<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Tokens on Samarthya</title><link>https://samarthya.me/tags/tokens/</link><description>Recent content in Tokens on Samarthya</description><generator>Hugo -- gohugo.io</generator><language>en</language><copyright>© Samarthya 2018-2026. All rights reserved.</copyright><lastBuildDate>Mon, 26 Oct 2026 18:20:00 -0400</lastBuildDate><atom:link href="https://samarthya.me/tags/tokens/index.xml" rel="self" type="application/rss+xml"/><item><title>AI in practice, part 3: Tokens and tokenomics</title><link>https://samarthya.me/posts/ai-in-practice-3-tokens-and-tokenomics/</link><pubDate>Mon, 26 Oct 2026 18:20:00 -0400</pubDate><guid>https://samarthya.me/posts/ai-in-practice-3-tokens-and-tokenomics/</guid><description>&lt;div class="lead text-neutral-500 dark:text-neutral-400 !mb-9 text-xl"&gt;&#10; Language models don&amp;rsquo;t read words, and they don&amp;rsquo;t bill by the request. They read and charge in tokens. Once you&#10;can count them, both the model&amp;rsquo;s limits and the monthly invoice stop being surprises.&#10;&lt;/div&gt;&#10;&#10;&#10;&lt;h2 class="relative group"&gt;What a token is&#10; &lt;div id="what-a-token-is" class="anchor"&gt;&lt;/div&gt;&#10; &#10; &lt;span&#10; class="absolute top-0 w-6 transition-opacity opacity-0 -start-6 not-prose group-hover:opacity-100 select-none"&gt;&#10; &lt;a class="text-primary-300 dark:text-neutral-700 !no-underline" href="#what-a-token-is" aria-label="Anchor"&gt;#&lt;/a&gt;&#10; &lt;/span&gt;&#10; &#10;&lt;/h2&gt;&#10;&lt;p&gt;Before a model sees your text, a &lt;strong&gt;tokenizer&lt;/strong&gt; cuts it into pieces from a fixed vocabulary, typically tens of&#10;thousands to a few hundred thousand entries. Common words are one token; rarer words are split into&#10;pieces; spaces and punctuation are tokens too. Most modern tokenizers use variants of &lt;em&gt;byte-pair encoding&lt;/em&gt;,&#10;which learns the most frequent fragments from training text.&lt;/p&gt;</description></item><item><title>PAT: Personal Access Token</title><link>https://samarthya.me/posts/pat-personal-access-token/</link><pubDate>Fri, 24 Feb 2023 12:00:00 +0000</pubDate><guid>https://samarthya.me/posts/pat-personal-access-token/</guid><description>&lt;figure&gt;&lt;img&#10; class="my-0 rounded-md"&#10; loading="lazy"&#10; decoding="async"&#10; fetchpriority="auto"&#10; alt="Steps for a Git personal access token. Generate it, keep it safe, configure git, use the macOS keychain credential helper, then push"&#10; width="1200"&#10; height="630"&#10; src="https://samarthya.me/posts/pat-personal-access-token/an-lint_hu_23adda14bfe86d6e.png"&#10; srcset="https://samarthya.me/posts/pat-personal-access-token/an-lint_hu_23adda14bfe86d6e.png 800w, https://samarthya.me/posts/pat-personal-access-token/an-lint.png 1280w"&#10; sizes="(min-width: 768px) 50vw, 65vw"&#10; data-zoom-src="https://samarthya.me/posts/pat-personal-access-token/an-lint.png"&gt;&lt;figcaption&gt;Recreated in October 2026 from the post; the original image was lost&lt;/figcaption&gt;&lt;/figure&gt;&#10;&lt;p&gt;Using Git Personal Access Tokens (PATs) is a secure and convenient way to authenticate with Git service providers. In this blog post, we will discuss how to use Git PATs on both Mac and CentOS 7 systems.&lt;/p&gt;</description></item></channel></rss>