education

I made a small game where RL algorithms break

Breaking Bellman is a free browser game for learning reinforcement learning. Real algorithms train in your browser and fail, and you fix them one word at a time.