{"type":"ScholarlyArticle","title":"LLMs as Parametric Knowledge Sources for Knowledge Graph Completion","authors":[{"name":"Deyu Chen","slug":"deyu-chen","given":"Deyu","family":"Chen","affiliation":""},{"name":"Qiyuan Li","slug":"qiyuan-li","given":"Qiyuan","family":"Li","affiliation":""},{"name":"Jinguang Gu","slug":"jinguang-gu","given":"Jinguang","family":"Gu","affiliation":""},{"name":"Meiyi Xie","slug":"meiyi-xie","given":"Meiyi","family":"Xie","affiliation":""},{"name":"Hong Zhu","slug":"hong-zhu","given":"Hong","family":"Zhu","affiliation":""}],"abstract":"Knowledge graphs (KGs) serve as crucial symbolic knowledge sources for many downstream applications but often suffer from incompleteness. Recent efforts have attempted to integrate pre-trained language models and KGs for the knowledge graph completion (KGC) task. However, existing methods introduce substantial model complexity or rely on prompt-based solutions that inject priors at a shallow level, which do not exploit the factual knowledge inherently encoded in model parameters. To address these issues, we propose a novel framework that treats LLMs as parametric knowledge sources for KGC. Our framework performs LLM knowledge elicitation to extract factual knowledge from the model’s internal representations. Then, a cross-granularity representation alignment method transforms sentence-level representations into entity-level representations and aligns them within a unified space. Finally, a dynamic learning schedule balances alignment and expressiveness throughout training. Extensive experiments on multiple benchmark datasets show that our proposed method can be integrated with diverse KGC baselines and consistently improves link prediction in both standard and lifelong settings.","doi":"10.24963/ijcai.2026/421","url":"https://ijcai.org/proceedings/2026/421","pdf":"https://ijcai.org/proceedings/2026/0421.pdf","markdown":"https://ijcai.org/proceedings/2026/421.md","bibtex":"https://ijcai.org/proceedings/2026/bibtex/421","volume":{"slug":"2026","title":"Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence","short_title":"IJCAI 2026","url":"https://ijcai.org/proceedings/2026/","year":2026,"location":"Bremen, Germany","dates":"15-21 August 2026","editors":"Diego Calvanese","isbn":"978-1-956792-09-6","publisher":"International Joint Conferences on Artificial Intelligence Organization","doi_prefix":"10.24963/ijcai.2026"},"section":"Knowledge Representation and Reasoning","track":"Main Track","number":421,"pages":{"first":3783,"last":3791},"keywords":["Knowledge Representation and Reasoning: Learning and reasoning","Natural Language Processing: Applications","Knowledge Representation and Reasoning: Semantic Web"],"videos":[],"references":[],"license_note":"Copyright International Joint Conferences on Artificial Intelligence."}