Sebastian Raschka is an LLM research engineer and the author of Build a Large Language Model (From Scratch) and Build a Reasoning Model (From Scratch). He publishes technical articles through Ahead of AI, coding lectures on YouTube, and open-source implementations on GitHub.
X (Twitter)BlogLatest we gathered
Latest 15We gather link previews. The full text and video stay at the source.
Philosophical Discussion on the Far Future of AI
Language Models for Text Classification: From Bag-of-Words to Jev
Focusing on Post-Training
MiMo-V2.6 Pro Architecture and Training Notes
REVIEWS
Reviews of Sebastian Raschka
No reviews yet
Share your experience with Sebastian Raschka's content with stars and tags.
More creators to explore
- Sebastian RaschkaCurrent profile
TopicsAI, technology & development
Sebastian Raschka is an LLM research engineer and the author of Build a Large Language Model (From Scratch) and Build a Reasoning Model (From Scratch). He publishes technical articles through Ahead of AI, coding lectures on YouTube, and open-source implementations on GitHub.
- Eric ParkSame field
TopicsAI, technology & development
What do they make?Eric Park is a product designer and creator whose Korean-language work demonstrates AI-assisted UX/UI design, reference-led prototyping, and practical comparisons of generated outputs. His videos focus on concrete screen decisions, while his essays explain design-system ideas through everyday examples. - GaebalbadakSame field
TopicsAI, technology & development
What do they make?A Korean developer talk-show channel hosted by Hyangro and Hodolman. It publishes conversations, developer interviews, and community posts about software development, learning, and careers. - Dingcodingco (Hyunjoon Park)Same field
TopicsAI, technology & development
What do they make?A backend developer and educator who connects technical knowledge with design decisions, measurable results, and interview explanations. His inspected videos and articles use concrete examples and everyday analogies, and discuss testing and refactoring as checks for AI-assisted coding. - UNDERkgSame field
TopicsAI, technology & development
What do they make?UNDERkg is a Korean-language technology review team and channel covering smartphones and related devices. The inspected videos, hosted by F717, discuss product use alongside practical details such as Korean mobile-network and app compatibility, earbud ANC and white noise, and projector setup and power limitations.
Earlier posts
It's Easy to Dismiss Jev as Just a Classifier
It’s easy to dismiss Jev it as “just a classifier”.

Inference scaling part 1.
Interesting new insights into the old AdamW vs Muon debate: Muon seems to do better because it reduces memorization.

Some food for thought when designing benchmarks...
Pacing != Pacing Development
AI Reasoning Models Course on LinkedIn Learning
I put together a mega write-up on GPT-6 Astra & looped transformers.
GPT-6 Astra, Looped Transformers, and Hidden Reasoning
OpenAI Astra and Looped Transformers
Who to introduce next
Can't find a creator?
Tell us the name or channel link of a creator you'd like us to introduce.
