Decoupling Value and Policy for Generalization in Reinforcement Learning.

Roberta Raileanu, Rob Fergus

Browse the full ICML paper archive.