AI Paper Atlas

Research map / Large language models / Retrieval and knowledge

ACE: Attribution-Controlled Knowledge Editing for Multi-hop Factual Recall

Jiayu Yang, Yuxuan Fan, Songning Lai, Shengen Wu, Jiaqi Tang, Chun Kang, Zhijiang Guo, Yutao Yue

ICLR 2026 · Large language models › Retrieval and knowledge · approach: Knowledge editing in LLMs · 3 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
use ACE (edit query and value layers) instead of PMETGPT-J-6B · MQuAKE-3K subsetmulti-hop accuracy: 37.01 -> 46.45Improved
use ACE instead of PMETQwen3-8B · MQuAKE-3K subsetmulti-hop accuracy: 20.78 -> 58.24Improved
use ACE instead of MEMIT in 1-edit caseGPT-J-6B · MQuAKE-3K subset, 1-editmulti-hop accuracy: 64.30 -> 45.26Hurt
use ACE instead of PMET on specificityGPT-J-6B · MQuAKE-3K subsetspecificity: 74.6 -> 79.2Improved
add query-layer editing (ACE) vs PMET on paraphraseQwen3-8B · MQuAKE-3K subsetparaphrase: 68.9 -> 94.2Improved
add edits to top-1 query layer (ablation)ACE · MQuAKE-3K (GPT-J)multi-hop accuracy: 43.26 -> 46.45Improved
add edits to top-3 query layers (ablation)ACE · MQuAKE-3K (GPT-J)multi-hop accuracy: 38.78 -> 46.45Improved
add edits to top-2 value layers (ablation)ACE · MQuAKE-3K (GPT-J)multi-hop accuracy: 33.97 -> 46.45Improved
add edits to top-2 value layers (ablation)ACE · MQuAKE-3K (Qwen3-8B)multi-hop accuracy: 34.68 -> 58.24Improved
add edits to top-3 query layers (ablation)ACE · MQuAKE-3K (Qwen3-8B)multi-hop accuracy: 45.26 -> 58.24Improved
add few-shot in-context prompts (vs zero-shot) (ablation)ACE · MQuAKE-3K (GPT-J)multi-hop accuracy: 9.4% degradation withoutImproved
edit more layers with ROME/MEMIT/PMET (9 vs 7) (ablation)baselines · MQuAKE-3K (GPT-J)multi-hop accuracy: 37.27 -> 37.98No clear effect

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