Stanford Seminar - Robot Learning Without Action Chunking

May 23, 2025 Student Speaker - Yuejiang Liu, Stanford University Recent advances in robot learning have mirrored the progress of large language models in many ways—yet, one key distinction remains: action chunking. In this talk, I will begin with an analysis of action chunking, highlighting its inherent tradeoff between long-term consistency and short-term reactivity. I will then introduce two methods to address this tradeoff: (i) Bidirectional Decoding: an inference algorithm that jointly optimizes consistency and reactivity using additional compute at test time; (ii) Past-Token Prediction: an auxiliary training objective that encourages diffusion policies to capture temporal dependencies in long-context observations. Together, these methods offer a promising path toward memory-aware robot policies without action chunking. About the speaker: https://sites.google.com/view/yuejiangliu/home More about the course can be found here: https://stanfordasl.github.io/robotics_seminar/ View the entire AA289 Stanford Robotics and Autonomous Systems Seminar playlist: https://www.youtube.com/playlist?list=PLoROMvodv4rMeercb-kvGLUrOq4HR6BZD ► Check out the entire catalog of courses and programs available through Stanford Online: https://online.stanford.edu/explore View our Robotics and Autonomous Systems Graduate Certificate: https://online.stanford.edu/programs/robotics-and-autonomous-systems-graduate-certificate