通过过程监督优化决策与执行，提升智能体RAG的复杂任务处理能力。
DecEx-RAG: Boosting Agentic Retrieval-Augmented Generation with Decision and Execution Optimization via Process Supervision
- 将RAG建模为包含决策与执行的马尔可夫决策过程
- 在6个数据集上平均性能提升6.2%，显著优于基线
- 提出高效剪枝策略，数据构建效率提升近6倍，适合高阶RAG研究者
智能体检索增强生成（Agentic RAG）通过动态检索与自适应工作流提升复杂任务处理能力。近期进展（如Search-R1）表明，基于结果监督的强化学习表现优异，但仍存在探索效率低、奖励信号稀疏及全局反馈模糊等问题。为此，我们提出DecEx-RAG，将RAG建模为包含决策与执行的马尔可夫决策过程（MDP），并引入高效剪枝策略以优化数据扩展。通过全面的过程级策略优化，DecEx-RAG显著提升了大语言模型（LLMs）在自主任务分解、动态检索和高质量答案生成方面的能力。实验显示，DecEx-RAG在六个数据集上平均绝对性能提升6.2%，显著优于现有基线。此外，剪枝策略使数据构建效率提升近6倍，为过程监督型RAG训练提供了高效解决方案。代码已开源：https://github.com/sdsxdxl/DecEx-RAG。
原文摘要 · Abstract (English)
Agentic Retrieval-Augmented Generation (Agentic RAG) enhances the processing capability for complex tasks through dynamic retrieval and adaptive workflows. Recent advances (e.g., Search-R1) have shown that outcome-supervised reinforcement learning demonstrate strong performance. However, this approach still suffers from inefficient exploration, sparse reward signals, and ambiguous global reward feedback. To address these challenges, we propose DecEx-RAG, which models RAG as a Markov Decision Process (MDP) incorporating decision-making and execution, while introducing an efficient pruning strategy to optimize data expansion. Through comprehensive process-level policy optimization, DecEx-RAG significantly enhances the autonomous task decomposition, dynamic retrieval, and high-quality answer generation capabilities of large language models (LLMs). Experiments show that DecEx-RAG achieves an average absolute performance improvement of $6.2\%$ across six datasets, significantly outperforming existing baselines. Moreover, the pruning strategy improves data construction efficiency by nearly $6 \times$, providing an efficient solution for process-supervised RAG training. The code is available at https://github.com/sdsxdxl/DecEx-RAG.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。