<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Anthropic on Yushi's Blog</title><link>https://blog.yushi91.com/blog/anthropic/</link><description>Recent content in Anthropic on Yushi's Blog</description><generator>Hugo</generator><language>en-US</language><copyright>Copyright © 2025, Yushi Cui.</copyright><lastBuildDate>Thu, 09 Apr 2026 10:00:00 +1300</lastBuildDate><atom:link href="https://blog.yushi91.com/blog/anthropic/index.xml" rel="self" type="application/rss+xml"/><item><title>Prompt Engineering Is Not About Memorizing Tricks</title><link>https://blog.yushi91.com/blog/prompt-engineering-evaluation/</link><pubDate>Thu, 09 Apr 2026 10:00:00 +1300</pubDate><guid>https://blog.yushi91.com/blog/prompt-engineering-evaluation/</guid><description>&lt;p>&lt;img src="https://blog.yushi91.com/prompt-engineering-evaluation.webp" alt="Prompt engineering evaluation" loading="lazy">
&lt;/p>
&lt;h2 id="the-black-box-problem">The black box problem&lt;/h2>
&lt;p>Everyone knows prompts matter. You&amp;rsquo;ve seen the Twitter threads. The Reddit posts. The &amp;ldquo;here&amp;rsquo;s my secret formula&amp;rdquo; guides. I&amp;rsquo;ve read most of them, and I actually learned from them. Step-by-step instructions instead of vague goals. A few examples so the model knows what you expect. A persona if it helps frame the task. These techniques are real and I use them daily.&lt;/p>
&lt;p>But I was always missing something: I had no idea if I was getting better.&lt;/p></description></item></channel></rss>