Lessons · Field guide · AI tutor
Machine learning,
from epoch zero.
Plain-language lessons, a browsable field guide to model architectures, and an AI tutor that reads the chapter you're on.
- 65 chapters
- 4 tracks
- 198 architectures
- free
Stuck? Ask the AI tutor.
It reads the chapter you're on, so you can ask about a formula, a quiz question or your code without copying anything over. Free with an account.
Three ways in
New to programming
Learn Python first
Epoch Zero Python: a free 12-week course on your own computer, from print() to a working Pac-Man.
12 weeks · beginnerThe Python course →
New to machine learning
Start the Core Course
Begin at chapter 1: what a model is, explained with a function and a few knobs.
17 chapters · beginner to intermediateChapter 1 →
Already write code
Jump to the math or the modern stack
The Math Behind ML covers the derivations; Modern Architectures builds attention, MoE and diffusion.
The Math Behind ML →8 chapters
Modern Architectures →6 chapters
Explore the catalog
198 architectures, from linear models to state-space hybrids and reasoning LLMs. Start from your problem, or let the decision wizard narrow it down.
- 1805Linear regression Linear modelsFits a straight-line (hyperplane) relationship by least squares. The reference point every other model is measured against.
- 2016XGBoost Trees & ensemblesA regularized, second-order, sparsity-aware implementation of gradient boosting that dominated tabular competitions.
- 2017Transformer Attention & transformersStacked blocks of multi-head self-attention plus a position-wise feed-forward network, with residual connections and normalization.
- 2023Mamba / selective SSM State-space & hybridsMakes the SSM parameters input-dependent (selective), so the model can choose what to remember or forget per token.