<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
<channel><title>Dolittle writing</title><link>https://dolittle.plainsight-systems.com/writing/</link><description>Explanations, explorations, and milestones from building Dolittle.</description><language>en-us</language><atom:link href="https://dolittle.plainsight-systems.com/writing/index.xml" rel="self" type="application/rss+xml" />

<item><title>Two views of a thinking model</title><link>https://dolittle.plainsight-systems.com/writing/two-views-of-a-thinking-model/</link><guid>https://dolittle.plainsight-systems.com/writing/two-views-of-a-thinking-model/</guid><pubDate>Fri, 16 Oct 2026 09:00:00 &#43;0000</pubDate><description>What generated thinking tells us, what an activation verbalizer adds, and how to read the places where they disagree.</description><category>explanation</category></item>

<item><title>Choosing a model we can run</title><link>https://dolittle.plainsight-systems.com/writing/choosing-a-model-we-can-run/</link><guid>https://dolittle.plainsight-systems.com/writing/choosing-a-model-we-can-run/</guid><pubDate>Fri, 16 Oct 2026 08:00:00 &#43;0000</pubDate><description>Comparing two Qwen3/NLA pairs on a 3090 Ti, and why the 8B pair is the starting point for Dolittle.</description><category>exploration</category></item>
</channel></rss>
