Understanding when Dynamics-Invariant Data Augmentations Benefit Model-free Reinforcement Learning Updates.

Nicholas Corrado, Josiah P. Hanna

Browse the full ICLR paper archive.