Exploration in Meta Learning

In your paper exploration is clearly a central issue especially for few shot meta RL. Can you talk about the importance of exploration in meta RL and the main methods you use to explore? So Exploration is kind of a central concept that makes meta RL distinct from just meta learning in general. You have to explore to figure out what task you're in and what data you need to identify the task itself. And that's kind of one of the central challenges in the few shot Meta RL setting.