<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/">
  <channel>
    <title>Simon&#39;s Journal</title>
    <link>https://bearisland.dev/</link>
    <description>Recent content on Simon&#39;s Journal</description>
    <generator>Hugo</generator>
    <language>en</language>
    <lastBuildDate>Sun, 07 Jun 2026 00:00:00 +0000</lastBuildDate>
    <atom:link href="https://bearisland.dev/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Pre-training: An Overview</title>
      <link>https://bearisland.dev/posts/pretraining-overview/</link>
      <pubDate>Sat, 06 Jun 2026 00:00:00 +0000</pubDate>
      <guid>https://bearisland.dev/posts/pretraining-overview/</guid>
      <description>A flyover of the pre-training pipeline. Text in, trained model out. Each stage gets its own deep dive in the rest of the series.</description>
    </item>
    <item>
      <title>Tokens and Tokenization</title>
      <link>https://bearisland.dev/posts/tokens-and-tokenization/</link>
      <pubDate>Sun, 07 Jun 2026 00:00:00 +0000</pubDate>
      <guid>https://bearisland.dev/posts/tokens-and-tokenization/</guid>
      <description>How LLMs split text into tokens, the BPE algorithm, and why &amp;#39;strawberry&amp;#39; has 3 r&amp;#39;s the model can&amp;#39;t see.</description>
    </item>
  </channel>
</rss>
