arXivDaily arXiv每日学术速递 周一至周五更新

大厂专区

NVIDIA(英伟达)

2026-05-04 至 2026-05-04 共收录 4
2510.18900 2026-05-04 physics.chem-ph cond-mat.mtrl-sci cs.LG

Foundation Models for Discovery and Exploration in Chemical Space

化学空间发现与探索中的基础模型

Alexius Wadell, Anoushka Bhutani, Victor Azumah, Austin R. Ellis-Mohr, Andrew J. Stier, Kareem Hegazy, Alexander Brace, Hancheng Zhao, Celia Kelly, Anuj K. Nayak, Yuhan Chen, Dimitrios Simatos, Hongyi Lin, Murali Emani, Venkatram Vishwanath, Kevin Gering, Melisa Alkan, Tom Gibbs, Jack Wells, Wesley W. Qian, Richard C. Gerkin, Benjamin Amorelli, Alexander B. Wiltschko, Lav R. Varshney, Bharath Ramsundar, Karthik Duraisamy, Michael W. Mahoney, Arvind Ramanathan, Venkatasubramanian Viswanathan

机构 * Department of Mechanical Engineering, University of Michigan(密歇根大学机械工程系) Department of Chemical Engineering, University of Michigan(密歇根大学化学工程系) Department of Electrical and Computer Engineering, University of Illinois at Urbana-Champaign(伊利诺伊大学厄巴纳-香槟分校电子与计算机工程系) The Santa Fe Institute(圣菲研究所) International Computer Science Institute(国际计算机科学研究所) Department of Statistics, University of California, Berkeley(加州大学伯克利分校统计学系) Department of Computer Science, University of Chicago(芝加哥大学计算机科学系) Argonne National Laboratory(阿贡国家实验室) Idaho National Laboratory(爱达荷国家实验室) NVIDIA Corporation(英伟达公司) Osmo Labs, PBC AI Innovation Institute, Stony Brook University(石溪大学AI创新研究所) Brookhaven National Laboratory(布鲁赫斯研究所) Deep Forest Sciences, Palo Alto, CA(帕洛阿尔托的Deep Forest Sciences) Department of Aerospace Engineering, University of Michigan(密歇根大学航空航天工程系) Lawrence Berkeley National Laboratory(伯克利劳伦斯国家实验室)

AI总结 本文提出MIST模型,通过大规模无标签数据训练,实现对化学空间中多种分子性质的预测,展示了其在多目标电解质溶剂筛选和立体化学推理中的应用,以及在超参数感知贝叶斯神经缩放定律下的高效训练能力。

Comments Main manuscript: 30 pages (including references), 7 tables and 5 figures. Supplementary information: 158 pages (including references), 15 tables and 128 figures

详情

展开后加载摘要…

URL PDF HTML 收藏
2605.00513 2026-05-04 cs.CL cs.LG

ControBench: An Interaction-Aware Benchmark for Controversial Discourse Analysis on Social Networks

ControBench:一种考虑交互的争议性 discourse 分析社交网络基准

Ta Thanh Thuy, Jiaqi Zhu, Xuan Liu, Lin Shang, Reihaneh Rabbany, Guillaume Rabusseau, Lihui Chen, Zheng Yilun, Sitao Luan

机构 * Nanyang Technological University(南洋理工大学) NVIDIA(NVIDIA公司) Nanjing University(南京大学) Mila - Quebec AI Institute(魁北克AI研究所) McGill University(麦吉尔大学) University of Montreal(蒙特利尔大学)

AI总结 ControBench结合异质社交交互图与丰富文本语义,通过Reddit讨论数据构建,用于研究争议性 discourse 分析中的政治极化、虚假信息和内容审核问题。

详情

展开后加载摘要…

URL PDF HTML 收藏
2603.02641 2026-05-04 cs.SD

Rethinking Training Targets, Architectures and Data Quality for Universal Speech Enhancement

重新思考训练目标、架构和数据质量以实现通用语音增强

Szu-Wei Fu, Rong Chao, Xuesong Yang, Sung-Feng Huang, Ryandhimas E. Zezario, Rauf Nasretdinov, Ante Jukić, Yu Tsao, Yu-Chiang Frank Wang

机构 * NVIDIA

AI总结 本文针对通用语音增强中的训练目标选择、失真与感知权衡及数据质量问题,提出改进方法,通过时间移位的无回声干净语音作为学习目标,提出两阶段框架以最小化失真并提升感知质量,并分析训练数据规模与质量的权衡,取得SOTA性能。

详情

展开后加载摘要…

URL PDF HTML 收藏
2509.24276 2026-05-04 cs.AI

G-reasoner: Foundation Models for Unified Reasoning over Graph-structured Knowledge

G-reasoner:用于统一图结构知识推理的基础模型

Linhao Luo, Zicheng Zhao, Junnan Liu, Zhangchi Qiu, Junnan Dong, Serge Panev, Chen Gong, Thuy-Trang Vu, Gholamreza Haffari, Dinh Phung, Alan Wee-Chung Liew, Shirui Pan

机构 * Monash University(莫纳什大学) Nanjing University of Science and Technology(南京理工大学) Griffith University(格里菲斯大学) Shanghai Jiao Tong University(上海交通大学) Tencent Youtu Lab(腾讯优图实验室) NVIDIA(英伟达)

AI总结 G-reasoner通过整合图与语言基础模型,实现对多样化图结构知识的高效推理,采用标准化四层抽象QuadGraph统一异构知识源,并通过混合精度训练和分布式消息传递提升效率与泛化能力。

Comments Accepted by ICLR 2026

详情

展开后加载摘要…

URL PDF HTML 收藏