CASE: Conflict-assessed Knowledge-sensitive Neuron Tuning for Lifelong Model Editing
Open Access
- 12 April 2026
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 3984-3995
Abstract
Large Language Models (LLMs) inevitably encounter factual hallucinations and knowledge obsolescence, necessitating lifelong knowledge editing to sustain reliability and factual advancement. While mainstream lifelong editing paradigms aim to alleviate knowledge forgetting through allocating and updating isolated parameter subspaces, they often overlook conflict assessment among distinct editing processes, leading to unjustified subspace allocation and indiscriminate neuron tuning. To address these issues, we propose the Conflict-Assessed Sensitive Editing (CASE) framework, which integrates a Conflict-Assessed Editing Allocation (CAA) module and a Knowledge-sensitive Neuron Tuning (KNT) strategy. The CAA module quantitatively assesses editing conflicts to enable justified subspace allocation, thereby reducing globally significant conflicts and routing errors. The KNT strategy adaptively identifies and tunes knowledge-sensitive neurons through a calibrated sensitivity threshold, effectively eliminating local conflicts and enhancing subspace stability. Extensive experiments on standard lifelong editing benchmarks demonstrate that CASE achieves state-of-the-art performance, improving average editing accuracy by nearly 10% after 1,000 sequential edits. Overall, CASE substantially mitigates editing conflicts and enhances knowledge retention, offering a scalable and conflict-resilient solution for lifelong model editing.Keywords
Funding Information
- National Key Research and Development Program of China (2023YFC2506800)
This publication has 7 references indexed in Scilit:
- MELO: Enhancing Model Editing with Neuron-Indexed Dynamic LoRAProceedings of the AAAI Conference on Artificial Intelligence, 2024
- WISE: Rethinking the Knowledge Memory for Lifelong Model Editing of Large Language Models2024
- Aging with GRACE: Lifelong Model Editing with Discrete Key-Value Adaptors2023
- Merging Models with Fisher-Weighted Averaging2022
- On Continual Model Refinement in Out-of-Distribution Data Streams2022
- Overcoming catastrophic forgetting in neural networksProceedings of the National Academy of Sciences, 2017
- Zero-Shot Relation Extraction via Reading Comprehension2017