<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Understanding R1-Zero-Like Training: A Critical Perspective · HackerPapers discussions</title><link>https://hackerpaper.kaipfister.de/p/2503.20783/</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/2503.20783/feed.xml" rel="self" type="application/rss+xml" /><item><title>Understanding R1-Zero-Like Training: A Critical Perspective — mentioned by cubefox</title><link>https://news.ycombinator.com/item?id=43679716</link><guid isPermaLink="false">urn:hackerpapers:mention:2503.20783:43679716</guid><pubDate>Mon, 14 Apr 2025 10:09:12 GMT</pubDate><description>&lt;p&gt;I wonder whether they implemented the GRPO correction from this paper, which fixes overly long response lengths: https://arxiv.org/abs/2503.20783 I guess probably not, as they don&amp;#x27;t mention it.&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/2503.20783/"&gt;Paper and discussion history&lt;/a&gt;. A mention is not an endorsement.&lt;/p&gt;</description><category>cs.LG</category></item></channel></rss>
