Shuyuan Zhang

SCAR: Shapley Credit Assignment for More Efficient RLHF

Learning subgoal representations from state graphs in goal-conditioned hierarchical reinforcement learning