Bingao Chen | 陈炳翱

Hi, I am a senior undergraduate student at Tsinghua University. I am currently working with Prof. C. Karen Liu at Stanford University.

Previously, I had the privilege of working with Prof. Xiangyu Yue at CUHK MMLab. Last summer, I was a visiting student with Prof. Chen Feng at NYU's AI4CE Lab.

My research lies in Robot Learning, with a focus on agentic robotics. I am interested in what must be built around the robot foundation model — planning, memory, active perception, skill transfer — and in making those capabilities hold up across tasks and environments. My goal is to build general embodied agents that solve real-world problems.

I am currently seeking a Ph.D. position starting in Fall 2027, and I am always happy to chat about research fit or potential collaborations. Feel free to reach out!

× WeChat QR code

Scan to add me on WeChat
Bingao Chen

News

  • 2026.09•📝 New blog post: Can Astra Solve RoboMME without Breaking the Bank?
  • 2026.07•🛫 Our embodied foundation model τ0-VLA has been released! Check out our long-horizon demos!
  • 2026.06•🌲 I joined The Movement Lab at Stanford University as a visiting student. Many thanks to Guy and Karen for the opportunity!
  • 2026.06•🏆 Our team won 1st Place in the RoboMME Challenge, and I gave a contributed talk on our approach at the FMEA workshop @ CVPR 2026.
  • 2025.10•🎓 I was awarded the National Scholarship, the highest honor for undergraduate students in China.

Research

Three-tier system: GPT-6 Astra plans subtasks, fine-tuned Qwen3-VL-4B checks completion, and fine-tuned pi0.5 acts

Can Astra Solve RoboMME without Breaking the Bank?
Bingao Chen, Haoquan Fang, C. Karen Liu
Blog Post, 2026
Blog Code

A three-tier system for memory-augmented manipulation: GPT-6 Astra plans grounded subtasks, a fine-tuned π0.5 executes them, and a small fine-tuned VLM monitors subtask completion to decide when to call Astra again. On 800 RoboMME test episodes, it reaches 79.13% success with only 3.63 Astra calls per episode, close to the 84.08% oracle-guided upper bound.

tau0-VLA teaser: pretraining data, high-level policy with test-time computation, and capabilities

τ0-VLA: A Hierarchical Robot Foundation Model with World-Model-Guided Test-Time Computation
Xiaowei Cai, Yunuo Cai, Bingao Chen, Jingxiao Chen, Zhi Chen, Siyuan Feng, Tengyu Hou, Jingshun Huang, Han Jiang, Runkun Ju, et al.
Technical Report, 2026
Paper Code Model Project page

A hierarchical robot foundation model that treats high-level subtask generation as a compute-scalable inference problem: it proposes candidate subtasks, predicts their outcomes with a world model, and searches over them before committing. Trained on 40,115 hours of real-world data, it improves success on long-horizon manipulation tasks.

Experience

Stanford University

Stanford University

U.S. 2026.06 - Present

Visiting Student at The Movement Lab

Advisor: Prof. C. Karen Liu
AgiBot

AgiBot

China 2026.02 - 2026.06

Research Intern

Advisor: Prof. Xiangyu Yue
New York University

New York University

U.S. 2025.06 - 2025.09

Visiting Student at the AI4CE Lab

Advisor: Prof. Chen Feng
Tsinghua University

Tsinghua University

China 2023.09 - Present

Undergraduate Student

Selected Awards

  • 2026: 1st Place, RoboMME Challenge @ CVPR 2026 FMEA Workshop
  • 2025: National Scholarship, the highest honor for undergraduate students in China

Webpage design inspired by Anqi Li.
Last updated: September, 2026