<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>预训练模型 on AI内参</title>
    <link>https://www.neican.ai/tags/%E9%A2%84%E8%AE%AD%E7%BB%83%E6%A8%A1%E5%9E%8B/</link>
    <description>Recent content in 预训练模型 on AI内参</description>
    <generator>Hugo</generator>
    <language>zh-cn</language>
    <lastBuildDate>Tue, 26 May 2026 11:40:03 +0800</lastBuildDate>
    <atom:link href="https://www.neican.ai/tags/%E9%A2%84%E8%AE%AD%E7%BB%83%E6%A8%A1%E5%9E%8B/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>效率的范式转移：HRM-Text 如何重构大模型预训练的“贫富差距”</title>
      <link>https://www.neican.ai/insights/hrm-text--20260526114003388-2/</link>
      <pubDate>Tue, 26 May 2026 11:40:03 +0800</pubDate>
      <guid>https://www.neican.ai/insights/hrm-text--20260526114003388-2/</guid>
      <description>HRM-Text 通过引入分层循环模型架构，极大地降低了预训练算力开销，打破了Scaling Law下的资源垄断。该技术预示着AI开发将从“暴力计算”转向“架构优化”，为高性能模型的平民化与边缘部署提供了新的路径。</description>
    </item>
  </channel>
</rss>
