AI Paper Atlas

Research map / Architectures and efficiency / Attention and transformers

Sparser Block-Sparse Attention via Token Permutation

Xinghao Wang, Pengyu Wang, Dong Zhang, Chenkun Tan, Shaojun Zhou, Zhaoxiang Liu, Shiguo Lian, Fangxu Liu, Kai Song, Xipeng Qiu

ICML 2026 · Architectures and efficiency › Attention and transformers · approach: Sparse attention for long-context LLMs

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 segmented key permutation to MeanPooling block selection (ablation)MeanPooling block-sparse attention · LongBench v2 (Qwen-2.5-7B-1M)accuracy: 26.24 -> 34.39 (31% relative)Improved
add segmented key permutation to MeanPooling block selection (ablation)MeanPooling block-sparse attention · RULER (Llama-3.1-8B)accuracy: 82.76 -> 85.97 (+3.21 avg, +7.66 at 128K)Improved
add segmented key permutation to MeanPooling block selection (ablation)MeanPooling block-sparse attention · LongBench (Qwen-2.5-7B-1M)score: 31.83 -> 36.37Improved
add segmented key permutation to MeanPooling block selection (ablation)MeanPooling block-sparse attention · LongBench (Llama-3.1-8B)score: 36.67 -> 37.37Improved
add permutation to XAttention antidiagonal block scoring (PBS-Attn+) (ablation)XAttention · RULERaccuracy: +1.41 on Llama-3.1-8B, +1.05 on Qwen-2.5-7B-1MImproved
add permutation before block selection (ablation)MeanPooling block selection · block-level densityrelative sparsity: 7% absolute improvement at 8K contextImproved
add permutation before block selection (ablation)Llama-3.1-8B · selected blocksselected-block count: reduces selected blocks by 10.7% at 8K and 14.4% at 128KImproved
permute keys randomly instead of by global importance (ablation)segmented permutation · sparsity gain at 97.5% coverage (NarrativeQA)absolute sparsity gain: -6.11 (Random) vs 7.90 (Random-Query-Attn)Hurt
sort keys by global attention importance instead of greedy local query alignment (ablation)segmented permutation · sparsity gain at 97.5% coverage (NarrativeQA)absolute sparsity gain: 3.00 -> 7.77 (Greedy vs Last-Block-Query-Attn)Improved
also permute queries in addition to keys (ablation)PBS-Attn · LongBench vs densityscore-density trade-offNo clear effect
apply permutation across heads of Llama-3.1-8B at 32K (ablation)PBS-Attn · niah_single head analysishead sparsity: 70.8% heads helped, 24.0% unchanged, 5.2% harmedMixed
use block size B=128 rather than 64 or 256 (ablation)PBS-Attn · LongBench vs densityLongBench scoreImproved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 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