Amy Zhang
TalkRL: The Reinforcement Learning Podcast
00:00
Invariant Causal Prediction in Reinforcement Learning
This chapter examines a recent paper on Invariant Causal Prediction tailored for Block MDPs, emphasizing its implications for causal inference within reinforcement learning. It highlights the challenges of spurious correlations and the importance of designing training environments that enhance out-of-distribution generalization.
Transcript
Play full episode
