Yunkai Zhan | 詹云凯
Hey there 👋
Yunkai Zhan | 詹云凯
Think as a Researcher, Ship as an Engineer
I am pursuing a B.S. in Computer Science and Applied Math at the University of Southern California. I'm interested in scalable, transferable, and data-driven methods for solving real-world problems. My work applies Reinforcement Learning and Agentic Systems to areas such as Coding Models, Robot Learning Data, and Social Simulation. At USC, I work closely with Wang Bill Zhu and Yue Wang.
🎓 I will graduate in May 2027 and am actively looking for internship and full-time Applied Researcher / Research Engineer positions. Feel free to reach out!
LinkedIn GitHub Google Scholar X Email
Publications

PreciseCoder: Faithful and Generalizable Debugging Beyond Unit-Test Rewards
Coding Models

Yunkai Zhan†, Miaosen Chai, Shangshang Wang, Jike Zhong, Yuqing Yang, Deqing Fu, Serina Chang, Wang Bill Zhu†

arXiv preprint, 2026

Website Paper

Humanoid Everywhere: Large-Scale Real-to-Sim Environment and Data Generator for Humanoid Loco-Manipulation
Robotics

Junying Wang, Xinjie Chen, Yunkai Zhan, Junsoo Kim, Songlin Wei, Xulang Guan, Vitor Campagnolo Guizilini, Chuhang Zou, Jiageng Mao, Yue Wang

RSS 2026 Workshop on Data-Centric Robotics: What Data Do Robots Really Need?, 2026

Website Paper
Latest from blog
When Research No Longer Needs Researchers

When Research No Longer Needs Researchers

How coding agents are taking over the AI research loop, the AI-native routes to self-improvement, and what recursive self-improvement means for science, academia, safety, and the role of humans.
© 2026 Yunkai Zhan | 詹云凯