Code repo for Neurips paper Improving Model-Based Reinforcement Learning by Converging to Flatter Minimas - View it on GitHub