Learning to Reason before you speak is how @OpenAI o1 generates its response, but how could this work? The Quiet-STaR paper explores if Language Models can teach themselves to think before speaking. 👀 Quiet-STaR teaches LMs to generate thoughts for each token to improve their… https://t.co/cgqoNYOr5t https://t.co/X4nRuCtsLV
— Philipp Schmid (@_philschmid) Sep 14, 2024
from Twitter https://twitter.com/_philschmid
September 14, 2024 at 07:57AM
via IFTTT
No comments:
Post a Comment