Yanxiao 赵砚潇
Research Intern at Z.ai
Beijing, China
Summary
🤩
Rockstar🎓
Top SchoolYanxiao 赵砚潇 is a PhD student in computer science at UCAS with eight years of hands-on experience applying deep reinforcement learning to real-world problems such as autonomous driving and LLM reasoning. Currently a research intern at Z.ai, he focuses on RL-driven agents that enable LLMs to interact with computer environments, building on recent internships at Li Auto, SenseTime, and ZERON. He contributes actively to open-source RL tooling—helping maintain cleanrl with bug fixes, Jax and PyTorch implementations, Gymnasium support, and Hugging Face integration—which demonstrates both research rigor and production-oriented engineering. Yanxiao combines academic depth with applied impact, moving from algorithmic development to evaluation and deployment in complex systems. Based in Beijing, he brings a pragmatic research mindset and a track record of shipping reproducible, single-file RL implementations that accelerate experimentation.
8 years of coding experience
Doctor of Philosophy - PhD, Computer Science, Doctor of Philosophy - PhD, Computer Science at University of Chinese Academy of SciencesBachelor of Engineering - BE, Computer Science, Bachelor of Engineering - BE, Computer Science at Xiamen University Tan Kah Kee College
Github Skills (18)
pytorch10
python10
gymnasium10
machine-learning10
reinforcement-learning10
dqn10
tdd10
ddpg10
deep-reinforcement-learning10
jax10
ppp9
huggingface-hub9
deeplearning-ai9
huggingface9
deep-learning9
Programming languages (1)
Github contributions (5)
vwxyzjn/cleanrl
Oct 2022 - Nov 2022
High-quality single file implementation of Deep Reinforcement Learning algorithms with research-friendly features (PPO, DQN, C51, DDPG, TD3, SAC, PPG)
Role in this project:
ML Engineer
ML EngineerContributions:8 reviews, 2 commits, 23 PRs in 13 days
Contributions summary:Yanxiao primarily contributed to the development and maintenance of reinforcement learning algorithms within the repository. Their work included fixing bugs in DDPG and DQN models, integrating Hugging Face for model saving and uploading, and adding support for Gymnasium. Additionally, the user implemented and refactored code for different reinforcement learning algorithms like C51, DDPG, and TD3. The user's contributions encompassed both Jax and PyTorch implementations and covered model evaluation and testing.
pythondeep-reinforcement-learninggomokutd3reinforcement
pa-zyx/training
May 2019 - Nov 2019
Contributions:8 commits, 6 pushes, 1 branch in 6 months
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.