Offline Multi-Task Reinforcement Learning

Shared Dynamics Model

Single transition model learned from multi-task batch data capturing common and specific dynamics of environments.

← Back