TLDR
大语言模型 (LLMs) 日益作为事实知识的通用接口，但其参数并不会自动反映预训练后更新的信息。知识编辑 (KE) 提供了一种针对高成本重训练的替代方案，通过修改特定知识并保留无关知识与通用能力。传统 KE 使用结构化事实三元组，而非结构化 KE (UKE) 使用包含多个事实的自由文本段落。然而，现有 UKE 编辑器表现出一种称为 context reliance 的失败模式：被编辑的 LLM 经常Large language models (LLMs) increasingly serve as general-purpose interfaces to factual knowledge, but their parameters do not automatically reflect information that changes after pretraining. Knowledge editing (KE) provides a targeted alternative to costly retraining by modifying selected knowledge and preserving unrelated knowledge and general capabilities. Conventional KE uses structured factual triples, whereas unstructured KE (UKE) uses free-form passages containing multiple facts. Nonetheless, existing UKE editors exhibit a failure mode known as context reliance: edited LLMs can often r