<?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>Technical Research on Matthieu</title>
    <link>https://matthieuMoullec.com/research/technical/</link>
    <description>Recent content in Technical Research on Matthieu</description>
    <generator>Hugo -- gohugo.io</generator>
    <language>en</language>
    <managingEditor>Matthieu</managingEditor>
    <lastBuildDate>Sun, 28 Jun 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://matthieuMoullec.com/research/technical/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Feature Learning in Deep Learning</title>
      <link>https://matthieuMoullec.com/research/technical/feature-learning-deep-learning/</link>
      <pubDate>Sun, 28 Jun 2026 00:00:00 +0000</pubDate>
      <author>Matthieu</author>
      <guid>https://matthieuMoullec.com/research/technical/feature-learning-deep-learning/</guid>
      <description>A research project on how deep learning models discover, organize, and use internal features.</description>
    </item>
    
    <item>
      <title>Semantic Geometry of Sentence Embeddings</title>
      <link>https://matthieuMoullec.com/research/technical/semantic-geometry-sentence-embeddings/</link>
      <pubDate>Thu, 25 Jun 2026 00:00:00 +0000</pubDate>
      <author>Matthieu</author>
      <guid>https://matthieuMoullec.com/research/technical/semantic-geometry-sentence-embeddings/</guid>
      <description>A framework for identifying interpretable feature directions and semantic regions in sentence embedding spaces.</description>
    </item>
    
    <item>
      <title>The Canonical Representation of a Task</title>
      <link>https://matthieuMoullec.com/research/technical/canonical-representation-task/</link>
      <pubDate>Wed, 24 Jun 2026 00:00:00 +0000</pubDate>
      <author>Matthieu</author>
      <guid>https://matthieuMoullec.com/research/technical/canonical-representation-task/</guid>
      <description>A theory of generalization in modular arithmetic tasks through task-derived canonical representations.</description>
    </item>
    
    <item>
      <title>The Representation of Space in LLMs</title>
      <link>https://matthieuMoullec.com/research/technical/representation-of-space-in-llms/</link>
      <pubDate>Tue, 23 Jun 2026 00:00:00 +0000</pubDate>
      <author>Matthieu</author>
      <guid>https://matthieuMoullec.com/research/technical/representation-of-space-in-llms/</guid>
      <description>A research note on how large language models may encode spatial structure and spatial relations.</description>
    </item>
    
    <item>
      <title>Canonical Representations of Language Representations</title>
      <link>https://matthieuMoullec.com/research/technical/canonical-representations-language-representations/</link>
      <pubDate>Mon, 22 Jun 2026 00:00:00 +0000</pubDate>
      <author>Matthieu</author>
      <guid>https://matthieuMoullec.com/research/technical/canonical-representations-language-representations/</guid>
      <description>A research project on task-derived structure and canonical forms in learned language representations.</description>
    </item>
    
  </channel>
</rss>
