Abstract: Applications of reinforcement learning (RL) in real-world scenarios are often limited by its generalizability across multiple different environments. Contextual RL offers a principled ...