<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Homogeneous Keys, Heterogeneous Values: Exploiting Local KV Cache Asymmetry for Long-Context LLMs · HackerPapers discussions</title><link>https://hackerpaper.kaipfister.de/p/2506.05410/</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/2506.05410/feed.xml" rel="self" type="application/rss+xml" /><item><title>Homogeneous Keys, Heterogeneous Values: Exploiting Local KV Cache Asymmetry for Long-Context LLMs — mentioned by yorwba</title><link>https://news.ycombinator.com/item?id=45493889</link><guid isPermaLink="false">urn:hackerpapers:mention:2506.05410:45493889</guid><pubDate>Mon, 06 Oct 2025 17:36:26 GMT</pubDate><description>&lt;p&gt;Yes, for example https://arxiv.org/pdf/2506.05410 merges two neighboring tokens with the lowest sum of past attention scores, and this method would enable using future expected attention instead.&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/2506.05410/"&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>Homogeneous Keys, Heterogeneous Values: Exploiting Local KV Cache Asymmetry for Long-Context LLMs — HN submission</title><link>https://news.ycombinator.com/item?id=44397503</link><guid isPermaLink="false">urn:hackerpapers:submission:2506.05410:44397503</guid><pubDate>Fri, 27 Jun 2025 15:22:27 GMT</pubDate><description>&lt;p&gt;Exploiting Local KV Cache Asymmetry for Long-Context LLMs&lt;/p&gt;&lt;p&gt;&lt;/p&gt;&lt;p&gt;6 points; 0 discussion replies on HN. &lt;a href="https://hackerpaper.kaipfister.de/p/2506.05410/"&gt;Paper and source history&lt;/a&gt;. A submission is not an endorsement.&lt;/p&gt;</description><category>cs.CL</category></item></channel></rss>
