<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Contextual Safety on Xinjie Shen</title><link>https://xinjie-shen.com/tag/contextual-safety/</link><description>Recent content in Contextual Safety on Xinjie Shen</description><generator>Hugo -- gohugo.io</generator><language>en-us</language><lastBuildDate>Mon, 09 Nov 2026 00:00:00 -0500</lastBuildDate><atom:link href="https://xinjie-shen.com/tag/contextual-safety/index.xml" rel="self" type="application/rss+xml"/><item><title>Shortcuts &amp; Boundaries: RL Misalignment and Contextual Safety</title><link>https://xinjie-shen.com/post/rlmisalignment/</link><pubDate>Mon, 09 Nov 2026 00:00:00 -0500</pubDate><guid>https://xinjie-shen.com/post/rlmisalignment/</guid><description>How agentic RL can improve real work while learning contextual violations, and how verifiable feedback supports useful, transferable safety.</description></item></channel></rss>