<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Alias-Free Generative Adversarial Networks · HackerPapers discussions</title><link>https://hackerpaper.kaipfister.de/p/2106.12423/</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/2106.12423/feed.xml" rel="self" type="application/rss+xml" /><item><title>Alias-Free Generative Adversarial Networks — mentioned by godelski</title><link>https://news.ycombinator.com/item?id=44834073</link><guid isPermaLink="false">urn:hackerpapers:mention:2106.12423:44834073</guid><pubDate>Fri, 08 Aug 2025 06:20:57 GMT</pubDate><description>&lt;p&gt;&amp;gt; Something like GPT-2 I got 2 things to say here 1) There&amp;#x27;s plenty of things that can achieve similar performance to GPT-2 these days. We mentioned Mamba, they compared to GPT-3 in their first paper[0]. They compare with the open sourced version and you&amp;#x27;ll also see some other architectures referenced there like Hyena and H3. It&amp;#x27;s the GPT-Neo and GPT-J models. Remember GPT-3 is pretty much just a scaled up GPT-2. 2) I think you are underestimating the costs to train some of these things. I know Karpathy said you can now train GPT-2 for like $1k[1] but a single training run is a small portion…&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/2106.12423/"&gt;Paper and discussion history&lt;/a&gt;. A mention is not an endorsement.&lt;/p&gt;</description><category>cs.CV</category></item><item><title>Alias-Free Generative Adversarial Networks — mentioned by soraki_soladead</title><link>https://news.ycombinator.com/item?id=27882913</link><guid isPermaLink="false">urn:hackerpapers:mention:2106.12423:27882913</guid><pubDate>Mon, 19 Jul 2021 14:21:53 GMT</pubDate><description>&lt;p&gt;This problem was recently addressed in detail here[1] which serves as a good overview. 1. https://arxiv.org/abs/2106.12423 (project page: https://nvlabs.github.io/alias-free-gan/)&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/2106.12423/"&gt;Paper and discussion history&lt;/a&gt;. A mention is not an endorsement.&lt;/p&gt;</description><category>cs.CV</category></item></channel></rss>
