提出首个针对大模型知识图谱推理幻觉的检测方法，提升推理可靠性。
Detecting Hallucinations for Large Language Model-based Knowledge Graph Reasoning
- 融合注意力分数、语义相似性与图结构信息进行幻觉检测
- 在9个数据集上超越15种基线方法，准确率显著提升
- 适合关注大模型推理安全性的研究人员和应用开发者
知识图谱推理从现有事实中推断新知识，广泛应用于问答、推荐和决策支持。随着大语言模型（LLM）的快速发展，基于检索知识图谱信息的LLM推理框架日益流行。然而，模型幻觉仍是关键问题：即使引入相关知识图谱信息，仍可能生成错误输出，导致误导性结论。现有检测方法或关注模型内部状态，或验证与检索上下文的一致性，但均忽视了知识图谱的结构信息，导致性能受限。为此，我们提出LUCID——首个面向基于大模型的知识图谱推理框架的幻觉检测方法。LUCID联合利用大模型注意力分数、知识图谱语义与结构信息：从注意力分数和语义相似度中提取节点与边特征，并通过图神经网络融合图结构。我们还构建了人工标注的基准数据集。在9个数据集上的实验表明，相比15种基线方法，LUCID达到最先进性能。
原文摘要 · Abstract (English)
Knowledge graph (KG) reasoning infers new knowledge from existing facts and is widely applied in question answering, recommendation, and decision support. With the rapid development of large language models (LLMs), LLM-based KG reasoning frameworks have become increasingly popular by leveraging retrieved KG information. However, hallucinations in LLMs remain a critical issue. Even when relevant KG knowledge is incorporated, models may still generate incorrect outputs, leading to misinformation and unreliable decisions. Existing hallucination detection methods either focus on LLM internal states or verify consistency with retrieved contexts, but both overlook the structural information in KGs, resulting in suboptimal performance. To address this gap, we propose LUCID, the first halLUcination deteCtIon method for LLM-based knowleDge graph reasoning frameworks. LUCID jointly leverages LLM attention scores, KG semantics, and structural information. Specifically, it extracts node and edge features from attention scores and semantic similarities, and integrates them with KG structure using a graph neural network. We also construct manually annotated benchmark datasets for evaluation. Experiments on nine datasets show that LUCID achieves state of the art performance compared to 15 baselines.
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。