NeurIPS 2020

An Improved Analysis of (Variance-Reduced) Policy Gradient and Natural Policy Gradient Methods

Meta Review

The results in this paper were deemed interesting. The reviewers were left somewhat uncertain, because it wasn't always clear exactly how significant these results will turn out to be (how limiting are the assumptions, how does this pan out in practice), but were still okay with accepting the paper as is, thus allowing the research community to build further to help answer some of these questions in the future.