<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Information Geometry of Evolution of Neural Network Parameters While Training · HackerPapers discussions</title><link>https://hackerpaper.kaipfister.de/p/2406.05295/</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/2406.05295/feed.xml" rel="self" type="application/rss+xml" /><item><title>Information Geometry of Evolution of Neural Network Parameters While Training — mentioned by griffzhowl</title><link>https://news.ycombinator.com/item?id=45816430</link><guid isPermaLink="false">urn:hackerpapers:mention:2406.05295:45816430</guid><pubDate>Tue, 04 Nov 2025 22:10:28 GMT</pubDate><description>&lt;p&gt;There&amp;#x27;s a field known as information geometry. I don&amp;#x27;t know much about it myself as I&amp;#x27;m more into physics, but here&amp;#x27;s a recent example of applying geometrical analysis to neural networks. Looks interesting as they find a phenomenon analogous to phase transitions during training Information Geometry of Evolution of Neural Network Parameters While Training https://arxiv.org/abs/2406.05295&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/2406.05295/"&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>
