SebastianRaschka.com is the personal blog of Sebastian Raschka, PhD, an LLM research engineer whose work bridges academia and industry in AI and machine learning. On his blog and notes section he publishes deep, well-documented articles on topics such as LLMs (large language models), reasoning models, machine learning in Python, neural networks, data science workflows, and deep learning architecture. Recent posts explore advanced themes like “reasoning LLMs”, comparisons of modern open-weight transformer architectures, and guides for building, training, or analyzing neural networks and model internals.
130 articles from this blog
Analysis of Gemma 4 architecture, benchmarks, and licensing compared to Gemma 3 and other models.
New LLM Architecture Gallery diff tool lets you compare model architecture stacks side by side, highlighting key differences.
A visual guide to attention variants in modern LLMs, covering MHA, GQA, MLA, sparse attention, and hybrid architectures.
A gallery showcasing and comparing architecture diagrams and technical details of recent open-weight Large Language Models (LLMs).
Analysis of NVIDIA's Nemotron 3 Super 120B-A12B model focusing on its accuracy-throughput trade-off design and efficiency features.
A 4.5-hour interview discussing the state of AI in 2026, covering LLMs, geopolitics, training, open vs. closed models, AGI timelines, and industry implications.
An overview of inference-time scaling methods for improving LLM reasoning, categorizing techniques and highlighting recent research.
A 2025 year-in-review analysis of large language models (LLMs), covering key developments in reasoning, architecture, costs, and predictions for 2026.
A curated list of notable LLM research papers from the second half of 2025, categorized by topics like reasoning, training, and multimodal models.
A timeline of beginner-friendly 'Hello World' examples in machine learning and AI, from Random Forests in 2013 to modern RLVR models in 2025.
A technical analysis of the DeepSeek model series, from V3 to the latest V3.2, covering architecture, performance, and release timeline.
Analysis of DeepSeek V3.2's architecture, sparse attention mechanism, and RL updates compared to its predecessor and proprietary models.
Author's method for effectively reading technical books, including multiple read-throughs, coding along, and doing exercises.
An overview of alternative LLM architectures beyond standard transformers, including linear attention hybrids, text diffusion models, and code world models.
Compares DGX Spark and Mac Mini for local PyTorch development, focusing on LLM inference and fine-tuning performance benchmarks.
A guide to the four main methods for evaluating Large Language Models, including code examples and practical implementation details.
A hands-on guide to understanding and implementing the Qwen3 large language model architecture from scratch using pure PyTorch.
Analysis of OpenAI's new gpt-oss models, comparing architectural improvements from GPT-2 and examining optimizations like MXFP4 and Mixture-of-Experts.
A detailed comparison of architectural developments in major large language models (LLMs) released in 2024-2025, focusing on structural changes beyond benchmarks.
A curated list of key LLM research papers from Jan-June 2025, organized by topic including reasoning models, RL methods, and efficient training.
