<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>The Elements of Differentiable Programming · HackerPapers discussions</title><link>https://hackerpaper.kaipfister.de/p/2403.14606/</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/2403.14606/feed.xml" rel="self" type="application/rss+xml" /><item><title>The Elements of Differentiable Programming — mentioned by mswphd</title><link>https://news.ycombinator.com/item?id=48832517</link><guid isPermaLink="false">urn:hackerpapers:mention:2403.14606:48832517</guid><pubDate>Wed, 08 Jul 2026 14:33:40 GMT</pubDate><description>&lt;p&gt;note that similar concepts appear in mathematics. Generally the term for it is a &amp;quot;mollified&amp;quot; function. applied to the step function, you would get a smooth cutoff function https://en.wikipedia.org/wiki/Mollifier#Smooth_cutoff_functi... this is also related somewhat to the notion of differentiable programming. RELU is (roughly) the same as x * step(x). In differentiable programming one can replace it with smooth approximations, cf &amp;quot;softplus&amp;quot; https://arxiv.org/pdf/2403.14606 That book also has a chapter on control flow, which is very similar to what you&amp;#x27;re talking about. Unrolling an if…&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/2403.14606/"&gt;Paper and discussion history&lt;/a&gt;. A mention is not an endorsement.&lt;/p&gt;</description><category>cs.LG</category></item><item><title>The Elements of Differentiable Programming — HN submission</title><link>https://news.ycombinator.com/item?id=39793191</link><guid isPermaLink="false">urn:hackerpapers:submission:2403.14606:39793191</guid><pubDate>Fri, 22 Mar 2024 18:08:18 GMT</pubDate><description>&lt;p&gt;The Elements of Differentiable Programming&lt;/p&gt;&lt;p&gt;&lt;/p&gt;&lt;p&gt;131 points; 77 discussion replies on HN. &lt;a href="https://hackerpaper.kaipfister.de/p/2403.14606/"&gt;Paper and source history&lt;/a&gt;. A submission is not an endorsement.&lt;/p&gt;</description><category>cs.LG</category></item></channel></rss>
