Human behavior displays hierarchical structure: simple actions cohere into subtask sequences, which work together to accomplish overall task goals. Although the neural substrates of such hierarchy have been the target of increasing research, they remain poorly understood. We propose that the computations supporting hierarchical behavior may relate to those in hierarchical reinforcement learning (HRL), a machine-learning framework that extends reinforcement-learning mechanisms into hierarchical domains. To test this, we leveraged a distinctive prediction arising from HRL. In ordinary reinforcement learning, reward prediction errors are computed when there is an unanticipated change in the prospects for accomplishing overall task goals. HRL entails that prediction errors should also occur in relation to task subgoals. In three neuroimaging studies we observed neural responses consistent with such subgoal-related reward prediction errors, within structures previously implicated in reinforcement learning. The results reported support the relevance of HRL to the neural processes underlying hierarchical behavior.
Starting in 2022, the lab has decided to post here only the archival, open-access version of our publications. This is part of the movement to emphasize quality and content over the impact factor or prestige of the journal a paper is published in. Full citations (for referencing papers in your own work) can be found on pubmed and/or within the archival version, which will be updated once a paper is accepted for publication after peer review.