<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Exploiting Cloze Questions for Few Shot Text Classification and Natural Language Inference · HackerPapers discussions</title><link>https://hackerpaper.kaipfister.de/p/2001.07676/</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/2001.07676/feed.xml" rel="self" type="application/rss+xml" /><item><title>Exploiting Cloze Questions for Few Shot Text Classification and Natural Language Inference — mentioned by cs702</title><link>https://news.ycombinator.com/item?id=24708179</link><guid isPermaLink="false">urn:hackerpapers:mention:2001.07676:24708179</guid><pubDate>Wed, 07 Oct 2020 13:51:14 GMT</pubDate><description>&lt;p&gt;The headline is clickbait (GPT-3 is about language modeling/generation, and there&amp;#x27;s ZERO mention of that in this work), but the actual work is interesting and the paper is worth a read. To the OP: please consider changing the headline to &amp;quot;Exploiting Cloze Questions for Few Shot Text Classification and Natural Language Inference&amp;quot; and link to the original paper at https://arxiv.org/abs/2001.07676 instead of this PR puff piece.&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/2001.07676/"&gt;Paper and discussion history&lt;/a&gt;. A mention is not an endorsement.&lt;/p&gt;</description><category>cs.CL</category></item></channel></rss>
