Review for NeurIPS paper: Steady State Analysis of Episodic Reinforcement Learning