arXivDaily arXiv每日学术速递 周一至周五更新

高校专区

University of Chinese Academy of Sciences(中国科学院大学)

2026-05-01 至 2026-05-01 共收录 12
2604.27998 2026-05-01 cs.LG cs.CL

Latent-GRPO: Group Relative Policy Optimization for Latent Reasoning

Latent-GRPO:用于潜在推理的组相对策略优化

Jingcheng Deng, Zihao Wei, Liang Pang, Junhong Wu, Shicheng Xu, Zenghao Duan, Huawei Shen

机构 * State Key Laboratory of AI Safety, Institute of Computing Technology, Chinese Academy of Sciences(中国科学院人工智能安全国家重点实验室,计算技术研究所) University of Chinese Academy of Sciences(中国科学院大学) Institute of Automation, Chinese Academy of Sciences(中国科学院自动化研究所)

AI总结 本文提出Latent-GRPO,通过解决潜在推理中的三个瓶颈问题,提升推理稳定性与效率,在低难度和高难度任务中均取得显著性能提升。

Comments This is an actively developing work, and we will continue to update the arXiv version

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.25326 2026-05-01 cs.AR cs.AI

AHASD: Asynchronous Heterogeneous Architecture for LLM Adaptive Drafting Speculative Decoding on Mobile Devices

AHASD:异步异构架构用于移动设备上的LLM自适应草案推测解码

Ma Zirui, Fan Zhihua, Li Wenxing, Wu Haibin, Zhang Fulin, Ye Xiaochun, Li Wenming

机构 * State Key Lab of Processors, Institute of Computing Technology, Chinese Academy of Sciences(中国科学院计算技术研究所处理器国家重点实验室) University of Chinese Academy of Sciences(中国科学院大学)

AI总结 AHASD提出一种异步异构架构,通过任务级DLM-TLM解耦和熵历史感知草案控制,提升移动设备上LLM自适应草案推测解码的吞吐量和能效。

Comments 7 pages, 9 figures, accepted by DAC 2026, repo: https://github.com/MAdrid1011/AHASD

详情

展开后加载摘要…

URL PDF HTML 收藏
2602.02980 2026-05-01 eess.AS cs.CL eess.SP

WST-X Series: Wavelet Scattering Transform for Interpretable Speech Deepfake Detection

WST-X系列:用于可解释性语音深度伪造检测的小波散射变换

Xi Xuan, Davide Carbone, Wenxin Zhang, Ruchi Pandey, Tomi H. Kinnunen

机构 * Computational Speech Group(计算语音组) University of Eastern Finland(东芬兰大学) Laboratoire de Physique de l’Ecole Normale Supérieure, Université PSL, CNRS, Sorbonne Université, Université de Paris(巴黎高等师范大学物理实验室,巴黎大学,CNRS,索邦大学,巴黎大学) University of Chinese Academy of Sciences(中国科学院大学) University of Toronto(多伦多大学)

AI总结 本文提出WST-X系列,结合小波散射变换提升语音深度伪造检测的可解释性与性能,通过多尺度特征提取在多个数据集上取得显著优势。

Comments IEEE Signal Processing Letters

详情

展开后加载摘要…

URL PDF HTML 收藏
2504.14602 2026-05-01 cs.RO cs.AI cs.HC

K2MUSE: A human lower-limb multimodal walking dataset spanning task and acquisition variability for rehabilitation robotics

K2MUSE:一种涵盖任务和采集变异性的下肢多模态行走数据集,用于康复机器人

Jiwei Li, Bi Zhang, Xiaowei Tan, Wanxin Chen, Zhaoyuan Liu, Juanjuan Zhang, Weiguang Huo, Jian Huang, Lianqing Liu, Xingang Zhao

机构 * State Key Laboratory of Robotics and Intelligent Systems, Shenyang Institute of Automation, Chinese Academy of Sciences(机器人与智能系统国家重点实验室,沈阳自动化研究所,中国科学院) University of Chinese Academy of Sciences(中国科学院大学) College of Artificial Intelligence, Tianjin Key Laboratory of Intelligent Robotics, Nankai University(人工智能学院,天津智能机器人重点实验室,南开大学) School of Artificial Intelligence and Automation, Huazhong University of Science and Technology(人工智能与自动化学院,华中科技大学)

AI总结 K2MUSE数据集通过多模态数据揭示下肢康复机器人与生物力学信息的关系,为康复机器人开发提供大规模步态样本和数据驱动方法支持。

Comments 34 pages, 30 figures,7 tables

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.27582 2026-05-01 cs.CV

Assessing Pancreatic Ductal Adenocarcinoma Vascular Invasion: the PDACVI Benchmark

评估胰腺导管腺癌血管侵袭:PDACVI基准

M. Riera-Marín, O. K. Sikha, J. Rodríguez-Comas, M. S. May, T. Kirscher, X. Coubez, P. Meyer, S. Faisan, Z. Pan, X. Zhou, X. Liang, C. Hémon, V. Boussot, J. -L. Dillenseger, J. -C. Nunes, K. -C. Kahl, C. Lüth, J. Traub, P. -H. Conze, M. M. Duh, A. Aubanell, R. de Figueiredo Cardoso, S. Egger-Hackenschmidt, J. García-López, M. A. González-Ballester, A. Galdran

机构 * Sycai Technologies SL, Scientific Technical Department, Barcelona, Spain Universit\" a tsklinikum Erlangen, Department of Radiology of the Uniklinikum Erlangen (UKER), Erlangen, Germany University Hospital Erlangen, Imaging Science Institute, Erlangen, Germany ICUBE Laboratory, CNRS UMR-7357, University of Strasbourg, Strasbourg, France Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences, Shenzhen, China University of Chinese Academy of Sciences, Beijing, China German Cancer Research Center (DKFZ), Medical Image Computing (E230), Heidelberg, Germany Hospital de Matar\' o , Matar\' o , Spain Hospital de Sant Pau i la Santa Creu, Diagnostic Imaging Department, Barcelona, Spain Institut de Recerca Sant Pau - CERCA, Advanced Medical Imaging, Artificial Intelligence Imaging-Guided Therapy Research Group, Barcelona, Spain TECNALIA, Basque Research

AI总结 本文提出PDACVI基准,通过密集标注数据集评估胰腺癌血管侵袭,揭示传统体积重叠指标的局限性,强调概率模型在术前决策中的重要性。

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.27488 2026-05-01 cs.CL

Skills-Coach: A Self-Evolving Skill Optimizer via Training-Free GRPO

Skills-Coach:一种通过无训练GRPO实现的自我进化技能优化器

Yu Tian, Jiawei Chen, Lifan Zheng, Mingxiang Tao, Xinyi Zeng, Zhaoxia Yin, Hang Su, Xian Sun

机构 * University of Chinese Academy of Sciences(中国科学院大学) East China Normal University(华东师范大学) Zhongguancun Academy(中关村学院) Southeast University(东南大学) Hainan University(海南大学) Tsinghua University(清华大学)

AI总结 Skills-Coach通过无训练GRPO框架提升LLM代理技能自我进化能力,包含四个核心模块,通过Skill-X基准测试展示其在多种技能类别中的显著性能提升。

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.27467 2026-05-01 cs.SE cs.CL

ScaleBox: Enabling High-Fidelity and Scalable Code Verification for Large Language Models

ScaleBox:实现大规模语言模型高保真和可扩展的代码验证

Jiasheng Zheng, Xin Zheng, Boxi Cao, Pengbo Wang, Zhengzhao Ma, Qiming Zhu, Jiazhen Jiang, Yaojie Lu, Hongyu Lin, Xianpei Han, Le Sun

机构 * Chinese Information Processing Laboratory(中国科学院信息处理实验室) Institute of Software, Chinese Academy of Sciences(中国科学院软件研究所) University of Chinese Academy of Sciences(中国科学院大学)

AI总结 ScaleBox通过自动化特殊判题生成与管理、细粒度并行执行和配置驱动的评估套件,提升大规模代码训练的验证准确性和效率,显著优于基线方法。

Comments Accepted to ACL 2026 Demo. Our project is available at https://github.com/icip-cas/ScaleBox

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.27419 2026-05-01 cs.AI cs.CL

InteractWeb-Bench: Can Multimodal Agent Escape Blind Execution in Interactive Website Generation?

InteractWeb-Bench: 多模态代理能否在交互式网站生成中避免盲目执行?

Qiyao Wang, Haoran Hu, Longze Chen, Hongbo Wang, Hamid Alinejad-Rokny, Yuan Lin, Min Yang

机构 * Shenzhen Institute of Advanced Technology Chinese Academy of Sciences(深圳先进技术研究院中国科学院) University of Chinese Academy of Sciences(中国科学院大学) Dalian University of Technology(大连理工大学) UNSW Sydney(悉尼大学) Shenzhen University of Advanced Technology(深圳先进技术大学)

AI总结 本文提出InteractWeb-Bench,首个针对非专家低代码用户的多模态交互基准,通过模拟用户行为中的模糊性、冗余性和矛盾性,揭示前沿多模态大语言模型在意图识别和适应性交互上的局限。

Comments 21 pages, 13 figures, 7 tables

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.27361 2026-05-01 cs.CV cs.GR

CasLayout: Cascaded 3D Layout Diffusion for Indoor Scene Synthesis with Implicit Relation Modeling

CasLayout:基于级联3D布局扩散的室内场景合成与隐式关系建模

Yingrui Wu, Youkang Kong, Mingyang Zhao, Weize Quan, Dong-Ming Yan, Yang Liu

机构 * MAIS, Institute of Automation, Chinese Academy of Sciences and School of Artificial Intelligence, University of Chinese Academy of Sciences(MAIS,自动化研究所,中国科学院,人工智能学院,中国科学院大学) Tsinghua University(清华大学) SKLMS, Academy of Mathematics and Systems Science, Chinese Academy of Sciences(系统科学研究所,中国科学院) Microsoft Research Asia(微软亚洲研究院)

AI总结 CasLayout通过级联扩散框架将场景生成分解为四个阶段,结合物理与语义约束,提升生成效率与可控性,实现高质量室内场景合成。

Comments SIGGARPH 2026 (Journal Track), Code: https://github.com/YingruiWoo/CasLayout

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.27357 2026-05-01 cs.LG cs.CV

AG-TAL: Anatomically-Guided Topology-Aware Loss for Multiclass Segmentation of the Circle of Willis Using Large-Scale Multi-Center Datasets

AG-TAL:基于解剖学的拓扑感知损失用于使用大规模多中心数据集的Willis环多类分割

Jialu Liu, Yue Cui, Shan Yu

机构 * Laboratory of Brain Atlas and Brain-inspired Intelligence, Institute of Automation, Chinese Academy of Sciences(脑图谱与脑启发智能实验室,自动化研究所,中国科学院) State Key Laboratory of Brain Cognition and Brain-inspired Intelligence Technology, Chinese Academy of Sciences(脑认知与脑启发智能技术国家重点实验室,中国科学院) School of Artificial Intelligence, University of Chinese Academy of Sciences(中国科学院大学人工智能学院) School of Future Technology, University of Chinese Academy of Sciences(中国科学院大学未来技术学院)

AI总结 本文提出AG-TAL损失函数,通过整合半径感知Dice损失、断裂感知clDice损失和邻接感知共现损失,提升Willis环多类分割的准确率与泛化能力。

Comments 11 pages, 5 figures, submitted to IEEE JBHI

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.27083 2026-05-01 cs.LG

Co-Evolving Policy Distillation

共进化策略蒸馏

Naibin Gu, Chenxu Yang, Qingyi Si, Chuanyu Qin, Dingyu Yao, Peng Fu, Zheng Lin, Weiping Wang, Nan Duan, Jiaqi Wang

机构 * Institute of Information Engineering, CAS(中国科学院信息工程研究所) School of Cyber Security, UCAS(中国科学院大学网络空间安全学院)

AI总结 本文提出CoPD,通过并行训练专家并引入OPD,在专家RLVR训练过程中实现双向知识传递,提升多模态推理能力,优于现有方法。

Comments Work in progress

详情

展开后加载摘要…

URL PDF HTML 收藏
2601.02845 2026-05-01 cs.CL cs.AI

TiMem: Temporal-Hierarchical Memory Consolidation for Long-Horizon Conversational Agents

TiMem:用于长时域对话代理的时序-层次记忆巩固

Kai Li, Xuanqing Yu, Ziyi Ni, Yi Zeng, Yao Xu, Zheqing Zhang, Xin Li, Jitao Sang, Xiaogang Duan, Xuelei Wang, Chengbao Liu, Jie Tan

机构 * Institute of Automation, CAS(中国科学院自动化研究所) School of Artificial Intelligence, UCAS(中国科学技术大学人工智能学院) AI Lab, AIGility Cloud Innovation(AIGility云创新AI实验室) North China Electric Power University(华北电力大学) Beijing Academy of Artificial Intelligence(北京人工智能研究院) Gaoling School of Artificial Intelligence, RUC(中国人民大学高陵人工智能学院) School of Biomedical Engineering, USTC(中国科学技术大学生物医学工程学院) Suzhou Institute for Advance Research, USTC(中国科学技术大学苏州市先进研究院) School of Computer Science and Technology, BJTU(北京理工大学计算机科学与技术学院) Hunan Central South Intelligent Equipment Co., Ltd.(湖南中南智能装备有限公司)

AI总结 TiMem通过时序记忆树实现对话记忆的系统性巩固,提升长时域个性化效果,实现75.30%和76.88%的基准测试准确率。

Comments ACL 2026 Findings

详情

展开后加载摘要…

URL PDF HTML 收藏