Neuron-Level Sequential Editing for Large Language Models

TL;DR

A new model editing method, namelyeuron-levelentialential NSE (NSE), tailored for supporting sequential model editing in large language models, which optimize the target layer's hidden states using the model's original weights to prevent model failure.

Abstract

This work explores sequential model editing in large language models (LLMs), a critical task that involves modifying internal knowledge within LLMs continuously through multi-round editing, each incorporating updates or corrections to adjust the model outputs without the need for costly retraining. Existing model editing methods, especially those that alter model parameters, typically focus on single-round editing and often face significant challenges in sequential model editing-most notably issues of model forgetting and failure. To address these challenges, we introduce a new model editing method, namely \textbf{N}euron-level \textbf{S}equential \textbf{E}diting (NSE), tailored for supporting sequential model editing. Specifically, we optimize the target layer's hidden states using the model's original weights to prevent model failure. Furthermore, we iteratively select neurons in multiple layers for editing based on their activation values to mitigate model forgetting. Our empirical experiments demonstrate that NSE significantly outperforms current modifying parameters model editing methods, marking a substantial advancement in the field of sequential model editing. Our code is released on \url{https://github.com/jianghoucheng/NSE}.

Figures (8)

• Figure 1: Example of sequential editing. (a) shows model forgetting and model failure issues in sequential editing using ROME/MEMIT, while (b) shows the accurate editing capabilities of our method without such issues.
• Figure 2: Overview of sequential model editing with NSE. (a) describes the process of weights rewinding for value computation. (b) illustrates the neuron selection and neuron-level weights update. (c) shows the process of iterative multi-layer editing.
• Figure 3: Editing performance of NSE and baselines with varying numbers of edits (batch size 100) in sequential editing, evaluated on the Counterfact dataset. Score is the harmonic mean of Efficacy, Generalization, and Specificity.
• Figure 4: Editing performance of NSE and MEMIT with different batch size, evaluated on Llama3 (8B). The red line and the blue line represent MEMIT and NSE, respectively.
• Figure 5: Performance on general tasks of edited models using NSE, ROME and MEMIT, with sequential editting on Llama3 (8B).