Weakly-Supervised Reinforcement Learning for Controllable Behavior
Lisa Lee, Benjamin Eysenbach, Russ Salakhutdinov, Shixiang (Shane) Gu, Chelsea Finn
Poster Session 7 (more posters)
on 2020-12-10T21:00:00-08:00 - 2020-12-10T23:00:00-08:00
GatherTown: Deep Learning ( Town D0 - Spot A0 )
on 2020-12-10T21:00:00-08:00 - 2020-12-10T23:00:00-08:00
GatherTown: Deep Learning ( Town D0 - Spot A0 )
Join GatherTown
Only iff poster is crowded, join Zoom . Authors have to start the Zoom call from their Profile page / Presentation History.
Only iff poster is crowded, join Zoom . Authors have to start the Zoom call from their Profile page / Presentation History.
Toggle Abstract Paper (in Proceedings / .pdf)
Abstract: Reinforcement learning (RL) is a powerful framework for learning to take actions to solve tasks. However, in many settings, an agent must winnow down the inconceivably large space of all possible tasks to the single task that it is currently being asked to solve. Can we instead constrain the space of tasks to those that are semantically meaningful? In this work, we introduce a framework for using weak supervision to automatically disentangle this semantically meaningful subspace of tasks from the enormous space of nonsensical "chaff" tasks. We show that this learned subspace enables efficient exploration and provides a representation that captures distance between states. On a variety of challenging, vision-based continuous control problems, our approach leads to substantial performance gains, particularly as the complexity of the environment grows.