<!DOCTYPE html><html lang="en"> <head> <meta charset="utf-8"> <meta name="viewport" content="width=device-width initial-scale=1" /> <meta http-equiv="X-UA-Compatible" content="IE=edge"> <meta name="author" content="Sebastian Raschka"> <meta property="og:title" content=" Sebastian Raschka "> <meta property="og:description" content=" I'm an LLM Research Engineer with over a decade of experience in artificial intelligence. My work bridges academia and industry, with roles including senior staff at an AI company and a statistics professor. My expertise lies in LLM research and the development of high-performance AI systems, with a deep focus on practical, code-driven implementations. "> <meta property="og:url" content="https://sebastianraschka.com/" /> <meta property="og:site_name" content="Sebastian Raschka, PhD" /> <meta property="og:locale" content="en_US" /> <meta property="og:type" content="article" /> <meta property="article:published_time" content="" /> <meta property="og:image" content="https://sebastianraschka.com/images/logos/hero.jpg" /> <meta name="twitter:site" content="@rasbt" /> <meta name="twitter:creator" content="@rasbt" /> <meta name="twitter:title" content="Sebastian Raschka" /> <meta name="twitter:description" content=" I'm an LLM Research Engineer with over a decade of experience in artificial intelligence. My work bridges academia and industry, with roles including senior staff at an AI company and a statistics professor. My expertise lies in LLM research and the development of high-performance AI systems, with a deep focus on practical, code-driven implementations. " /> <meta name="twitter:card" content="summary_large_image"> <meta name="twitter:image" content="https://sebastianraschka.com/images/logos/hero.jpg"> <title>Sebastian Raschka</title> <meta name="description" content="I'm an LLM Research Engineer with over a decade of experience in artificial intelligence. My work bridges academia and industry, with roles including senior staff at an AI company and a statistics professor. My expertise lies in LLM research and the development of high-performance AI systems, with a deep focus on practical, code-driven implementations."> <link rel="stylesheet" href="/css/combined_direct_no_sass.css"> <link rel="stylesheet" href="/css/fork-awesome.min.css"> <meta property='og:title' content="Sebastian Raschka"><meta property="og:type" content="article"><meta property="og:url" content="https://sebastianraschka.com/"> <meta property="og:image" content=""> <meta property="og:description" content=""><meta property="og:site_name" content="Sebastian Raschka, PhD"><meta property="og:locale" content="en_US"> <meta property="fb:admins" content=""><meta property="fb:app_id" content=""> <link rel="canonical" href="https://sebastianraschka.com/"> <link rel="icon" type="image/png" sizes="32x32" href="/images/favicons/favicon-32x32.png"> <link rel="icon" type="image/png" sizes="16x16" href="/images/favicons/favicon-16x16.png"> <link rel="manifest" href="/site.webmanifest"> <link rel="mask-icon" href="/images/favicons/safari-pinned-tab.svg" color="#5bbad5"> <meta name="msapplication-TileColor" content="#ffc40d"> <meta name="theme-color" content="#ffffff"> <body> <img src="/images/logos/ahead-of-ai-icon.png" alt="Ahead of AI logo" style="display: none;"> <header class="site-header"> <div class="site-title" style="text-decoration: none; margin-top: 2em;"> <a href="/"><span style="color:black">Sebastian</span> <span style="color:#c5050c">Raschka</span> </a> <a href="https://x.com/rasbt"><img src="/images/logos/twitter-bw.jpg" height="20" style="padding-left:20px;" alt="Twitter/X icon"></a> <!--<a href="https://threads.net/@sebastianraschka"><img src="/images/logos/threads-logo-alt-small.png" height="20" style="padding-left:5px;" alt="Threads icon"></a>--> <a href="https://www.linkedin.com/in/sebastianraschka/"><img src="/images/logos/linkedin-bw.jpg" height="20" style="padding-left:5px;" alt="LinkedIn Icon"></a> <a href="https://github.com/rasbt"><img src="/images/logos/github-bw.jpg" height="20" style="padding-left:5px;" alt="GitHub icon"></a> </div> <!-- <div style="width:100%;height:50;float:left;margin-bottom:10px;"> <a href="https://twitter.com/rasbt"><img src="/images/logos/twitter-bw.jpg" height="20"></a>--> <!-- <a href="https://www.buymeacoffee.com/rasbt"><img src="/images/logos/coffee-bw.jpg" height="20"></a>--> <!-- <a href="https://mastodon.social/@SebRaschka"><img src="/images/logos/mastodon-bw.jpg" height="20"></a>--> <!-- </div>--> <div class="wrapper"> <nav class="site-nav"> <a href="#" class="menu-icon"> <svg viewBox="0 0 18 15"> <path fill="#424242" d="M18,1.484c0,0.82-0.665,1.484-1.484,1.484H1.484C0.665,2.969,0,2.304,0,1.484l0,0C0,0.665,0.665,0,1.484,0 h15.031C17.335,0,18,0.665,18,1.484L18,1.484z"/> <path fill="#424242" d="M18,7.516C18,8.335,17.335,9,16.516,9H1.484C0.665,9,0,8.335,0,7.516l0,0c0-0.82,0.665-1.484,1.484-1.484 h15.031C17.335,6.031,18,6.696,18,7.516L18,7.516z"/> <path fill="#424242" d="M18,13.516C18,14.335,17.335,15,16.516,15H1.484C0.665,15,0,14.335,0,13.516l0,0 c0-0.82,0.665-1.484,1.484-1.484h15.031C17.335,12.031,18,12.696,18,13.516L18,13.516z"/> </svg> </a> <div class="trigger"> <!--<script type="text/javascript"> var total_images = 2; var random_number = Math.floor((Math.random()*total_images)); var random_img = new Array(); random_img[0] = '<a href="https://twitter.com/rasbt"><img src="/images/logos/twitter-1.jpg" height="20"></a>'; random_img[1] = '<a href="https://linkedin.com/in/sebastianraschka"><img src="/images/logos/linkedin-1.jpg" height="20"></a>'; document.write(random_img[random_number]); </script>--> <span style="padding-left:0px;margin-left:0px;"></span> <a class="page-link" href="https://magazine.sebastianraschka.com"><span style="color:#c5050c;"><img src="/images/logos/ahead-of-ai-icon.png" alt="Ahead of AI Logo" height="20"> Blog</span></a> <!--<a class="page-link" href="/blog/index.html">Blog</a>--> <a class="page-link" href="/books">Books</a> <!--<a class="page-link" href="/newsletter">AI Newsletter</a>--> <a class="page-link" href="/teaching">Courses</a> <a class="page-link" href="https://github.com/rasbt/LLMs-from-scratch">LLMs From Scratch</a> <!--<a class="page-link" href="/publications">Research</a>--> <a class="page-link" href="/elsewhere">Talks</a> <a class="page-link" href="/contact">Contact</a> <a class="page-link" href="/resources">More</a> </div> </nav> </div> </header> <div class="page-content"> <div class="wrapper"> <div style="display: flex; align-items: center; justify-content: center; flex-wrap: wrap; max-width: 800px; margin: 40px auto; gap: 30px;"> <div style="flex: 1; min-width: 300px; text-align: left;"> <h1 style="font-size: 1.8em; font-weight: 600; color: #222; margin-bottom: 20px;"> Hello, I'm Sebastian Raschka, PhD </h1> <p style="font-size: 1.1em; line-height: 1.8; color: #555; margin-bottom: 20px;"> I'm an LLM Research Engineer with over a decade of experience in artificial intelligence. My work bridges academia and industry, with roles including senior staff at an AI company and a statistics professor. </p> <p style="font-size: 1.1em; line-height: 1.8; color: #555; margin-bottom: 20px;"> My expertise lies in LLM research and the development of high-performance AI systems, with a deep focus on practical, code-driven implementations. (For my most up-to-date CV details, please visit my <a href="https://www.linkedin.com/in/sebastianraschka">LinkedIn profile</a>.) </p> </div> <div style="flex: 0 0 200px; text-align: center;"> <picture> <source type="image/webp" srcset="/images/logos/photo-2021-08-25_compressed.webp"> <source type="image/jpeg" srcset="/images/logos/photo-2021-08-25_compressed.jpg"> <img src="/images/logos/photo-2021-08-25_compressed.jpg" alt="Sebastian Raschka Profile Picture" style="width: 200px; height: 200px; object-fit: cover; object-position: 100% 10%; border-radius: 50%; box-shadow: 0 4px 10px rgba(0,0,0,0.1);"> </picture> </div></div> <!-- Subscribe Section --><div class="rss" style="display: flex; align-items: center; justify-content: center; width: 96%; margin: 20px auto; max-width: 800px;"> <a href="/rss_feed.xml" style="display: inline-flex; align-items: center; margin-right: 20px;"> <i class="fa fa-rss fa-1x" style="padding-right: 5px;"></i> <span>RSS Feed (Blog and Notes)</span> </a> <a class="page-link" href="https://magazine.sebastianraschka.com/subscribe" style="display: inline-flex; align-items: center;"> <img src="/images/logos/ahead-of-ai-icon.png" alt="Ahead of AI Logo" height="20" style="padding-right: 5px;"> <span>Subscribe via Email (AI Magazine)</span> </a></div> <div style="height: 50px;"></div> <!-- Blog Entries Section --><div style="max-width: 800px; margin: 30px auto;"> <blockquote> <!-- Content about Sebastian --> </blockquote> <div class="post"> <header class="post-header"> <h1 class="post-title">Recent Notes and Blog Entries</h1> </header> <div style="margin-bottom: 45px; text-align: center;"> See <a href="/blog/">Blog and Notes Archive</a> for all entries. </div> <article class="post-content posts-by-year"> <div class="post-entry"> <div class="post-hero-image"> <img src="/images/blog/2025/from-gpt-2-to-gpt-oss/hero.jpg" alt="From GPT-2 to gpt-oss: Analyzing the Architectural Advances"> </div> <div class="post-details"> <h2 class="post-title"> <a href="/blog/2025/from-gpt-2-to-gpt-oss.html">From GPT-2 to gpt-oss: Analyzing the Architectural Advances</a> </h2> <p class="post-date">Aug 9, 2025</p> <p class="post-description">OpenAI just released their new open-weight LLMs this week: gpt-oss-120b and gpt-oss-20b, their first open-weight models since GPT-2 in 2019. And yes, thanks to some clever optimizations, they can run locally. I spent the past few days reading through the code and technical reports to summarize the most interesting details.</p> </div> </div> <div class="post-entry"> <div class="post-hero-image"> <img src="/images/blog/2025/the-big-llm-architecture-comparison/hero.jpg" alt="The Big LLM Architecture Comparison"> </div> <div class="post-details"> <h2 class="post-title"> <a href="/blog/2025/the-big-llm-architecture-comparison.html">The Big LLM Architecture Comparison</a> </h2> <p class="post-date">Jul 19, 2025</p> <p class="post-description">It has been seven years since the original GPT architecture was developed. At first glance, looking back at GPT-2 (2019) and forward to DeepSeek-V3 and Llama 4 (2024-2025), one might be surprised at how structurally similar these models still are. Comparing LLMs to determine the key ingredients that contribute to their good (or not-so-good) performance is notoriously challenging: datasets, training techniques, and hyperparameters vary widely and are often not well documented. However, I think that there is still a lot of value in examining the structural changes of the architectures themselves to see what LLM developers are up to in 2025.</p> </div> </div> <div class="post-entry"> <div class="post-hero-image"> <img src="/images/blog/2025/llm-research-papers-the-2025-list-january-to-june/hero.jpeg" alt="LLM Research Papers: The 2025 List (January to June)"> </div> <div class="post-details"> <h2 class="post-title"> <a href="/blog/2025/llm-research-papers-the-2025-list-january-to-june.html">LLM Research Papers: The 2025 List (January to June)</a> </h2> <p class="post-date">Jul 1, 2025</p> <p class="post-description">The latest in LLM research with a hand-curated, topic-organized list of over 200 research papers from 2025.</p> </div> </div> <div class="post-entry"> <div class="post-hero-image"> <img src="/images/blog/2025/coding-the-kv-cache-in-llms/hero.jpg" alt="Understanding and Coding the KV Cache in LLMs from Scratch"> </div> <div class="post-details"> <h2 class="post-title"> <a href="/blog/2025/coding-the-kv-cache-in-llms.html">Understanding and Coding the KV Cache in LLMs from Scratch</a> </h2> <p class="post-date">Jun 17, 2025</p> <p class="post-description">KV caches are one of the most critical techniques for efficient inference in LLMs in production. KV caches are an important component for compute-efficient LLM inference in production. This article explains how they work conceptually and in code with a from-scratch, human-readable implementation.</p> </div> </div> <div class="post-entry"> <div class="post-hero-image"> <img src="/images/blog/2025/coding-llms-from-the-ground-up-a-complete-course/hero.jpg" alt="Coding LLMs from the Ground Up: A Complete Course"> </div> <div class="post-details"> <h2 class="post-title"> <a href="/blog/2025/coding-llms-from-the-ground-up-a-complete-course.html">Coding LLMs from the Ground Up: A Complete Course</a> </h2> <p class="post-date">May 10, 2025</p> <p class="post-description">Why build an LLM from scratch? It's probably the best and most efficient way to learn how LLMs really work. Plus, many readers have told me they had a lot of fun doing it.</p> </div> </div> </article> </div></div> <!-- Styles --><style> .post-entry { display: flex; align-items: flex-start; margin-bottom: 20px; border-bottom: 1px solid #ddd; padding-bottom: 20px; } .post-hero-image { flex: 0 0 120px; margin-right: 20px; } .post-hero-image img { width: 100%; height: auto; } .post-details { flex: 1; } .post-title { margin: 0; font-size: 1.5em; color: #333; } .post-title a { color: #333; text-decoration: none; } .post-title a:hover { text-decoration: underline; color: #0056b3; } .post-date { color: #666; font-size: 0.9em; margin: -4px 0; } .post-description { margin: 0; } h1.post-title { text-align: center; } h2.post-title, h2.post-title a { all: unset; display: block; font-size: 1.0em; color: #333; text-decoration: none; } h2.post-title a:hover { color: #0056b3; text-decoration: underline; } a.anchorjs-link { display: none !important; }</style> </div> </div> <footer class="site-footer"> <div class="wrapper"> <div class="footer-col-wrapper"> <div class="footer-col social-col"> <a href="https://magazine.sebastianraschka.com"><span><i class="fa fa-rss fa-2x"></i></span> </a> <a href="/contact"><span><i class="fa fa-envelope fa-2x"></i></span> </a> <a href="https://twitter.com/rasbt"> <span><i class="fa fa-twitter fa-2x"></i></span> </a> <a href="https://youtube.com/c/SebastianRaschka"><span><i class="fa fa-youtube fa-2x"></i></span> </a> <a href="https://github.com/rasbt"><span><i class="fa fa-github fa-2x"></i> </span></a> <a href="https://scholar.google.com/citations?user=X4RCC0IAAAAJ&hl=enrasbt"><span><i class="fa fa-google fa-2x"></i> </span></a> <a href="https://linkedin.com/in/sebastianraschka"><span><i class="fa fa-linkedin fa-2x"></i> </span></a> </div> <div class="footer-col copyright-col"> <p>© 2013-2025 Sebastian Raschka</p> </div> </div></div> <!-- Google tag (gtag.js) --><script async src="https://www.googletagmanager.com/gtag/js?id=G-BYQXBRPK81"></script><script> window.dataLayer = window.dataLayer || []; function gtag(){dataLayer.push(arguments);} gtag('js', new Date()); gtag('config', 'G-BYQXBRPK81');</script></footer></div> <script src="/js/anchor.min.js" type="text/javascript"></script> <script> var selector = 'h2, h3, h4, h5, h6'; /* anchors.options = { icon: '#', visible: 'always', placement: 'left', class: 'bb-anchor' } */ anchors.add(selector); </script> </body></html> 
