How to Scale Model-Based Learning to Solve Complex Tasks

This paper is example driven model-based RL for solving long horizon visual motor tasks and from either stand basically as you said the complexity of Robonat it's even hard to scale this model-basedRL to really complicated stuff that involves especially multiple segments. And then moving on to one last paper which was a bit of a follow up to this high-profile variational art encoder or led by the same guy Bohan who is in my lab as well so I've chatted with him about this. Yeah we definitely could have recollected more data in the Packard building but we were a little bit short on time and how much he was leading the paper deck. He didn't have