Alex Dimakis
@AlexGDimakis · 19. Juni 2024
This paper seems very interesting: say you train an LLM to play chess using only transcripts of games of players up to 1000 elo. Is it possible that the model plays better than 1000 elo? (i.e. "transcends" the training data performance?). It seems you get something from nothing,
Elon Musk
@elonmusk
Interesting
14:59 · 19. Juni 2024 · 43.632 Aufrufe
56
17
563