<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Beyond Autoregression: Discrete Diffusion for Complex Reasoning and Planning · HackerPapers discussions</title><link>https://hackerpaper.kaipfister.de/p/2410.14157/</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/2410.14157/feed.xml" rel="self" type="application/rss+xml" /><item><title>Beyond Autoregression: Discrete Diffusion for Complex Reasoning and Planning — mentioned by mountainriver</title><link>https://news.ycombinator.com/item?id=44064625</link><guid isPermaLink="false">urn:hackerpapers:mention:2410.14157:44064625</guid><pubDate>Thu, 22 May 2025 17:47:29 GMT</pubDate><description>&lt;p&gt;A big discussion on this happened here as well https://news.ycombinator.com/item?id=44057820 There is quite a bit of evidence diffusion models work better at reasoning because they don&amp;#x27;t suffer from early token bias. https://github.com/HKUNLP/diffusion-vs-ar https://arxiv.org/html/2410.14157v3&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/2410.14157/"&gt;Paper and discussion history&lt;/a&gt;. A mention is not an endorsement.&lt;/p&gt;</description><category>cs.CL</category></item><item><title>Beyond Autoregression: Discrete Diffusion for Complex Reasoning and Planning — mentioned by mdp2021</title><link>https://news.ycombinator.com/item?id=44058762</link><guid isPermaLink="false">urn:hackerpapers:mention:2410.14157:44058762</guid><pubDate>Thu, 22 May 2025 04:22:17 GMT</pubDate><description>&lt;p&gt;I.e.: https://arxiv.org/html/2410.14157v3 # Beyond Autoregression: Discrete Diffusion for Complex Reasoning and Planning&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/2410.14157/"&gt;Paper and discussion history&lt;/a&gt;. A mention is not an endorsement.&lt;/p&gt;</description><category>cs.CL</category></item></channel></rss>
