https://jaehun.me/en/tags/sparse-attention/Recent content in Sparse Attention on Jaehun's BlogHugoen-USTue, 08 Sep 2026 13:09:59 +0000https://jaehun.me/en/posts/paper-review-radial-attention-on-log-n-sparse-attention-with-energy-decay-for-long-video-generation/Mon, 15 Dec 2025 00:00:00 +0900https://jaehun.me/en/posts/paper-review-radial-attention-on-log-n-sparse-attention-with-energy-decay-for-long-video-generation/<p><a href="https://arxiv.org/abs/2506.19852v1"target="_blank" class="inline-flex items-center gap-1" >Paper<svg class="h-3 w-3 flex-shrink-0" id="external-link" xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24"><path fill="none" stroke="currentColor" stroke-linecap="round" stroke-linejoin="round" stroke-width="2" d="M15 3h6v6m-11 5L21 3m-3 10v6a2 2 0 0 1-2 2H5a2 2 0 0 1-2-2V8a2 2 0 0 1 2-2h6"/></svg> </a></p> <h2 id="radial-attention-on-log-n-sparse-attention-with-energy-decay--generating-long-videos-cheaply">Radial Attention: O(n log n) Sparse Attention with Energy Decay — Generating &ldquo;Long Videos&rdquo; Cheaply<a href="#radial-attention-on-log-n-sparse-attention-with-energy-decay--generating-long-videos-cheaply" class="heading-anchor" aria-label="Link to this section"><svg class="h-4 w-4" aria-hidden="true" xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24"><g fill="none" stroke="currentColor" stroke-linecap="round" stroke-linejoin="round" stroke-width="2"><path d="M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71"/><path d="M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71"/></g></svg></a></h2><h2 id="one-line-summary-tldr">One-Line Summary (TL;DR)<a href="#one-line-summary-tldr" class="heading-anchor" aria-label="Link to this section"><svg class="h-4 w-4" aria-hidden="true" xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24"><g fill="none" stroke="currentColor" stroke-linecap="round" stroke-linejoin="round" stroke-width="2"><path d="M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71"/><path d="M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71"/></g></svg></a></h2><p>Based on the observation that post-softmax attention energy decays <strong>exponentially</strong> with distance (in time/space), the authors design a <strong>static mask</strong> that decays compute density in the same way, substantially cutting the training/inference cost of long-video generation. (source: §4.1–§4.2, Eq.3–Eq.4)</p>