Paper

Reward-Estimated Hypergradient for Bilevel Reinforcement Learning with Black-Box Follower · ByteTrail