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

+479新增论文
+361其中 AI 相关
+64已有论文修订
+95新增 AI 导读

arXiv 最近完成扫描:2026-11-16 01:34 · 全站最近入库:2026-11-16 21:56各信源进度 →

统计口径与数据新鲜度

统计 2026-11-16 00:10 — 2026-11-17 00:10(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:2026-11-16 01:34最近发现新增版本:2026-11-16 21:56库内论文最新更新时间:2026-11-14 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。

23,827 已收录论文19,941 AI 相关6,922 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 23,827 篇论文 · 按版本更新时间排序上轮扫描起点 2026-11-16 01:33(北京时间)
版本更新 · v5arXiv:2605.14322

Are Agents Ready to Teach? A Multi-Stage Benchmark for Real-World Teaching Workflows

Zixin Chen, Peng Liu, Rui Sheng, Haobo Li, Jianhong Tu, Xiaodong Deng, Kashun Shum, Dayiheng Liu 等 9 位作者

原文摘要

Language agents are increasingly deployed in professional workflows, yet tutoring remains a high-stakes capability that existing evaluations only partially capture. Effective tutor agents require more than producing correct answers or executing accurate tool calls: they must infer a warranted teaching decision from evidence, adapt support as learner state changes, and carry an instructor's request through a learning-management system (LMS) to a completed, verified intervention. We introduce TeachArena, a source-grounded benchmark that jointly evaluates three complementary surfaces of teaching work: professional pedagogical judgment, situated multi-turn tutoring, and end-to-end LMS teaching workflows. Its 354 tasks are each built around a pedagogical insight, grounded in evidence, and evaluated with matched verifiers over observable turn-level responses, tutoring trajectories, and persistent artifacts or environment states. Across a comprehensive evaluation of frontier models, our findings reveal that current models are generally capable of bounded pedagogical judgment, but still fall short of professional teaching standards in situated tutoring and end-to-end teaching-workflow execution. By unifying teacher judgment, adaptive tutoring, and institutional action in one benchmark, TeachArena provides a measurement foundation for developing tutor agents that can support realistic teaching work.

cs.AI

新论文 · v1arXiv:2611.11149

ZarrFM: A Canonical Tensor Storage Architecture for EEG Foundation-Model Pretraining

Dheeraj Chintapalli, Sunkalp Chandra

原文摘要

Statistical models for brain-computer interfaces (BCIs) increasingly require large, heterogeneous EEG corpora, yet existing datasets vary in sampling rates, channel layouts, window lengths, and storage formats, creating substantial preprocessing and I/O overhead. We propose ZarrFM, an execution-oriented representation that complements BIDS by storing a corpus as a single Zarr v3 hierarchy with indexed metadata, canonical channel spaces, explicit masks, hash-stable subject splits, verified event tables, and block-aligned sharding. On real PhysioNet recordings and synthetic corpora, we compare ZarrFM with EDF, FIF, NPY, HDF5, and Parquet. Across 200 recordings, its two-level index reduces metadata queries from 601 requests to 14. Preserving native signal codes reduces storage from 42.4 MB/hour in EDF to 22.7 MB/hour, while block-aligned sampling reaches 3,131 windows/s, up to 11 times faster than Parquet. Reads remain bit-identical across independent Zarr implementations, and concurrent failure-injected appends recover without data loss. These results show that ZarrFM can provide a compact, interoperable, and execution-efficient storage layer for large-scale EEG pretraining.

cs.LG

版本更新 · v2arXiv:2608.17162

OraclePhys: A Systematic Framework for LLM Fine-Tuning on Structural Mechanics

Mingyu Li, Guorui Song, Jing Lin, Haoqian Wang

原文摘要

What fine-tuning teaches a model is usually diagnosed after the fact. We make it an experimental variable. OraclePhys has three parts. OraclePhys-Bench is a structural-mechanics benchmark whose finite-element oracle grades every answer and counterfactual edit. OraclePhys-30K holds seven answer forms over byte-identical structure descriptions. A controlled training study compares the forms and verifier roles. Two findings follow. First, within this task family the answer form of the label determines what fine-tuning teaches. Ranking supervision lifts held-out localization of the most-deforming story on taller structures from 0.25 to 0.86, a peak-value label yields at best a partial predictor, and a pass/fail label none. This ranking-versus-peak gap replicates on a second domain, a second model family and a paraphrased surface. Second, the gap between our standard SFT and GRPO runs is mostly a recipe effect. With learning rate and step budget matched, the two reach held-out accuracies within 0.03 of each other. At higher rates, GRPO falls below its start while SFT does not. Both also teach the model to apply edits given as instructions. The trained 8B matches a specialist and exceeds a frontier LLM. Both domains are linear, steady-state systems with synthetic text inputs.

cs.LG

版本更新 · v2arXiv:2610.39070

When the Same Prompt Gives Different Answers (and the Same Answer Isn't Always Right): A Taxonomy of Failure Mechanisms, Testing, and Mitigation Strategies for LLM-Assisted Evidence Synthesis Read-and-Evaluate Tasks

Carole Lunny, Ian Goodman, Stefano Berrone, Connor TA Brenna, Emma K Reid, Ivan Perez-Neri, Shailesh Kolekar, Nityanand Jain 等 15 位作者

原文摘要

LLM read-and-evaluate tasks include study screening, data extraction, risk-of-bias assessment, and critical appraisal in the evidence synthesis field. These task differ in the amount of LLM interpretation they require. Some tasks involve retrieving relatively simple information, such as an author name, sample size, or publication year. Others require the LLM to read source material, interpret it against predefined criteria, and make a judgement. Researchers should assess whether LLM judgements are accurate, supported by appropriate source evidence, and reproducible under specified conditions. A correct judgement does not necessarily indicate that the underlying evaluation process is reliable. We developed a taxonomy for identifying mechanisms that may compromise these properties and for informing the selection and reporting of appropriate evaluation methods. The taxonomy outlines potential ways that LLM judgments from read-and-evaluate tasks can fail, called TRAILS (Taxonomy of Reliability And Instability in LLM-assisted evidence Synthesis). The TRAILS taxonomy addresses two complementary questions. First, what mechanisms can cause the same read-and-evaluate task to produce different judgements across repeated LLM runs? Second, what mechanisms can systematically affect the judgement or required output in a way that is not apparent from the judgement itself? This study presents 38 mechanisms and identifies the deployment settings in which each can operate, approaches for testing them, and mitigation strategies. We also show why repeatability or accuracy does not rule out a hidden judgement failure and why human evaluation should assess whether a judgement is grounded in the supplied evidence, independent of inappropriate external influence, and produced through a process that can be examined and verified.

cs.AI · cs.PF · stat.ME

新论文 · v1arXiv:2611.11148

MOSU2: Long-range Navigation System with Multi-modal Scene Understanding

Jing Liang, Zhenyu Hou, Senthurbavan Kirubaharan, Md Safwan Mondal, Ashok Gudivada, Xuesu Xiao, Dinesh Manocha

原文摘要

Robots that travel long distances through public outdoor spaces must do more than avoid obstacles: they must remain on terrain suited to their mobility and decide when to proceed, yield, or detour around pedestrians, cyclists, vehicles, traffic signals, and temporary restrictions. These decisions are difficult because their correctness often depends on how a scene evolves, such as whether an approaching car is slowing or a cyclist will cross the robot's path. Vision-language models (VLMs) can interpret such cues, but they reason over images that lack metric geometry and motion. Supplying this evidence at every step is too slow for real-time navigation, and hand-crafted triggers cannot anticipate the open-ended variety of outdoor situations. We present MOSU2, a long-range navigation system with two-stage trajectory selection, in which first-stage contextual reasoning determines when additional evidence is needed and what to retrieve. Only when an interaction remains unresolved does the first stage trigger the second stage to retrieve geometric and dynamic information and align it with scene context to make navigation decisions. Building on MOSU, MOSU2 follows routes from GPS-based route planning and filters candidate trajectories through multi-modal traversability analysis before selecting a trajectory to move the robot. In real-world experiments spanning stop gestures, construction sites, crosswalks, and roads shared with vehicles, cyclists, and pedestrians, MOSU2 raises mean navigation success from 61.8% to 94.2% and condition-detection accuracy from 72.2% to 91.4% over its single-image predecessor, at about 2.2 times its mean decision latency. Project website: https://jinggm.github.io/mosu2/

cs.RO

版本更新 · v2arXiv:2606.03014

MOSAIC: Efficient Mixture-of-Agent Scheduling via Adaptive Aggregation and Inference Concurrency

Saptarshi Mitra, Yifan Zhang, Rachid Karami, Sreetama Sarkar, Phyo Pyae Moe Aung, Nazmul Takbir, Souvik Kundu, Sitao Huang

原文摘要

Mixture-of-Agents (MoA) systems improve reasoning accuracy by routing each query to multiple expert LLMs and aggregating their outputs. Efficiently executing this workload on limited GPU resources has bottlenecks. Skill-based routing creates skewed expert demand, and combining instruction-tuned LLMs with long-reasoning models results in extreme variability in generation lengths. Consequently, traditional scheduling strategies suffer from significant GPU idling and throughput collapse due to load imbalances. We present MOSAIC, a scheduling framework for accelerating batch-oriented MoA workloads. MOSAIC uses two-stage workload-aware scheduling: an expert placement stage, where an integer linear program (ILP) decides expert placement, followed by a query-sequencing stage that schedules requests based on predicted output lengths. Additionally, MOSAIC uses confidence-aware adaptive aggregation, leveraging inter-expert agreement to bypass the heavy aggregator LLM for consensus queries. On a 4-GPU system, MOSAIC achieves 2.7-3.4x end-to-end speedups over round-robin scheduling and reduces the mean time-to-answer by 3.9-7.8x, while matching accuracy within +/-1pp.

cs.LG · cs.AR

新论文 · v1arXiv:2611.11147

Not All Layers Are Equal in Image-to-Video Transfer

Thinesh Thiyakesan Ponbagavathi, Constantin Seibold, Alina Roitberg

原文摘要

Parameter-efficient image-to-video transfer adapts frozen image foundation models to video via lightweight trainable adapters. Existing methods insert the same adapter at every transformer layer, implicitly assuming all layers require equal temporal modeling. A layer-wise sensitivity analysis across four video-native model families and two pretraining paradigms reveals that this assumption is wrong: temporal sensitivity increases with depth, and the pretraining paradigm determines its growth profile. Vision-only models exhibit a gradual rise, while vision-language models show early suppression followed by late recovery. This finding motivates STRAP, a training-free adapter placement rule that derives a spatial-to-temporal transition boundary from a paradigm-matched video model. Spatial adapters are assigned before this boundary and temporal adapters after it, without modifying the adapter architecture itself. On four fine-grained action recognition benchmarks, STRAP improves accuracy by up to 11.8% over uniform placement, reduces trainable parameters by 29 to 48%, and lowers inference time by up to 10%. On standard benchmarks, it matches accuracy with fewer parameters. Code is available at https://github.com/th-nesh/STRAP.

cs.CV

版本更新 · v2arXiv:2610.30946

Tracing Hallucinations in LLM Reasoning for UML Class Diagram Generation: An Empirical Taxonomy and Mitigation Study

Mojdeh Rahmanian, Foutse Khomh, Yanchao Yu, Ashkan Sami

原文摘要

Large Language Models (LLMs) can generate UML class diagrams from natural language requirements, but hallucinations may compromise the correctness and completeness of the resulting diagrams. Existing evaluations mostly assess the final diagram, without examining the intermediate traces of reasoning-capable LLMs, which can reveal where these hallucinations arise. We develop an empirical taxonomy of eleven hallucination sub-types organised into three dimensions: Prompt Faithfulness (deviations between requirements and reasoning), Code Faithfulness (discrepancies between reasoning and generated PlantUML code), and Reasoning Quality (problems within the reasoning). The taxonomy was derived from 60 manually annotated DeepSeek-R1 traces, applied to 120 further DeepSeek-R1 traces through a dual-LLM annotation pipeline, and examined on 20 DeepSeek-V4-Pro traces. Code Faithfulness was the most frequent dimension for DeepSeek-R1 in both datasets, whereas Prompt Faithfulness dominated for DeepSeek-V4-Pro, suggesting model-dependent failure patterns. In the manually annotated traces, reasoning-trace length showed no significant monotonic association with hallucination frequency, whereas requirement complexity was negatively associated with it, particularly for Prompt Faithfulness errors. In a mitigation experiment on 13 Golden UML ModelSet cases with three Qwen3-based models, adding Prompt Faithfulness and Reasoning Quality constraints to the generation prompt raised the pooled mean F1 score by 4.0 percentage points, whereas a post-hoc Code Faithfulness review provided no consistent further benefit. Relationship modelling remained the lowest-scoring element throughout. These results suggest that localising hallucinations within the generation process can inform targeted, prompt-based mitigation. Whether these findings generalise beyond the DeepSeek and Qwen3 model families remains to be tested.

cs.SE · cs.CL

版本更新 · v3arXiv:2511.07365

Private Sparse Sketches for Linear Regression

Shrutimoy Das, Anirban Dasgupta

原文摘要

Sketching compresses a large dataset into a smaller summary, allowing regression on the sketch to approximate the full-data objective. A sketch released under differential privacy can be reused with any non-private regression solver, any number of times, without further privacy loss. We study private sparse sketches for $\ell_2$ and $\ell_1$ regression. We first prove lower bounds on the noise required by common sparse sketches for both problems when the sketching matrix is public. Since regression on a sketch can be solved without revealing the sketching matrix, we construct sketches whose randomness is kept secret. Hiding how individual records enter the sketch provides an additional source of privacy, allowing less noise for the same privacy guarantee. We prove differential privacy guarantees for both constructions, and experiments on real datasets show improved utility over public-sketch baselines.

cs.LG · stat.ML

新论文 · v1arXiv:2611.11145

DISCOVER: Online Variance-Guided Data Discovery for Budgeted Multimodal GRPO

Peyman Gholami, Shivam Chandhok, Saber Malekmohammadi, Shayan Shams, Robert Xiao, Leonid Sigal

原文摘要

Recent advances in reinforcement learning have made verifiable rewards a central post-training paradigm for improving multimodal reasoning. However, methods such as GRPO remain data- and compute-inefficient, spending rollout and annotation budget on prompts that provide little learning signal. We propose DISCOVER, an online data-discovery framework for budgeted multimodal GRPO that enables vision-language models to dynamically identify prompts that are most useful for the current policy. DISCOVER first organizes the multimodal sample pool into semantic clusters and queries a small set of density-diverse representative samples to obtain initial reward observations. During training, it estimates the utility of unannotated samples by transferring sparse reward feedback from queried samples. Since sample utility changes as the policy evolves, DISCOVER dynamically updates these estimates with recency-weighted rollout observations. It then selects only high-utility prompts for annotation using a policy-conditioned score that accounts for reward variance and difficulty. Once the discovery budget is exhausted, DISCOVER continues with utility-based replay. Experiments on ViRL39K and VLAA-Thinking with Qwen3-VL-2B show that DISCOVER substantially improves multimodal mathematical reasoning under a strict 10% unique-sample budget, outperforming budget-matched random selection, hard-example sampling, full-pool GRPO under the same training-step budget, and online selection baselines. Ablations further validate the importance of semantic coverage, online reward transfer, and stable discovery cadence, establishing DISCOVER as a practical framework for data-efficient multimodal RL post-training.

cs.LG · cs.AI

版本更新 · v3arXiv:2601.02915

ChemBART: Pre-training a Reaction Language Model for Comprehensive Synthesis Planning

Kenan Li, Yijian Zhang, Jin Wang, Haipeng Gan, Zeying Sun, Xiaoguang Lei, Hao Dong

原文摘要

Recent advances in large language models (LLMs) have demonstrated transformative potential across diverse fields. However, in computer-aided synthesis planning (CASP), current LLM methodologies are often restricted to single-molecule representations, limiting their application to isolated, single-task scenarios such as precursor prediction. Addressing this gap, we present ChemBART, a BART-based language model that adopts a reaction-centric pre-training strategy. By leveraging a mask-filling pre-training task on holistic reaction expressions (Reactant > Reagent > Product), ChemBART effectively solves a variety of chemical problems, including precursor/reagent generation, temperature-yield regression, and molecular property classification. The reaction-centric pre-training further enables the integration of ChemBART with Monte Carlo tree search (MCTS) for multi-step synthesis route design, where the model serves to optimize the policy and value functions within a reinforcement learning framework. Beyond single-molecule pre-trained LLMs that are constrained to specific applications, ChemBART unifies diverse chemical tasks to facilitate synthesis planning with comprehensive implementation information. Most significantly, the proposed multi-step synthesis routes and reaction conditions were validated in the wet lab, confirming the discovery of shorter synthetic pathways with a >20% yield improvement over established literature benchmarks. Our work highlights the potential of reaction-focused pre-training and showcases the broad utility of ChemBART in closing the loop between in silico prediction and wet-lab synthesis.

cs.LG · cs.CL

新论文 · v1arXiv:2611.11141

TALK: One-Shot Batch Design for Protein Variant Effect Prediction

Qiao Huang, Ao Shen, Jie Du, Manning Wang

原文摘要

Protein variant effect prediction is central to understanding protein function and guiding protein engineering, yet experimental labels remain costly. In many protein experiments, the variants that can be assayed are dictated by protocol constraints, while the variants that need accurate prediction may lie in a different or broader region of the landscape. This assay-prediction mismatch makes candidate-local criteria such as predicted function, uncertainty, or exploration value insufficient for measurement design. Given a protocol-defined accessible space and a limited budget, we ask which variants should be assayed so that the resulting data best improves landscape prediction. We propose TALK, which estimates each candidate's contribution through posterior coupling between accessible and target variants and reduces redundant information within the batch. Across retrospective ProteinGym and FLIP protocols, TALK generally improves target-space prediction under limited measurement budgets. These results point to a practical two-stage workflow: first, spend a limited assay budget on variants selected to be informative for the target landscape and train a predictive model from the measured labels; then use this model to score and prioritize high-potential variants. Code is available at https://github.com/joehuang618/TALK.

cs.CE · cs.LG

版本更新 · v3arXiv:2604.08884

Beyond RGB: Benchmarking and Enhancing MLLMs for Hyperspectral Image Understanding via Training-Free Reasoning Framework

Xinyu Zhang, Zurong Mai, Qingmei Li, Xiaoya Fan, Zjin Liao, Haoyuan Liang, Yibin Wen, Yuhang Chen 等 16 位作者

原文摘要

Multimodal Large Language Models (MLLMs) have achieved strong performance on RGB image understanding, yet their ability to use spectral evidence beyond the visible range remains largely unexplored. Hyperspectral imagery (HSI) provides dense spectral measurements that reveal material and environmental cues unavailable in RGB, but current MLLMs cannot directly ingest high-dimensional HSI. To study this gap, we introduce HM-Bench, an evidence-grounded benchmark for hyperspectral image understanding with MLLMs. HM-Bench contains 19,337 question--answer pairs from 2,178 hyperspectral samples across 13 task categories, covering general perception, spectral reasoning, and spatial--spectral reasoning. To make HSI accessible to the native image--text interfaces of existing MLLMs, we further propose VSR$^{2}$, a training-free Visual--Spectral--Report Reasoning framework. VSR$^{2}$ represents each HSI sample with three aligned views: an RGB image for visual semantics and spatial context, a PCA-based image for dominant spectral variation, and a structured report for quantitative spectral--spatial evidence. Under a controlled RGB-only versus VSR$^{2}$ evaluation protocol, experiments on 17 representative MLLMs show that HSI information improves average accuracy from 38.90% to 40.95%, with gains varying substantially across models and task categories. These results indicate that HSI information (beyond RGB) is useful for current MLLMs, while robust hyperspectral reasoning remains an open challenge.

cs.CV · cs.AI

版本更新 · v2arXiv:2611.08140

FATE: Fault-Aware Transfer Error for Unsupervised Anomaly Detection

Tao Wang

原文摘要

We study unsupervised industrial anomaly detection in the single-class setting, where a detector learns only from normal images of one product category. Reconstruction-based detectors model normal features from a frozen encoder, but their target layers are usually chosen by architectural convention. We instead ask what information one block can pass to another. Treating depth as a sequence of discrete transitions, we train eight directed, token-wise predictors from block 1 and blocks 21-23 to blocks 7 and 14. Adjacent-layer cosine similarity drops at four repeatable fault lines in DINOv2 ViT-L, separating stable depth segments. The deepest blocks retain semantic context but lose local detail, so the residual of a source-to-sensitive prediction measures information that cannot be transported. The fault structure constrains the source and target roles, and the same rule transfers to ViT-B and RADIO-B. A bounded foreground prior, calibrated only on normal training images, suppresses background residuals while retaining a positive floor for evidence outside the estimated foreground. With the same eight channels and a common training-only calibration procedure in every single-category model, category-averaged image AUROC reaches 99.75 on MVTec-AD, 99.42 on VisA and 94.27 on Real-IAD; object-level AUROC on Real-IAD is 96.26. For the 448x448 runs, category-averaged P-AUROC/P-AUPRO on a common 256x256 evaluation grid is 98.46/95.97 on MVTec-AD, 99.21/96.12 on VisA and 99.26/96.59 on Real-IAD.

cs.CV

版本更新 · v3arXiv:2610.01425

Why Does Train-Validation Separation Emerge? Update-Pressure Density Dynamics in Pretrained Backbones

Yuchen Li, Zongqi Fan, Ken-Tye Yong, Nguyen H. Tran

原文摘要

During fine-tuning, training accuracy often continues to improve after validation accuracy levels off, widening the empirical gap. We propose a support-transfer account: early updates reuse broadly shared structure, while later fitting relies increasingly on narrower structure with less benefit for held-out examples. From fixed training probes, we normalize per-example gradient magnitudes and measure their pooled dispersion, $D_{\mathrm{den}}$, and their probe-count-weighted within-class dispersion, $D_{\mathrm{within}}^{\mathrm{wt}}$. A local model gives conditions for dispersion and directional cross-entropy separation to grow together, separating uniform logit scaling; additional decision-boundary conditions characterize accuracy-gap growth. A controlled five-level hierarchy tests the effect of sample-private support. A 240-branch intervention matrix crosses training-progress anchors, batch sizes, private shares, and gradient-differentiation doses. The negative gap dose response is clearer at later training stages and is retained across the four batch sizes. A further 48 branches replicate its direction across additional seeds at a fixed stage and batch. Both normalized readouts are positively associated with the accuracy gap in all 90 language fine-tuning runs across three backbones and six tasks. The same direction holds in all 20 vision runs across four datasets. Together, the controlled hierarchy, targeted intervention, and real-model trajectories support a connection between evolving gradient differentiation and train-validation separation.

cs.LG

新论文 · v1arXiv:2611.11140

SurgWorld: Learning Surgical Representations Beyond Observation via Latent Prediction

Jingkun Yue, Pengxiang Wang, Siqi Zhang, Zinan Jia, Yiyao Zhang, Huihuan Xu, Yichang Hao, Xiaohong Liu 等 10 位作者

原文摘要

Surgical procedures are inherently partially observable, requiring surgeons to continuously infer evolving anatomical dynamics and future surgical states beyond the current observation. However, existing self-supervised learning methods primarily learn from observed surgical videos through masked reconstruction or predicting representations within the camera's current field of view, lacking the capability to explicitly model the out-of-sight surgical dynamics. Inspired by latent predictive world models, we introduce SurgWorld, a video-native, self-supervised framework for surgical representation learning, which treats a specified spatial or temporal shift as a pseudo-action and learns to predict the target latent representation from a partial observation. Specifically, we propose Out-of-Sight Dynamics Modeling (OSDM) as the pretraining objective for predicting representations conditioned on relative spatio-temporal shifts. We further regularize the learned latent dynamics from OSDM with a Geometric Visual Similarity (GVS) objective by aligning patch-level relational structures between predicted and target representations, preserving spatial geometry and local visual coherence for dense perception. These two objectives are integrated through an agglomerative training framework that consolidates complementary supervisory signals from multiple teachers to learn a unified latent representation of surgical dynamics and visual geometry. Extensive transfer learning experiments on 17 surgical benchmarks show that SurgWorld consistently outperforms existing vision foundation models across surgical workflow recognition, fine-grained action understanding, and dense visual perception. We further conduct tailored qualitative and quantitative analyses, demonstrating that SurgWorld effectively captures surgical dynamics beyond the directly observed context through latent prediction.

cs.CV

版本更新 · v2arXiv:2605.23522

Precise: SDE-Consistent Stochastic Sampling for RL Post-Training of Flow-Matching Models

Jade Zou, Tao Huang, Weijie Kong, Junzhe Li, Yue Wu, Qi Tian, Jiangfeng Xiong, Jianwei Zhang 等 10 位作者

原文摘要

Reinforcement learning (RL) has become an effective way to improve prompt alignment and perceptual quality in diffusion and flow-matching generators. A critical step for applying online RL to flow matching is turning the deterministic sampling trajectory into a stochastic policy, typically by replacing the reverse-time Ordinary Differential Equation (ODE) with a Stochastic Differential Equation (SDE). The stochastic sampler, controlling the exploration behavior and denoising dynamics, is thus part of the policy, and its design can significantly affect the reward optimization performance. We break down the sampler design into two interdependent components: choosing the right amount of stochastic exploration, and discretizing the resulting SDE faithfully at the small step counts used in RL. To address the first component, we analyze the inherent tension between exploration and stability in denoising and derive an SDE schedule that balances the two. Turning to the discretization challenge, we use a toy example to show that existing samplers can deviate from the flow-matching process, either by introducing excessive discretization noise or by relying on heuristic rules that do not guarantee convergence to the data distribution. To address these issues, we propose Precise, a new stochastic sampler that balances effective exploration with stability. Crucially, Precise keeps the denoising trajectory SDE-consistent through a novel approximation that freezes the clean-latent posterior mean, resolving the excess noise issue in standard samplers. Extensive experiments demonstrate that this formulation leads to significantly faster and more stable reward optimization via reinforcement learning, achieving state-of-the-art alignment scores (e.g., PickScore, HPSv2.1) while requiring 13.1-53.2% less wall-clock training time to match the best in-domain performance of prior samplers.

cs.LG · cs.AI · cs.CV

新论文 · v1arXiv:2611.11139

Graphene: A Lightweight Transformer with Ising Mean-Field Refinement for Image Super-Resolution

Seungki Hong, Taekwang Jang

原文摘要

This paper proposes Graphene, a lightweight image super-resolution architecture with two complementary feature refinement mechanisms: Ising-model-inspired local feature refinement using an Ising Refinement Block (IRB) and stage-history-based feature refinement using a History Attention (HistAttn) module. The IRB reuses attention output features and pre-softmax attention logits to construct signed pairwise token couplings within each local window, applies unrolled mean-field updates, and refines features through a gated residual correction. These signed couplings allow alignment and opposition between latent spins to modulate local feature refinement. HistAttn attends over the current stage output and earlier representations at the same token position and adaptively blends the current stage output with a history-weighted feature aggregate without adding spatial pairwise attention. On Urban100 at $\times4$, Graphene and Graphene-light improve over IET and IET-light by 0.24 dB and 0.38 dB, respectively, with 34.3%/13.7% fewer parameters, 44.8%/44.4% fewer FLOPs, 78.4%/71.5% lower latency, and 76.0%/85.5% lower peak memory. Real-world SR comparisons further show that Graphene recovers fine details and preserves textures across diverse degradations. Code and reference checkpoints are available at https://github.com/skethz/Graphene

cs.CV

版本更新 · v2arXiv:2606.12616

PersonaDrive: Human-Style Retrieval-Augmented VLA Agents for Closed-Loop Driving Simulation

Mahmoud Srewa, Praneetsai Vasu Iddamsetty, Mohammad Abdullah Al Faruque, Salma Elmalaki

原文摘要

Closed-loop driving simulators mostly fill their environments with traffic agents that behave alike, whether they are rule-based or learned. Recent methods that control driving style learn it from labels inferred from existing driving recordings, from synthetic data generated to reflect each style, or from reward weights estimated by a large language model (LLM). All of these are indirect estimates of a style, not recordings of people who were instructed to drive in it. We introduce PersonaDrive, a vision-language-action (VLA) driving agent that takes its style from retrieved human demonstrations. The demonstrations come from a new dataset in which 15 participants drove CARLA routes on a driving simulator under conservative, neutral, and aggressive instructions. PersonaDrive trains a lightweight retrieval head on features from a frozen image encoder and on control signals over per-style databases, and fine-tunes one VLA backbone over the retrieved demonstrations. Switching style only means searching a different FAISS index, with no retraining. On the Bench2Drive benchmark without style conditioning, PersonaDrive improves the Driving Score by 4.6% over SimLingo and by 2.5% over HiP-AD, a gain that fine-tuning without retrieval does not reproduce. With style conditioning, PersonaDrive outperforms the strongest baseline, Drive My Way, in every style, with a Driving Score 6.7% to 11.6% higher. Its driving also follows the instruction, since average speed rises by 19.8% from the conservative to the aggressive style. Each traffic agent is evaluated individually as a driving policy. The dataset and model weights are available on the project page: https://pervasiveautonomylab.github.io/personadrive-landpage/.

cs.AI · cs.CL

新论文 · v1arXiv:2611.11137

GlobalTap: Globally Diverse Beat Tracking Benchmark via Crowdsourced Tapping

William T. Botticelli-Wells, Manuel Anglada-Tort, Elif Celen, Rena Far, Harin Lee, Peter M. C. Harrison, Nori Jacoby

原文摘要

Existing beat-tracking benchmarks draw from a narrow range of musical cultures, predominantly Western, and are difficult to scale due to reliance on expert annotation. We address both limits with GlobalTap, a benchmark of 200 popular-music excerpts spanning 59 countries, annotated by 746 participants recruited across ten countries. The benchmark is created using a scalable online tapping pipeline that records taps via audio for high temporal precision, aggregates pooled tap times into consensus beat annotations, and computes intrinsic quality measures that flag perceptually difficult excerpts without requiring a reference. In a separate study on a different 240-excerpt validation set, tapped by an additional 733 participants, an independent group of 130 listeners rated how well the resulting beat tracks lined up with the music, and we detected no overall difference between our annotations and existing reference annotations. Crowd-derived annotations occasionally diverged from the references in phase and metrical interpretation, yet listeners did not consistently prefer the reference versions, suggesting that crowd-derived and existing reference annotations can reflect different perceptual or structural interpretations. The disagreements were often systematic rather than random, and metrical-level choices differed across the ten recruitment countries. We release GlobalTap together with open-source pipeline code as a listener-derived resource for evaluating beat-tracking systems on culturally diverse music.

cs.SD · eess.AS

新论文 · v1arXiv:2611.11136

From Game Generation to Game Playing: Fully Automated Ability Training for Vision-Language-Action Models

Changran Liu, Qi Wu, Wenchao Ma, Sharon X. Huang, Haomiao Jiang

原文摘要

Training vision-language-action (VLA) models to follow gameplay instructions across diverse games is limited by the cost of collecting demonstrations. We introduce a framework that automatically generates game environments and expert demonstrations. An LLM-powered agent uses game-building tools to construct executable environments and tasks targeting specified skills. Programmatic experts then use privileged simulator state to generate aligned instructions, rendered observations, and keyboard and mouse actions. We train our game policy, Script2Play, by post-training a 4B vision-language model with behavior cloning on 150.3 hours of synthetic Minecraft and Roblox gameplay, without human gameplay demonstrations or reinforcement learning. Evaluation on multiple public game benchmarks demonstrates that Script2Play follows gameplay instructions effectively and generalizes across game engines, highlighting the effectiveness of targeted synthetic supervision.

cs.CV

版本更新 · v2arXiv:2611.01077

Multi-Turn Semantic Caching for Low-Cost LLM Serving

Baran Atalar, Selly Beyene, Xutong Liu, Carlee Joe-Wong

原文摘要

Semantic caching lowers the cost of serving large language models (LLMs) by reusing stored responses for similar requests. In multi-turn conversations, whether a cached response is acceptable depends on both conversations' histories, which embedding similarity of requests and histories predicts poorly (AUC 0.62--0.74 on five workloads). We therefore learn the \emph{reuse risk}, the probability that a cached response is unacceptable, from judge feedback, and choose a budgeted set of conversation states to cache by weighing the generation cost a reuse avoids against a price for unacceptable reuse. We propose CUCB-SC-MT, which learns a cache from logged conversations with a suboptimality guarantee, and CLCB-LS-MT, which learns online and has $\tilde O(\sqrt N)$ regret over $N$ requests when conversation states recur. On math-tutoring (SocraTeach) and movie-ticketing (Taskmaster-3) conversations, both achieve lower loss than prior semantic caching approaches: ContextCache, multi-turn variants of ElastiCache and vCache, and single-turn semantic caching at every cache budget. With 1,000 cache entries, CLCB-LS-MT avoids 35--41\% of the generation cost through acceptable reuses, against at most 22--25\% for the baselines, and lowers estimated serving latency by 19--35\%.

cs.LG · cs.CL · cs.DC

版本更新 · v3arXiv:2604.24093

The Power of Menus in Dynamic Pricing: Optimal Regret and Equivalence Results

Shiliang Zuo

原文摘要

We study dynamic pricing where a seller repeatedly interacts with a strategic, non-myopic buyer who has a fixed private valuation and discounts future utility. Prior work focused exclusively on posted-price mechanisms, where the seller gives a take-it-or-leave-it offer. For our first result, under geometric discounting $γ_t = γ^{t-1}$, we show that menu mechanisms consisting of allocation-payment contracts achieve $O(T_γ)$ regret, where $T_γ$ is the buyer's effective discounted time horizon. We also establish a $Ω(T_γ)$ lower bound, demonstrating the bound is tight. Considering the geometric discounting buyer with a constant discount factor, our bound is $O(1)$, while prior bounds using posted-price mechanisms incur an unavoidable $Ω(\log\log T)$ factor in regret. Our second contribution is more conceptual in nature. The problem of dynamic pricing sits at the intersection of two paradigms: learning with strategic agents in computer science / machine learning and revelation-principle-based mechanism design in economics, yet their relationship has remained unclear. We establish an equivalence between these two classes of mechanisms in our single-buyer setting: any adaptive indirect menu mechanism can be converted into an incentive-compatible direct mechanism with no less revenue, and conversely, any such direct mechanism can be implemented by an adaptive indirect menu mechanism with no less revenue. Consequently, the two classes achieve identical optimal regret. The adaptive, data-driven algorithms of online learning and explicit type elicitation from direct revelation type mechanisms are two languages towards solving the same problem.

cs.GT

版本更新 · v2arXiv:2610.34248

HNSecW: A Logarithmic-Time MPC Protocol for Approximate Nearest Neighbor Search

Fahong Zhang, Weijie Jin, Lixu Wang, Ning Luo, Xiao Wang, Wei Dong

原文摘要

As a core component of modern RAG systems, approximate nearest neighbor (ANN) search has been extensively studied for decades. In practice, deploying ANN often requires symmetric privacy, where the querier's query and the database owner's data must remain hidden from each other, necessitating a secure multi-party computation (MPC) solution. So far, most MPC-based ANN protocols follow a cluster-based approach: they evaluate distances for the query over a large candidate set and rely on heavyweight primitives such as distributed ORAM to hide access patterns. As a result, each search typically requires $O(\sqrt{N})$ time. We propose HNSecW, a novel MPC protocol for ANN search that achieves the $O(\log N)$ search time for the first time. We first study the single-query setting. Starting from HNSW, the state-of-the-art ANN algorithm, we redesign its control flow and data structures to achieve fully data-independent execution. To further achieve logarithmic search time, we employ offline reshuffling and partial database duplication. Furthermore, we extend our protocol to support multiple sequential queries by introducing a "double visited" caching structure that eliminates redundant cross-query memory accesses, thereby avoiding per-query online shuffling. In addition, we develop a batch-query protocol that exploits higher parallelism to reduce both round complexity and the circuit's critical path. Our experiments show that HNSecW preserves the high accuracy of plaintext HNSW ($\geq 90\%$ recall) and delivers over a $20\times$ efficiency improvement over prior work on million-scale datasets.

cs.CR · cs.DB

版本更新 · v2arXiv:2608.07555

You Don't Need To Stay in The Loop: An Agentic Robotics Loop for Robot-Policy Improvement

Hang Yu

原文摘要

Coding agents close the software loop for training a robot policy without having a human in the loop: a main agent manages the training, subagents analyze and execute, treat robot policies and hardware as tools and do the work. We port this architecture to robot-policy improvement, where one difference dominates the design: robotic tools such as trained policies, training pipelines, and data collection fail routinely, so a tool's quality must be measured, recorded at every call, and expired when the artifact behind it changes. AgenticRobotics is a backend-independent control plane in which an LLM controller drives disposable workers through durable train--evaluate--improve transactions: an immutable objective, controller-owned measurement, commit-keyed crash recovery, an evidence-graded skill library, and a tool registry with a standardized, recorded call surface. The title is an operational claim, not a selection claim: the operator can leave because promotion is evidence-gated, state is recoverable, and capability quality is derived from records, not because the loop picks better checkpoints than a human. On the one lineage we measured, it does not. The gates provide measured false-promotion control (0.001 false promotions per run with the hardened margin, against 0.005--0.021 with the shipped rule), anytime-valid decisions under optional stopping, zero lost or duplicate effects under kill injection, and a signed verifier that caught six of six artifact-tampering classes.

cs.RO

新论文 · v1arXiv:2611.11133

Online Correlated Exponential Clocks

Zhiyi Huang, Kalen Patton, Peilin Yang, Jiahao Zhao

原文摘要

In online resource allocation, we seek to allocate an incoming sequence of requests $j \in J$ among a set of offline resources $i \in I$, where decisions must be made immediately without knowledge of the future. State-of-the-art results typically follow from relax-and-round methods. For many problems, recent advances have come from improved methods of online rounding with negative correlations. However, such rounding schemes have largely been implemented on a case-by-case basis, making it difficult to transfer improvements and new ideas between problem settings. This leads us to ask the question: Can we develop unified methods of online rounding with negative correlations which can obtain state-of-the-art guarantees across different resource allocation settings? We answer this question affirmatively by introducing the framework of Online Correlated Exponential Clocks (OCEC). In this framework, fractional allocations $x_{ij}$ are assigned exponential clocks $Y_{ij} \sim \mathrm{Exp}(x_{ij})$, and each request $j$ is allocated to the resource $i$ with smallest $Y_{ij}$. The key idea is that the sequence $(Y_{ij})_j$ for each fixed $i$ is negatively correlated via an OCEC generator, resulting in better-than-baseline guarantees. Our contributions are two-fold: First, we introduce OCEC rounding as a problem-independent interface which any online algorithm can use. This isolates the challenge of online rounding into the self-contained problem of designing an effective OCEC generator. Second, we demonstrate the flexibility of this method by giving a single OCEC generator which yields improved competitive bounds in three distinct problems: 0.556 for Display Ads, 0.530 for AdWords, and 4.69 for squared-load balancing. Previous best results were obtained by Blanc and Charikar (FOCS 2021), Chen, Huang, and Sun (FOCS 2024), and Borst and Kashaev (SODA 2027), respectively.

cs.DS

版本更新 · v2arXiv:2610.00896

Seamless Reconfiguration for DAG-BFT (Technical Report)

Michael Yiqing Hu, Leander Jehl, Paul Franke-Bergmann, Jialin Li, Christian Berger

原文摘要

Byzantine Atomic Broadcast in Asynchronous networks has been studied extensively for decades. In a fully asynchronous system with even one possible crash failure, the FLP impossibility result rules out deterministic consensus, motivating randomized protocols that achieve termination with probability one. More recently, DAG-Rider popularized a new abstraction in which processes continuously reliably broadcast blocks to construct a Directed Acyclic Graph (DAG) and then locally derive a total order using a randomized perfect coin. This separation of dissemination from ordering has renewed interest in asynchronous DAG-based Byzantine Fault Tolerance (BFT) and inspired numerous protocols that improve expected latency or throughput. Reconfiguration, however, remains largely unexplored. Existing DAG-based BFT protocols generally assume a static set of processes, whereas deployed systems must occasionally add or remove nodes in response to operational demands. Existing solutions are not seamless; they either pause the system or delay membership changes until a predetermined epoch boundary. This may be impractical and can introduce downtime. A single agreed switching round is difficult to use seamlessly: by the time processes learn it, the dissemination layer may already have advanced beyond it. In this work, we present a seamless reconfiguration protocol for Asynchronous DAG-based Byzantine Atomic Broadcast. Our protocol allows the DAG to continue growing while the membership and corresponding quorum thresholds change, without halting the system. To the best of our knowledge, ours is the first protocol to provide seamless reconfiguration for DAG-based Byzantine atomic broadcast in a fully asynchronous network.

cs.DC

新论文 · v1arXiv:2611.11132

Confidence-driven adaptive semantic transmission for task-oriented communication

Suman Das, Prajnyajit Mohanty, Siddhartha S. Borkotoky

原文摘要

Image-based remote monitoring over wireless networks faces a trade-off between communication cost and classification reliability. Transmitting complete images is communication-intensive, whereas relying solely on lightweight edge classifiers may compromise accuracy. To address this challenge, we propose Confidence-Driven Adaptation (CDA) for image transmission and classification. An edge device (ED) captures images and applies a lightweight classifier before transmission to a remote server. Based on its classification confidence, the ED selects among three transmission modes with varying levels of information: a short text string reporting its predicted class when confidence is high, a compressed Region of Interest (ROI) at intermediate confidence levels, or an uncompressed ROI when confidence is low. The server accepts the predicted class directly in the first case and performs classification using a more powerful classifier in the other two cases. Numerical evaluations under bandwidth-limited, lossy links and latency-constrained settings demonstrate that CDA achieves high task accuracy at the receiver compared with non-adaptive baselines that transmit only the edge-classifier outcome or the uncompressed ROI.

cs.NI · cs.CV

新论文 · v1arXiv:2611.11131

CrossLiDAR: Controllable Multi-Sensor LiDAR Sequence Generation

Khanh Nguyen, Thomas Walker, Hichem Telli, Carlos de la Torre-Ortiz, Stefano Fiorini, Salah Eddine Bekhouche, Mohammed-En-Nadhir Zighem, Wassim Swaileh 等 9 位作者

原文摘要

High-fidelity LiDAR simulation remains a bottleneck for scalable development and validation of autonomous systems. Existing generators can synthesize realistic scans, but they are typically constrained to a single sensor configuration, offer limited control over structure and geometry, and drift in long-horizon 4D generation. We present CrossLiDAR, a unified transformer-based framework capable of generating controllable and spatio-temporally consistent multi-sensor LiDAR sequences. Our approach leverages a rectified flow objective with a Multi-Modal Diffusion Transformer (MM-DiT) backbone, within a compact latent space of range images. To support multiple LiDAR sensors, an Inter-sensor Module enables interaction across sensors, while a ControlNet-style branch provides structural conditioning (e.g., BEV priors or 3D bounding boxes). For modeling the temporal consistency, CrossLiDAR generates short $N$-frame sequences and extends them autoregressively to improve long-horizon coherence. Our model significantly outperforms state-of-the-art baselines, achieving an average improvement of 29.56\% in JSD and 15.36\% in MMD evaluated on nuScenes and the Waymo Dataset. CrossLiDAR bridges the gap to modern autonomous-driving simulation by enabling calibration-consistent multi-sensor LiDAR rollouts over long horizons for controllable, scenario-based testing.

cs.CV · cs.AI · cs.RO

新论文 · v1arXiv:2611.11127

From Festival Video to the Stage: A Sim-to-Real Ladder for Long-Horizon Expressive Whole-Body Dance on a Commercial Humanoid

Hesham Shehata

原文摘要

We present a pipeline for transferring Awa Odori, a traditional Japanese festival dance, from monocular video to a commercial Unitree G1 EDU humanoid. The pipeline combines human mesh recovery, kinematic retargeting, reinforcement learning for motion tracking in simulation, and onboard deployment via ONNX. Training ran on a single RTX 5070 GPU. An eleven-version deployment ladder evaluated on hardware examines three interventions. We detect and repair silent corruption of deployment artifacts, model state-estimator drift during training, and edit the reference motion to support stable execution and a controlled ending. The final policies achieved three consecutive rope-free completions of a 43-second routine, a 46-second performance with automatic handoff, and a continuous 9.7-minute rope-free performance composed from the captured motion. In a live handover, a single remote request let the robot start from and return to the manufacturer's walking controller on stage, with a measured torque-free window of 45-140 ms. We report the training dynamics and hardware outcomes and discuss failure boundaries involving compliant flooring and occlusion in monocular capture.

cs.RO