<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Compute on Neat Guy Coding</title><link>https://neatguycoding.com/tags/compute/</link><description>Recent content in Compute on Neat Guy Coding</description><generator>Hugo -- gohugo.io</generator><language>en</language><copyright>© 2026 NeatGuyCoding</copyright><lastBuildDate>Mon, 18 May 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://neatguycoding.com/tags/compute/index.xml" rel="self" type="application/rss+xml"/><item><title>Judge-Time Compute: When LLM Evaluation Moves from a Single Score to a Composable Pipeline</title><link>https://neatguycoding.com/posts/2026-05-18-weaviate-podcast-haize-labs-with-leonard-tang-weaviate-podcast-121/</link><pubDate>Mon, 18 May 2026 00:00:00 +0000</pubDate><guid>https://neatguycoding.com/posts/2026-05-18-weaviate-podcast-haize-labs-with-leonard-tang-weaviate-podcast-121/</guid><description>Judge-time compute: stacking structured, composable weak-model calls at evaluation time instead of assuming one expensive judge pass is enough—Verdict, agreement metrics, and production guardrails, with evidence boundaries called out.</description><media:content xmlns:media="http://search.yahoo.com/mrss/" url="https://neatguycoding.com/posts/2026-05-18-weaviate-podcast-haize-labs-with-leonard-tang-weaviate-podcast-121/cover.png"/></item></channel></rss>