<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Differentiable Programming on Programmer.ie: Modern AI programming</title>
    <link>http://programmer.ie/tags/differentiable-programming/</link>
    <description>Recent content in Differentiable Programming on Programmer.ie: Modern AI programming</description>
    <generator>Hugo</generator>
    <language>en-us</language>
    <lastBuildDate>Mon, 10 Aug 2026 20:31:00 +0100</lastBuildDate>
    <atom:link href="http://programmer.ie/tags/differentiable-programming/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Cellular Automata From First Principles 36: Make the Automaton Differentiable</title>
      <link>http://programmer.ie/post/cellular-automata-from-first-principles-36/</link>
      <pubDate>Mon, 10 Aug 2026 20:31:00 +0100</pubDate>
      <guid>http://programmer.ie/post/cellular-automata-from-first-principles-36/</guid>
      <description>&lt;h1 id=&#34;cellular-automata-from-first-principles-36-make-the-automaton-differentiable&#34;&gt;Cellular Automata From First Principles 36: Make the Automaton Differentiable&lt;/h1&gt;&#xA;&lt;p&gt;So far every rule in this book has been chosen by us.&lt;/p&gt;&#xA;&lt;p&gt;Even Lenia still asks us to decide the neighborhood kernel, the growth function and the parameters that connect them.&lt;/p&gt;&#xA;&lt;p&gt;What if we stop designing the local rule directly?&lt;/p&gt;&#xA;&lt;p&gt;What if we define only the &lt;strong&gt;goal&lt;/strong&gt;, then let gradient descent discover a local update rule that achieves it?&lt;/p&gt;&#xA;&lt;p&gt;That requires one major change:&lt;/p&gt;</description>
    </item>
  </channel>
</rss>
