Continual Reinforcement Learning with Multi-Timescale Successor FeaturesCOSYNE 2022 · McGill UniversitySaveMar 19, 2022Neuroscience