WilKE: Wise-Layer Knowledge Editor for Lifelong Knowledge Editing

arXiv:2402.10987v1 Announce Type: new
Abstract: Knowledge editing aims to rectify inaccuracies in large language models (LLMs) without costly retraining for outdated or erroneous knowledge. However, current knowledge editing methods primarily focus on single editing, failing to meet the requirements for lifelong editing. In this paper, lifelong editing is synonymous with lifelong knowledge editing. This study reveals a performance degradation encountered by knowledge editing in lifelong editing, characterized by toxicity buildup and toxicity flash, with the primary cause identified as pattern …

abstract arxiv cs.ai cs.cl current editing editor focus knowledge language language models large language large language models layer llms paper requirements retraining type wise