arXivDaily arXiv每日学术速递 周一至周五更新

高校专区

University of Toronto(多伦多大学)

共收录 1022
2511.21143 2025-11-27 cs.HC cs.CV

STAR: Smartphone-analogous Typing in Augmented Reality

STAR:增强现实中的智能手机类打字

Taejun Kim, Amy Karlson, Aakar Gupta, Tovi Grossman, Jason Wu, Parastoo Abtahi, Christopher Collins, Michael Glueck, Hemant Bhaskar Surale

机构 * Realiy Labs Research, Meta Reality Labs Research, Meta University of Toronto(多伦多大学)

AI总结 STAR通过在手部皮肤上叠加虚拟键盘,实现AR环境中的智能手机类打字,实验显示其打字速度为智能手机的56%,错误率低,具有实用性。

Comments ACM UIST 2023

详情

展开后加载摘要…

URL PDF HTML 收藏
2505.18513 2025-11-27 cs.LG

Enhancing Training Data Attribution with Representational Optimization

通过表征优化增强训练数据归因

Weiwei Sun, Haokun Liu, Nikhil Kandpal, Colin Raffel, Yiming Yang

机构 * Carnegie Mellon University(卡内基梅隆大学) University of Toronto(多伦多大学) Vector Institute(向量研究所)

AI总结 AirRep通过优化表征提升训练数据归因效率,实现与先进梯度方法相当的性能,且推理效率提高近两数量级。

Comments NeurIPS 2025

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.19987 2025-11-26 cs.CL cs.IR

$\text{R}^2\text{R}$: A Route-to-Rerank Post-Training Framework for Multi-Domain Decoder-Only Rerankers

R²R:多领域解码器-only重排序框架的路由到重排序后训练方法

Xinyu Wang, Hanwei Wu, Qingchen Hu, Zhenghan Tai, Jingrui Tian, Lei Ding, Jijun Chi, Hailin He, Tung Sum Thomas Kwok, Yufei Cui, Sicheng Lyu, Muzhi Li, Mingze Li, Xinyue Yu, Ling Zhou, Peng Lu

机构 * McGill University(麦吉尔大学) University of Toronto(多伦多大学) University of Manitoba(曼尼托巴大学) Mila CUHK(中国科技大学) Université de Montréal(蒙特利尔大学) CG Matrix(CG矩阵)

AI总结 R²R通过动态专家路由和两阶段训练策略,提升多领域解码器重排序器的领域适应能力与跨领域鲁棒性。

Comments 13 pages, including 3 figures and 3 tables

详情

展开后加载摘要…

URL PDF HTML 收藏
2510.20820 2025-11-26 cs.CV

LayerComposer: Multi-Human Personalized Generation via Layered Canvas

LayerComposer: 通过分层画布实现多人类个性化生成

Guocheng Gordon Qian, Ruihang Zhang, Tsai-Shien Chen, Yusuf Dalva, Anujraaj Argo Goyal, Willi Menapace, Ivan Skorokhodov, Meng Dong, Arpit Sahni, Daniil Ostashev, Ju Hu, Sergey Tulyakov, Kuan-Chieh Jackson Wang

机构 * Snap Inc. University of Toronto(多伦多大学) UC Merced(加州大学默塞德分校) Virginia Tech(弗吉尼亚理工大学)

AI总结 LayerComposer通过分层画布实现多人类个性化生成,提供直观的人类注入和高效 scalable 的生成框架。

Comments 17 pages including appendix, preprint. Project page: https://snap-research.github.io/layercomposer/

详情

展开后加载摘要…

URL PDF HTML 收藏
2510.01508 2025-11-26 cs.LG

Realistic CDSS Drug Dosing with End-to-end Recurrent Q-learning for Dual Vasopressor Control

面向真实临床环境的CDSS药物剂量优化:基于端到端递归Q学习的双血管加压素控制

Will Y. Zou, Jean Feng, Alexandre Kalimouttou, Jennifer Yuntong Zhang, Christopher W. Seymour, Romain Pirracchio

机构 * University of California, San Francisco(加州大学旧金山分校) Engineering Science, University of Toronto(多伦多大学工程科学系) University of Pittsburgh(匹兹堡大学)

AI总结 本文提出一种端到端递归Q学习方法,用于ICU中双血管加压素的剂量优化,通过设计动作空间提升可解释性与临床实用性。

Comments 13 pages, 5 figures. Neurips 2025 Workshop Learning from Time Series for Health

详情

展开后加载摘要…

URL PDF HTML 收藏
2508.10973 2025-11-26 cs.RO cond-mat.mtrl-sci

Developing and Validating a High-Throughput Robotic System for the Accelerated Development of Porous Membranes

开发和验证一种高通量机器人系统用于多孔膜的加速开发

Hongchen Wang, Sima Zeinali Danalou, Jiahao Zhu, Kenneth Sulimro, Chaewon Lim, Smita Basak, Aimee Tai, Usan Siriwardana, Jason Hattrick-Simpers, Jay Werber

机构 * Department of Materials Science and Engineering, University of Toronto(材料科学与工程系,多伦多大学) Department of Chemical Engineering, University of Toronto(化学工程系,多伦多大学) Department of Engineering Science, University of Toronto(工程科学系,多伦多大学) Department of Mechanical Engineering, University of Waterloo(机械工程系,滑铁卢大学)

AI总结 本研究提出了一种自动化平台,通过非溶剂诱导相分离技术实现多孔膜的高通量制备与表征,提升实验效率和一致性。

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.19330 2025-11-25 cs.LG

Targeted Manipulation: Slope-Based Attacks on Financial Time-Series Data

定向操纵:基于斜率的金融时间序列数据攻击

Dominik Luszczynski

机构 * University of Toronto Department of Computer Science(多伦多大学计算机科学系)

AI总结 本文提出两种基于斜率的攻击方法,用于操纵N-HiTS模型的股票预测,同时展示了恶意软件注入攻击以提升模型安全性

Comments 13 pages, 6 figures, 4 tables, preprint; Total including Appendix: 21 pages, 11 figures, 7 tables

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.18630 2025-11-25 cs.LG cs.AI cs.CL stat.ME stat.ML

Majority of the Bests: Improving Best-of-N via Bootstrapping

多数最佳:通过Bootstrap改进最佳-N

Amin Rakhsha, Kanika Madan, Tianyu Zhang, Amir-massoud Farahmand, Amir Khasahmadi

机构 * University of Toronto(多伦多大学) Vector Institute(向量研究所) Autodesk(Autodesk公司) Polytechnique Montréal(蒙特利尔理工学院) Mila - Quebec AI Institute(魁北克人工智能研究所)

AI总结 本文提出MoB方法,通过Bootstrap估计最佳-N的输出分布并选择其模式,以提高在不完美奖励下的准确率。

详情

展开后加载摘要…

URL PDF HTML 收藏
2510.24591 2025-11-25 cs.CL astro-ph.IM

ReplicationBench: Can AI Agents Replicate Astrophysics Research Papers?

ReplicationBench: 人工智能代理能否复制天体物理学研究论文?

Christine Ye, Sihan Yuan, Suchetha Cooray, Steven Dillmann, Ian L. V. Roque, Dalya Baron, Philipp Frank, Sergio Martin-Alvarez, Nolan Koblischke, Frank J Qu, Diyi Yang, Risa Wechsler, Ioana Ciuca

机构 * Stanford University(斯坦福大学) University of Toronto(多伦多大学)

AI总结 ReplicationBench旨在评估人工智能代理复制天体物理学研究论文的能力,通过测试代理在实验设置、推导、数据分析和代码库等任务上的表现,揭示代理在科学研究中的可靠性及挑战。

详情

展开后加载摘要…

URL PDF HTML 收藏
2509.21354 2025-11-25 cs.CV cs.AI

KV-Efficient VLA: A Method to Speed up Vision Language Models with RNN-Gated Chunked KV Cache

KV-Efficient VLA: 一种加速视觉语言模型的方法通过RNN门控分块KV缓存

Wanshun Xu, Long Zhuang, Lianlei Shan

机构 * University of Toronto(多伦多大学) Tsinghua University(清华大学)

AI总结 KV-Efficient VLA通过RNN门控分块KV缓存机制,有效降低视觉语言模型的计算和内存开销,提升推理效率。

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.18120 2025-11-25 cs.CV

MVS-TTA: Test-Time Adaptation for Multi-View Stereo via Meta-Auxiliary Learning

MVS-TTA: 通过元辅助学习实现多视图立体的测试时间适应

Hannuo Zhang, Zhixiang Chi, Yang Wang, Xinxin Zuo

机构 * Gina Cody School of Engineering and Computer Science, Concordia University(吉娜·柯迪工程与计算机科学学院,康科迪亚大学) The Edward S. Rogers Sr. Department of Electrical and Computer Engineering, University of Toronto(埃德华·S·罗杰斯Sr.电子与计算机工程系,多伦多大学)

AI总结 MVS-TTA通过元辅助学习结合优化与学习方法,提升多视图立体的测试时间适应性能。

Comments 8 pages, 7 figures

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.17760 2025-11-25 cond-mat.mtrl-sci cs.LG

When Active Learning Fails, Uncalibrated Out of Distribution Uncertainty Quantification Might Be the Problem

当主动学习失效时,未校准的分布外不确定性量化可能是问题所在

Ashley S. Dale, Kangming Li, Brian DeCost, Hao Wan, Yuchen Han, Yao Fehlis, Jason Hattrick-Simpers

机构 * University of Toronto(多伦多大学) King Abdullah University of Science and Technology(卡布斯大学) National Institute of Standards and Technology(美国国家标准与技术研究院) Artificial, Inc.(Artificial公司)

AI总结 研究发现,在主动学习中,未校准的分布外不确定性量化可能影响模型泛化能力,需关注特征输入空间中的经验不确定性。

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.17078 2025-11-24 cs.LG

Hash Collisions in Molecular Fingerprints: Effects on Property Prediction and Bayesian Optimization

分子指纹中的哈希碰撞:对性质预测和贝叶斯优化的影响

Walter Virany, Austin Tripp

机构 * University of Toronto(多伦多大学) Valence Labs(Valence实验室)

AI总结 本研究探讨了分子指纹中哈希碰撞对性质预测和贝叶斯优化的影响,发现精确指纹在性质预测中略有提升,但在贝叶斯优化中未见显著改进。

Comments NeurIPS 2025 AI4Science workshop. Code: https://github.com/wvirany/molcollisions Openreview: https://openreview.net/forum?id=POgOHi8a7t

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.15996 2025-11-21 cs.IR cs.CL

QueryGym: A Toolkit for Reproducible LLM-Based Query Reformulation

QueryGym: 一个用于可重复LLM查询重述的工具包

Amin Bigdeli, Radin Hamidi Rad, Mert Incesu, Negar Arabzadeh, Charles L. A. Clarke, Ebrahim Bagheri

机构 * University of Waterloo(滑铁卢大学) Mila – Quebec AI Institute(魁北克人工智能研究所) University of Toronto(多伦多大学) University of California, Berkeley(加州大学伯克利分校)

AI总结 QueryGym是一个用于可重复LLM查询重述的开源工具包,提供统一框架和多种集成支持,促进方法比较与基准测试。

Comments 4 pages

详情

展开后加载摘要…

URL PDF HTML 收藏
2408.07757 2025-11-20 cs.RO

Inverse k-visibility for RSSI-based Indoor Geometric Mapping

Junseo Kim, Matthew Lisondra, Yeganeh Bahoo, Sajad Saeedi

机构 * Department of Cognitive Robotics, Delft University of Technology(认知机器人系,代尔夫特理工大学) Department of Mechanical and Industrial Engineering, University of Toronto(机械与工业工程系,多伦多大学) Department of Computer Science, Toronto Metropolitan University(计算机科学系,多伦多 Metropolitan 大学) Department of Computer Science, University College London(计算机科学系,伦敦大学学院)

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.14848 2025-11-20 cs.CV

Gaussian See, Gaussian Do: Semantic 3D Motion Transfer from Multiview Video

Yarin Bekor, Gal Michael Harari, Or Perel, Or Litany

机构 * Technion - Israel Institute of Technology(技术学院-以色列理工学院) NVIDIA(英伟达) University of Toronto(多伦多大学) Vector Institute(向量研究所)

Comments SIGGRAPH Asia 2025

详情

展开后加载摘要…

URL PDF HTML 收藏
2505.20429 2025-11-19 cs.CL cs.CV

PreP-OCR: A Complete Pipeline for Document Image Restoration and Enhanced OCR Accuracy

Shuhao Guan, Moule Lin, Cheng Xu, Xinyi Liu, Jinman Zhao, Jiexin Fan, Qi Xu, Derek Greene

机构 * University College Dublin(都柏林大学) Trinity College Dublin(都柏林三一学院) University of Toronto(多伦多大学) Shanghai University(上海大学)

Comments ACL 2025 main

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.12363 2025-11-18 cs.CV

Explainable AI-Generated Image Detection RewardBench

Michael Yang, Shijian Deng, William T. Doan, Kai Wang, Tianyu Yang, Harsh Singh, Yapeng Tian

机构 * The University of Texas at Dallas(德克萨斯大学达拉斯分校) University of Toronto(多伦多大学) University of Notre Dame(诺特大学) Stony Brook University(石溪大学)

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.12342 2025-11-18 cs.CV cs.AI cs.LG

Ground Plane Projection for Improved Traffic Analytics at Intersections

Sajjad Pakdamansavoji, Kumar Vaibhav Jha, Baher Abdulhai, James H Elder

机构 * Center of Vision Research (CVR) and Centre for AI & Society (CAIS) at York University(约克大学视觉研究中心(CVR)和人工智能与社会研究中心(CAIS)) Vector Institute for Artificial Intelligence(人工智能向量研究所) Region of York(约克地区) York University(约克大学) University of Toronto(多伦多大学)

Journal ref 2024 IEEE International Conference on Intelligent Transportation Systems (ITSC), Edmonton, Canada, pp. 741-748, 2024

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.12278 2025-11-18 stat.ML cs.LG

PCA++: How Uniformity Induces Robustness to Background Noise in Contrastive Learning

Mingqi Wu, Qiang Sun, Yi Yang

机构 * McGill University(麦吉尔大学) Mila University of Toronto(多伦多大学) MBZUAI

Comments 14 pages main, 26 pages appendix

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.09287 2025-11-18 cs.AI cs.CY cs.LG

From Model Training to Model Raising

Roland Aydin, Christian Cyron, Steve Bachelor, Ashton Anderson, Robert West

机构 * Hamburg University of Technology(汉堡理工大学) Helmholtz-Zentrum Hereon(海德堡研究中心) University of Toronto(多伦多大学) EPFL(苏黎世联邦理工学院)

Comments Accepted for publication in Communications of the ACM (CACM), Opinion section

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.11576 2025-11-18 cs.LG cs.AI

DAOpt: Modeling and Evaluation of Data-Driven Optimization under Uncertainty with LLMs

WenZhuo Zhu, Zheng Cui, Wenhan Lu, Sheng Liu, Yue Zhao

机构 * Zhejiang University(浙江大学) University of Toronto(多伦多大学) Peking University(北京大学)

详情

展开后加载摘要…

URL PDF HTML 收藏
2506.18588 2025-11-17 cs.LG cs.AI stat.ML

Optimization-Induced Dynamics of Lipschitz Continuity in Neural Networks

Róisín Luo, James McDermott, Christian Gagné, Qiang Sun, Colm O'Riordan

机构 * University of Galway(Galway大学) Irish National Centre for Research Training in AI (CRT-AI)(爱尔兰人工智能研究培训中心) Université Laval(拉瓦尔大学) Canada-CIFAR AI Chair(加拿大CIFAR人工智能 chair) Mila - Québec AI Institute(魁北克人工智能研究所) University of Toronto(多伦多大学) MBZUAI, UAE(阿联酋MBZUAI)

详情

展开后加载摘要…

URL PDF HTML 收藏
2411.18634 2025-11-14 cs.CL

Semantic, Orthographic, and Phonological Biases in Humans' Wordle Gameplay

Jiadong Liang, Adam Kabbara, Jiaying Liu, Ronaldo Luo, Kina Kim, Michael Guerzhoy

机构 * Division of Engineering Science, University of Toronto(工程科学系,多伦多大学)

Comments Findings of the ACL: IJCNLP-AACL 2025 (accepted)

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.09665 2025-11-14 cs.LG

Generalization Can Emerge in Tabular Foundation Models From a Single Table

Junwei Ma, Nour Shaheen, Alex Labach, Amine Mhedhbi, Frank Hutter, Anthony L. Caterini, Valentin Thomas

机构 * University of Toronto(多伦多大学) Polytechnique Montréal(蒙特利尔大学) Mila – Quebec AI Institute(魁北克人工智能研究所) Layer 6 AI(Layer 6 AI公司) Prior Labs(Prior实验室) ELLIS Institute(ELLIS研究所) Tübingen University of Freiburg(弗赖堡大学)

详情

展开后加载摘要…

URL PDF HTML 收藏
2503.06161 2025-11-14 cs.CV cs.AI

Feature-EndoGaussian: Feature Distilled Gaussian Splatting in Surgical Deformable Scene Reconstruction

Kai Li, Junhao Wang, William Han, Ding Zhao

机构 * University of Toronto(多伦多大学) Carnegie Mellon University(卡内基梅隆大学)

Comments 17 pages, 5 figures; Accepted to ML4H 2025

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.09228 2025-11-13 cs.CV cs.CL

Taming Object Hallucinations with Verified Atomic Confidence Estimation

Jiarui Liu, Weihao Xuan, Zhijing Jin, Mona Diab

机构 * CMU(卡内基梅隆大学) The University of Tokyo(东京大学) The University of Toronto(多伦多大学)

详情

展开后加载摘要…

URL PDF HTML 收藏
2510.23749 2025-11-13 astro-ph.IM cs.LG

Re-envisioning Euclid Galaxy Morphology: Identifying and Interpreting Features with Sparse Autoencoders

John F. Wu, Michael Walmsley

机构 * Space Telescope Science Institute(太空望远镜科学研究所) Johns Hopkins University(约翰霍普金斯大学) University of Toronto(多伦多大学)

Comments Authors contributed equally to this work. Accepted to NeurIPS Machine Learning and the Physical Sciences Workshop. See trained model at https://huggingface.co/mwalmsley/euclid-rr2-mae, HuggingFace demo at https://huggingface.co/spaces/mwalmsley/euclid_masked_autoencoder, and code at https://github.com/jwuphysics/euclid-galaxy-morphology-saes

详情

展开后加载摘要…

URL PDF HTML 收藏
2508.06894 2025-11-13 cs.AI cs.LG

Pushdown Reward Machines for Reinforcement Learning

Giovanni Varricchione, Toryn Q. Klassen, Natasha Alechina, Mehdi Dastani, Brian Logan, Sheila A. McIlraith

机构 * Utrecht Universiteit(乌得勒支大学) University of Toronto(多伦多大学) Vector Institute(向量研究所) Open Universiteit(开放大学) University of Aberdeen(阿伯丁大学)

Comments Extended version of a paper accepted for publication at the 22nd International Conference on Principles of Knowledge Representation and Reasoning (KR 2025)

详情

展开后加载摘要…

URL PDF HTML 收藏
2506.08065 2025-11-13 astro-ph.IM cs.LG

Dynamic Diffusion Schrödinger Bridge in Astrophysical Observational Inversions

Ye Zhu, Duo Xu, Zhiwei Deng, Jonathan C. Tan, Olga Russakovsky

机构 * LIX, École Polytechnique, IP Paris(巴黎高等理工学院LIX研究所) Canadian Institute for Theoretical Astrophysics (CITA), University of Toronto(多伦多大学加拿大理论天体物理研究所) Google DeepMind(谷歌DeepMind) Department of Astronomy, University of Virginia(弗吉尼亚大学天文学系) Department of Space, Earth & Environment, Chalmers University of Technology(楚科夫斯基技术大学空间、地球与环境系)

Comments Accepted to NeurIPS 2025. Code available at https://github.com/L-YeZhu/AstroDSB. Updated funding information from last version

详情

展开后加载摘要…

URL PDF HTML 收藏