PulseAugur
中
实时 06:56:33
English(EN) Token2Tile: Token-Wise Sparse Attention with Tile Execution through a Low-Rank Space for Video DiTs

Token2Tile 通过稀疏注意力加速视频扩散 Transformer

研究人员开发了 Token2Tile,一种新颖的稀疏注意力算子,旨在加速视频扩散 Transformer。这种无需训练的方法结合了逐查询评分和瓦片执行,利用在线低秩空间来有效地聚类和选择相关的关键 Token。Token2Tile 已在 MiniMax-H3 和 Wan2.1-14B 等模型上实现了显著的端到端加速,生成速度提高了 1.48-1.54 倍,同时保持了具有竞争力的 VBench 分数。在特定数据集上,与现有方法相比,它在 PSNR 方面也显示出显著的改进。 AI

影响 这项研究提供了一种显著加速视频扩散 Transformer 推理的方法,有可能降低计算成本并实现更广泛的应用。

排序理由 这是一篇详细介绍加速 AI 模型新颖技术方法的论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.CV 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。 我们如何撰写摘要 →

Token2Tile 通过稀疏注意力加速视频扩散 Transformer

本文如何被排名

Signal score
37 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
这是一篇详细介绍加速 AI 模型新颖技术方法的论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration
Single-source cluster
Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics
paper, infra
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
Breaking (< 6h)
Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准。

报道来源 [1]

  1. arXiv cs.CV TIER_1 English(EN) · Xuecheng Wu, Aichen Feng, Jinyan Chen, Qingjie Zhao, Julien Lai ·

    Token2Tile:通过低秩空间进行分块执行的逐令牌稀疏注意力,用于视频 DiTs

    arXiv:2611.17930v1 Announce Type: new Abstract: Attention is essential for modeling spatio-temporal dependencies in video diffusion transformers, but its quadratic cost makes it a major inference bottleneck. Its high sparsity offers an opportunity to accelerate generation by focu…