RB
Replay buffers in reinforcement learning
Topic
What experts have said about Replay buffers in reinforcement learning
1 statement · 1 positive
Replay buffers should contain on-policy states plus off-policy recovery states.
“your replay buffer really should have the states that your policy would visit, plus some distribution of states that you might drift to and then how to return back to your optimal states”
Open the episode · Eric Jang – Building AlphaGo from scratchListen at 2:04:31
Statements are attributed to the speaker as said on the episode and reflect their view at the time, not PodLume's. They are not advice.