Emergent transition from code to natural language for reasoning tasks when RL tuning a language model for math. Interesting to consider implications for "Language of Thought" style theories in cognition.

https://hkust-nlp.notion.site/simplerl-reason
Post image

Comments

Posting Rules

  • Be respectful to others
  • No spam or self-promotion
  • Stay on topic
  • Follow Bluesky's terms of service