Neural networks
Official9 units44 modulesDeep
How a network turns pixels into a guess, how it learns by gradient descent and backpropagation, and how transformers and attention power language models.
No account needed: your progress is kept in this browser. Create a free account to keep it on every device and get review reminders.
Your progress, unit by unit
Mastered
You finished it and got its questions right the first time. An article you've read through counts as mastered.
Proficient
You finished it and got nearly everything right, a few on a second try.
Familiar
You finished it, but some answers took a few tries. Worth another look before you move on.
Attempted
You've started it but haven't finished, or most of its answers were wrong. Pick it up where you left off.
Not started
This is where you'll start. Read through the unit if it's new to you, or jump to its quiz if you feel sure of it.
Quiz
The questions at the end of each unit. They check the unit's goal, and you can take one any time to see what you already know.