AI Paper Atlas

Research map / Architectures and efficiency / Attention and transformers

Vision-MoR: Scaling Vision Transformer via Patch-Level Mixture-of-Recursions

Yunhong He, Zhengqing Yuan, Weixiang Sun, Yiyang Li, Yixin Liu, Yanfang Ye, Lichao Sun

AAAI 2026 · Architectures and efficiency › Attention and transformers · approach: Efficient and robust vision transformers · 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
use Vision-MoR Small instead of SHViT-S1Vision-MoR · ImageNet-1K val (224px, 300 epochs, from scratch)Top-1 accuracy (%): 72.8 -> 74.6Improved
use Vision-MoR X-Large instead of SHViT-S4Vision-MoR · ImageNet-1K val (224px, 300 epochs, from scratch)Top-1 accuracy (%): 79.4 -> 80.4Improved
use Vision-MoR X-Large backbone instead of SHViT-S4Vision-MoR · COCO 2017 detection (RetinaNet)AP: 38.8 -> 39.1Improved
use Vision-MoR instead of SHViT at smallest scaleVision-MoR · GH200 GPU, smallest scale (batch 1)latency (ms): 7.03 -> 3.63Improved
replace Linear router with spatial router (shifted-window attention) (ablation)Vision-MoR · ImageNet-1K ablation, expert-choice routerTop-1 accuracy (%): 70.1 -> 77.1Improved
use auxiliary loss instead of auxiliary router (ablation)Vision-MoR · ImageNet-1K ablation, expert-choice routerTop-1 accuracy (%): 55.4 -> 77.1Improved
add Z-loss (ablation)Vision-MoR (expert-choice) · ImageNet-1K ablation, expert-choice routerTop-1 accuracy (%): 76.5 -> 77.1Improved
add Z-loss (ablation)Vision-MoR (token-choice) · ImageNet-1K ablation, token-choice routerTop-1 accuracy (%): 74.2 -> 73.5Hurt
replace sigmoid with tanh (ablation)Vision-MoR (expert-choice) · ImageNet-1K ablation, expert-choice routerTop-1 accuracy (%): 77.1 -> 76.4No clear effect
increase input resolution from 224 to 1120 (ablation)Vision-MoR Small · ImageNet-1K valTop-1 accuracy (%): 74.6 -> 79.6Improved

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