<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Holophrasm: a neural Automated Theorem Prover for higher-order logic · HackerPapers discussions</title><link>https://hackerpaper.kaipfister.de/p/1608.02644/</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/1608.02644/feed.xml" rel="self" type="application/rss+xml" /><item><title>Holophrasm: a neural Automated Theorem Prover for higher-order logic — mentioned by raphlinus</title><link>https://news.ycombinator.com/item?id=34129462</link><guid isPermaLink="false">urn:hackerpapers:mention:1608.02644:34129462</guid><pubDate>Sun, 25 Dec 2022 18:02:38 GMT</pubDate><description>&lt;p&gt;Do not nerd snipe me! There&amp;#x27;s tons of juice left on the CPU side. Again I recommend Metamath Zero (which is somewhat of an opposite direction as HoTT), as it&amp;#x27;s able to check the entire Metamath proof database in a few hundred milliseconds. This is orders of magnitude faster than most other proof systems. If you did want to manipulate trees on the GPU, then my stack monoid work[1] might be a good basis for it. Of course, a whole nother direction is to use AI to develop proofs, which is a fascinating and growing area (see Holophrasm[2] and DeepMind work[3]). That of course runs on the GPU. [1]…&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/1608.02644/"&gt;Paper and discussion history&lt;/a&gt;. A mention is not an endorsement.&lt;/p&gt;</description><category>cs.AI</category></item><item><title>Holophrasm: a neural Automated Theorem Prover for higher-order logic — mentioned by drchewbacca</title><link>https://news.ycombinator.com/item?id=21108360</link><guid isPermaLink="false">urn:hackerpapers:mention:1608.02644:21108360</guid><pubDate>Sun, 29 Sep 2019 17:33:07 GMT</pubDate><description>&lt;p&gt;It&amp;#x27;s not really like that. NN&amp;#x27;s are only used when searching for new proofs or steps which can be used to fill in details, the verification of a finished proof is done by a classical program by a series of substitutions which are very straightforward and spelled out in detail. Here is an example. https://arxiv.org/pdf/1608.02644.pdf&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/1608.02644/"&gt;Paper and discussion history&lt;/a&gt;. A mention is not an endorsement.&lt;/p&gt;</description><category>cs.AI</category></item><item><title>Holophrasm: a neural Automated Theorem Prover for higher-order logic — mentioned by drchewbacca</title><link>https://news.ycombinator.com/item?id=20788870</link><guid isPermaLink="false">urn:hackerpapers:mention:1608.02644:20788870</guid><pubDate>Sat, 24 Aug 2019 18:37:06 GMT</pubDate><description>&lt;p&gt;And another https://arxiv.org/pdf/1608.02644.pdf I&amp;#x27;d really love to see someone like Deepmind get a hold of this. They are perfectly placed to apply a lot of resources to it and it&amp;#x27;s right in their wheelhouse. Don&amp;#x27;t know how to contact them about it though.&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/1608.02644/"&gt;Paper and discussion history&lt;/a&gt;. A mention is not an endorsement.&lt;/p&gt;</description><category>cs.AI</category></item></channel></rss>
