<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>A Primer in BERTology: What we know about how BERT works · HackerPapers discussions</title><link>https://hackerpaper.kaipfister.de/p/2002.12327/</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/2002.12327/feed.xml" rel="self" type="application/rss+xml" /><item><title>A Primer in BERTology: What we know about how BERT works — mentioned by sigtstp</title><link>https://news.ycombinator.com/item?id=43892913</link><guid isPermaLink="false">urn:hackerpapers:mention:2002.12327:43892913</guid><pubDate>Mon, 05 May 2025 08:17:15 GMT</pubDate><description>&lt;p&gt;I feel this makes some fundamental conceptual mistakes and is just riding the LLM wave. &amp;quot;Semantics&amp;quot; is literally behavior under execution. This is syntactical analysis by a stochastic language model. I know the NLP literature uses &amp;quot;semantics&amp;quot; to talk about representations but that is an assertion which is contested [1]. Coming back to testing, this implicitly relies on the strong assumption of the LLM correctly associating the code (syntax) with assertions of properties under execution (semantic properties). This is a very risky assumption considering, once again, these things are stochastic…&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/2002.12327/"&gt;Paper and discussion history&lt;/a&gt;. A mention is not an endorsement.&lt;/p&gt;</description><category>cs.CL</category></item><item><title>A Primer in BERTology: What we know about how BERT works — mentioned by MacsHeadroom</title><link>https://news.ycombinator.com/item?id=35352214</link><guid isPermaLink="false">urn:hackerpapers:mention:2002.12327:35352214</guid><pubDate>Wed, 29 Mar 2023 03:52:18 GMT</pubDate><description>&lt;p&gt;I don&amp;#x27;t understand what you want me to ask it about. So I asked it something, as usual every citation and URL is real despite not having internet access or training data after 2021: &amp;lt;PROMPT&amp;gt;List some citations of recent Machine Learning papers discussing reasoning ability in Machine Learning Models such as HellaSwag. [REPLY]&amp;quot; As of my knowledge cutoff in September 2021, I can provide you with some citations for ML papers that discuss reasoning abilities in models and datasets like HellaSwag. However, you may want to search for more recent papers using a platform like arXiv or Google Scholar.…&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/2002.12327/"&gt;Paper and discussion history&lt;/a&gt;. A mention is not an endorsement.&lt;/p&gt;</description><category>cs.CL</category></item><item><title>A Primer in BERTology: What we know about how BERT works — HN submission</title><link>https://news.ycombinator.com/item?id=25043280</link><guid isPermaLink="false">urn:hackerpapers:submission:2002.12327:25043280</guid><pubDate>Tue, 10 Nov 2020 06:23:01 GMT</pubDate><description>&lt;p&gt;A Primer in BERTology: What We Know About How Bert Works&lt;/p&gt;&lt;p&gt;&lt;/p&gt;&lt;p&gt;81 points; 16 discussion replies on HN. &lt;a href="https://hackerpaper.kaipfister.de/p/2002.12327/"&gt;Paper and source history&lt;/a&gt;. A submission is not an endorsement.&lt;/p&gt;</description><category>cs.CL</category></item><item><title>A Primer in BERTology: What we know about how BERT works — HN submission</title><link>https://news.ycombinator.com/item?id=22500095</link><guid isPermaLink="false">urn:hackerpapers:submission:2002.12327:22500095</guid><pubDate>Fri, 06 Mar 2020 00:51:13 GMT</pubDate><description>&lt;p&gt;A Primer in BERTology: What we know about how BERT works&lt;/p&gt;&lt;p&gt;&lt;/p&gt;&lt;p&gt;2 points; 0 discussion replies on HN. &lt;a href="https://hackerpaper.kaipfister.de/p/2002.12327/"&gt;Paper and source history&lt;/a&gt;. A submission is not an endorsement.&lt;/p&gt;</description><category>cs.CL</category></item></channel></rss>
