<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Least Squares on Programmer.ie: Modern AI programming</title>
    <link>http://programmer.ie/tags/least-squares/</link>
    <description>Recent content in Least Squares on Programmer.ie: Modern AI programming</description>
    <generator>Hugo</generator>
    <language>en-us</language>
    <lastBuildDate>Mon, 07 Sep 2026 14:00:00 +0000</lastBuildDate>
    <atom:link href="http://programmer.ie/tags/least-squares/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Can One Embedding Space Be Translated Into Another?</title>
      <link>http://programmer.ie/books/embeddings-from-first-principles/18-chapter/</link>
      <pubDate>Mon, 07 Sep 2026 14:00:00 +0000</pubDate>
      <guid>http://programmer.ie/books/embeddings-from-first-principles/18-chapter/</guid>
      <description>&lt;p&gt;&lt;em&gt;Part VI — Crossing Embedding Spaces&lt;/em&gt;&lt;/p&gt;&#xA;&lt;h2 id=&#34;the-question-stated-carefully&#34;&gt;The question, stated carefully&lt;/h2&gt;&#xA;&lt;p&gt;We have two encoders, &lt;code&gt;E_A&lt;/code&gt; and &lt;code&gt;E_B&lt;/code&gt;, and a set of objects &lt;code&gt;x&lt;/code&gt;. Each object has two representations: &lt;code&gt;E_A(x)&lt;/code&gt; in space A, &lt;code&gt;E_B(x)&lt;/code&gt; in space B. Chapter 16 established these live in unrelated coordinate systems.&lt;/p&gt;&#xA;&lt;p&gt;The question:&lt;/p&gt;&#xA;&lt;blockquote&gt;&#xA;&lt;p&gt;&lt;strong&gt;Does there exist a map &lt;code&gt;T&lt;/code&gt; such that &lt;code&gt;T(E_A(x)) ≈ E_B(x)&lt;/code&gt; for objects &lt;code&gt;x&lt;/code&gt; the map was never trained on — and by what standard do we judge &amp;ldquo;≈&amp;rdquo;?&lt;/strong&gt;&lt;/p&gt;</description>
    </item>
  </channel>
</rss>
