causal
Prior knowledge is the key to making RL work for robots
Although RL is in principle better than imitation learning, current models use imitation/supervised learning first because effective learning from one's own experience requires already knowing something about the task; supervised learning builds the prior-knowledge foundation that lets the model later figure things out quickly via RL, mirroring how LLMs go from next-token prediction to RL.
causalpending
Speaker
Sergey LevineEvidence Quote
“LLMs start off being trained purely with next token prediction. That provided an excellent starting point, first for all sorts of synthetic data generation and then for RL.”
Source
Fully autonomous robots are much closer than you think – Sergey Levine— Dwarkesh Patel PodcastCreated: 6/13/2026, 3:29:18 AM
Your notes
Loading notes...