Knowledge editing aims to efficiently correct factual errors in language models. Widely used locate-then-edit methods update an MLP layer by adjusting its weights to change the mapping between the layer's input vector (key) and output vector (value), thereby editing the model's knowledge. As this update is driven by key and value vectors, obtaining these vectors without careful constraints causes significant model perturbations beyond the targeted edit, a common issue in many prior knowledge editing methods. To address this, we propose Subspace Knowledge Edit (SUIT), which computes key and value vectors only within the subspace of critical features relevant to the edit. Our empirical results on LLaMA3, GPT-J, and Qwen2.5 models show that SUIT dramatically improves knowledge preservation over strong baselines while maintaining high editing performance. These results support the claim that SUIT successfully identifies the critical subspace for the edit. Beyond quantitative gains, our analyses show that SUIT reduces unintended perturbations in hidden states while confining updates to directions that are more effective for editing. Taken together, these findings establish edit-critical subspace identification as a key principle for reliable, low-perturbation knowledge editing. Our code is available at https://github.com/holi-lab/SUIT.

翻译：知识编辑旨在高效修正语言模型中的事实性错误。广泛使用的"定位-编辑"方法通过调整MLP层权重来改变该层输入向量（键）与输出向量（值）之间的映射关系，从而实现知识编辑。由于这种更新由键向量和值向量驱动，在缺乏严格约束的情况下获取这些向量会导致模型产生超出目标编辑范围的显著扰动，这是许多现有知识编辑方法的常见问题。为解决这一问题，我们提出子空间知识编辑方法（SUIT），该方法仅在编辑相关的关键特征子空间内计算键向量和值向量。我们在LLaMA3、GPT-J和Qwen2.5模型上的实验结果表明，SUIT在保持高水平编辑性能的同时，相比现有强基线方法显著提升了知识保留能力。这些结果证实了SUIT能够成功识别编辑所需的关键子空间。除量化指标外，我们的分析表明SUIT能减少隐藏状态中的非预期扰动，同时将更新限制在更有效的编辑方向上。综合来看，这些发现确立了编辑关键子空间识别作为实现可靠、低扰动知识编辑的核心原则。代码已开源：https://github.com/holi-lab/SUIT。