Noise or Nuance: An Investigation Into Useful Information and Filtering For LLM Driven AKBC
This work addresses the problem of enhancing LLM outputs for knowledge base completion in constrained settings, though it appears incremental as it builds on existing triple completion tasks.
The paper investigated how to improve LLM performance on knowledge base completion tasks under constrained conditions where standard techniques like RAG and fine-tuning are restricted, finding that additional information boosts generation quality and LLMs can effectively filter poor-quality triples.
RAG and fine-tuning are prevalent strategies for improving the quality of LLM outputs. However, in constrained situations, such as that of the 2025 LM-KBC challenge, such techniques are restricted. In this work we investigate three facets of the triple completion task: generation, quality assurance, and LLM response parsing. Our work finds that in this constrained setting: additional information improves generation quality, LLMs can be effective at filtering poor quality triples, and the tradeoff between flexibility and consistency with LLM response parsing is setting dependent.