<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Companion Systems | Yiyang "Diana" Wang</title><link>https://hello-diana.github.io/tags/companion-systems/</link><atom:link href="https://hello-diana.github.io/tags/companion-systems/index.xml" rel="self" type="application/rss+xml"/><description>Companion Systems</description><generator>Hugo Blox Builder (https://hugoblox.com)</generator><language>en-us</language><lastBuildDate>Tue, 20 Jan 2026 00:00:00 +0000</lastBuildDate><image><url>https://hello-diana.github.io/media/icon_hu_3520ea6f5cfedd63.png</url><title>Companion Systems</title><link>https://hello-diana.github.io/tags/companion-systems/</link></image><item><title>MASCOT: Towards Multi-Agent Socio-Collaborative Companion Systems</title><link>https://hello-diana.github.io/publication/mascot/</link><pubDate>Tue, 20 Jan 2026 00:00:00 +0000</pubDate><guid>https://hello-diana.github.io/publication/mascot/</guid><description>&lt;h2 id="abstract">Abstract&lt;/h2>
&lt;p>Multi-agent systems are emerging as promising socio-collaborative companions for emotional and cognitive support. However, existing systems frequently suffer from &lt;em>persona collapse&lt;/em>, where agents revert to generic assistant behaviors, and &lt;em>social sycophancy&lt;/em>, where agents produce redundant, non-constructive dialogue. We propose &lt;strong>MASCOT&lt;/strong>, a multi-agent framework for multi-perspective socio-collaborative companions, with a bi-level optimization strategy that harmonizes individual identities and collective discourse. MASCOT improves persona consistency by up to +14.1 and social contribution by up to +10.6 across in-domain and out-of-domain settings.&lt;/p></description></item></channel></rss>