Yihao Wu
Yihao Wu is a robotics researcher with a joint affiliation between Tsinghua University and Tencent Robotics X. He is best known as the lead author of FlowPRO, a reward-free reinforced fine-tuning framework for flow-matching Vision-Language-Action models published in June 2025, which introduces the RPRO (Robotic Flow-matching Proximalized Preference Optimization) algorithm. His research focuses on improving robot manipulation policies through human-guided failure correction without hand-designed reward functions, with demonstrated results on precision bimanual tasks.
“A new fine-tuning framework that uses human-guided failure correction — without designing reward functions — to push robot manipulation policies from 'good demo performance' to near-deployment-grade reliability, achieving 92–99% success rates on hard bimanual tasks.”
Source→“Yihao Wu — Tsinghua University / Tencent Robotics X. Lead author, joint appointment between one of China's premier engineering universities and Tencent's robotics division.”
Source→AI-extracted from podcast / newsletter / paper summaries. May contain errors.