AI Paper Atlas

Research map / Architectures and efficiency / Attention and transformers

Spatial-Aware Reduction Framework: Towards Efficient and Faithful Visual State Space Models

Jindi Lv, Aoyu Li, Yuhao Zhou, Zheng Zhu, Xiaofeng Wang, Qing Ye, Yueqi Duan, Wentao Feng, Jiancheng Lv

ICML 2026 · Architectures and efficiency › Attention and transformers · approach: Block-sparse attention kernels · 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
add structured spatial reduction (row/column) instead of global ToMe merging (ablation)ToMe · ImageNet-1K VMamba-S 29% reductiontop-1 accuracy: 17.6 -> 71.4Improved
add local window constraint to structured reduction (ablation)STORM (ToMe) · ImageNet-1K VMamba-S 36% reductiontop-1 accuracy: 65.3 -> 73.3Improved
apply STORM to ToMe on VMamba-TToMe · ImageNet-1K VMamba-Ttop-1 accuracy: 31.5 -> 78.8Improved
apply STORM to EViT on VMamba-TEViT · ImageNet-1K VMamba-Ttop-1 accuracy: 15.2 -> 78.5Improved
apply STORM to ToMe on VMamba-BToMe · ImageNet-1K VMamba-Btop-1 accuracy: 35.7 -> 82.6Improved
apply STORM to ToMe on PlainMamba-L1ToMe · ImageNet-1K PlainMamba-L1top-1 accuracy: 69.6 -> 75.4Improved
apply STORM to ToMe on PlainMamba-L3ToMe · ImageNet-1K PlainMamba-L3top-1 accuracy: 76.1 -> 80.9Improved
apply STORM to EViT for detection on VMamba-BEViT · COCO detection VMamba-Bbox AP: 3.7 -> 46.2Improved
use content-aware ToMe selection instead of random selection within STORM (ablation)STORM · ImageNet-1K VMamba-S 29% reductiontop-1 accuracy: 74.4 -> 78.0Improved
apply STORM to ToMe with 5 epochs fine-tuningToMe · ImageNet-1K VMamba-Stop-1 accuracy: 48.1 -> 82.1Improved
use STORM instead of Mamba-specific HSAVMamba-S · ImageNet-1K training-freetop-1 accuracy: 46.2 -> 82.0Improved

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