<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>A Comparison of Code Embeddings and Beyond · HackerPapers discussions</title><link>https://hackerpaper.kaipfister.de/p/2109.07173/</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/2109.07173/feed.xml" rel="self" type="application/rss+xml" /><item><title>A Comparison of Code Embeddings and Beyond — mentioned by seanhunter</title><link>https://news.ycombinator.com/item?id=38207987</link><guid isPermaLink="false">urn:hackerpapers:mention:2109.07173:38207987</guid><pubDate>Thu, 09 Nov 2023 17:28:13 GMT</pubDate><description>&lt;p&gt;The problem with this line of thinking is that the specific embedding chosen has a big impact on task performance[1], and the person producing the piece of content doesn&amp;#x27;t necessarily know what you&amp;#x27;re going to be wanting to use that content for a lot of the time. The second thing is that certain embedding formats are very specific to particular model architectures. From a practical perspective there are some standard embedding formats that people use a lot so if you&amp;#x27;re performing a normal sort of task there&amp;#x27;s probably a standard format for that particular task (eg worth checking out spacy&amp;#x27;s…&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/2109.07173/"&gt;Paper and discussion history&lt;/a&gt;. A mention is not an endorsement.&lt;/p&gt;</description><category>cs.SE</category></item></channel></rss>
