Active Hierarchical Exploration with Stable Subgoal Representation Learning
This work addresses a bottleneck in hierarchical reinforcement learning for sparse-reward environments, offering an incremental improvement over existing GCHRL methods.
The paper tackles the challenge of goal-conditioned hierarchical reinforcement learning (GCHRL) in temporally extended tasks with sparse rewards, where dynamic subgoal representation changes hinder high-level exploration, and proposes a regularization method for stable subgoal representation learning and an active hierarchical exploration strategy, resulting in significant outperformance over state-of-the-art baselines in continuous control tasks.
Goal-conditioned hierarchical reinforcement learning (GCHRL) provides a promising approach to solving long-horizon tasks. Recently, its success has been extended to more general settings by concurrently learning hierarchical policies and subgoal representations. Although GCHRL possesses superior exploration ability by decomposing tasks via subgoals, existing GCHRL methods struggle in temporally extended tasks with sparse external rewards, since the high-level policy learning relies on external rewards. As the high-level policy selects subgoals in an online learned representation space, the dynamic change of the subgoal space severely hinders effective high-level exploration. In this paper, we propose a novel regularization that contributes to both stable and efficient subgoal representation learning. Building upon the stable representation, we design measures of novelty and potential for subgoals, and develop an active hierarchical exploration strategy that seeks out new promising subgoals and states without intrinsic rewards. Experimental results show that our approach significantly outperforms state-of-the-art baselines in continuous control tasks with sparse rewards.