‹ BackHN Continuity

Thread

Fixing GRPO's credit assignment problem without evaluating every step

23 points · 3 comments · mrkn1

  1. vatsachak · · focus · HN ↗
    How do frontier labs choose which ideas to include in a training run? It seems that there are too many to choose from
    1. simianwords · · focus · HN ↗
      Why do you think they are buying compute and accelerating to RSI? It’s for this.
Open on Hacker News to reply ↗

Unofficial Hacker News client; not affiliated with Y Combinator.