<rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Hacker News: Dewey5001</title><link>https://news.ycombinator.com/user?id=Dewey5001</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Sat, 18 Apr 2026 05:38:24 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=Dewey5001" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by Dewey5001 in "Don't use cosine similarity carelessly"]]></title><description><![CDATA[
<p>These are very interesting and well-founded arguments. However, I believe the solution should not be around LLMs. As rightly pointed out in the article, the proposed solution is prohibitively expensive. How could I instead reformulate the approach to align with my specific definition of similarity?</p>
]]></description><pubDate>Thu, 16 Jan 2025 16:46:56 +0000</pubDate><link>https://news.ycombinator.com/item?id=42727672</link><dc:creator>Dewey5001</dc:creator><comments>https://news.ycombinator.com/item?id=42727672</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=42727672</guid></item><item><title><![CDATA[New comment by Dewey5001 in "Don't use cosine similarity carelessly"]]></title><description><![CDATA[
<p>I believe the intention here was to highlight a use case where cosine similarity falls short, leading into the next section that introduces alternatives. That said, I would appreciate more detail in the 'Extracting the right features' section, if someone has an example I would love to see it.</p>
]]></description><pubDate>Thu, 16 Jan 2025 16:44:00 +0000</pubDate><link>https://news.ycombinator.com/item?id=42727634</link><dc:creator>Dewey5001</dc:creator><comments>https://news.ycombinator.com/item?id=42727634</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=42727634</guid></item></channel></rss>