<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Lens on Thede Technologies</title><link>https://thedetech.com/tags/lens/</link><description>Recent content in Lens on Thede Technologies</description><generator>Hugo</generator><language>en-us</language><lastBuildDate>Fri, 16 Oct 2026 19:52:17 -0500</lastBuildDate><atom:link href="https://thedetech.com/tags/lens/index.xml" rel="self" type="application/rss+xml"/><item><title>Getting good output from Lens</title><link>https://thedetech.com/blog/2026-05-05-getting-good-output-from-lens/</link><pubDate>Tue, 05 May 2026 00:00:00 +0000</pubDate><guid>https://thedetech.com/blog/2026-05-05-getting-good-output-from-lens/</guid><description>&lt;p>&lt;figure class="side">&lt;img src="https://thedetech.com/images/blog/2026-05-05-getting-good-output-from-lens/mbp-garbled-output.png" alt="" width="1884" height="2648" loading="lazy" decoding="async">&lt;figcaption>A Lens daily summary derailed by reasoning artifacts: &lt;code>**People:**&lt;/code>, &lt;code>**Patterns:**&lt;/code>, &lt;code>**Drafting structure:**&lt;/code>, &amp;ldquo;Constraint Check&amp;rdquo;, bleeding into the narrative&lt;/figcaption>&lt;/figure>&lt;/p>
&lt;p>People pick up &lt;a href="https://tractorandsilo.com/lens" target="_blank" rel="noopener noreferrer">Lens&lt;/a>, connect it to LM Studio, point it at whatever model they downloaded first, and expect polished daily snapshots. Sometimes they get them. Sometimes the same setup on a different machine produces output that reads like the model thinking out loud, or worse, fabricating events that didn&amp;rsquo;t happen.&lt;/p>
&lt;p>Lens itself doesn&amp;rsquo;t change between those outcomes. The model, the quantization, and a few LM Studio settings do. This is the field guide I wish I&amp;rsquo;d shipped with the first build.&lt;/p></description></item></channel></rss>