Tao Yu

Tao Yu 俞涛

Tech Lead, LimX Dynamics

I'm the Tech Lead at LimX Dynamics, where I built our motion-mimic system and whole-body tracking system (System 0) from the ground up. I'm broadly interested in reinforcement learning, robot learning, multimodal models, and generative AI — all in service of one long-term goal: building robots that perceive, reason, and act like humans.

Previously I was a Senior Researcher at Tencent (2023–2025), working on reinforcement learning and game AI. I received my Ph.D. in 2023 from the University of Science and Technology of China, advised by Prof. Zhibo Chen and Prof. Weiping Li. During my Ph.D., I was a Visiting Ph.D. at the National University of Singapore (2022–2023), advised by Prof. Xinchao Wang, and a Research Intern at Microsoft Research Asia (2020–2022) with Dr. Cuiling Lan and Prof. Wenjun Zeng.

Hiring We're hiring at LimX Dynamics — full-time researchers/engineers and interns in robot learning and humanoid foundation models. Drop me a CV at tao@limxdynamics.com.

Selected Publications

Full list on my Google Scholar. (* equal contribution)

Selected Open-Source Projects

More on GitHub.

Experience

2025.03 – now
Tech Lead, LimX Dynamics · Shenzhen
Humanoid whole-body control and humanoid foundation models.
2023.07 – 2025.02
Senior Researcher, Tencent · Shenzhen
Reinforcement learning and game AI.
2022.07 – 2022.08
Research Intern, Tencent · Shenzhen
Reinforcement learning and game AI.
2020.07 – 2022.01
Research Intern, Microsoft Research Asia · Beijing
Data-efficient reinforcement learning, representation learning.

Education

2018 – 2023
Ph.D., Electronic Engineering & Information Science, University of Science and Technology of China
Advised by Prof. Zhibo Chen and Prof. Weiping Li.
2022 – 2023
Visiting Ph.D., Electrical and Computer Engineering, National University of Singapore
Advised by Prof. Xinchao Wang.
2014 – 2018
B.Eng., Electronics & Information Engineering, Anhui University

Selected Awards

Service

Reviewer for NeurIPS, ICML, ICLR, CVPR, ICCV, ECCV, AAAI, ICRA, TPAMI, IJCV, TMM.

Contact

The fastest way to reach me is by email — tao@limxdynamics.com for work, yutao666@mail.ustc.edu.cn for personal. I'm also on GitHub, Google Scholar, LinkedIn, Twitter/X, Zhihu, and WeChat geekyutao.