AI Paper Atlas

Research map / Large language models / Retrieval and knowledge

The Labyrinth and the Thread: Rethinking Regularizations in Sequential Knowledge Editing for Large Language Models

Zheng Wang, Kaixuan Zhang, Wanfang Chen, Jingwen Zhang, Xiaonan Lu

ICML 2026 · Large language models › Retrieval and knowledge · approach: Knowledge editing in LLMs · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
keep the K0 K0^T P preservation terms instead of dropping them via null-space simplification (ablation)sequential editing, Memorize-the-Latest task · GLUE SST (LLaMA-3)accuracy: 0.000 -> 0.846Improved
keep the K0 K0^T P preservation terms instead of dropping them via null-space simplification (ablation)sequential editing, Memorize-the-Latest task · MMLU (LLaMA-3)accuracy: 0.014 -> 0.548Improved
add error-correction term (Algorithm 1) to RECT post-processing (ablation)RECT SE · CounterFactGeneralization: 83.60 -> 94.34Improved
add error-correction term (Algorithm 1) to RECT post-processing (ablation)RECT SE · CounterFactSpecificity: 84.86 -> 81.56Hurt
add error-correction term (Algorithm 1) to PRUNE post-processing (ablation)PRUNE SE · CounterFactEfficacy/Generalization/Specificity: Eff. 99.82 -> 99.87No clear effect
enforce OTE alignment of sequential updates (ablation)PRUNE SE · CounterFactEfficacy: 56.30 -> 99.87Improved
enforce OTE alignment of sequential updates (ablation)RECT SE · CounterFactEfficacy: 60.35 -> 99.88Improved
apply post-processing regularization (RECT) to aggregated update in one-time editingRECT OTE · CounterFactediting performanceHurt
add PRUNE regularization on top of aligned MEMIT in one-time editingMEMIT (aligned) · CounterFactediting performanceNo clear effect
use OTE-aligned sequential editing with different regularizersMEMIT/AlphaEdit/PRUNE/RECT (aligned) · GLUE tasks (LLaMA-3)F1No clear effect
enforce OTE-SE alignment on PRUNE and RECT (ablation)PRUNE, RECT · hidden representation shift (LLaMA-3)distribution shiftImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy