用强化学习让大模型自主查知识图谱，又快又准还通用。
Efficient and Transferable Agentic Knowledge Graph RAG via Reinforcement Learning
- 单个智能体通过强化学习边查图边推理，不再依赖多模块流水线。
- 用30亿参数模型在问答任务上比更大模型更省生成token，准确率更高。
- 训练后可直接用在新知识图谱上，无需重新训练，适合实际部署。
知识图谱增强生成（KG-RAG）将大语言模型与结构化、可验证的知识图谱结合，以减少幻觉并提供推理路径。然而，现有系统通常依赖多个大模型模块组成的固定流程（如规划、推理、响应），导致推理成本高且性能受限于特定图谱模式。为此，我们提出KG-R1，一种基于强化学习的代理框架，优化KG-RAG。与模块化工作流不同，KG-R1采用单一智能体作为环境交互者，每一步学习检索信息，并在统一过程中融合到推理与生成。在知识图谱问答（KGQA）基准测试中，使用Qwen 2.5-3B模型，KG-R1在生成更少token的情况下，比采用更大基础或微调模型的多模块方法取得更高准确率。此外，KG-R1具备强即插即用能力：训练后可在未见知识图谱上保持准确率，无需再训练。这些特性使其成为真实场景部署的有力候选。代码已公开于github.com/junhongmit/KG-R1/。
原文摘要 · Abstract (English)
Knowledge-graph retrieval-augmented generation (KG-RAG) couples large language models (LLMs) with structured, verifiable knowledge graphs (KGs) to reduce hallucination and provide reasoning traces. However, current KG-RAG systems often rely on fixed pipelines of multiple LLM modules (e.g., planning, reasoning, and responding), which inflate inference costs and tie performance to specific graph schemas. To address this, we introduce KG-R1, an agentic framework that optimizes KG-RAG through reinforcement learning (RL). Unlike modular workflows, KG-R1 uses a single agent that interacts with KGs as its environment, learning to retrieve information at each step and incorporating it into its reasoning and generation in a unified process. Across Knowledge-Graph Question Answering (KGQA) benchmarks, KG-R1 demonstrates both efficiency and transferability-using Qwen 2.5-3B, KG-R1 improves answer accuracy with fewer generation tokens than prior multi-module workflow methods that use much larger foundation or fine-tuned models. Furthermore, KG-R1 exhibits strong plug-and-play capability: after training, maintaining accuracy on unseen KGs without retraining. These properties make KG-R1 a promising KG-RAG framework for real-world deployment. Our code is publicly available at github.com/junhongmit/KG-R1/.
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。