Perry Dong
Perry Dong is a PhD researcher at Stanford University whose work focuses on reinforcement learning for robotics. He is best known as the lead author of EXPO and EXPO-FT, which address stable, sample-efficient online RL fine-tuning of expressive and vision-language-action policies, achieving high task reliability with minimal real-world robot interaction time.
“Stanford researchers have cracked a critical bottleneck in physical AI deployment: how to take a pretrained robot foundation model and push it to 100% task reliability in under 20 minutes of real robot time.”
Source→“We build on the recently proposed EXPO algorithm, which provides a principled foundation for RL fine-tuning in this regime.”
Source→AI-extracted from podcast / newsletter / paper summaries. May contain errors.