<?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>Benchmarks on e4developer</title>
    <link>https://e4developer.com/tags/benchmarks/</link>
    <description>Recent content in Benchmarks on e4developer</description>
    <image>
      <title>e4developer</title>
      <url>https://e4developer.com/images/headshot.png</url>
      <link>https://e4developer.com/images/headshot.png</link>
    </image>
    <generator>Hugo</generator>
    <language>en-us</language>
    <lastBuildDate>Wed, 08 Jul 2026 12:00:00 +0000</lastBuildDate>
    <atom:link href="https://e4developer.com/tags/benchmarks/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>D20bench - Benchmarking LLMs with Dungeons &amp; Dragons Combat</title>
      <link>https://e4developer.com/posts/d20bench-benchmarking-llms-with-dungeons-and-dragons/</link>
      <pubDate>Wed, 08 Jul 2026 12:00:00 +0000</pubDate>
      <guid>https://e4developer.com/posts/d20bench-benchmarking-llms-with-dungeons-and-dragons/</guid>
      <description>How I turned a D&amp;amp;D combat engine into an LLM benchmark, why the models only looked stupid, and how a language model ended up beating my scripted AI. Open source, and you can help.</description>
    </item>
  </channel>
</rss>
