AI Paper Atlas

Research map / Architectures and efficiency / Attention and transformers

Multi-Head Normalization for Wide Vision Transformers

Guoyizhe Wei, Feng Wang, Alan Yuille, Rama Chellappa

ECCV 2026 · Architectures and efficiency › Attention and transformers · approach: Hybrid and interpretable ViT designs · 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
replace global RMSNorm with MH-RMSNorm on ViT-HViT-H · ImageNet-1Ktop-1 accuracy: 85.2 -> 85.5Improved
replace global RMSNorm with MH-RMSNorm on ViT-BViT-B · ImageNet-1Ktop-1 accuracy: 83.8 -> 83.9No clear effect
replace global RMSNorm with MH-RMSNorm on ViT-HViT-H · ADE20KmIoU: 50.6 -> 51.0Improved
replace RMSNorm with MH-RMSNorm in DiT-XL/2 (7M steps)DiT-XL/2 · ImageNet-256FID: 2.20 -> 1.95Improved
replace RMSNorm with MH-RMSNorm in DiT-XL/2 (800k steps)DiT-XL/2 · ImageNet-256FID: 13.8 -> 13.0Improved
partition width-4096 features into 8 to 32 chunks instead of global RMSNorm (ablation)ViT width 4096 · ImageNet-1Ktop-1 accuracy: 61.1 -> 84.4Improved
use too-small chunks (d=64 instead of 256) (ablation)MH-RMSNorm width 4096 · ImageNet-1Ktop-1 accuracy: 84.4 -> 83.5Hurt
use MH-LayerNorm instead of MH-RMSNorm (ablation)MH-Norm width 4096 · ImageNet-1Ktop-1 accuracy: 84.4 -> 84.2Hurt
use head-wise instead of shared affine parameters in MH-RMSNorm (ablation)MH-RMSNorm width 4096 · ImageNet-1Ktop-1 accuracy: 84.0 -> 84.4Improved
apply MH-RMSNorm to both attention and MLP branches instead of attention only (ablation)ViT width 2048 · ImageNet-1Ktop-1 accuracy: 84.0 -> 84.2Improved
apply MH-RMSNorm to all layers instead of only deep layers 9-12 (ablation)ViT width 2048 · ImageNet-1Ktop-1 accuracy: 83.9 -> 84.2Improved
apply MH-RMSNorm to only shallow layers 1-4 (ablation)ViT width 2048 · ImageNet-1Ktop-1 accuracy: 82.2 -> 83.1Mixed

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