Qiayuan Liao

I'm a third-year Ph.D. student at UC Berkeley, advised by Koushil Sreenath and Pieter Abbeel.

Previously, I was a research intern at Boston Dynamics, where I was fortunate to work with Scott Kuindersma on reinforcement-learning control for the E-Atlas.

I'm a true full-stack roboticist dedicated to creating general-purpose humanoid systems. Toward this goal, I focus on (a) agile and robust whole-body loco-manipulation and (b) co-design of data, hardware, and control.

Scholar      Github      LinkedIn     

Workhorse preprint is out!
BeyondMimic is published in Science Robotics!
MuJoCo Playground got ★ Best Demo Paper Award, RSS 2025 ★
Four papers were accepted to RSS 2025, many thanks to my collaborators...

Workhorse: Learning Robust Whole-Body Humanoid Loco-Manipulation from Human Data

Songbo Hu*, Qiayuan Liao*, Yufeng Chi, Kevin Zakka, Yakun Sophia Shao, Pieter Abbeel, Koushil Sreenath
Webpage  •   Paper  •   Video

BeyondMimic: From Motion Tracking to Versatile Humanoid Control via Guided Diffusion

Qiayuan Liao*, Takara E. Truong*, Xiaoyu Huang*, Yuman Gao, Guy Tevet, Koushil Sreenath†, C. Karen Liu†
Science Robotics, 2026
Webpage  •   Paper  •   Code

MuJoCo Playground

Kevin Zakka*, Baruch Tabanpour*, Qiayuan Liao*, Mustafa Haiderbhai*, Samuel Holt*, Jing Yuan Luo, Arthur Allshire, Erik Frey, Koushil Sreenath, Lueder A. Kahrs, Carmelo Sferrazza, Yuval Tassa, Pieter Abbeel
Robotics: Science and Systems (RSS) 2025
★ Best Demo Paper Award, RSS ★
Webpage  •   Paper  •   Code

Berkeley Humanoid A Research Platform for Learning-based Control

Qiayuan Liao, Bike Zhang, Xuanyu Huang, Xiaoyu Huang, Zhongyu Li, Koushil Sreenath
International Conference on Robotics and Automation (ICRA), 2025
Webpage  •   Paper  •   Code

Walking in Narrow Spaces: Safety-critical Locomotion Control for Quadrupedal Robots with Duality-based Optimization

Qiayuan Liao, Bike Zhang, Xuanyu Huang, Xiaoyu Huang, Zhongyu Li, Koushil Sreenath
International Conference on Intelligent Robots and Systems (IROS), 2023
Paper  •   Code

I develop open-source projects used widely in both industry and academia, spanning model-based and data-driven control libraries, and robot platforms. I value rigorous, reproducible research and solid engineering—If these code doesn't work as advertised, feel free to publicly roast me.
whole_body_tracking - Scalable whole body motion tracking training for humanoid robots
legged_control - NMPC, WBC, state estimation, and sim2real framework for legged robots based on OCS2 and ros-controls Retired
mjlab - Combines Isaac Lab's proven API with best-in-class MuJoCo physics to provide lightweight, modular abstractions for RL robotics research and sim-to-real deployment
Berkeley-Humanoid-Lite - sub-$5,000 humanoid robot
MuJoCo Playground - Train and deploy robot skills in minutes with GPU-accelerated MuJoCo Retired
“If you fail to abide by the terms of this license, then your conscience
will haunt you for the rest of your life.” — System Enhancement Associates, ARC shareware license (1985)