mapo-mixed-advantage-policy-optimization

【效率与协作】Dynamically reweight advantage functions based on trajectory certainty to improve policy optimization in foundation models. Addresses advantage reversion and mirror problems by mixing standardized and mean-normalized advantage formulations. Enables more stable gradient signals across high- and low-certainty samples.

独立 GitHub 快照 · 自动分类 · 以原作者文档为准