arXiv cs.AI· Haoben Huang, Shuxin Liu, Ou Wu, Di Gao·· 3 天前AI 评分20
JNO：通过压力感知联合邻域优化重新审视知识编辑中的涟漪效应
Revisiting Ripple Effects in Knowledge Editing through Pressure-Aware Joint Neighborhood Optimization
AI 导读
针对大语言模型单次编辑引发的涟漪效应，研究者提出知识编辑框架 JNO，将可编辑侧协调与保留侧泄漏两种耦合设计压力在目标规划阶段联合建模。JNO 通过压力感知协调（PAC）在耦合约束下联合优化邻域目标表示，并引入语义预执行门在参数执行前拒绝高风险目标方案。在 RippleEdits 上，JNO 将传播与保留指标提升至少 7.0%，同时保持跨骨干编辑稳定性。
正文 · 原文
Abstract:Single-edit updates in large language models can trigger ripple effects across local knowledge neighborhoods: desirable propagation to related facts and unintended perturbation of preserved ones. Existing methods address these two effects separately, without explicitly modeling their coupling. We challenge this separation through an analysis of ripple responses across typical baselines, identifying two coupled design pressures: editable-side coordination and preserved-side leakage. We propose Joint Neighborhood Optimization (JNO), a new knowledge-editing framework to formalize and jointly address both pressures at the target-planning stage. JNO instantiates this principle through Pressure-Aware Coordination (PAC), which jointly optimizes neighborhood target representations under coupled constraints, and a semantic pre-execution gate that rejects high-risk target plans before parameter execution. Experiments on RippleEdits show JNO improves propagation and preservation metrics by at least 7.0% while preserving cross-backbone editing stability.
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2606.01610 [cs.AI] |
| (or arXiv:2606.01610v2 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2606.01610 arXiv-issued DOI via DataCite |
Submission history
From: Shuxin Liu [view email]
[v1]
Mon, 1 Jun 2026 03:00:13 UTC (9,067 KB)
[v2]
Tue, 13 Oct 2026 12:25:15 UTC (10,102 KB)
来源：arXiv cs.AI · arxiv.org