AI Paper Atlas

Research map / Architectures and efficiency / Attention and transformers

CLUENet: Cluster Attention Makes Neural Networks Have Eyes

Xiangshuai Song, Jun-Jie Huang, Tianrui Liu, Ke Liang, Chang Tang

AAAI 2026 · Architectures and efficiency › Attention and transformers · approach: Theory of self-attention dynamics · 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 global soft feature aggregation (incl. TCosAttn and gate) (ablation)CLUENet (micro) · Mini-ImageNet valTop-1 accuracy (%): 76.97 -> 78.75Improved
use temperature-scaled cosine attention instead of standard attention (ablation)CLUENet (micro) · Mini-ImageNet valTop-1 accuracy (%): -0.59 Top-1 when removed (text)Improved
add gated residual fusion (ablation)CLUENet (micro) · Mini-ImageNet valTop-1 accuracy (%): 78.46 -> 78.75Improved
add shared dispatching across blocks (ablation)CLUENet (micro) · Mini-ImageNet valTop-1 accuracy (%): 78.32 -> 78.75Improved
add learnable positional embedding (ablation)CLUENet (micro) · Mini-ImageNet valTop-1 accuracy (%): 78.21 -> 78.75Improved
replace FEC cluster pooling with improved cluster pooling (pooling in similarity space, 2-layer MLP) (ablation)CLUENet (micro) · Mini-ImageNet valTop-1 accuracy (%): 78.07 -> 78.75Improved
use 2-layer MLP instead of single layer for projv (ablation)CLUENet (micro) · Mini-ImageNet valTop-1 accuracy (%): 78.10 -> 78.75Improved
increase projv depth from two to three layers (ablation)CLUENet (micro) · Mini-ImageNet valTop-1 accuracy (%): 78.75 -> 78.71No clear effect
add cluster-center-guided semantic query (ablation)CLUENet (micro) · Mini-ImageNet valTop-1 accuracy (%): 77.72 -> 78.75Improved
use multi-head instead of single-head center query (ablation)CLUENet (micro) · Mini-ImageNet valTop-1 accuracy (%): 78.41 -> 78.75Improved
use global clustering with 49 clusters instead of window partition (ablation)CLUENet (micro) · Mini-ImageNet valTop-1 accuracy (%): 78.30 -> 78.75Improved
increase cluster number from 49 to 64 (ablation)CLUENet (micro) · Mini-ImageNet valTop-1 accuracy (%): 78.75 -> 78.70No 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