<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>llm on Digitable Courses</title>
    <link>https://courses.digitable.life/en/tags/llm/</link>
    <description>Recent content in llm on Digitable Courses</description>
    <generator>Hugo -- gohugo.io</generator>
    <language>en-US</language>
    <lastBuildDate>Tue, 01 Sep 2026 10:00:00 +0300</lastBuildDate><atom:link href="https://courses.digitable.life/en/tags/llm/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Nine measurements of prompting and an agent cluster: where a technique gives &#43;48 points and where nothing can be told apart</title>
      <link>https://courses.digitable.life/en/post/prompting-study/</link>
      <pubDate>Tue, 01 Sep 2026 10:00:00 +0300</pubDate>
      
      <guid>https://courses.digitable.life/en/post/prompting-study/</guid>
      <description>Nine of our measurements in a row: eight well-known prompting techniques against a plain prompt, a detailed brief against a short one, a cluster of agents against a single one. Self-consistency gives +28.8 points on arithmetic and +48.0 on tasks that were in nobody&amp;rsquo;s training set; on code the techniques only do harm; on recall questions seven techniques out of eight lose. Three measurements ended in an honest “cannot be told apart”, one was never finished. Everything is brought together here — together with the list of what we did not measure.</description>
    </item>
    
    <item>
      <title>Everything an IT professional needs to know about AI: algorithms, prompting, and reasoning.</title>
      <link>https://courses.digitable.life/en/video/ai-basics/</link>
      <pubDate>Sun, 31 May 2026 15:58:07 +0300</pubDate>
      
      <guid>https://courses.digitable.life/en/video/ai-basics/</guid>
      <description>In this video:</description>
    </item>
    
  </channel>
</rss>
