Training on wrong answers outpaces training on correct ones. 10 times more learning emerges from plausible errors than from truths. Large language models refine their accuracy slowly when they learn only from correct examples. This paper introduces Likra, which trains one https://t.co/iQQ0Q8RQpL
— Rohan Paul (@rohanpaul_ai) Jun 9, 2025
from Twitter https://twitter.com/rohanpaul_ai
June 09, 2025 at 09:39AM
via IFTTT
No comments:
Post a Comment