As LLMs advance, post-training reinforcement learning (RL) increasingly relies on multi-dimensional rewards to cultivate comprehensive capabilities. This shift demands new algorithms capable of optimizing diverse and potentially competing objectives simultaneously. To address this, existing methods such as Group reward-Decoupled Policy Optimization (GDPO) decompose the overall score into independent reward groups, then compute the RL loss separately within each group. However, this strategy still encounters multi-reward conflicts: a single rollout can yield positive advantages on certain reward dimensions but negative ones on others, causing opposing signals to cancel each other out during aggregation, further hindering RL training efficiency. Inspired by Dynamic sAmpling Policy Optimization (DAPO), which improves RL training efficiency by filtering out ineffective rollouts with near-zero advantages, we propose Group-Dynamic reward-Decoupled Policy Optimization (GD$^2$PO). Specifically, GD$^2$PO employs a conflict-aware filtering mechanism to mask out rollouts suffering from severe reward-wise disagreement. By preventing conflicting signals from canceling each other out, this masking strategy preserves and enhances the magnitude of effective RL advantages, thereby significantly accelerating learning efficiency. Furthermore, we introduce query-level reweighting to dynamically adjust the update intensity of each query based on its overall reward consensus. Experiments on various multi-reward scenarios, including tool calling and human preference alignment, demonstrate that GD$^2$PO consistently and significantly outperforms existing baselines. The code is available at https://github.com/Qwen-Applications/GD2PO.

翻译：随着大语言模型（LLM）的进步，后训练强化学习（RL）日益依赖多维奖励以培养综合能力。这一转变要求新算法能够同时优化多样化且可能相互竞争的目标。为此，现有方法如群组奖励解耦策略优化（GDPO）将总体得分分解为独立的奖励群组，随后在各群组内分别计算RL损失。然而，该策略仍面临多奖励冲突：单次 rollout 可在某些奖励维度上产生正向优势，却在其他维度上产生负向优势，导致聚合过程中对立信号相互抵消，进一步阻碍RL训练效率。受动态采样策略优化（DAPO）启发——该方法通过过滤近零优势的无效 rollout 提升RL训练效率——我们提出群组动态奖励解耦策略优化（GD$^2$PO）。具体而言，GD$^2$PO采用冲突感知过滤机制，掩码处理遭受严重奖励维度分歧的 rollout。通过防止冲突信号相互抵消，该掩码策略保留并增强了有效RL优势的幅度，从而显著加速学习效率。此外，我们引入查询级重加权机制，依据每个查询的整体奖励共识动态调整其更新强度。在工具调用与人类偏好对齐等多种多奖励场景下的实验表明，GD$^2$PO始终显著优于现有基线方法。代码已开源在https://github.com/Qwen-Applications/GD2PO。