A sequenced path ยท math included

Learn AI from the math
up to agents.

12 phases that build on each other, from vectors and probability to transformers, RAG and agentic systems. Every lesson pairs derivations with videos, practice and a quick self-test.

0%
The whole path
116 lessons ยท 25 live demos ยท 112 papers
Sign in to sync across devices
0 hof 364 h studied
0day streak
0study days

Your roadmap

Each phase builds on the previous ones ยท open a phase to see its lessons
  1. How this path works, the map of the field, and your lab setup
  2. Linear algebra, calculus, probability, information and optimization โ€” the pillars every model stands on
  3. The minimum stack that lets you run every idea in this path
  4. Learning from data, with the math that tells you when it works
  5. Stack dot products and nonlinearities, then learn every weight with the chain rule
  6. Teaching machines to see โ€” and the architecture ideas that shaped all of deep learning
  7. From tokens and n-grams to recurrent networks and the first attention
  8. One architecture, derived line by line, from attention to frontier-scale inference
  9. How frontier models are built, aligned, prompted, sampled, evaluated and understood
  10. Ground models in your data and ship LLM apps that hold up in production
  11. Models that act: tools, memory, planning, protocols, evaluation and safety
  12. Read the papers, ship the projects, steer the career
Three passes per lessonWatch for intuition, read for the math, practice in code.
An AI tutor on every lessonOne-click prompts that explain, derive, quiz or set exercises.
Remembers for youSpaced-repetition flashcards, quiz scores and notes.