2021 (modified: 06 Jun 2022)ICML 2021Readers: Everyone
Abstract:Learning composable policies for environments with complex rules and tasks is a challenging problem. We introduce a hierarchical reinforcement learning framework called the Logical Options Framewor...