summary. On essentially every summary, a small, shared set of heads, partly specific to each kind of choice, is necessary and sufficient; blocking what the tested heads read removes their effect. MISTILL builds on this: besides the teacher's text, it supervises where the student attends at the teacher's decision tokens. In a cross-family student, its own heads, found after training, restore nearly twice as much of the choice between similar codes as before, where chain-of-thought distillation alone adds under a third; one that already carried most of it gains a little, consistently across seeds, and both gains hold on the students' own completed traces. On two corpora with coder-marked evidence, a MISTILL student's decision heads attend to it more, and more specifically, than a chain-of-thought-distilled student's. Controls locate the gain in supervising attention at the decision tokens; in the cross-family student, a target built from unselected teacher heads loses it. No task difference from chain-of-thought distillation is detected at matched precision. The structure recurs in a second reasoner; ranked heads beat random ones on legal-document labelling and code generation; and without retraining the students keep their gain on Spanish case reports and a QA benchmark.

cs.CL · cs.AI · cs.LG

新论文 · v1arXiv:2611.25619

Cyber-Physical RSI for AI-RAN: A Case Study

Binghan Wu, Kai Zeng, Zhiqi Li, Zhigang Wang, Yang Liu, Ya-Qin Zhang, Joseph Sifakis, Ye Ouyang

原文摘要

Learned models for radio access networks (RANs) are often developed using simulation, yet their predictions can shift when exposed to physical propagation conditions, protocol behavior, and equipment implementations. This paper presents a cyber-physical recursive self-improvement (RSI) framework for AI-RAN that connects a task-oriented student model, an experiment-planning coach agent, a calibratable virtual environment, and a physical host agent. Physical measurements are organized as evidence for both model adaptation and subsequent experiment planning. We examine an implementation of the physical-feedback pathway using ping-pong handover as a case study. The system completes 12 physical executions organized into six baseline--candidate pairs, updates the student from measured outcomes, and evaluates the revised model on reserved physical configurations. In the observed short-window tests, increasing the time-to-trigger from 320 to 1280\,ms changes the ping-pong count from three to zero. The updated student achieves a grouped-validation mean absolute error (MAE) of 2.673 round trips/min, while a common three-configuration physical evaluation yields MAEs of 1.240 and 1.601 round trips/min for the pre-feedback and post-feedback models, respectively. These results characterize both the operational feedback loop and the parameter-dependent generalization behavior revealed by new physical experiments.

cs.AI

新论文 · v1arXiv:2611.25614

Don't Lose the Future: Latent-Aware 4-Bit Quantization of World Action Models

Yichen Guo, Yanlong Chen, Xinyue Zhang, Mohan Liu, Song Zhang, Yawei Li

原文摘要

World action models (WAMs) predict how a scene will evolve and condition their actions on this predicted future. Deploying them on a robot's onboard computer calls for 4-bit weights, activations and key-value caches (W4A4KV4). Existing 4-bit quantizers for WAMs preserve the generated action, and estimate sensitivity one denoising step at a time. However, they fail to preserve the fidelity of the predicted future: once the action pathway is protected, the remaining action error enters mostly through a degraded future that none of them deliberately protects. We propose LAQ, a latent-aware quantization framework that preserves the predicted future where it is written, stored and read. Specifically, we introduce a reader metric, which measures future error by its effect on the action, and a read count, which weights each rounding error by the number of denoising steps that read it. Together they give the allocation of precision over activations and caches a closed-form solution. We further develop gaze-guided refinement, which gives the tokens the action attends to a second 4-bit pass in the denoising loop and in the cache. All weights stay at 4 bits, with no high-precision branch and no per-task recalibration. On the LIBERO-Plus benchmark, LAQ reaches 76.4% and 83.3% average success on Fast-WAM and Cosmos-Policy, within 1.2 percentage points of the 16-bit models and all above the baselines with 16-bit KV caches. It compresses the model by 3.8x, reduces the peak GPU memory of the whole policy by 2.3x, and runs 1.66x faster than the 16-bit model. On a real AgileX PiPER arm, it reaches 47.0% average success over three tasks, against 52.0% for the 16-bit model and 30.0% higher than QuaRot.

cs.AI

新论文 · v1arXiv:2611.25613

Can an LLM Fix a Language Model Without Training It?

Bharath Kumar Bolla, Raghavender Surya Upadhayula, Vishnu Surya Reddy Nandi, Ramesh Nandipalli

原文摘要

Although an external language model can select and test repairs to another model's weights, behavior can also be recovered through fixed numerical edits. We assessed the value of interactive selection in a controlled study of Transformer repair. After inspecting predictions, activation summaries, and weight statistics, the editor selects projection recipes to apply to the target without gradient updates. We analyzed 20 recorded sessions involving five masked and causal language models, comparing the selected edits with fixed projections, scripted and matched-budget black-box search, and low-rank adaptation. Under the core two-layer fault, live selection matches black-box search on SmolLM2 and Qwen2.5. For Pythia, live selection achieves 65.62% completion accuracy, compared with 0% for black-box search and 66.15% for a fixed projection. Under a separate compound fault affecting attention and feed-forward weights, both scripted and LLM-guided selection increase the descriptive five-checkpoint mean from 24.89% to 71.25% using the same operator family. On unmodified checkpoints, scripted projection yields small gains or losses on a separate set of framed questions. The experiments identify cases in which feedback changes an edit decision, operator coverage accounts for recovery, or supervised adaptation performs better. Combined with measurements of parameter restoration and preservation, these comparisons establish the scope of controller-guided repair for the tested faults and completion tasks.

cs.AI

版本更新 · v4arXiv:2509.21961

FlowDrive: moderated flow matching with data balancing for trajectory planning

Lingguang Wang, Ömer Şahin Taş, Marlon Steiner, Nick Le Large, Christoph Stiller

原文摘要

Learning-based planners are sensitive to the long-tailed distribution of driving data. Common maneuvers dominate datasets, while dangerous or rare scenarios are sparse. This imbalance can bias models toward the frequent cases and degrade performance on critical scenarios. To tackle this problem, we compare balancing strategies for sampling training data and find reweighting by trajectory pattern an effective approach. We then present FlowDrive, a flow-matching trajectory planner that learns a conditional rectified flow to map noise directly to trajectory distributions with few flow-matching steps. We further introduce moderated, in-the-loop guidance that injects small perturbation between flow steps to systematically increase trajectory diversity while remaining scene-consistent. On nuPlan and the interaction-focused interPlan benchmarks, FlowDrive achieves state-of-the-art results among learning-based planners and approaches methods with rule-based refinements. After adding moderated guidance and light post-processing (FlowDrive*), it achieves overall state-of-the-art performance across nearly all benchmark splits. Our code is available at https://github.com/einsteinguang/flow_drive_planner.

cs.RO · cs.AI · cs.LG

版本更新 · v3arXiv:2605.00668

SENECA: Small-Sample Discrete Entropy Estimation via Self-Consistent Missing Mass

Lucas H. McCabe, H. Howie Huang

原文摘要

Discrete entropy estimation is a classic information theory problem, wherein the average information content of a discrete random variable is estimated from samples alone. Naive approaches, such as the plugin method, fail to account for the probability mass associated with members of the random variable's support that are unobserved in a given sample, known as the "missing mass." The resulting systemic underestimation is particularly problematic when data is time-consuming or costly to gather. We propose SENECA, an entropy estimation scheme based on a novel "self-consistent" missing mass calculation. Extensive numerical experiments indicate that our approach outperforms many state-of-the-art alternatives overall in the small-sample setting. We then apply SENECA to two practical use cases, namely biodiversity estimation and the detection of incorrect large language model responses, where our method is competitive with domain-specific approaches. Our work advances SENECA as an effective drop-in replacement for small-sample entropy estimation, with broad utility across several domains.

cs.IT · stat.CO

新论文 · v1arXiv:2611.25610

Hierarchical Spatiotemporal World Modeling with Content-Adaptive Spatial Addressing

Tiantian Li, Junfeng Lu, Yifei Wang, Zhaoyang Zhang, Lei Wang, Tianyi Qian

原文摘要

Long-horizon world modeling in partially observable environments requires compact memories that preserve spatial structure despite accumulated localization drift. Inspired by hippocampal episodic and spatial memory and the multi-periodic coding of entorhinal grid cells, we propose HiSTFormer, a hierarchical spatiotemporal world model, which adopts a layer-wise organization of local temporal RoPE and multi-periodic spatial RoPE, together with content-conditioned address offsets, to capture recent temporal context and long-range spatial structure. A fixed-capacity memory keeps KV storage and per-step attention cost bounded as the sequence grows. On MemoryMaze and HM3D, HiSTFormer outperforms strong world-model baselines within the training sequence length and generalizes substantially better to prediction horizons far beyond those seen during training, while using a compact model. The hierarchical spatiotemporal design remains effective across deeper Transformer stacks, with consistent gains as depth increases. These results show that hierarchical spatiotemporal organization and content-adaptive spatial addressing provide effective inductive biases for efficient long-horizon world modeling.

cs.NE · cs.LG

新论文 · v1arXiv:2611.25608

MetaScanner: An Empirical Study and Proactive Defense against Deceptive UI Attacks in Social VR

Junhee Lee, Seungwon Woo, Minseok Kim, Jongseop Kim, Kiyoung Kim, Hwanjo Heo, Jinwoo Kim

原文摘要

Social virtual reality (VR) platforms have surged in popularity but continue to face significant security challenges, including privacy invasion and user manipulation. Yet, the security risks posed by deceptive content remain largely unexplored. In this paper, we introduce four novel deceptive UI attacks that covertly manipulate social VR users into performing unintended actions without their awareness. We show how adversarial content creators can embed seemingly benign scripts and interactive elements within carefully designed virtual worlds, deceiving users into unknowingly carrying out harmful behaviors. We implemented these attacks on VRChat, a leading social VR platform, and an IRB-approved user study involving 30 participants confirmed their effectiveness. To mitigate these threats, we propose MetaScanner, a proactive analysis system that efficiently scans virtual world assets, including scripts and interactive objects, to identify suspicious or deceptive elements before users are exposed to them. We demonstrate that MetaScanner can successfully detect all threats in test worlds where the four attack scenarios were implemented. Additionally, MetaScanner shows its lightweight nature, with an average scanning time of 5 seconds across 38 real-world assets.

cs.CR · cs.HC

新论文 · v1arXiv:2611.25607

PSTO: A PDE-structured transport operator with shared implicit updates for multiscale CDR systems

Kaiwen Li, Dezhi Zhou, Kun Wu, Xingcai Lu

原文摘要

Multiscale convection-diffusion-reaction (CDR) systems couple spatial transport with local stiff reaction kinetics, making their numerical simulation computationally demanding. Conventional neural surrogates require costly coupled CDR training data, and their parameter counts scale with the number of transported variables, limiting scalability to high-dimensional CDR systems. Here we propose a PDE-Structured Transport Operator (PSTO) that learns convection--diffusion dynamics solely from reaction-free data. By embedding the PDE structure into the neural architecture, PSTO generalizes to full CDR systems at inference without retraining. A single transport operator block is shared across all transported variables and refined through relaxed fixed-point iterations, performing an implicit-like correction in latent space. This improves stability during recursive multi-step prediction without adding trainable parameters. We evaluate PSTO on three benchmarks: a canonical 2D CDR system, a multiscale Robertson counterflow flame, and a multiscale hydrogen planar jet flame. In every case, PSTO reduces the mean relative $L_2$ error after 100 coupled steps by at least 50% compared to the strongest neural baseline. Ablation studies on the fixed-point iterations and detailed error decomposition further confirm that PSTO achieves both high one-step accuracy and long-term stability in coupled rollouts. Time-step sensitivity analysis shows that PSTO maintains accuracy even with a tenfold larger time step, whereas baseline accuracies decay rapidly as the time step increases. This framework enables efficient, modular modeling of multiscale CDR dynamics by separating transport learning from the treatment of local kinetics.

cs.LG

版本更新 · v3arXiv:2605.26089

Channel-wise Vector Quantization: Native 1D Visual Tokenization with Learnable Ordering

Wei Song, Tianhang Wang, Yitong Chen, Tong Zhang, Zuxuan Wu, Min Li, Jiaqi Wang, Kaicheng Yu

原文摘要

We present Channel-wise Vector Quantization (CVQ), a novel image tokenization paradigm that replaces patch-wise tokens with channel-wise tokens. Unlike conventional vector quantization, which assigns a discrete token to each patch feature vector, CVQ quantizes each channel of the feature map. This represents an image as a 1D channel sequence with learnable coarse-to-fine ordering, rather than a 2D patch grid. Based on CVQ, we introduce a new visual autoregressive framework with "next-channel prediction". Instead of rendering images patch by patch in raster order, our Channel-wise Autoregressive (CAR) model predicts image channels sequentially, producing progressively enriched visual details. Specifically, it first sketches global structure and then refines fine-grained attributes, akin to a human artist's workflow. Empirically, we show that: (1) CVQ exhibits strong codebook scalability, maintaining 95%+ utilization with up to $2^{16}$ entries while delivering growing reconstruction gains over VQ as the codebook scales up; and (2) CAR attains a DPG score of 86.7 and a GenEval score of 0.79, demonstrating strong effectiveness for text-to-image generation.

cs.CV · cs.AI

新论文 · v1arXiv:2611.25605

Copilot or Culprit? How AI Is Changing Open-Source Bug Reporting and Maintenance

Yinuo Liu, Yunlong Lyu, Zichao Luo, Tian Dong, Zian Luo, Peng Chen, Hao Chen

原文摘要

AI-generated bug reports have drawn criticism for unreliable claims and faulty reproducers that burden maintainers. However, as LLM capabilities improve, some maintainers have begun to express more positive views on AI use in bug reporting and maintenance. Understanding this evolution requires examining whether AI-involved reports are becoming more reliable, which uses of AI are associated with valid findings, and how maintainers investigate these reports and act on them. To this end, we conduct the first large-scale study of 120,918 bug reports from 101 open-source projects between January 2024 and September 2026, tracing AI involvement from discovery and reporting through validation and resolution. Our findings show that AI-involved reports become more complete over time, with a growing share of reports confirmed as valid. In bug reporting, AI for reproduction and diagnosis are associated with higher validity, whereas report writing is not. Extends into maintenance, maintainers increasingly use AI to investigate code and recommend fixes. However, they devote more effort to AI-assisted reports of human-discovered bugs than to reports of AI-discovered bugs. These findings highlight both AI's growing participation in bug investigation and the continuing need for verification and follow-up. Drawing on these findings, we provide practical guidance for turning AI-generated findings into verifiable reports: emphasizing reproducibility, verification, and accountability. We followed this guidance when submitting 53 reports on AI-discovered bugs, of which 69.81% were accepted by maintainers as of writing.

cs.SE · cs.AI

新论文 · v1arXiv:2611.25604

PAD: Policy-Aware Data Valuation for Autonomous Driving

Chengkai Xu, Yicheng Guo, Jiaqi Liu, Yiming Cui, Shiyu Fang, Peng Hang, Hao Zhang, Jian Sun

原文摘要

End-to-end autonomous driving increasingly depends on scaling training demonstrations, yet data growth does not guarantee proportional gains. However, additional data do not guarantee better policies: a demonstration that benefits one planner may be redundant or harmful to another. Existing curation methods typically infer data value from intrinsic sample properties or training-dependent proxies that do not measure downstream policy utility. To address this, we introduce \textbf{Policy-Aware Data Valuation (PAD)}, a framework that assigns each candidate demonstration a utility specific to the target driving policy. PAD combines scene-level calibration feedback with projected policy gradients and a regularized approximation of local training geometry, producing scalable influence-based values for the policy. A controlled group-removal study establishes the directional fidelity of these values, while cross-planner analysis shows that the same candidate pool induces markedly different rankings across planner families. On DiffusionDrive, PAD improves EPDMS over real-only training by $3.5$ points on NavTest and $5.0$ points on NavHard. Policy-specific curation consistently outperforms real-only and full-pool training across diffusion-, regression-, and flow-matching-based planners while substantially reducing scene-level regressions. These results establish policy-aware data valuation as a basis for replacing indiscriminate data accumulation with targeted policy improvement, enabling more efficient and reliable scaling of autonomous-driving systems.

cs.RO

版本更新 · v2arXiv:2603.18604

xAppGen: LLM-driven Natural Language Programming for Agile xApp Development

Xin Li, Shiming Yu, Leming Shen, Jianing Zhang, Yuanqing Zheng, Yaxiong Xie

原文摘要

Traditional RAN systems are closed and monolithic, stifling innovation. The openness and programmability enabled by Open Radio Access Network (O-RAN) are envisioned to revolutionize cellular networks with control-plane applications--xApps. The development of xApps (typically by third-party developers), however, remains time-consuming and cumbersome, often requiring months of manual coding and integration, which hinders the roll-out of new functionalities in practice. To lower the barrier of xApp development for both developers and network operators, we present xAppGen, the first LLM-driven natural language programming framework for agile xApps that automates the entire xApp development pipeline. In a nutshell, xAppGen turns high-level user intents into swiftly deployable xApps within minutes, eliminating the need for manual coding or testing. To this end, AutORAN builds a fully automated xApp generation pipeline, which integrates multiple functional modules (from user requirement elicitation, AI/ML function design and validation, to xApp synthesis and deployment). We design, implement, and comprehensively evaluate xAppGen on representative xApp tasks. Results show xAppGen-generated xApps can achieve similar or even better performance than the best known hand-crafted baselines. xAppGen drastically accelerates the xApp development cycle (from user intent elicitation to roll-out), streamlining O-RAN innovation.

cs.NI · cs.AI

版本更新 · v2arXiv:2611.16310

Selective Generative Restoration for Video Deraining via Uncertainty-Aware Diffusion

Yi'ang Chen, Han Li, Laibin Chang, Hui Liu, Kui Jiang, Junjun Jiang

原文摘要

Video deraining is challenging due to rain streaks, occlusions, and atmospheric effects. Existing methods mainly follow a deterministic restoration paradigm, assuming that clean scenes can be uniquely recovered from rainy observations. However, rain degradation is spatially heterogeneous. Reliable regions require faithful reconstruction, whereas severely occluded regions contain ambiguous content that require stronger prior-driven inference. Therefore, conventional deterministic restoration often produces over-smoothed results, while global generative restoration may introduce unnecessary hallucinations. We introduce a Selective Generative Restoration paradigm for video deraining, where generative priors are adaptively activated according to restoration uncertainty. Based on this principle, we propose Selective Generative Video Deraining (SGVD), an uncertainty-aware diffusion framework that enables adaptive fidelity preservation and generative recovery. Specifically, SGVD exploits stochastic diversity between dual one-step diffusion predictions to estimate region-level restoration uncertainty, which determines where observation fidelity should be preserved and where generative inference should be activated. Furthermore, an anchor-guided trajectory consistency mechanism is proposed to regularize stochastic diffusion evolution across frames, reducing temporal generation drift and improving video coherence. To achieve practical efficiency, SGVD adopts one-step diffusion inference while maintaining uncertainty-aware generative capability. Extensive experiments demonstrate that SGVD achieves superior restoration performance, perceptual quality, and temporal consistency compared with state-of-the-art video deraining methods.

cs.CV

新论文 · v1arXiv:2611.25603

On the Informativeness of Structural Stability Bounds for GNNs: GINs versus GATs

Alan Gany, Paul Landrier, Bogdan Cautis, Laks V. S. Lakshmanan, Silviu Maniu

原文摘要

Stability bounds for graph neural networks limit how much node representations can drift when the graph changes, but they are rarely compared with the simplest alternative: a bound on the distance between any two outputs, whatever the graph. We derive structural stability bounds for Graph Isomorphism Networks (GINs) and single-head GATv1 networks under deterministic edge perturbations that need not be small, and compare each with such a perturbation-independent diameter bound. For GINs, we characterize exactly when the stability bound is tighter, through a condition linking perturbation size, depth and maximum degree; it holds with a wide margin on Cora, but fails beyond a few layers on the low-degree MUTAG graphs. For GATv1, the bound is never tighter, which calls for analyses based on attention masses.

cs.LG · cs.AI · cs.LO

新论文 · v1arXiv:2611.25599

Tether: Verified Stateful Policy Engine for System-Level Monitoring of Agents

Meghana Aparna Sistla, Alaia Solko-Breslin, Somesh Jha, Aditya Kanade, Krishnamurthy Dj Dvijotham

原文摘要

Autonomous AI agents are increasingly deployed for long-running tasks that execute arbitrary shell commands or programs. However, these actions can result in unauthorized network requests or persistent, malicious changes to the user's file system, processes, or privileges. Existing defense mechanisms rely on either LLM-based monitors, which lack provable safety guarantees, or stateless policy enforcement layers, which miss multi-step data flows and execution side-effects unmodeled by the policy language (e.g., network calls initiated inside a Python script). We introduce Tether, a framework for runtime enforcement of stateful security policies over the observed side-effects of an agent's actions. Tether executes each proposed action in an ephemeral, isolated sandbox, inferring abstract action semantics from observed system calls. Tether then enforces stateful policies over these inferred semantics, permitting actions to proceed only if they comply with the policy. Policies and action semantics are defined in an embedded DSL in Lean 4 capable of expressing diverse constraints, including information flow control (IFC), access control, and network security policies. Because actions can propagate labels from sources to destinations, a single rule can enforce fine-grained IFC policies across multi-step execution flows of arbitrary length. Expressing policies in Lean enables formal verification that they satisfy global security invariants (e.g., preventing network access after reading a sensitive file). We evaluate Tether on realistic coding-agent benchmarks across both benign and adversarial trajectories. With a single policy shared across all tasks, Tether blocks 100% of in-scope attacks while maintaining a benign false-positive rate under 5%.

cs.CR · cs.AI

版本更新 · v2arXiv:2611.24757

Variance-Optimal Counterfactual Policy Design for Sample-Efficient Online A/B Testing

Shashank Gupta, Subhadeep Maji

原文摘要

Online A/B testing is the standard approach for evaluating changes to user-facing systems, such as a new recommendation or ranking policy, before deployment. While the difference-in-means estimator of a standard A/B test is unbiased, it often suffers from high variance. When the treatment and production policies choose the same actions with similar probabilities, the resulting observations can add noise while providing little information about the difference between the policies. Consequently, A/B tests can require large amounts of traffic and long durations to detect small treatment effects. Existing variance-reduction methods mainly improve the estimator after the data-collection scheme has been fixed, and recent overlap-aware designs only optimize a single traffic split between the two policies; thereby, they cannot direct traffic to the actions on which the policies disagree. In this paper, we jointly optimize how actions are sampled and how the resulting observations are used to estimate the treatment effect. Our contributions are two-fold. First, building on pairwise off-policy evaluation, we derive the variance-optimal action-level logging policy for estimating the average treatment effect between two policies, which has a closed-form solution for any fixed control variate. Second, we use historical logs to jointly optimize the logging policy and a control variate to further reduce variance. Our experimental results in simulated policy A/B tests show that our designs estimate the average treatment effect more accurately than existing variance reduction methods on every evaluated setting, while generally improving statistical power and reducing how often the worse policy is selected. As a result, our designs enable practitioners to reach reliable deployment decisions with considerably less online traffic.

cs.LG · cs.IR

新论文 · v1arXiv:2611.25595

LoopedWAM: Looped World Action Models for Elastic Inference

Anh Dao, Xinxin Liu, Hang Yu, Jeffrey A Chan Santiago, Yuzhang Shang

原文摘要

World action models (WAMs) couple generative video modeling with robotic action prediction but typically require large video and action transformers, limiting deployment efficiency. We present LoopedWAM, a compact WAM that recurrently applies a shared six-block core in each expert, executing 30 transformer layers while storing only 12 blocks per expert. To account for their asymmetric inference costs, the video and action experts can operate at different depths: visual features are computed once per query, whereas action computation repeats at every sampling step. We further propose elastic self-distillation, which trains a single elastic checkpoint, LoopedWAM-E, across action-depth and sampling-step budgets. With only 0.58B policy parameters and no robot pretraining, LoopedWAM achieves 96.1% success on LIBERO, 91.9% on RoboTwin 2.0, and 52.2% on RoboCasaGR1, while using $7.4\times$ less peak GPU memory than Fast-WAM under matched inference measurements. On LIBERO, LoopedWAM-E achieves 96.0% success in 27.6 ms or 95.2% in 24.7 ms on an H100, providing multiple accuracy-latency operating points from a single checkpoint; with two sampling steps and four action loops, it stays within 0.5 points of LoopedWAM on all three benchmarks. Ablations demonstrate that weight-shared depth preserves competitive performance, while aggressive action-depth reduction requires explicit budget-aware training. On a real AgileX PiPER arm, LoopedWAM succeeds in 71.3% of 80 trials across four tasks.

cs.CV · cs.RO

新论文 · v1arXiv:2611.25594

Bradley-Terry Training Can Distort Outcome Probabilities of Initially Accurate Language World Models

Xinze Shi, Litian Zhang, Zhenjie Su, Zelong Yu, Yuting Zhang, Qihua Feng, Zifeng Kang, Chaozhuo Li 等 9 位作者

原文摘要

Comparative feedback, in which a judge says which of two candidate outcomes is closer to what happened, is a practical way to train language world models that assign probabilities to an action's outcomes. We show that, fitted with a Bradley-Terry objective, it need not teach the true outcome probabilities and can distort accurate ones. On a simulated robot push with enumerated outcome distributions, this objective takes a model that is accurate on unseen layouts (hazard error 2.4 percentage points, pp) and moves 33 to 43 pp of its probability onto outcomes that none of a condition's enumerated branches produces. The cause is the objective, not the comparisons: its optimum, computed before training, predicts where the model ends up, and a likelihood that knows the judge's rule learns from the same comparisons without the shift. The shift is robust: it recurs on unseen conditions, under five candidate laws and a three-fold denser enumeration, and about 29 pp remain without the physically impossible events. Its cost is conditional: with those events it inflates hazard probabilities, and a threshold inherited from the pre-feedback model costs about 3 pp of utility per decision on trained and 1 to 2 pp on unseen conditions, through false rejects; without them the utility loss averages below 1 pp over six seeds, though a small regret in action choice remains. Audits of readout, response and decision trace the effect from probabilities to actions.

cs.LG · cs.AI · cs.CV

版本更新 · v2arXiv:2611.24138

Ego2X: Contact-aware Transfer from Egocentric Human Data to Any Robot

Xinke Lyu, Bowen Yang, Zhicheng Wang

原文摘要

Embodied models depend on real interaction data, yet collecting it on physical robots is slow and labor-intensive. Recent ego-to-robot pipelines retarget first-person view human videos to robot data via transferring hand and arm motion. They judge fidelity mainly by visual appearance. However, trajectories often fail in terms of physics: fingers hover beside objects, and grippers close on empty space. We argue that how the object moves and the contacts that drive the motion count. Thus, we present Ego2X, an object-centric framework that converts monocular egocentric video into physically executable robot data across dexterous hands and parallel grippers. Ego2X centers on CORE (Contact-centric Object-interaction REpresentation), which records each contact on the object by its surface anchor, as well as relative force, grasp mode, and interaction state. Throughout the pipeline, CORE corrects hand-object drift via object-relative constraints, steers inverse kinematics toward contacts the robot must realize, and preserves held objects during arm erasure and robot compositing. On HOT3D, contact-driven correction reduces in-contact hand depth error by 33.7%, while on EgoDex, contact-driven retargeting realizes 62.9% of annotated contacts, surpassing the official HuRo pipeline (27.9%). In RoboTwin 2.0, post-training X-VLA on Ego2X episodes alone, without simulator demonstrations, raises task success rate from 40.5% zero-shot to 67.0% across 200 paired scenes. We will publicly release our conversion pipeline and CORE-annotated dataset for dexterous hands and parallel grippers.

cs.RO · cs.CV

新论文 · v1arXiv:2611.25592

From Data Alignment to Parameter Generation: Few-Shot Cross-City Traffic Prediction with Conditional Diffusion

Shengchun Wang, Yuhang Lin, Tong Li, Pengyuan Zhou, Andy Chi Lok Tai, Wenjie Chen, Zhu Xiao, Zhu Han

原文摘要

Accurate mobile traffic forecasting is crucial for network service management, supporting proactive capacity provisioning, congestion mitigation, and energy-efficient operation. However, many cities and regions are data-scarce (e.g., newly deployed or poorly instrumented areas), thus making traffic prediction a few-shot learning problem and limiting the practicality of standard data-driven models. Existing cross-city solutions primarily rely on data-space alignment to match traffic distributions between data-rich and data-scarce cities, but such alignment is often brittle under real-world distribution shifts, measurement noise, and city-specific bias. To address these limitations, we propose a Conditional Diffusion-based Parameter Generation (CDPG) framework. Instead of aligning raw data, CDPG learns the conditional distribution of high-performing predictor parameters for source cities and directly generates node-specific model parameters for the target city. To support reliable adaptation for network traffic prediction, CDPG conditions parameter generation on multi-modal spatio-temporal context. Specifically, CDPG captures spatial environment cues by combining satellite imagery with large language model (LLM)-derived semantic descriptions, and summarizes complex temporal behaviors via unsupervised pattern clustering. These contextual components guide CDPG in producing parameters tailored to each base station's operational context, thereby improving robustness under cross-city and cross-region shifts. Extensive experiments on three large-scale datasets show that CDPG consistently outperforms state-of-the-art cross-city few-shot baselines by up to 9.72% on standard forecasting metrics and remains robust for few-shot multi-step prediction. The code is available at https://github.com/LingoodBoy/CDPG.

cs.LG