f-grpo-divergence-alignment

【效率与协作】Unify LLM alignment methods through f-divergence theory. f-GRPO extends GRPO to handle any divergence measure (KL, Jensen-Shannon, Hellinger), enabling tailored alignment objectives. f-HAL combines on-policy and off-policy preference learning to prevent reward hacking while maintaining safety alignment.

独立 GitHub 快照 · 自动分类 · 以原作者文档为准