Bowen Wang

I'm a

2nd

year Ph.D. student at the University of Hong Kong, advised by Prof. Tao Yu. Previously I got my bachlor's degree of Computer Science and Technology at Tsinghua University 
My research focuses on building universal digital agents. I have primarily worked on computer-use agents (CUA), developing open foundations through mid-training and post-training and building open-sourced CUA models. My current research interests lie in digital agentic models, particularly in post-training methodologies for creating generalizable and intelligent agents that can reliably operate across diverse digital environments including coding, computer-use, terminal-use and more.
I hope to make computer-use a core capability of frontier models. I believe universal digital agents (UDA) are a path toward AGI. If you would like to talk about research or life, feel free to reach out. I'm open to opportunities.

What's New

  • 06/2026: Gave talks at NVIDIA Research, the Deep Learning Group @ MSR, Mila-Quebec AI Institute, WebAgentLab, and CAMEL-AI. Check out the slides!
  • 05/2026: We're thrilled to release CUA-Gym, check it out! We'll gradually release the full data and models, stay tuned.
  • 05/2026: New blog post: We Spent a Year Building Verifiable Training Tasks for GUI Agents.
  • 02/2026: Computer Agent Arena was accepted to ICLR'26 as a poster (top 3%). See you in Brazil 🇧🇷!
  • 10/2025: OpenCUA was accepted to NeurIPS 2025 as a Spotlight.
  • 09/2025: OpenCUA received the Best Paper Award at the COLM AIA workshop.

Education

  1. Company
    The University of Hong Kong
    Role
    Ph.D. Student
    Date
  2. Company
    Tsinghua University
    Role
    B.E. in Computer Science
    Date

Selected Publications

CUA-Gym: Scaling Verifiable Training Environments and Tasks for Computer-Use Agents

Bowen Wang, Dunjie Lu, Junli Wang, Tianyi Bai, Shixuan Liu, Zhipeng Zhang, Haiquan Wang, Hao Hu, Tianbao Xie, Shuai Bai, Dayiheng Liu, Que Shen, Junyang Lin, Tao Yu

Computer Agent Arena: Evaluating Computer-Use Agents via Crowdsourcing from Real Users

Bowen Wang*, Xinyuan Wang*, Jiaqi Deng*, Tianbao Xie, Ryan Li, Yanzhe Zhang, Zicheng Gong, Gavin Li, Toh Jing Hua, Ion Stoica, Wei-Lin Chiang, Diyi Yang, Yu Su, Yi Zhang, Zhiguo Wang, Victor Zhong, Tao Yu

ICLR'26 (Top 3%)[Website][Paper]

OpenCUA: Open Foundations for Computer-Use Agents

Xinyuan Wang*, Bowen Wang*, Dunjie Lu, Junlin Yang, Tianbao Xie, Junli Wang, Jiaqi Deng, Xiaole Guo, Yiheng Xu, Chen Henry Wu, Zhennan Shen, Zhuokai Li, Ryan Li, Xiaochuan Li, Junda Chen, Boyuan Zheng, Peihang Li, Fangyu Lei, Ruisheng Cao, Yeqiao Fu, Dongchan Shin, Martin Shin, Jiariu Hu, Yuyan Wang, Jixuan Chen, Yuxiao Ye, Danyang Zhang, Yipu Wang, Heng Wang, Diyi Yang, Victor Zhong, Y. Charles, Zhilin Yang, Tao Yu

NeurIPS'25 (Spotlight)COLM'25 AIA Workshop (Best Paper Award)[Website][Paper][Code]

Honors & Awards

  1. 2025
    Workshop Best Paper Award
    COLM 2025 AIA workshop
  2. 2025
    Research Postgraduate Student Innovation Award
    The University of Hong Kong
  3. 2023
    Research and Innovation Scholarship
    Tsinghua University
  4. 2022
    2-nd Prize, Challenge-Cup Competition Award
    Tsinghua University

Talks

  1. 2026
    How Do We Scale Computer-Use Agent Post-training?
    NVIDIA Research
    MSR Deep Learning Group
    Mila-Quebec AI Institute
    WebAgentLab
    CAMEL-AI
  2. 2025
    Language Agents
    University of Tasmania