用合成步骤监督提升大模型的文本图检索能力
Enhancing Agentic Textual Graph Retrieval with Synthetic Stepwise Supervision
- 用合成数据生成每一步的黄金子图，提供密集奖励信号
- 在三个数据集上平均提升15.6%准确率和17.2% F1分数
- 适合需要多跳推理的复杂问答任务研究者
将文本图融入大型语言模型（LLM）有望实现基于图的复杂问答。然而，关键瓶颈在于检索出信息丰富且紧凑的子图以适配LLM上下文。现有检索器常因依赖浅层嵌入相似性或高成本的交互策略而表现不佳，后者需大量人工标注。为此，我们提出一种基于LLM的代理式文本图推理框架，其检索器通过合成步骤监督进行训练。不同于仅依赖最终答案奖励（信号稀疏且不稳定），我们通过对比每一步与离线提取的黄金子图来优化检索器。该方法通过专用数据合成管道提炼黄金子图，构建密集奖励信号，支持两阶段训练，有效学习交互式图探索策略。在三个常见数据集上与七种强基线对比的实验表明，本方法平均准确率提升15.6%，F1得分提升17.2%，在更复杂的多跳推理任务中优势更显著。
原文摘要 · Abstract (English)
Integrating textual graphs into Large Language Models (LLMs) is promising for complex graph-based QA. However, a key bottleneck is retrieving informative yet compact subgraphs that fit the LLM context. Existing retrievers often struggle, relying either on shallow embedding similarity or costly interactive policies that require excessive supervision. To address these challenges, we introduce an agentic textual graph reasoning framework featuring an LLM-based retriever trained with synthetic stepwise supervision. Rather than relying on final answer rewards which often yield sparse and unstable signals, we optimize the retriever by evaluating each step against offline-extracted golden subgraphs. Our approach distills golden subgraphs via a specialized data synthesis pipeline to formulate dense rewards, facilitating a two-stage training scheme that effectively learns the interactive graph exploration policy. Based on extensive experiments on three common datasets in comparison with seven strong baselines, our approach achieves an average improvement of 15.6% in accuracy and 17.2% in F1 score. The advantage is even higher in more complicated multi-hop reasoning tasks.
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。