Max Schwarzer
TalkRL: The Reinforcement Learning Podcast
BBF: A Better Learning Algorithm Than Rainbow
I absolutely believe that some RL squared or meta RL algorithm could come up with a dramatically more efficient learning algorithm than BBF. The other thing is BBF is really not designed for ultra-hard exploration problems. I would think of BBF as essentially a bigger, better, and faster version of Rainbow. But that means probably no continuous control unless you figure out action space to screen discretization for your problem.
00:00
Transcript
Play full episode
Remember Everything You Learn from Podcasts
Save insights instantly, chat with episodes, and build lasting knowledge - all powered by AI.
