The goal of this project was to investigate meta-learning in reinforcement learning under highly stochastic environments using mathematically tractable and autonomous Braitenberg vehicles. In ...