Transformers as Decision Makers: Provable In-Context Reinforcement Learning via Supervised Pretraining

Open in new window