AI Paper Atlas

Research map / Architectures and efficiency / Attention and transformers

VecAttention: Vector-wise Sparse Attention for Accelerating Long Context Inference

Anmin Liu, Ruixuan Yang, Huiqiang Jiang, Bin Lin, Minmin Sun, Yong Li, Chen Zhang, Tao Xie

CVPR 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
use vertical-vector sparse attention instead of coarse line/block patternsQwen2.5-VL-7B-Instruct · video understanding avgavg accuracy: 47.2 -> 49.9 at 73.6 -> 78.5% sparsity (vs XAttention)Improved
use VecAttention on InternVL-3.5-8BInternVL-3.5-8B · video understanding avgavg accuracy: 52.7 -> 51.1Mixed
VecAttention vs XAttention and SVG on video generationWan2.1 and HunyuanVideo · VBenchPSNR: Hunyuan 21.8 -> 22.8 PSNR; LPIPS 0.326 -> 0.330 vs SVGMixed
replace row-wise topP with sorting-free minS filter (ablation)vector selection · 64K context, sparsity 0.9selection latency: 3.77x speedupImproved
fuse selection into tiled GEMM (TilingSelect) (ablation)minS selection · 64K context, sparsity 0.9HBM access / latency: 18.3 GB -> 1.8 GB; 2.42x speedupImproved
use vertical-vector instead of horizontal-vector pattern (ablation)attention sparsity pattern · VLM attention recall at matched sparsityrecall: 0.89 -> 0.95 (V-Vec vs H-Vec at ~69% sparsity)Improved
increase vector size Pq beyond 64 (ablation)VecAttention · VideoMMEaccuracy/latencyMixed
increase K-block group size Gk to enable cross-tile minS accumulation (ablation)VecAttention · VideoMMEaccuracyImproved
use VecAttention for end-to-end prefillVLM · VideoMME TTFTTTFT speedup: 1.17x end-to-end vs 2.65x attention-onlyImproved

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