Fetching the paper…

Transformers Can Learn Temporal Difference Methods for In-Context Reinforcement Learning · Around