<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>One Token Away from Collapse: The Fragility of Instruction-Tuned Helpfulness · HackerPapers discussions</title><link>https://hackerpaper.kaipfister.de/p/2604.13006/</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/2604.13006/feed.xml" rel="self" type="application/rss+xml" /><item><title>One Token Away from Collapse: The Fragility of Instruction-Tuned Helpfulness — mentioned by tensegrist</title><link>https://news.ycombinator.com/item?id=48910760</link><guid isPermaLink="false">urn:hackerpapers:mention:2604.13006:48910760</guid><pubDate>Tue, 14 Jul 2026 18:02:55 GMT</pubDate><description>&lt;p&gt;things like that indeed tend to be fraught some relevant links https://arxiv.org/abs/2408.02442 https://arxiv.org/abs/2510.15061 https://arxiv.org/abs/2604.13006&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/2604.13006/"&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>
