<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Metacognition on ROBOCO</title>
    <link>https://roboco.io/en/tags/metacognition/</link>
    <description>Recent content in Metacognition on ROBOCO</description>
    <generator>Hugo</generator>
    <language>en-us</language>
    <lastBuildDate>Thu, 27 Nov 2025 07:57:49 +0900</lastBuildDate>
    <atom:link href="https://roboco.io/en/tags/metacognition/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>LLM Prompting: Persona Assignment vs. Metacognition</title>
      <link>https://roboco.io/en/posts/persona-vs-metacognition/</link>
      <pubDate>Thu, 27 Nov 2025 07:57:49 +0900</pubDate>
      <guid>https://roboco.io/en/posts/persona-vs-metacognition/</guid>
      <description>&lt;blockquote&gt;&#xA;&lt;p&gt;On walking the tightrope between confidence and caution&lt;/p&gt;&#xA;&lt;/blockquote&gt;&#xA;&lt;figure&gt;&lt;img src=&#34;https://roboco.io/posts/images/Dohyun.png&#34;&gt;&lt;figcaption&gt;&#xA;      &lt;h4&gt;Dohyun Jung - Principal Consultant, ROBOCO&lt;/h4&gt;&#xA;    &lt;/figcaption&gt;&#xA;&lt;/figure&gt;&#xA;&#xA;&lt;h2 id=&#34;introduction&#34;&gt;Introduction&lt;/h2&gt;&#xA;&lt;p&gt;What is the difference between telling a model &amp;ldquo;answer like a senior engineer&amp;rdquo; and telling it &amp;ldquo;first think about what you are, then answer&amp;rdquo;? On the surface they look similar, but internally two completely different mechanisms are at work. This post examines the mathematical difference between the two approaches and what it means in practice.&lt;/p&gt;</description>
    </item>
  </channel>
</rss>
