ReMA: Learning to Meta-Think for LLMs with Multi-agent Reinforcement Learning

Devs

ReMA: Learning to Meta-Think for LLMs with Multi-agent Reinforcement Learning | Read Paper on Bytez