blockwise-advantage-estimation

【效率与协作】Improve credit assignment in multi-objective RL by decomposing advantages into segment-specific values. Use Outcome-Conditioned Baselines to reduce cross-objective interference without expensive rollouts, enabling better training signals for multi-step completions with different reward functions per segment.

独立 GitHub 快照 · 自动分类 · 以原作者文档为准