Modularity in Reinforcement Learning via Algorithmic Independence in Credit Assignment.

Michael Chang, Sidhant Kaushik, Sergey Levine, Tom Griffiths

Browse the full ICML paper archive.