I'm a final-year Computer Science PhD student at the Institute for Interdisciplinary Information Sciences (IIIS), Tsinghua University, advised by Professor Huazhe Xu. I'm currently interning at Amazon's Frontier AI & Robotics (FAR) team. I received my Bachelor's Degree at the Department of Electronic Engineering, Tsinghua University.
My research focuses on how robots can acquire diverse skills and generalize in the real world. To that end, I've explored the dynamics and learned priors of generative models, data scaling, and real-world learning.
I was fortunate to work with Professors C. Karen Liu and Shuran Song during my visit to the Movement Lab at Stanford University. I'm more than glad to work with Prof. Karen Liu again at FAR!
I'm always curious about the intersection of different fields. If you're up to something weird, I'm all ears! I love watching movies and have got my hands on (very premature) documentary making. Here are some of my works.
I'm actively exploring industry opportunities. Feel free to invite me for a coffee chat!
Updates
- Oct. 2026 We released BURGER, a benchmark for understanding VLA scaling through drawing!
- Aug. 2026 I started my internship at Amazon's Frontier AI & Robotics (FAR), happy to be back at the bay area!
- July 2026 I led the organization of the Diffusion for Robot Learning workshop at RSS 2026 in Sydney!
- Mar. 2026 AffordGen has been accepted to CVPR 2026!
- Jan. 2026 Failure-Aware RL has been accepted to ICRA 2026!
- Nov. 2025 Robot Trains Robot has been accepted to CoRL 2025!
- Mar. 2025 I finished my visit and returned to Tsinghua! Chat is always welcome!
- Feb. 2025 Two papers Stem-Ob and DenseMatcher has been accepted to ICLR 2025 as spotlight!
- Sept. 2024 Make-An-Agent has been accpeted to Neurips 2024.
- Aug. 2024 I started my visiting at Stanford University! Advised by Prof. C. Karen Liu.
- July 2024 Robo-ABC has been accpeted to ECCV 2024.
- May 2024 Rethinking Transformers in Solving POMDPs has been accpeted to ICML 2024.
- Jan. 2024 Uni-O4 has been accpeted to ICLR 2024 with a high score of 6688!
- Sept. 2023 LfVoid and RL-ViGen has been accpeted to Neurips 2023.
- Jan. 2023 DeFog has been accpeted to ICLR 2023.
- Jan. 2023 EIL has been accpeted to ICRA 2023.
- Sept. 2022 I started my PhD at Tsinghua University, advised by Prof. Huazhe Xu.
Community
Diffusion for Robot Learning: From Generation to Action
RSS 2026 Workshop · July 17, 2026 · Sydney, Australia
Organizers: Kaizhe Hu, Brent Yi, Haoyu Xiong, Mitsuhiko Nakamoto, Takara Truong, Utkarsh Mishra, Xiaomeng Xu, Huazhe Xu, Shuran Song
Speakers: Yilun Du, Negar Mehr, Max Simchowitz, Ian Manchester
Website • OpenReview • Twitter
Research
Xiangyu Xu*,
Bangshuai Peng*,
Ziyi Liu*,
Mengjie Zhao*,
Chenhao Lu,
Zhengyang Fan,
Huazhe Xu†,
Kaizhe Hu†
arXiv preprint, Oct. 2026
* Equal contribution † Corresponding authors
Website •
ArXiv •
Data
AffordGen: Generating Diverse Demonstrations for Generalizable Object Manipulation with Affordance Correspondence
Jiawei Zhang*,
Kaizhe Hu*,
Yingqian Huang,
Yuanchen Ju,
Zhengrong Xue,
Huazhe Xu
CVPR 2026, Poster, Apr. 2026
Website •
ArXiv •
Code •
Twitter
Failure-Aware RL: Reliable Offline-to-Online Reinforcement Learning with Self-Recovery for Real-World Manipulation
Huanyu Li*,
Kun Lei*,
Sheng Zang,
Kaizhe Hu,
Yongyuan Liang,
Bo An,
Xiaoli Li,
Huazhe Xu
ICRA 2026, Poster, Jan. 2026
Website •
ArXiv •
Twitter
Robot Trains Robot: Automatic Real-World Policy Adaptation and Learning for Humanoids
Kaizhe Hu*,
Haochen Shi*,
Yao He,
Weizhuo Wang,
C. Karen Liu†,
Shuran Song†
CoRL 2025, Poster, Oct. 2025
Website •
ArXiv •
Code •
Video •
Twitter
Stem-Ob: Generalizable Visual Imitation Learning with Stem-Like Convergent Observation through Diffusion Inversion
Kaizhe Hu*,
Zihang Rui*,
Yao He,
Yuyao Liu,
Pu Hua,
Huazhe Xu
ICLR 2025, Spotlight (Top 5.1%) , Nov. 2024
Website •
ArXiv •
Code •
Video •
Twitter
DenseMatcher: Learning 3D Semantic Correspondence for Category-Level Manipulation from One Demo
Junzhe Zhu*,
Yuanchen Ju*,
Junyi Zhang,
Muhan Wang,
Zhecheng Yuan,
Kaizhe Hu,
Huazhe Xu
ICLR 2025, Spotlight (Top 5.1%) , Dec 2024
Website •
ArXiv •
Code •
Data •
Twitter
Make-An-Agent: A Generalizable Policy Network Generator with Behavior-Prompted Diffusion
Yongyuan Liang,
Tingqiang Xu,
Kaizhe Hu,
Guangqi Jiang,
Furong Huang,
Huazhe Xu
Neurips 2024, Poster, July 2024
Website •
ArXiv •
Code •
Twitter •
Models&Dataset
Robo-ABC: Affordance Generalization Beyond Categories via Semantic Correspondence for Robot Manipulation
Yuanchen Ju*,
Kaizhe Hu*,
Guowei Zhang,
Gu Zhang,
Mingrun Jiang,
Huazhe Xu
European Conference on Computer Vision (ECCV) 2024, Poster, Jan. 2024
Website •
ArXiv •
Code •
Twitter
Rethinking Transformers in Solving POMDPs
Chenhao Lu,
Ruizhe Shi*,
Yuyao Liu*,
Kaizhe Hu,
Simon Shaolei Du,
Huazhe Xu
The International Conference on Machine Learning (ICML) 2024, Poster, May
ArXiv •
Code
Uni-O4: Unifying Online and Offline Deep Reinforcement Learning with Multi-Step On-Policy Optimization
Lei Kun,
Zhengmao He*,
Chenhao Lu*,
Kaizhe Hu,
Gao Yang,
Huazhe Xu
International Conference on Learning Representations (ICLR) 2024, Poster, Nov. 2023
Website •
ArXiv •
Code •
Twitter
LfVoid: Can Pre-Trained Text-to-Image Models Generate Visual Goals for Reinforcement Learning?
Jialu Gao*,
Kaizhe Hu*,
Guowei Xu,
Huazhe Xu
Conference on Neural Information Processing Systems (Neurips) 2023, Poster, Jul. 2023
Website •
ArXiv •
Code •
Twitter
RL-ViGen: A Reinforcement Learning Benchmark for Visual Generalization
Zhecheng Yuan*,
Sizhe Yang*,
Pu Hua,
Can Chang,
Kaizhe Hu,
Xiaolong Wang,
Huazhe Xu
Conference on Neural Information Processing Systems (Neurips) 2023 Datasets and Benchmarks Track, Poster, Jul. 2023
Website •
ArXiv •
Code
DeFog: Decision Transformer under Random Frame Dropping
Kaizhe Hu*,
Ray Chen Zheng* ,
Yang Gao,
Huazhe Xu
International Conference on Learning Representations (ICLR) 2023, Poster, Mar. 2023
Website •
ArXiv •
Code
EIL: Extraneousness-Aware Imitation Learning
Ray Chen Zheng*,
Kaizhe Hu*,
Zhecheng Yuan,
Boyuan Chen,
Huazhe Xu,
International Conference on Robotics and Automation (ICRA) 2023, Poster, Oct. 2022
Website •
ArXiv •
Code