Diffusion Subgoal Planning for Long-Horizon Offline Goal-Conditioned Reinforcement Learning
Offline goal-conditioned reinforcement learning (GCRL) learns goal-directed policies from reward-free data, but in long-horizon tasks, goal-conditioned value functions often provide unstable guidance due to sparse rewards and discounting. Hierarchical methods partially mitigate this issue via subgoal decomposition; how...
Heng-Rui Zhang, Yu-Hu Cheng, C. L. Philip Chen et al.
· 1 citation