no code implementations • 5 Oct 2021 • Haoya Li, Samarth Gupta, HsiangFu Yu, Lexing Ying, Inderjit Dhillon
This paper proposes an approximate Newton method for the policy gradient algorithm with entropy regularization.