AI Paper Atlas

Research map / Large language models / Retrieval and knowledge

Energy-Regularized Sequential Model Editing on Hyperspheres

Qingyuan Liu, Jia-Chen Gu, Yunzhi Yao, Hong Wang, Nanyun Peng

ICLR 2026 · Large language models › Retrieval and knowledge · approach: Knowledge editing in LLMs · 1 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
add SPHERE sparse-space projection to AlphaEditAlphaEdit, LLaMA3-8B · CounterFactefficacy: 4.37 -> 52.89Improved
add SPHERE projection to AlphaEditAlphaEdit, LLaMA3-8B · ZsREspecificity: 28.78 -> 45.40Improved
add SPHERE projection to AlphaEditAlphaEdit, Qwen2.5-7B · ZsREefficacy: 42.01 -> 70.04Improved
add SPHERE projection to AlphaEditAlphaEdit, Qwen2.5-7B · CounterFactefficacy: 43.92 -> 60.76Improved
add SPHERE projection to AlphaEditAlphaEdit, Qwen2.5-7B · CounterFactconsistency: PRUNE 29.50 vs SPHERE 14.74Hurt
add SPHERE projection to AlphaEditAlphaEdit, LLaMA3-8B · CounterFactfluency: 482.36 -> 551.51Improved
insert SPHERE projection into baseline editing methodsFT, MEMIT, RECT, PRUNE, AlphaEdit, LLaMA3-8B · 3000 sequential editsefficacy: 49.05%Improved
insert SPHERE projection into baseline editing methodsbaseline editors, LLaMA3-8B · 3000 sequential editsgeneralization: 42.64%Improved
insert SPHERE projection into baseline editing methodsbaseline editors, LLaMA3-8B · 3000 sequential editsspecificity: 24.44%Improved
add SPHERE projectionedit methods, LLaMA3-8B · general abilities after 15k edits (GSM8K, RTE, NQ, BoolQ)general performanceImproved
vary cumulative ratio eta and suppression strength alpha of SPHERE (ablation)MEMIT + SPHERE, LLaMA3-8B · 3000 sequential editsediting performanceImproved

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