<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Multimodal Few-Shot Learning with Frozen Language Models · HackerPapers discussions</title><link>https://hackerpaper.kaipfister.de/p/2106.13884/</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/2106.13884/feed.xml" rel="self" type="application/rss+xml" /><item><title>Multimodal Few-Shot Learning with Frozen Language Models — mentioned by gwern</title><link>https://news.ycombinator.com/item?id=27900509</link><guid isPermaLink="false">urn:hackerpapers:mention:2106.13884:27900509</guid><pubDate>Tue, 20 Jul 2021 23:25:27 GMT</pubDate><description>&lt;p&gt;&amp;gt; I would be interested to see what GPT-3 could come up with if explicity tasked to do this type of re-phrasing. First, this is definitely not GPT-3, see my other comment. However, as one would expect, GPT-3 is capable of all sorts of fascinating &amp;#x27;rewriting&amp;#x27; tasks. GPT-2 could do summarization via &amp;#x27;tldr&amp;#x27; but GPT-3 can so so much more! I enjoyed &amp;quot;overcomplicated explanations&amp;quot; ( https://www.gwern.net/GPT-3-nonfiction#overcomplicated-expla... ) as the reverse of &amp;#x27;explain it to a second grader&amp;#x27; ( https://andrewmayneblog.wordpress.com/2020/06/13/openai-api-... ) but it can also do memes…&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/2106.13884/"&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>
