Feiyang Wu

I am on the 2027 postdoc / research scientist job market. If you’re hiring or know of a good fit, I’d love to hear from you — reach out via email.

My research interests lie in the intersection of optimization, reinforcement learning, and robotics. I develop algorithms to efficiently train robots, ideally with theoretical guarantee.

I am an ML PhD student at Georgia Tech, co-advised by Prof. Anqi Wu and Prof. Ye Zhao. Additionally, I had the privilege to collaborate with Prof. George Lan on stochastic optimization and RL.

News

Nov 2026 Spectral Skills won a second award: the Best Paper Award at the Compositional Generalization workshop, CoRL 2026!
Oct 2026 Our Humanoid SRL paper is now on arXiv!
Sep 2026 Our Spectral Skills workshop paper won the ROBOTIS Best Paper Award (1st place) at CompRobotis, IROS 2026!
Sep 2026 Our paper Spectral Skills is on arXiv: a latent command space for humanoid control that tracks, chains, and composes skills. Blog.
Earlier news
May 2026 Our Distributional Inverse Reinforcement Learning paper was upgraded to an oral presentation at ICML 2026!
May 2026 Our RA-L paper Learn to Teach was selected for an oral presentation at ICRA 2026!
Apr 2026 We deployed the L2T policy on Digit aboard the Navy’s Stiletto. Field notes and videos.
Apr 2026 Our paper Distributional Inverse Reinforcement Learning was accepted as an ICML 2026 spotlight!
Jan 2026 Two papers accepted to ICRA 2026!
Sep 2025 Our work on IRL with reward distributions is released on arXiv.
Aug 2025 Our paper L2T got accepted at RA-L. Check out here.
May 2025 I passed my qualifying exam.
May 2025 One paper accepted to ICML 2025. Congrats Jingyang!
Apr 2025 One paper submitted to RA-L. Check out lidar-learn-to-teach.github.io.
Sep 2024 Visiting student Xavier Nal finished his master’s project in our lab.
May 2024 Interned at Georgia Tech Research Institute on generalizing diffusion policies for robot manipulation.
Dec 2023 Our paper on inverse reinforcement learning accepted to ICRA 2024.
Oct 2023 Our paper on average reward IRL accepted as poster at NeurIPS 2023.

Selected publications

Humanoid robots with neck- and back-mounted extra arms performing manipulation and support tasks
Humanoid SRL
arXiv 2026
An extra arm for humanoid loco-manipulation.
arXiv Blog
Spectral Skills
arXiv 2026
Composable skill commands for humanoid control.
Project page arXiv Blog
GM-Loco
arXiv 2026
Humanoid walking on sand and gravel.
Project page arXiv Blog
Learn to Teach
RA-L 2025 · ICRA 2026 Oral
Sample-efficient teacher-student locomotion on Digit.
Project page arXiv Blog Video Field test
SEEC: Stable End-Effector Control
ICRA 2026
Steady hands while a humanoid walks.
Project page arXiv Blog
RL-augmented Adaptive MPC
ICRA 2026
RL-corrected MPC for bipedal walking.
Project page arXiv Blog
Neural dopamine reward recovery distributions and empirical CDFs
Distributional Inverse Reinforcement Learning
ICML 2026 Oral
Reward distributions from demonstrations.
arXiv Blog

All publications

★ Humanoid SRL: Augmenting Humanoids with Supernumerary Limbs for Coordinated Loco-Manipulation
Fukang Liu, Feiyang Wu, Paul Barsa, Jaehwi Jang, Akshat Jain, Zhigen Zhao, Yipu Chen, Zhaoyuan Gu, Nikhil Vemuri, Anshul Chandaliya, Kaijie Zhu, Sean Lublinsky, Matthew Zhu, Tony G. Chen, Ye Zhao
arXiv preprint, 2026
★ Learning Expressive and Compositional Motion Representation via Spectral Skills
Feiyang Wu*, Chenxiao Gao*, Chen Yang, Ye Zhao, Bo Dai, Anqi Wu
arXiv preprint, 2026
Best Paper Award — Compositional Generalization for Real-World Robot Learning, CoRL 2026 (workshop version)
ROBOTIS Best Paper Award (1st place) — CompRobotis, IROS 2026 (workshop version)
GM-Loco: Terrain-Adaptive Humanoid Locomotion on Granular Media
Junnosuke Kamohara, Feiyang Wu, Andy Ningan Zong, Daniel I. Goldman, Yashwanth Nakka, Seth Hutchinson, Ye Zhao
arXiv preprint, 2026
★ Distributional Inverse Reinforcement Learning
Feiyang Wu, Ye Zhao, Anqi Wu
ICML 2026 Oral (0.7% acceptance rate)
RL-augmented Adaptive Model Predictive Control for Bipedal Locomotion over Challenging Terrain
Junnosuke Kamohara, Feiyang Wu, Chinmayee Wamorkar, Seth Hutchinson, Ye Zhao
ICRA 2026
SEEC: Stable End-Effector Control with Model-Enhanced Residual Learning for Humanoid Loco-Manipulation
Jaehwi Jang, Zhuoheng Wang, Ziyi Zhou, Feiyang Wu, Ye Zhao
ICRA 2026
Inverse Reinforcement Learning with Switching Rewards and History Dependency for Characterizing Animal Behaviors
Jingyang Ke, Feiyang Wu, Jiyi Wang, Jeffrey Markowitz, Anqi Wu
ICML 2025
★ Learn to Teach: Sample-Efficient Privileged Learning for Humanoid Locomotion over Real-World Uneven Terrain
Feiyang Wu, Xavier Nal, Jaehwi Jang, Wei Zhu, Zhaoyuan Gu, Anqi Wu, Ye Zhao
RA-L 2025, ICRA 2026 Oral
★ Stochastic First-Order Methods for Average-Reward Markov Decision Processes
Tianjiao Li, Feiyang Wu, Guanghui Lan
Mathematics of Operations Research 2025
Infer and Adapt: Bipedal Locomotion Reward Learning from Demonstrations via Inverse Reinforcement Learning
Feiyang Wu, Zhaoyuan Gu, Hanran Wu, Anqi Wu, Ye Zhao
ICRA 2024
★ Inverse Reinforcement Learning with the Average Reward Criterion
Feiyang Wu, Jingyang Ke, Anqi Wu
NeurIPS 2023
SAniHead: Sketching Animal-Like 3D Character Heads Using a View-Surface Collaborative Mesh Generative Network
Dong Du, Xiaoguang Han, Hongbo Fu, Feiyang Wu, Yizhou Yu, Shuguang Cui, Ligang Liu
IEEE Transactions on Visualization and Computer Graphics (TVCG) 2020