Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
The paper introduces C2RAG, a framework for robust multi-hop knowledge graph retrieval-augmented generation that addresses retrieval drift and hallucinations caused by imperfect knowledge graph quality. C2RAG employs constraint-based retrieval using atomic constraint triples and fine-grained constraint anchoring to filter candidates, mitigating retrieval drift. It also incorporates a sufficiency check to prevent retrieval hallucinations by determining if the current evidence is sufficient for structural propagation, activating textual recovery when needed. Experiments on multi-hop benchmarks demonstrate that C2RAG outperforms existing baselines, achieving a 3.4% EM and 3.9% F1 improvement while showing improved robustness to KG quality issues.
Imperfect knowledge graphs can lead to retrieval drift and hallucinations in multi-hop reasoning, but C2RAG offers a robust solution that improves EM by 3.4% and F1 by 3.9% over existing methods.
Graph Retrieval-Augmented Generation enhances multi-hop reasoning but relies on imperfect knowledge graphs that frequently suffer from inherent quality issues. Current approaches often overlook these issues, consequently struggling with retrieval drift driven by spurious noise and retrieval hallucinations stemming from incomplete information. To address these challenges, we propose C2RAG (Constraint-Checked Retrieval-Augmented Generation), a framework aimed at robust multi-hop retrieval over the imperfect KG. First, C2RAG performs constraint-based retrieval by decomposing each query into atomic constraint triples, with using fine-grained constraint anchoring to filter candidates for suppressing retrieval drift. Second, C2RAG introduces a sufficiency check to explicitly prevent retrieval hallucinations by deciding whether the current evidence is sufficient to justify structural propagation, and activating textual recovery otherwise. Extensive experiments on multi-hop benchmarks demonstrate that C2RAG consistently outperforms the latest baselines by 3.4\% EM and 3.9\% F1 on average, while exhibiting improved robustness under KG issues.