cs.AINov 11, 2026

MuEdit: An efficient multi-task editing method towards inter-domain knowledge conflicts

Authors: Songshi Liang, Ting-En Lin, Zixuan Huang, Yuchuan Wu, Yongbin Li, Zihe Wang, Rui Yan

Organizations: Gaoling School of Artificial Intelligence, Renmin University of China · Alibaba Group · Beihang University · School of Artificial Intelligence, Wuhan University · 

Abstract

Large language models (LLMs) require frequent knowledge updates across heterogeneous domains, but existing editing methods fail when applied sequentially to multi-task settings. We trace this collapse to inter-domain null-space misalignment: heterogeneous tasks induce distinct Jacobian geometries, causing their common null space to shrink dramatically and leaving insufficient admissible directions for conflict-free updates. We propose the Conflict Index to quantify this geometric interference, and introduce Mu-Edit, which mitigates multi-task conflicts via (1) conflict-aware ordering to minimize cumulative interference, and (2) dynamic low-rank approximation to expand the null space when ordering alone is insufficient. Experiments across five functional domains and three backbones show that Mu-Edit significantly outperforms existing baselines in multi-task editing performance while preserving general capabilities, and remains effective under incremental task arrival.

Figures & tables

Explore similar work

CardsList
  1. ManiEdit: Sequential Unstructured Knowledge Editing for Language Models from a Manifold Perspective

    Sep 27, 2026Rui Liu, Chenheng Zhang, Haoxuan Li +1Knowledge EditingContinual Learning for LLMs

  2. Beyond Binary Edits Robust Multimodal Knowledge Editing with Adversarial Subspace Alignment

    May 22, 2026Haoyuan Wang, Xiaohao Liu, Jiajie Su +2Multimodal RobustnessAdversarial Robustness

  3. Capacity Allocation at the Source: Sparse Target Optimization for LLM Knowledge Editing

    Nov 8, 2026Dongyao Chen, Jiayun Lei, Zhiyuan Ji +3Knowledge EditingLanguage Model Editing