factual
The AlphaGo playout averaging trick was unnecessary
AlphaGo Lee averaged the value network estimate with a real policy-vs-policy playout to the end (α·Vθ + (1−α)·rollout) to ground value estimates, but all subsequent papers removed this; Jang's implementation also dropped it, which speeds things up greatly since you don't roll games out on every simulation.
factualpending
Speaker
Eric JangEvidence Quote
“In my implementation, I did the same, and it speeds things up a lot because you don't have to roll these games out on every single simulation.”
Created: 6/13/2026, 3:37:56 AM
Your notes
Loading notes...