1 paper
Michal Korniak, Kamil Dybek, Benjamin Eysenbach +2
While self-supervised approaches to reinforcement learning have achieved strong results by learning representations of states and actions, a key open question is the time scale ove…