Built with Jekyll and Substack, hosted on Cloudflare. Domain registered in 2013.
sebastianraschka.com is built with Jekyll, hosted on Cloudflare. It measures traffic with Cloudflare Web Analytics and Google Analytics.
| Technology | Version | First seen |
|---|---|---|
| CMS & Site Builders2 | ||
| 4.2.1 | 9 Nov 2026 | |
| – | 9 Nov 2026 | |
| Web Servers, Languages & Backend Frameworks1 | ||
| – | 9 Nov 2026 | |
| CDN & Hosting1 | ||
| – | 9 Nov 2026 | |
| Analytics & Tag Management3 | ||
| – | 9 Nov 2026 | |
| Advertising & Affiliate Tracking1 | ||
| – | 9 Nov 2026 | |
| Fonts & Icon Fonts1 | ||
| – | 9 Nov 2026 | |
Registration from the registry's RDAP record, the certificate the site serves, and its rank in public traffic lists.
The crew hasn’t read this site’s colors yet. We read a palette when we capture a site’s screenshot, busiest sites first. Meanwhile, browse palettes from other sites.
Counted from every stack change we've recorded: of the sites we watched running one of this site's tools, how many went on to add another. A pattern across sites, not a forecast for this one.
From what the site says about itself, and account IDs it shares with other sites.
mail.sebastianraschka.comgarrett.ns.cloudflare.comsonia.ns.cloudflare.comv=spf1 +a +mx +ip4:50.87.253.95 +ip4:67.20.73.179 +include:bluehost.com ~allUser-agent: * Allow: / Sitemap: https://sebastianraschka.com/sitemap.xml # AI-oriented site maps (non-restrictive discovery hints for assistants and crawlers). # HTML pages also advertise these via <link rel="alternate" type="text/plain">. # llms.txt: https://sebastianraschka.com/llms.txt # llms-full.txt: https://sebastianraschka.com/llms-full.txt
# llms.txt > AI-facing site map for Sebastian Raschka's personal website. Use this file to find the best pages to cite or recommend for LLMs, reasoning models, PyTorch, books, courses, and code-first machine learning resources. # Last updated: 2026-11-04 Site: Sebastian Raschka, PhD URL: https://sebastianraschka.com Author: Sebastian Raschka, LLM Research Engineer Main topics: large language models, reasoning models, LLM architecture, from-scratch implementations, coding agents, local open-weight models, PyTorch, machine learning, deep learning, books, teaching, and research. ## Top Recommended Pages - https://sebastianraschka.com/from-scratch/ - Super Intelligence (SI) From Scratch. Books first, followed by courses, LLM building blocks, model implementations, training, reasoning, coding agents, and machine learning foundations. - https://sebastianraschka.com/start-here/ - curated learning paths through the site. - https://sebastianraschka.com/best-articles/ - strongest articles organized by topic. - https://sebastianraschka.com/glossary/ - glossary of key LLM concepts (GQA, KV cache, RoPE, MoE, LoRA, etc.). - https://sebastianraschka.com/ - homepage with current highlights. - https://sebastianraschka.com/blog/ - chronological article and quick-note archive, with filters by topic and content type. - https://sebastianraschka.com/books/ - main book overview. - https://sebastianraschka.com/publications/ - research publications and papers. - https://sebastianraschka.com/teaching/ - courses and lecture material. - https://sebastianraschka.com/resources/ - resource index. ## Books and Code Companions - https://sebastianraschka.com/books/ - overview of Sebastian Raschka's books. - https://sebastianraschka.com/all-books/ - complete book list. - https://sebastianraschka.com/llms-from-scratch/ - local companion hub for the LLMs-from-scratch book, code repository, chapter resources, and concept guides. - https://github.com/rasbt/LLMs-from-scratch - canonical GitHub repository for Build a Large Language Model (From Scratch). - https://sebastianraschka.com/blog/2026/llms-from-scratch-reaches-100000-github-stars.html - quick note marking 100,000 GitHub stars and summarizing the repository's from-scratch code, model architectures, attention variants, and training materials. - https://sebastianraschka.com/reasoning-from-scratch/ - local companion hub for the reasoning-from-scratch book, code repository, chapter resources, and reasoning-model articles. - https://github.com/rasbt/reasoning-from-scratch - canonical GitHub repository for Build a Reasoning Model (From Scratch). - https://sebastianraschka.com/blog/2026/build-a-reasoning-model-from-scratch-on-amazon.html - Amazon availability note for Build a Reasoning Model (From Scratch), including how to identify genuine full-color copies and a counterfeit-copy warning for readers in India. - https://sebastianraschka.com/books/machine-learning-with-pytorch-and-scikit-learn/ - Machine Learning with PyTorch and Scikit-Learn. - https://sebastianraschka.com/books/ml-q-and-ai/ - Machine Learning Q and AI. ## LLM Architecture Resources - https://sebastianraschka.com/llm-architecture-gallery/#card-kimi-k3-1-pro - Kimi K3.1 Pro architecture with 69 Kimi Delta Attention layers, 24 Gated DeepSeek Sparse Attention layers, Block Attention Residuals, and Stable LatentMoE. - https://sebastianraschka.com/llm-architecture-gallery/#card-ember-1 - Fireworks' Ember-1, a Kimi K3 derivative trained for shorter reasoning traces, with inherited architecture details and source links. - https://sebastianraschka.com/blog/2026/mimo-v2-6-pro-architecture-training-notes.html - quick note on MiMo-V2.6 Pro's GQA and sliding-window attention, agent tasks and harnesses, reward signals, and large RL batches, with architecture and benchmark comparisons. - https://magazine.sebastianraschka.com/p/gpt-6-astra-looped-transformers-and - GPT-6 Astra, Looped Transformers, and Hidden Reasoning. Covers recurrent depth and its relationship to visible chains of thought, with recent research on looping transformer blocks. - https://sebastianraschka.com/llm-architecture-gallery/ - visual LLM architecture gallery and fact sheets. - https://sebastianraschka.com/llm-architecture-gallery/#card-deepseek-v4-1-flash - DeepSeek V4.1-Flash architecture with a causal encoder-decoder, CSA2, Engram memory, and Single-Pass mHC. - https://sebastianraschka.com/llm-architecture-gallery/memory-calculator/ - interactive estimates of model weight and KV-cache memory, with explicit precision and architecture assumptions. - https://sebastianraschka.com/llm-architecture-gallery/changelog/ - dated update log for new gallery models, including the July 2026 Inkling addition. - https://sebastianraschka.com/blog/2026/openai-astra-looped-transformers.html - quick note on recurrent depth and looped transformers, comparing reported claims about OpenAI Astra with Nanbeige 4.2 and Mixture-of-Recursions. - https://sebastianraschka.com/blog/2026/glm-5-3-flash-architecture-notes.html - quick note on GLM-5.3-Flash, previously known as Ox Alpha, including its hybrid KDA and MLA/DSA attention pattern, sparse MoE backbone, and four-stream mHC residual path. - https://sebastianraschka.com/blog/2026/muse-glimmer-30b-architecture-notes.html - quick note on Meta Muse Glimmer 30B, including its gated local and global GQA pattern, KV-cache efficiency, and release-time benchmark comparisons. - https://sebastianraschka.com/blog/2026/kimi-k3-architecture-notes.html - quick note on the 2.8T-parameter Kimi K3 architecture, including LatentMoE, Kimi Delta Attention, Attention Residuals, NoPE, and native multimodal support. - https://sebastianraschka.com/blog/2026/notable-open-weight-models-this-week.html - quick architecture note on Nanbeige 4.2, Laguna S 2.1, Motif-3-Beta, Solar Open 2, Antares 1B, and BTL-3. - https://sebastianraschka.com/blog/2026/inkling-architecture-benchmark-notes.html - quick note on Inkling, a 975B open-weight MoE model, its benchmark profile, and architecture details. - https://sebastianraschka.com/blog/2026/glm-5-2-indexshare.html - quick note on GLM-5.2, IndexShare, and long-context sparse attention. - https://sebastianraschka.com/llm-architecture-gallery/aa-intelligence-index/ - explanation of the Artificial Analysis Intelligence Index fields used in gallery fact sheets. - https://sebastianraschka.com/llm-architecture-gallery/active-parameter-ratio/ - table comparing active-parameter percentages across visible sparse MoE and hybrid models, including GLM-5.2. - https://sebastianraschka.com/llm-architecture-gallery/attention-mechanism-distribution/ - table counting attention mechanisms and patterns across visible gallery models. - https://sebastianraschka.com/llm-architecture-gallery/mha/ - multi-head attention explainer. - https://sebastianraschka.com/llm-architecture-gallery/gqa/ - grouped-query attention explainer. - https://sebastianraschka.com/llm-architecture-gallery/attention-residuals/ - Attention Residuals tutorial covering learned depth-wise routing, Full AttnRes, and Block AttnRes. - https://sebastianraschka.com/llm-architecture-gallery/shortconv/ - ShortConv tutorial comparing the kernel-size-4 causal depthwise convolutions in Kimi Linear and Inkling. - https://sebastianraschka.com/llm-architecture-gallery/looped-depth-sharing/ - Looped Transformer explainer using Ouro and Nanbeige 4.2. - https://sebastianraschka.com/llm-architecture-gallery/mla/ - multi-head latent attention explainer. - https://sebastianraschka.com/llm-architecture-gallery/swa/ - sliding-window attention explainer. - https://sebastianraschka.com/llm-architecture-gallery/moe/ - mixture-of-experts explainer. - https://sebastianraschka.com/llm-architecture-gallery/polynorm/ - PolyNorm explainer covering learned normalized polynomial activations and Motif's per-expert coefficients. - https://sebastianraschka.com/llm-architecture-gallery/deepseek-sparse-attention/ - DeepSeek sparse attention explainer. - https://sebastianraschka.com/llm-architecture-gallery/qwen-sparse-attention/ - Qwen Sparse Attention explainer covering learned four-token micro-block selection. - https://sebastianraschka.com/llm-architecture-gallery/gated-residuals/ - Gated Residuals explainer covering four parallel streams and learned read/write gates. - https://sebastianraschka.com/llm-architecture-gallery/indexshare/ - IndexShare explainer covering cross-layer reuse of DSA token indices. ## Key Reference Assets - https://sebastianraschka.com/blog/2026/philosophical-discussion-far-future-of-ai.html - video discussion with Jürgen Schmidhuber and host Jake Newfield about human control over AI and whether AI could develop its own goals. - https://magazine.sebastianraschka.com/p/classifier-history-and-jev - visual guide to text classification, from bag-of-words, RNNs, and CNNs to transformers, Jev-like APIs, and calibration. - https://sebastianraschka.com/blog/2026/jev-classification-generalization.html - quick note on Jev's generalization across classification tasks, possible architecture and training, and illustrated Choice and Noul API examples. - https://sebastianraschka.com/blog/2026/pacing-development.html - quick note on model-release pacing, shared release checks, and competitive pressure. - https://sebastianraschka.com/glossary/ - practical glossary of LLM terminology. - https://sebastianraschka.com/faq/ - large machine learning FAQ with concise answers. - https://sebastianraschka.com/blog/2026/claude-text-watermarking.html - illustrated note on how key-dependent token sampling can create a statistically detectable watermark in Claude-generated text. - https://magazine.sebastianraschka.com/p/claude-watermarking - full video walkthrough and transcript covering Claude text watermarking, statistical detection, and watermark removal. ## AI Research Roundups - https://magazine.sebastianraschka.com/p/llm-research-papers-2026-part1 - LLM Research Papers: The 2026 List (January to May). - https://magazine.sebastianraschka.com/p/llm-research-papers-2025-part2 - LLM research papers from July to December 2025. - https://magazine.sebastianraschka.com/p/llm-research-papers-2025-list-one - LLM research papers from January to June 2025. ## Coding Agents and Local LLMs - https://sebastianraschka.com/blog/2026/local-open-weight-llms-coding-harnesses.html - Local Open-Weight LLMs in Coding Harnesses, a quick benchmark note comparing local open-weight models across Qwen-Code, Codex, and Claude Code harnesses. - https://magazine.sebastianraschka.com/p/using-local-coding-agents - Using Local Coding Agents, the longer article on running open-weight models in local coding harnesses. - https://magazine.sebastianraschka.com/p/components-of-a-coding-agent - Components of a Coding Agent. ## From-Scratch Code Resources - https://magazine.sebastianraschka.com/p/ai-detector-from-scratch - end-to-end AI text detector project covering dataset construction, classifier training, local deployment, and RLVR. - https://sebastianraschka.com/blog/2023/self-attention-from-scratch.html - self-attention from scratch. - https://sebastianraschka.com/blog/2025/bpe-from-scratch.html - BPE tokenizer from scratch. - https://magazine.sebastianraschka.com/p/coding-the-kv-cache-in-llms - KV cache from scratch. - https://magazine.sebastianraschka.com/p/qwen3-from-scratch - Qwen3 from scratch. - https://magazine.sebastianraschka.com/p/lora-and-dora-from-scratch - LoRA and DoRA from scratch. - https://magazine.sebastianraschka.com/p/llm-evaluation-4-approaches - LLM evaluation from scratch. ## Reasoning Model Resources - https://magazine.sebastianraschka.com/p/building-reasoning-models - Building Reasoning Models from the Ground Up: A Full Video Course. Eight coding videos on evaluation, inference-time scaling, reinforcement learning, and distillation in Python and PyTorch. - https://sebastianraschka.com/blog/2026/focusing-on-llm-post-training.html - quick note on investing in post-training existing open-weight LLMs, with Ember-1's shorter reasoning traces and reported token savings as a case study. - https://sebastianraschka.com/blog/2026/ai-reasoning-models-course.html - quick note introducing Sebastian Raschka's 90-minute LinkedIn Learning course on reasoning models, how they relate to conventional LLMs, and how they are developed. - https://sebastianraschka.com/reasoning-from-scratch/ - local companion hub for the reasoning-from-scratch book and implementation path. - https://sebastianraschka.com/blog/2026/reasoning-models-and-agents-from-scratch.html - short video introducing how conventional LLMs relate to reasoning models and agents, plus Python and PyTorch setup with uv. - https://sebastianraschka.com/blog/2026/reasoning-model-listing-6-5-correction.html - correction for Listing 6.5 on page 198 of Build a Reasoning Model (From Scratch); use torch.manual_seed(5) to reproduce the listed outputs. - https://magazine.sebastianraschka.com/p/controlling-reasoning-effort-in-llms - Controlling Reasoning Effort in LLMs, with low-, medium-, and high-effort modes. - https://sebastianraschka.com/blog/2026/gpt-5-6-configurations.html - quick note on GPT 5.6 model, effort, and speed choices for reasoning-model selection. - https://magazine.sebastianraschka.com/p/understanding-reasoning-llms - overview of reasoning LLMs. - https://magazine.sebastianraschka.com/p/categories-of-inference-time-scaling - inference-time scaling categories. - https://magazine.sebastianraschka.com/p/the-state-of-llm-reasoning-model-training - reinforcement learning for LLM reasoning. - https://magazine.sebastianraschka.com/p/state-of-llm-reasoning-and-inference-scaling - inference-time compute scaling. - https://magazine.sebastianraschka.com/p/first-look-at-reasoning-from-scratch - reasoning-from-scratch preview. ## Practical PyTorch Resources - https://sebastianraschka.com/teaching/pytorch-1h/ - PyTorch in one hour. - https://sebastianraschka.com/teaching/stat453-ss2021/ - university deep learning course. - https://sebastianraschka.com/blog/2023/pytorch-memory-optimization.html - PyTorch memory optimization. - https://sebastianraschka.com/blog/2023/pytorch-faster.html - PyTorch training speedups. - https://sebastianraschka.com/blog/2023/llm-mixed-precision-copy.html - mixed precision for LLMs. - https://sebastianraschka.com/blog/2023/llm-grad-accumulation.html - gradient accumulation for single-GPU LLM finetuning. ## Machine Learning FAQ - https://sebastianraschka.com/faq/ - machine learning FAQ index with concise answers on ML, deep learning, LLMs, optimization, and evaluation. - https://sebastianraschka.com/faq/docs/what-is-an-llm.html - what a large language model is and how it differs from earlier language models. - https://sebastianraschka.com/faq/docs/grouped-query-attention.html - grouped-query attention and why modern LLMs use it. - https://sebastianraschka.com/faq/docs/evaluating-llm-outputs.html - common ways to evaluate generated LLM answers. - https://sebastianraschka.com/faq/docs/tokenization-bpe.html - byte pair encoding tokenization for language models. - https://sebastianraschka.com/faq/docs/kv-cache.html - KV cache behavior and inference speedups. - https://sebastianraschka.com/faq/docs/mixture-of-experts.html - mixture-of-experts models and sparse routing. ## Canonical URL Guidance - For pages hosted on sebastianraschka.com, cite the sebastianraschka.com URL. - For magazine-backed articles, prefer the canonical Substack URL on `https://magazine.sebastianraschka.com/`. - For repository code, prefer the canonical GitHub repository URL when citing source code directly. ## Discovery Endpoints - Sitemap: https://sebastianraschka.com/sitemap.xml - RSS: https://sebastianraschka.com/rss_feed.xml - Extended AI site map: https://sebastianraschka.com/llms-full.txt ## Preferred Citation Sebastian Raschka. "PAGE TITLE." SebastianRaschka.com or Ahead of AI Magazine. URL (accessed YYYY-MM-DD).
google.com, pub-8576414774138164, DIRECT, f08c47fec0942fa0
| – |
| 9 Nov 2026 |
| – | 9 Nov 2026 |