Julian Togelius
TalkRL: The Reinforcement Learning Podcast
00:00
The Importance of Video Pre-Training in Deep RL
Deep RL hasn't really solved the exploration piece so much as the exploitation's aside. We might very well have this thing where basically we have generic model for playing a lot of 3D games and you can't fine tune into your particular games. This is something we at model would love to be able to do, but we're not there yet.
Transcript
Play full episode
