Mechanistic Circuit-Based Knowledge Editing in Large Language Models
This addresses the challenge of knowledge editing in dynamic environments for LLM users, representing a novel method for a known bottleneck.
The paper tackles the problem of updating knowledge in Large Language Models for multi-step reasoning by introducing MCircKE, a framework that identifies and edits causal circuits, achieving effective performance on the MQuAKE-3K benchmark.
Deploying Large Language Models (LLMs) in real-world dynamic environments raises the challenge of updating their pre-trained knowledge. While existing knowledge editing methods can reliably patch isolated facts, they frequently suffer from a "Reasoning Gap", where the model recalls the edited fact but fails to utilize it in multi-step reasoning chains. To bridge this gap, we introduce MCircKE (\underline{M}echanistic \underline{Circ}uit-based \underline{K}nowledge \underline{E}diting), a novel framework that enables a precise "map-and-adapt" editing procedure. MCircKE first identifies the causal circuits responsible for a specific reasoning task, capturing both the storage of the fact and the routing of its logical consequences. It then surgically update parameters exclusively within this mapped circuit. Extensive experiments on the MQuAKE-3K benchmark demonstrate the effectiveness of the proposed method for multi-hop reasoning in knowledge editing.