<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>TurboRAG: Accelerating Retrieval-Augmented Generation with Precomputed KV Caches for Chunked Text · HackerPapers discussions</title><link>https://hackerpaper.kaipfister.de/p/2410.07590/</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/2410.07590/feed.xml" rel="self" type="application/rss+xml" /><item><title>TurboRAG: Accelerating Retrieval-Augmented Generation with Precomputed KV Caches for Chunked Text — mentioned by Palmik</title><link>https://news.ycombinator.com/item?id=45559816</link><guid isPermaLink="false">urn:hackerpapers:mention:2410.07590:45559816</guid><pubDate>Sun, 12 Oct 2025 17:07:22 GMT</pubDate><description>&lt;p&gt;The observation about the &amp;quot;block-diagonal patterns&amp;quot; in RAG isn&amp;#x27;t new and has been exploited / explored before: - https://arxiv.org/abs/2410.07590 (literally titled &amp;quot;Block-Attention for Efficient RAG&amp;quot;) - https://arxiv.org/abs/2409.15355v3 - https://arxiv.org/abs/2212.10947 The REFRAG paper does not cite any of these.&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/2410.07590/"&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>
