<?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>N-Grams on Cognaptus</title>
    <link>https://cognaptus.com/tags/n-grams/</link>
    <description>Recent content in N-Grams on Cognaptus</description>
    <generator>Hugo -- 0.145.0</generator>
    <language>en-us</language>
    <lastBuildDate>Thu, 29 Oct 2026 00:00:00 +0000</lastBuildDate>
    <atom:link href="https://cognaptus.com/tags/n-grams/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>COUNT on RAM: A Different Way to Scale Language Models</title>
      <link>https://cognaptus.com/blog/2026-10-29-count-on-ram-a-different-way-to-scale-language-models/</link>
      <pubDate>Thu, 29 Oct 2026 00:00:00 +0000</pubDate>
      <guid>https://cognaptus.com/blog/2026-10-29-count-on-ram-a-different-way-to-scale-language-models/</guid>
      <description>COUNT shows how frozen corpus statistics stored in host RAM can improve transformer performance and support domain adaptation without additional learned n-gram embeddings.</description>
    </item>
  </channel>
</rss>
