<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Current Limitations of Language Models: What You Need is Retrieval · HackerPapers discussions</title><link>https://hackerpaper.kaipfister.de/p/2009.06857/</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/2009.06857/feed.xml" rel="self" type="application/rss+xml" /><item><title>Current Limitations of Language Models: What You Need is Retrieval — mentioned by Isinlor</title><link>https://news.ycombinator.com/item?id=27218897</link><guid isPermaLink="false">urn:hackerpapers:mention:2009.06857:27218897</guid><pubDate>Thu, 20 May 2021 08:49:56 GMT</pubDate><description>&lt;p&gt;I don&amp;#x27;t see anything impressive in this paper. They basically say what they want to do, no results. They are not even original with the idea for this type of paper. In September last year Aran Komatsuzaki posted: &amp;quot;Current Limitations of Language Models: What You Need is Retrieval&amp;quot; https://arxiv.org/abs/2009.06857 I have also a prediction market from August last year with this question: Will a language model with at least 100B parameters trained to do external information retrieval exist before 2023? Seems like people gave it 61% in March this year. I was at 75%.…&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/2009.06857/"&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>
