rStar-Math: Small LLMs Can Master Math Reasoning with Self-Evolved Deep Thinking

arxiv.org

36 points

roboboffin

2 days ago


6 comments

s-macke 21 hours ago

> Notably, no self-reflection training data or prompt was included, suggesting that advanced System 2 reasoning can foster intrinsic self-reflection.

They suggest, that self-reflection is an emergent phenomena of reasoning. Impressive. Can't wait to see the code.

helltone 19 hours ago

Off topic but how is MCTS usually implemented efficiently? It has a branching structure that doesn't seem parallelizable (GPU).

throwaway81523 a day ago

Abstract is impressive. I'm surprised this post hasn't gotten more attention.

  • roboboffin a day ago

    Yeah, that's what I thought.

dantodor a day ago

The repo gives 404?

  • funcDropShadow a day ago

    The abstract says the code will be available.