<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>FORTAP: Using Formulas for Numerical-Reasoning-Aware Table Pretraining · HackerPapers discussions</title><link>https://hackerpaper.kaipfister.de/p/2109.07323/</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/2109.07323/feed.xml" rel="self" type="application/rss+xml" /><item><title>FORTAP: Using Formulas for Numerical-Reasoning-Aware Table Pretraining — mentioned by antidnan</title><link>https://news.ycombinator.com/item?id=39934965</link><guid isPermaLink="false">urn:hackerpapers:mention:2109.07323:39934965</guid><pubDate>Thu, 04 Apr 2024 19:47:19 GMT</pubDate><description>&lt;p&gt;There&amp;#x27;s also a book on the subject: https://en.wikipedia.org/wiki/The_History_of_Mathematical_Ta... Interesting aside: AI models trained on spreadsheets need &amp;quot;good tables&amp;quot; such as column names, headers, etc. to understand context. Like Fortap: https://arxiv.org/abs/2109.07323&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/2109.07323/"&gt;Paper and discussion history&lt;/a&gt;. A mention is not an endorsement.&lt;/p&gt;</description><category>cs.IR</category></item></channel></rss>
