arXivDaily arXiv每日学术速递 周一至周五更新

高校专区

University of Michigan(密歇根大学安娜堡分校)

共收录 1280
2606.22454 2026-06-23 cs.CL cs.AI 新提交

CASPER in the Machine: Insights into Character Variety in LLM-Generated Stories

机器中的CASPER:LLM生成故事中角色多样性的洞察

Anneliese Brei, Abhisheik Sharma, Nicholas Sanaie, Lu Wang, Snigdha Chaturvedi

机构 * UNC Chapel Hill(北卡罗来纳大学教堂山分校) Georgia Institute of Technology(佐治亚理工学院) University of Michigan(密歇根大学)

AI总结 本文借用叙事学定义,从八个维度分析LLM与人类创作故事中角色的刻画,发现两者在角色类型和多样性上既有相似也有差异。

Comments Proceedings of ACL, 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2606.22171 2026-06-23 cs.LG 新提交

Beyond Time Series: Spatial Reasoning for Epidemic Forecasting via Multimodal Learning

超越时间序列:基于多模态学习的空间推理用于流行病预测

Diana Guadalupe Gomez, Chenwei Wu, Zhiyi Wang, Liyue Shen, Alexander Rodríguez

机构 * University of Michigan(密歇根大学)

AI总结 提出M-SPICE框架,通过注意力机制融合区域级时序数据与空间辅助信号,在COVID-19、流感和ILI预测任务上超越现有基线。

Comments To appear in the Proceedings of the 32nd ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD 2026), AI for Science Track

详情

展开后加载摘要…

URL PDF HTML 收藏
2606.21869 2026-06-23 cs.CL cs.AI 新提交

The Language-Energy Divide: Measuring Energy Costs of Multilingual LLM Inference

语言-能量鸿沟:衡量多语言大模型推理的能量成本

Naihao Deng, Alissa Shen, Yiming Feng, Joan Nwatu, Jae-Won Chung, Mosharaf Chowdhury, Yulong Chen, Rada Mihalcea

机构 * University of Michigan(密歇根大学) University of Cambridge(剑桥大学) University of Aberdeen(阿伯丁大学)

AI总结 系统研究多语言LLM推理的能量消耗,发现不同语言间每token能耗差异达8.3倍,总能耗差异达179倍,低资源语言能耗高且准确率低,建议将能量作为评估指标。

详情

展开后加载摘要…

URL PDF HTML 收藏
2606.21866 2026-06-23 cs.RO 新提交

SurGE: Surrogate Gradient-guided Evolution for Co-design of Legged Robots with Parallel Elasticity

SurGE: 基于代理梯度引导的并行弹性腿式机器人协同设计进化

Yulun Zhuang, Yue Qin, Justin Lu, Zelin Shen, Yichen Wang, Sicheng He, Yanran Ding

机构 * University of Michigan, Ann Arbor(密歇根大学安娜堡分校) University of Tennessee, Knoxville(田纳西大学诺克斯维尔分校)

AI总结 提出SurGE框架,通过可微运动动力学模型和设计感知控制策略计算代理梯度,并注入CMA-ES实现非可微协同设计,在跳跃机器人上降低37.65%的设计目标。

Comments 8 pages, 7 figures. Accepted for publication at IROS 2026. Website at https://arcad-lab-um.github.io/surge-codesign/

详情

展开后加载摘要…

URL PDF HTML 收藏
2606.21861 2026-06-23 cs.CV 新提交

Zero-Shot Vision-Language Models for Classroom Engagement Recognition: A Benchmark Study of Prompt Sensitivity and Cross-Dataset Generalization

零样本视觉语言模型用于课堂投入度识别:提示敏感性与跨数据集泛化的基准研究

Aman Goyal, Kshama Nitin Shah, Kemmannu Vineet Venkatesh Rao

机构 * Carnegie Mellon University(卡内基梅隆大学) University of Michigan, Ann Arbor(密歇根大学安娜堡分校) Magna International(麦格纳国际)

AI总结 本研究系统评估五种视觉语言模型在零样本条件下识别课堂投入度的表现,发现三个主要失败模式:个体学生识别近乎随机、类别崩溃和极端提示敏感性,但场景级分类效果较好。

Comments 11 pages, 6 figures, including supplementary material. Presented as a non-archival paper at the CV4Edu Workshop, CVPR 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2606.21830 2026-06-23 cs.LG 新提交

Mat-Pref: Verifiable-Reward Training Improves Compositional Reasoning in Inorganic Materials

Mat-Pref: 可验证奖励训练提升无机材料中的组合推理能力

Sarrah R. Mikhail Leung, Taehan Kim, Jeongbin Park

机构 * University of California, Berkeley(加州大学伯克利分校) University of Michigan(密歇根大学)

AI总结 提出Mat-Pref基准,通过可验证奖励强化学习(GRPO)提升无机材料组合推理,在结构泛化和属性迁移上超越大模型。

Comments 10 pages, 4 figures, Accepted at ICML AI4Physics 2026 Workshop

详情

展开后加载摘要…

URL PDF HTML 收藏
2606.21119 2026-06-23 cs.CV cs.AI 新提交

MammoExpert: Benchmarking Chain-of-Thought Reasoning in Mammography Diagnosis

MammoExpert:乳腺X线诊断中的思维链推理基准测试

Di Dai, Bo Liu, Youcheng Li, Haojun Yu, Zhouhang Bian, Quanlin Wu, Dong Wang, Sichen Meng, Hongye Xuan, Zijie Lan, Shenda Hong, Liwei Wang

机构 * State Key Laboratory of General Artificial Intelligence, School of Intelligence Science and Technology, Peking University(北京大学智能科学与技术学院通用人工智能国家重点实验室) School of Computer Science and Engineer, Beijing University of Aeronautics and Astronautics(北京航空航天大学计算机科学与工程学院) Center for Data Science, Peking University(北京大学数据科学中心) Yizhun co. ltd(医准有限公司) International School, Beijing University of Post and Telecommunications(北京邮电大学国际学院) School of Public Health, University of Michigan, Ann Arbor(密歇根大学安娜堡分校公共卫生学院) Future Technology College, Xi'an Jiaotong University(西安交通大学未来技术学院) Peking University(北京大学)

AI总结 提出首个包含思维链推理标注的乳腺X线数据集MammoExpert,覆盖2379张图像和67种病理亚型,通过三阶段推理提升病灶分类准确率,在多个数据集上验证了有效性。

Comments KDD 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2606.20905 2026-06-23 cs.RO cs.AI 新提交

Vesta: A Generalist Embodied Reasoning Model

Vesta: 一种通用具身推理模型

Johan Bjorck, Zhiqi Li, Yunze Man, Jing Wang, An-Chieh Cheng, Sifei Liu, Shihao Wang, Zhiding Yu, Abhishek Badki, Stan Birchfield, Valts Blukis, Yevgen Chebotar, Siyi Chen, Sicong Leng, Yu-Cheng Chou, Tianli Ding, Boyi Li, Zhengyi Luo, Hang Su, Jonathan Tremblay, Tingwu Wang, Bowen Wen, Jimmy Wu, Xianghui Xie, Hanrong Ye, Hongxu Yin, K. R. Zentner, Liangyan Gui, Yu-Xiong Wang, Yuke Zhu, Linxi "Jim" Fan, Jan Kautz

机构 * University of Illinois Urbana-Champaign(伊利诺伊大学厄巴纳-香槟分校) University of California, San Diego(加州大学圣地亚哥分校) The Hong Kong Polytechnic University(香港理工大学) University of Michigan(密歇根大学) Nanyang Technological University(南洋理工大学) Johns Hopkins University(约翰霍普金斯大学) University of Tübingen(图宾根大学)

AI总结 提出统一的基础模型Vesta,整合定位、空间推理、导航和长时规划能力,通过大规模空间感知语料库和简单多模态记忆机制,在多个基准上平均优于专用模型20%以上,在真实机器人任务中成功率提升35%。

详情

展开后加载摘要…

URL PDF HTML 收藏
2606.20772 2026-06-23 cs.RO eess.IV 新提交

Mind the Privileged-to-Camera Gap: Actor-Centric Sidecar Supervision for Camera-First Open-Loop Waypoint Prediction

注意特权到摄像头的差距:面向摄像头优先开环路径预测的以参与者为中心的侧车监督

Feeza Khan Khanzada, Jaerock Kwon

机构 * University of Michigan-Dearborn(密歇根大学迪尔伯恩分校)

AI总结 针对摄像头优先开环路径预测中缺乏对道路参与者显式监督的问题,提出侧车监督方法,利用模拟器生成的参与者标签(如相关性、短时运动)训练模型,在部署时无需额外输入,最终将终点位移误差降低32.6%。

详情

展开后加载摘要…

URL PDF HTML 收藏
2606.20752 2026-06-23 cs.CV cs.CR 新提交

Mirage: a Clean-Label Backdoor against LiDAR 3D Object Detection

Mirage:针对LiDAR 3D目标检测的干净标签后门攻击

Ziba Parsons, Ang Li

机构 * University of Michigan - Dearborn(密歇根大学迪尔伯恩分校)

AI总结 提出Mirage,一种黑盒、干净标签的后门攻击方法,通过注入少量标签一致的毒化样本,使LiDAR 3D目标检测模型学习触发器与目标类别的恶意关联,实现73%误分类成功率且仅需0.5%毒化率。

详情

展开后加载摘要…

URL PDF HTML 收藏
2606.20643 2026-06-23 cs.AI cs.CV 新提交

SPARC: A Multi-Agent System for Electrical Circuit Question Answering

SPARC:用于电路问答的多智能体系统

Mushtari Sadia, Zhenning Yang, Umme Habiba Lamia, Nishat Shawrin, Ang Chen, Amrita Roy Chowdhury

机构 * University of Michigan(密歇根大学) Bangladesh University of Engineering and Technology(孟加拉工程与技术大学)

AI总结 提出SPARC多智能体系统,通过可执行的物理仿真程序进行推理,在电路问答任务上实现83%准确率,较基线提升高达58%。

详情

展开后加载摘要…

URL PDF HTML 收藏
2606.23627 2026-06-23 stat.ML cs.LG math.ST stat.TH 新提交

Diffusion Models Adapt to Low-Dimensional Structure Under Flexible Coefficient Choices

扩散模型在灵活系数选择下适应低维结构

Changxiao Cai, Yuchen Jiao, Gen Li

机构 * Department of Industrial and Operations Engineering, University of Michigan(工业与运营工程系,密歇根大学) Department of Statistics and Data Science, Chinese University of Hong Kong(统计与数据科学系,中国香港大学)

AI总结 本文证明扩散模型在广泛系数选择下,仅需O(k/ε)次迭代即可生成ε-精确样本,独立于环境维度,从而验证了其适应低维结构的鲁棒性。

详情

展开后加载摘要…

URL PDF HTML 收藏
2606.21701 2026-06-23 astro-ph.EP astro-ph.IM cs.DB cs.LG 新提交

ARCO-Mars: A Unified Cloud-Optimized Archive of Mars Atmosphere Reanalysis

ARCO-Mars:统一云优化的火星大气再分析档案

Ananyo Bhattacharya

机构 * University of Michigan, Department of Climate and Space Sciences and Engineering(密歇根大学气候与空间科学与工程系)

AI总结 提出ARCO-Mars,一个统一的分析就绪云优化数据集,整合三种火星大气再分析产品(EMARS、MACDA、OpenMARS),覆盖火星年24-35,以Zarr v3格式存储于HuggingFace,支持高效云访问,并比较了系统差异。

详情

展开后加载摘要…

URL PDF HTML 收藏
2606.15980 2026-06-23 cs.LG cs.AI cs.CL 新提交

Do Activation Monitors Survive Model Updates? Benchmarking, Predicting, and Repairing Activation-Monitor Staleness

安全监控器在更新后是否仍可靠?激活监控器陈旧性的基准测试与预测

Evan Duan

机构 * University of Michigan(密歇根大学)

AI总结 研究语言模型更新后激活监控器是否仍可靠,发现量化更新影响小,微调更新常导致监控器失效,且可通过预部署特征预测退化。

详情

展开后加载摘要…

URL PDF HTML 收藏
2512.24075 2026-06-23 cs.LG

Evolutionary Physics-Informed Temporal Fusion for Lane-Change Intention Prediction

进化物理信息时间融合用于换道意图预测

Jiazhao Shi, Qiyang Xie, Ziyu Wang, Yichen Lin, Di Zhu, Chen Xie, Ziwei Wang, Haoyun Zhang, Enliang Li, Zetong Guan

机构 * Tandon School of Engineering(工程学院) New York University(纽约大学) Khoury College of Computer Science(计算机科学学院) Northeastern University(东北大学) School of Business(商学院) Wake Forest University(威克森林大学) Independent Researcher(独立研究者) University of Massachusetts Amherst(马萨诸塞大学阿默斯特分校) Carnegie Mellon University(卡内基梅隆大学) University of Pennsylvania(宾夕法尼亚大学) Qualcomm CDMA Technologies(高通CDMA技术) University of Michigan(密歇根大学)

AI总结 提出一种进化物理信息时间融合框架,通过融合从传统交通信号导出的时间描述符和从原始轨迹序列学习的时间嵌入,实现三分类换道意图预测,在highD和exiD数据集上取得高F1分数。

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.03219 2026-06-23 eess.AS cs.SD 版本更新

Unmixing The Crowd: Learning Persistent Speaker Representations from Mixture-Derived Multi-Speaker Embeddings

解混人群:从混合衍生多说话人嵌入中学习持久说话人表示

Sidharth Sidharth, Meysam Asgari, Hao-Wen Dong, Dhruv Jain

机构 * University of Michigan Ann Arbor, MI, USA Oregon Health \& Science University Portland, OR, USA

AI总结 提出教师-学生框架,仅使用短重叠片段和置换不变潜在监督学习混合衍生多说话人嵌入,结合轻量在线记忆机制实现长时说话人重识别。

Comments Submitted to IEEE SLT 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2507.12744 2026-06-23 cs.RO 版本更新

ASC-SW: A Lightweight Atrous Strip Convolution Network for DLOs Segmentation on Edge mobile Robots

ASC-SW:面向边缘移动机器人的轻量级空洞条带卷积DLOs分割网络

Cheng Liu, Fan Zhu, Yifeng Xu, Baoru Huang, Mohd Rizal Arshad

机构 * School of Robotics, Xi’an Jiaotong-Liverpool University(西安交通大学利物浦大学机器人学院) Department of Robotics, University of Michigan(密歇根大学机器人系) Department of Artificial Intelligence, University of Liverpool(利物浦大学人工智能系)

AI总结 针对移动机器人视角下细长可变形线性物体分割的挑战,提出轻量级几何感知框架ASC-SW,通过空洞条带卷积和滑动窗口优化,在边缘设备上实现74.1% mIoU和261 FPS。

Comments paper update: Rejected by IROS2026 on June 17th

详情

展开后加载摘要…

URL PDF HTML 收藏
2411.15919 2026-06-23 cs.LG 版本更新

Enhancing Symbolic Regression and Universal Physics-Informed Neural Networks with Dimensional Analysis

利用量纲分析增强符号回归和通用物理信息神经网络

Lena Podina, Diba Darooneh, Joshveer Grewal, Mohammad Kohandel

机构 * Cheriton School of Computer Science(切尔顿计算机科学学院) Department of Electrical and Computer Engineering(电气与计算机工程系) University of Waterloo(滑铁卢大学) Electrical Engineering and Computer Science Department(电气工程与计算机科学系) Department of Applied Mathematics(应用数学系) University of Michigan(密歇根大学)

AI总结 提出结合量纲分析(Buckingham Π定理和Ipsen方法)与符号回归及通用物理信息神经网络,通过无量纲化降低输入维度、减少过拟合,提高微分方程未知项恢复的准确性和计算效率。

详情

展开后加载摘要…

URL PDF HTML 收藏
2606.20206 2026-06-19 stat.ML cs.LG 新提交

Off-Policy Evaluation for Missingness-Aware Policies in MDPs with Rewards Missing Not at Random

马尔可夫决策过程中奖励非随机缺失的缺失感知策略的离线评估

Ziheng Wei, Annie Qu, Rui Miao

机构 * Department of Statistics, University of Michigan at Ann Arbor(密歇根大学安娜堡分校统计学系) Department of Statistics(统计学系) Applied Probability, University of California at Santa Barbara(加州大学圣巴巴拉分校应用概率系) Department of Mathematical Sciences, University of Texas at Dallas(德克萨斯大学达拉斯分校数学科学系)

AI总结 针对奖励非随机缺失的离线强化学习问题,提出基于未来状态作为影子变量的识别方法,并利用桥函数和min-max估计器恢复条件均值奖励,实现缺失感知策略的离线评估。

Comments Accepted at ICML 2026. 31 pages, 6 figures

详情

展开后加载摘要…

URL PDF HTML 收藏
2606.19539 2026-06-19 astro-ph.SR cs.AI 新提交

Review of Machine Learning Models for Solar Energetic Particle Prediction

太阳高能粒子预测的机器学习模型综述

Spiridon Kasapis, Pouya Hosseinzadeh, Kathryn Whitman, Ricky Egeland, Manolis Georgoulis, Angelos Vourlidas, Athanasios Papaioannou, Eleni Lavasa, Anastasios Anastasiadis, Giorgos Giannopoulos, Andres Munoz-Jaramillo, Bala Poduval, Irina N. Kitiashvili, Alexander G. Kosovichev, Viacheslav Sadykov, Soukaina Filali Boubrahimi, Tate T. Hutchins, Hameedullah A. Farooki, Manuel E. Cuesta, Leng Y. Khoo, Sungmin Pak, Robert Czarnota, Jamie S. Rankin, Jamey Szalay, Mitchell M. Shen, Georgios Livadiotis, Zigong Xu, David J. McComas, Nikolaos Sarlis, Dionissios Hristopulos, Arik Posner, Alec J. Engell, Mohammed AbuBakr Ali, Ali G. A. Abdelkawy, Abdelrazek M. K. Shaltout, M. M. Beheary, Christina O. Lee, Sigiava Aminalragia-Giamini, Constantinos Papadimitriou, Ingmar Sandberg, Savvas Raptis, Shah Muhammad Hamdi, Monica Laurenza, Mirko Stumpo, Sumanth A. Rotti, India Jackson, Aatiya Ali, Atilim Gunes Baydin, Nathan Schwadron, Subhamoy Chatterjee, Maher A. Dayeh, Gelu M. Nita, Patrick M. O'Keefe, Chun Jie Chong, Paul Kosovich, Russell D. Marroquin, Berkay Aydin, Petrus C. Martens, Lulu Zhao, Yang Chen, Yian Yu, Monica G. Bobra, Ward Manchester, Tamas Gombosi, Ming Zhang, Jesse Torres, Philip K. Chan, Mohamed Nedal, Kamen Kozarev, Peijin Zhang, Kimberly Moreland, Hazel M. Bain, Samuel Hart, Michael J. Starkey, Alan G. Ling, Simone Benella

机构 * Department of Astrophysical Sciences, Princeton University, Princeton, NJ, USA Computational Physics Branch, NASA Ames Research Center, Moffett Field, CA, USA Department of Computer Science, Utah State University, Logan, UT, USA Space Radiation Analysis Group, NASA Johnson Space Center, Houston, TX, USA Johns Hopkins Applied Physics Lab, 11100 Johns Hopkins Rd, Laurel, MD 20723, United States Research Center for Astronomy Applied Mathematics of the Academy of Athens, 4 Soranou Efesiou Street, Athens 11527, Greece Institute for Astronomy, Astrophysics, Space Applications Southwest Research Institute, Boulder, CO, USA Space Science Center, University of New Hampshire, Durham, NH, USA Department of Physics, New Jersey Institute of Technology, Newark, NJ, USA Astronomy Department, Georgia State University, Atlanta, GA, USA Department of Computer Science, Princeton University, Princeton, NJ, USA Department of Mathematics, Rowan University, Glassboro, NJ, USA Astronomy, California Institute of Technology, Pasadena, CA, USA Department of Physics, National Kapodistrian University of Athens, Athens, Greece School of Electrical Computer Engineering, Technical University of Crete, Chania, Greece Department of Astronomy Meteorology, Faculty of Science, Al-Azhar University, Cairo, Egypt Space Sciences Lab, University of California, Berkeley, CA, USA Research Consultancy, Athens, Greece Institute for Space Astrophysics Department of Physics Astronomy, Georgia State University, Atlanta, GA 30303, USA Aryabhatta Research Institute of Observational Sciences (ARIES), Manora Peak, Nainital-263001, Uttarakhand, India Department of Computer Science, Oxford University, Oxford, England Southwest Research Institute, San Antonio, TX, USA Computer Science Department, New Jersey Institute of Technology, Newark, NJ, USA Department of Physics, University of California San Diego, La Jolla, CA 92093, USA Department of Computer Science, Georgia State University, Atlanta, GA 30303, USA Department of Climate Engineering, University of Michigan, Ann Arbor, MI, USA Department of Statistics, University of Michigan, Ann Arbor, MI, USA Department of Electrical Engineering Computer Science, Florida Institute of Technology, Melbourne, FL, USA Astrophysics Section, School of Cosmic Physics, Dublin Institute for Advanced Studies, DIAS Dunsink Observatory, Dublin D15 XR2R, Ireland Institute of Astronomy of the Bulgarian Academy of Sciences, Sofia, Bulgaria Center for Solar-Terrestrial Research, New Jersey Institute of Technology, Newark, NJ 07102, USA Cooperative Programs for the Advancement of Earth System Science, University Corporation for Atmospheric Research, Boulder, CO, USA CIRES, University of Colorado Boulder, Boulder, CO, USA Space Weather Prediction Center, NOAA, Boulder, CO, USA Astronomy, College of Science, The University of Texas at San Antonio, San Antonio, TX, USA Space Weather Prediction Center, National Oceanic The University of Texas at San Antonio, San Antonio, TX, USA Environmental Research, Inc., MA, USA

AI总结 综述了用于太阳高能粒子预测的机器学习模型,包括数据集、架构、输入输出比较,并提出了未来研究建议。

Comments Review Paper, Maine text: 23 pages, References: 5 pages, Appendix: 42 pages

详情

展开后加载摘要…

URL PDF HTML 收藏
2606.17054 2026-06-19 cs.RO cs.AI cs.CV cs.LG 新提交

Human Universal Grasping

人类通用抓取

Kevin Yuanbo Wu, Tianxing Zhou, Isaac Tu, Billy Yan, Irmak Guzey, David Fouhey, Dandan Shan, Lerrel Pinto

机构 * New York University(纽约大学) Tsinghua University(清华大学) University of Michigan(密歇根大学)

AI总结 提出HUG模型,利用人类抓取数据(1M-HUG数据集)和流匹配方法,从单张RGB-D图像生成多样化抓取姿态,并重定向到机器人手,实现零样本抓取,在HUG-Bench上超越基线23%-34%。

Comments 28 pages, 20 figures, 7 tables

详情

展开后加载摘要…

URL PDF HTML 收藏
2605.28654 2026-06-19 cs.RO cs.SY eess.SY math.OC 版本更新

Integrated Exploration-Aware UAV Route Optimization and Path Planning

集成探索感知的无人机路径优化与轨迹规划

Jimin Choi, Grant Stagg, Cameron K. Peterson, Max Z. Li

机构 * Department of Aerospace Engineering, University of Michigan(密歇根大学航空航天工程系) Department of Electrical Engineering, Brigham Young University(BYU 电子工程系) Department of Aerospace Engineering, Department of Civil and Environmental Engineering, and Department of Industrial and Operations Engineering, University of Michigan(密歇根大学航空航天工程系、土木与环境工程系和工业与运营管理工程系)

AI总结 提出一种集成探索感知的无人机路径优化与轨迹规划框架,通过风险地图、不确定兴趣区域建模、B样条轨迹优化和在线重规划,在灾害监测中平衡报告点访问与新信息探索,实现平均KL散度降低15.9%。

详情

展开后加载摘要…

URL PDF HTML 收藏
2510.01565 2026-06-19 cs.LG cs.DC 版本更新

TetriServe: Efficiently Serving Mixed DiT Workloads

TetriServe: 高效服务混合DiT工作负载

Runyu Lu, Shiqi He, Wenxuan Tan, Shenggui Li, Ruofan Wu, Jeff J. Ma, Ang Chen, Mosharaf Chowdhury

机构 * University of Michigan(密歇根大学) University of Wisconsin-Madison(威斯康星大学麦迪逊分校) Nanyang Technological University(南洋理工大学)

AI总结 针对混合分辨率与截止时间的异构DiT工作负载,提出基于步骤级序列并行的TetriServe系统,通过轮次调度与自适应并行度,在保证图像质量下将SLO达成率提升32%。

详情

展开后加载摘要…

URL PDF HTML 收藏
2601.02322 2026-06-19 stat.ME cs.LG 版本更新

Environment-Adaptive Covariate Selection: Learning When to Use Spurious Correlations for Out-of-Distribution Prediction

环境自适应协变量选择:学习何时利用虚假相关进行分布外预测

Shuozhi Zuo, Yixin Wang

机构 * Department of Statistics, University of Michigan, Ann Arbor(统计系,密歇根大学,安阿伯分校)

AI总结 针对分布外预测中协变量选择问题,提出环境自适应算法,根据环境特征动态选择协变量集,在模拟和实际数据中优于静态方法。

详情

展开后加载摘要…

URL PDF HTML 收藏
2606.19233 2026-06-18 cs.RO 新提交

Mobile Pedipulation for Object Sliding via Hierarchical Control on a Wheeled Bipedal Robot

基于轮式双足机器人分层控制的移动式腿部操作物体滑动

Yue Qin, Yulun Zhuang, Zelin Shen, Yanran Ding

机构 * Department of Robotics, University of Michigan(密歇根大学机器人系)

AI总结 提出一种分层控制框架,使轮式双足机器人能用腿部滑动平面物体,通过简化三刚体动力学模型和轨迹优化运动规划器,在实验中成功实现1kg物体取回和4kg物体滑动。

Comments 8 pages, 7 figures

Journal ref IEEE Robotics and Automation Letters, vol. 11, no. 8, pp. 9787-9794, Aug. 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2606.18656 2026-06-18 cs.CL 新提交

The Wrong Kind of Right: Quantifying and Localizing Misfired Alignment in LLMs

错误的正确:量化和定位大语言模型中的失调对齐

Naihao Deng, Yiming Feng, Chimaobi Okite, Kaijian Zou, Lu Wang, Rada Mihalcea, Yulong Chen

机构 * University of Michigan(密歇根大学) University of Cambridge(剑桥大学) University of Aberdeen(阿伯丁大学)

AI总结 本文提出VETO基准和失调对齐率(MAR)指标,发现所有LLM在刻板印象相关问题上均存在非平凡的失调对齐,且人类为0%,机制分析表明对齐诱导的线索会放大该现象。

详情

展开后加载摘要…

URL PDF HTML 收藏
2606.18471 2026-06-18 cs.CL 新提交

Possible or Definite? A Benchmark for Evaluating Diagnostic Uncertainty Preservation in Clinical Text

可能还是确定?评估临床文本中诊断不确定性保留的基准

Hongbo Du, Zixin Lu, Jiaming Qu

机构 * Trine University(特里尼大学) University of Michigan(密歇根大学) Amazon(亚马逊)

AI总结 构建包含9184个不确定性标注的基准,评估LLM在临床文本中保留诊断不确定性的能力,发现LLM保留原始不确定性线索不足一半,且难以区分相邻级别。

详情

展开后加载摘要…

URL PDF HTML 收藏
2606.18438 2026-06-18 math.OC cs.LG 新提交

Sequential Hiring of Contingent Workers Through Learning-Based Optimization

基于学习优化的临时工顺序雇佣

Chris Lee, Xiuli Chao, Izak Duenyas

机构 * Department of Industrial and Operations Engineering, University of Michigan(工业与运营工程系,密歇根大学) Ross School of Business, University of Michigan(罗斯商学院,密歇根大学)

AI总结 针对临时工场景中工人产能和劳动力供给的不确定性,提出DR-UCB策略,通过学习周期顺序决策替换与雇佣,实现累积利润最大化,并证明其遗憾下界匹配。

详情

展开后加载摘要…

URL PDF HTML 收藏
2606.15633 2026-06-18 cs.LG 新提交

Formalizing and Mitigating Structural Distortion in LLM Attention for Graph Reasoning

形式化并缓解大语言模型注意力中的结构失真以实现零样本图推理

Donald Loveland, Puja Trivedi, Ari Weinstein, Edward W Huang, Danai Koutra

机构 * University of Michigan(密歇根大学) Amazon(亚马逊)

AI总结 本文形式化了大语言模型处理文本属性图时因图线性化导致的结构失真机制,并提出轻量级推理时修改方法GaLA,通过校正注意力偏差提升零样本图推理性能。

Comments Accepted to KDD 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2606.05409 2026-06-18 cs.CV cs.CL 版本更新

Would you still call this Dax? Novel Visual References in VLMs and Humans

你还会称它为Dax吗?VLM与人类中的新颖视觉参照

Ada Defne Tür, Gaurav Kamath, Joyce Chai, Siva Reddy, Benno Krojer

机构 * McGill University(麦吉尔大学) Mila Quebec AI Institute(魁北克人工智能研究所) University of Michigan - Ann Arbor(密歇根大学安娜堡分校) Canada CIFAR AI Chair(加拿大CIFAR人工智能主席)

AI总结 提出新颖视觉参照数据集(NVRD),通过对比VLM和人类对新颖视觉概念的泛化能力,发现模型在矛盾先验知识时难以习得新概念,且过度泛化。

详情

展开后加载摘要…

URL PDF HTML 收藏