AI Paper Atlas

Research map / Large language models / Retrieval and knowledge

TangleScore: Tangle-Guided Purge and Imprint for Unstructured Knowledge Editing

Hao-Xiang Xu, Ziqi Peng, Jun-Yu Ma, Yuhao Sun, Zhen-Hua Ling, Jia-Chen Gu

ICLR 2026 · Large language models › Retrieval and knowledge · approach: Knowledge editing in LLMs

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
use ratio form of TangleScore instead of numerator only (ablation)PIPE · UNKEBench LLaMA3-8B-InstructFactScore: 32.79 -> 50.91Improved
use ratio form of TangleScore instead of denominator only (ablation)PIPE · UNKEBench LLaMA3-8B-InstructFactScore: 36.55 -> 50.91Improved
use dynamic purge rate instead of fixed 0.001 (ablation)PIPE · UNKEBench LLaMA3-8B-InstructFactScore: 48.21 -> 50.91Improved
use dynamic purge rate instead of fixed 0.0001 (ablation)PIPE · UNKEBench LLaMA3-8B-InstructFactScore: 43.85 -> 50.91Improved
add Lpurge loss (ablation)PIPE · UNKEBench LLaMA3-8B-InstructFactScore: 47.15 -> 50.91Improved
add Lconsistency loss (ablation)PIPE · UNKEBench LLaMA3-8B-InstructFactScore: 26.61 -> 50.91Improved
edit layer 7 instead of layer 4 (ablation)PIPE · UNKEBench LLaMA3-8B-InstructFactScore: 48.27 -> 50.91Improved
edit layer 7 instead of layer 10 (ablation)PIPE · UNKEBench LLaMA3-8B-InstructFactScore: 48.11 -> 50.91Improved
use PIPE instead of AnyEditAnyEdit · UNKEBench LLaMA3-8B-InstructRouge-2 (paraphrased): 72.59 -> 78.03Improved
use PIPE instead of UNKEUNKE · KEBenchPara-Acc: 85.34 -> 88.23Improved
use PIPE instead of AnyEditAnyEdit · UNKEBench Qwen2.5-7B-InstructFactScore: 40.60 -> 42.47Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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