<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/"><channel><title>Local-Llm on zolty.systems</title><link>https://blog.zolty.systems/tags/local-llm/</link><description>Recent content in Local-Llm on zolty.systems</description><generator>Hugo</generator><language>en-us</language><lastBuildDate>Fri, 16 Oct 2026 10:00:00 -0400</lastBuildDate><atom:link href="https://blog.zolty.systems/tags/local-llm/index.xml" rel="self" type="application/rss+xml"/><item><title>Jev in 25 lines: building the decision model locally</title><link>https://blog.zolty.systems/posts/2026-10-16-jev-in-25-lines-local-logits/</link><pubDate>Fri, 16 Oct 2026 10:00:00 -0400</pubDate><guid>https://blog.zolty.systems/posts/2026-10-16-jev-in-25-lines-local-logits/</guid><description>Reading option-letter logits from a small local model gives you a Jev-shaped classifier for free. The sketch, the calibration catch, and why Ollama can&amp;#39;t do it.</description></item></channel></rss>