<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Twins: Revisiting the Design of Spatial Attention in Vision Transformers · HackerPapers discussions</title><link>https://hackerpaper.kaipfister.de/p/2104.13840/</link><description>Collected HN comments and submissions linking this paper.</description><language>en</language><lastBuildDate>Mon, 02 Nov 2026 11:16:22 GMT</lastBuildDate><atom:link href="https://hackerpaper.kaipfister.de/p/2104.13840/feed.xml" rel="self" type="application/rss+xml" /><item><title>Twins: Revisiting the Design of Spatial Attention in Vision Transformers — mentioned by EnricoShippole</title><link>https://news.ycombinator.com/item?id=32087249</link><guid isPermaLink="false">urn:hackerpapers:mention:2104.13840:32087249</guid><pubDate>Wed, 13 Jul 2022 18:58:05 GMT</pubDate><description>&lt;p&gt;An #opensource implementation of the Twins: Revisiting the Design of Spatial Attention in Vision Transformers research paper in Google&amp;#x27;s #JAX and #Flax. JAX/Flax Repository: https://github.com/conceptofmind/Twins-SVT-Flax Twins SVT Research Paper: https://arxiv.org/abs/2104.13840 Official Github Repository: https://github.com/Meituan-AutoML/Twins In collaboration with Dr. Phil &amp;#x27;Lucid&amp;#x27; Wang: https://github.com/lucidrains.&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/2104.13840/"&gt;Paper and discussion history&lt;/a&gt;. A mention is not an endorsement.&lt;/p&gt;</description><category>cs.CV</category></item></channel></rss>
