<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
<channel>
  <title>MATRIX · SenseNova-Loop-9B-Preview</title>
  <link>https://derekwang2002.github.io/ai-intelligence/en/projects/sensenova-loop-9b/</link>
  <description>Material changes in AI models, agents, engineering and infrastructure, with primary sources and recommendations.</description>
  <language>en</language>
  <lastBuildDate>Mon, 02 Nov 2026 16:56:00 GMT</lastBuildDate>
  <atom:link href="https://derekwang2002.github.io/ai-intelligence/en/feeds/projects/sensenova-loop-9b.xml" rel="self" type="application/rss+xml"/>
  <item>
    <title>SenseNova opens a looped multimodal model for long-context agents</title>
    <link>https://derekwang2002.github.io/ai-intelligence/en/events/ev-20261030-04/</link>
    <guid isPermaLink="false">ev-20261030-04-published</guid>
    <pubDate>Sat, 31 Oct 2026 15:56:40 GMT</pubDate>
    <category>New</category>
    <category>TRIAL</category>
    <description>&lt;p&gt;SenseNova released ungated Apache-2.0 weights and code for SenseNova-Loop-9B-Preview, a text-and-image model for long-context and agentic workflows. It applies a 24-layer block twice, producing 56 layer executions from 32 distinct decoder layers, and was trained on sequences up to 256K tokens. Deployment currently relies on LightLLM main or an unmerged SGLang pull request.&lt;/p&gt;&lt;p&gt;Recommendation: TRIAL&lt;/p&gt;&lt;p&gt;Sources: &lt;a href=&quot;https://huggingface.co/SenseTime/SenseNova-Loop-9B-Preview&quot;&gt;SenseNova-Loop-9B-Preview model card&lt;/a&gt; · &lt;a href=&quot;https://github.com/SenseTime-FVG/OpenSenseNova&quot;&gt;OpenSenseNova repository&lt;/a&gt;&lt;/p&gt;</description>
  </item>
</channel>
</rss>
