<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>An Investigation of Why Overparameterization Exacerbates Spurious Correlations · HackerPapers discussions</title><link>https://hackerpaper.kaipfister.de/p/2005.04345/</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/2005.04345/feed.xml" rel="self" type="application/rss+xml" /><item><title>An Investigation of Why Overparameterization Exacerbates Spurious Correlations — mentioned by pms</title><link>https://news.ycombinator.com/item?id=35284054</link><guid isPermaLink="false">urn:hackerpapers:mention:2005.04345:35284054</guid><pubDate>Fri, 24 Mar 2023 01:49:25 GMT</pubDate><description>&lt;p&gt;Have you provided the visual novel as a prompt or just asked about it? ChatGPT does seem to know more about books than songs, maybe because they are more often described in texts, while songs not, not sure. I wished more parameters gave a generalizable solution, but research suggests that&amp;#x27;s not the case [1, 2, 3]. OpenAI won&amp;#x27;t tell you this, cause they have a product to sell. [1] http://arxiv.org/abs/2211.08411 [2] http://arxiv.org/abs/2011.03395 [3] http://arxiv.org/abs/2005.04345&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/2005.04345/"&gt;Paper and discussion history&lt;/a&gt;. A mention is not an endorsement.&lt;/p&gt;</description><category>cs.LG</category></item></channel></rss>
