<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Online normalizer calculation for softmax · HackerPapers discussions</title><link>https://hackerpaper.kaipfister.de/p/1805.02867/</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/1805.02867/feed.xml" rel="self" type="application/rss+xml" /><item><title>Online normalizer calculation for softmax — mentioned by sailingparrot</title><link>https://news.ycombinator.com/item?id=45450252</link><guid isPermaLink="false">urn:hackerpapers:mention:1805.02867:45450252</guid><pubDate>Thu, 02 Oct 2025 14:40:55 GMT</pubDate><description>&lt;p&gt;&amp;gt; they set their hiring bar for engineers too high Not sure I agree, if you look at the head count growth of companies like OpenAI, Anthropic etc, it is super fast, its already pretty hard to keep everything working smoothly with that rate of employee growth, so going faster than that seems very risky. Ultimately I think it&amp;#x27;s mostly caused by the field still being so new. Everything still needs to be optimized and there just aren&amp;#x27;t that many very good CUDA programmers to start with, then you need to find one that also has deep knowledge of ML and transformers architectures, which further…&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/1805.02867/"&gt;Paper and discussion history&lt;/a&gt;. A mention is not an endorsement.&lt;/p&gt;</description><category>cs.PF</category></item></channel></rss>
