<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Computer-Use Agents | Yiyang "Diana" Wang</title><link>https://hello-diana.github.io/tags/computer-use-agents/</link><atom:link href="https://hello-diana.github.io/tags/computer-use-agents/index.xml" rel="self" type="application/rss+xml"/><description>Computer-Use Agents</description><generator>Hugo Blox Builder (https://hugoblox.com)</generator><language>en-us</language><lastBuildDate>Fri, 16 Oct 2026 00:00:00 +0000</lastBuildDate><image><url>https://hello-diana.github.io/media/icon_hu_9f6d321bccb06734.png</url><title>Computer-Use Agents</title><link>https://hello-diana.github.io/tags/computer-use-agents/</link></image><item><title>PersonaCUABench: Benchmarking Computer-Use Agents Serving Diverse Personas</title><link>https://hello-diana.github.io/publication/personacuabench/</link><pubDate>Fri, 16 Oct 2026 00:00:00 +0000</pubDate><guid>https://hello-diana.github.io/publication/personacuabench/</guid><description>&lt;p&gt;&lt;strong&gt;PersonaCUABench&lt;/strong&gt; evaluates computer-use agents through the people they serve. It holds tasks fixed while varying user personas, letting agents interact with requests that may omit requirements, add constraints, or call for approval before consequential actions.&lt;/p&gt;
&lt;p&gt;The benchmark distinguishes task completion from meeting a user&amp;rsquo;s expressed needs, examining both agent performance and user satisfaction.&lt;/p&gt;</description></item></channel></rss>