<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/"><channel><title>GRPO on Xu'Blog</title><link>https://xuquant.com/tags/grpo/</link><description>Recent content in GRPO on Xu'Blog</description><image><title>Xu'Blog</title><url>https://xuquant.com/og-default.png</url><link>https://xuquant.com/og-default.png</link></image><generator>Hugo -- 0.152.2</generator><language>zh</language><lastBuildDate>Thu, 21 May 2026 20:00:00 +0800</lastBuildDate><atom:link href="https://xuquant.com/tags/grpo/index.xml" rel="self" type="application/rss+xml"/><item><title>ATLAS：视觉推理的动作词表</title><link>https://xuquant.com/posts/foundation-models/atlas-one-word-visual-reasoning/</link><pubDate>Thu, 21 May 2026 20:00:00 +0800</pubDate><guid>https://xuquant.com/posts/foundation-models/atlas-one-word-visual-reasoning/</guid><description>解读 ATLAS: Agentic or Latent Visual Reasoning? One Word is Enough for Both：把画辅助线、框选区域、箭头指示、文本标注等中间视觉操作压缩成可训练的 functional tokens。</description></item></channel></rss>