AI RESEARCH
Smaller Models are Natural Explorers for Policy-Level Diversity in GRPO
arXiv CS.AI
•
ArXi:2605.30789v1 Announce Type: cross We identify a new dimension for enhancing rollout diversity in Group Relative Policy Optimization (GRPO) for LLMs. While GRPO relies on diverse rollouts, prevailing strategies primarily increase diversity by injecting token-level randomness, which may