<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Agent-Training on skulldc</title><link>https://skulldc.net/tags/agent-training/</link><description>Recent content in Agent-Training on skulldc</description><generator>Hugo</generator><language>en-us</language><lastBuildDate>Mon, 09 Nov 2026 10:24:46 +0000</lastBuildDate><atom:link href="https://skulldc.net/tags/agent-training/index.xml" rel="self" type="application/rss+xml"/><item><title>Harness-S trains one agent policy on rescues from specialised sibling harnesses</title><link>https://skulldc.net/2026/11/harness-s-trains-one-agent-policy-on-rescues-from-specialised-sibling/</link><pubDate>Mon, 09 Nov 2026 10:24:46 +0000</pubDate><guid>https://skulldc.net/2026/11/harness-s-trains-one-agent-policy-on-rescues-from-specialised-sibling/</guid><description>A paper proposes calling a fixed pool of specialised agent harnesses only when the base harness fails, then training the policy on those rescued rollouts so it improves under the original harness.</description></item></channel></rss>