On Credit Assignment in Hierarchical Reinforcement Learning
This work provides fundamental insights into hierarchical reinforcement learning, potentially advancing the field by distinguishing hierarchical backups as a basis for research, though it appears incremental in nature.
The paper tackled the challenge of hierarchical credit assignment in reinforcement learning by analyzing hierarchical backups as conventional multistep backups with skip connections, and developed a new algorithm, Hier$Q_k(λ)$, showing that hierarchical credit assignment alone can boost agent performance.
Hierarchical Reinforcement Learning (HRL) has held longstanding promise to advance reinforcement learning. Yet, it has remained a considerable challenge to develop practical algorithms that exhibit some of these promises. To improve our fundamental understanding of HRL, we investigate hierarchical credit assignment from the perspective of conventional multistep reinforcement learning. We show how e.g., a 1-step `hierarchical backup' can be seen as a conventional multistep backup with $n$ skip connections over time connecting each subsequent state to the first independent of actions inbetween. Furthermore, we find that generalizing hierarchy to multistep return estimation methods requires us to consider how to partition the environment trace, in order to construct backup paths. We leverage these insight to develop a new hierarchical algorithm Hier$Q_k(λ)$, for which we demonstrate that hierarchical credit assignment alone can already boost agent performance (i.e., when eliminating generalization or exploration). Altogether, our work yields fundamental insight into the nature of hierarchical backups and distinguishes this as an additional basis for reinforcement learning research.