Attention: Queries, Keys and Values

3 min
Audience
ages 10–15
Duration
30 to 40 min
Adventure
6/8
Badge to earn
Attention Navigator

Today's mission

Compute a mini attention with a query, keys, normalized weights and a sum of values.

Proof of success: explain "Attention: queries, keys and values", complete "The Weighted Mix" and name the limit of the analogy.

The story

In "The robot puts the cup away because it is empty", Amine looks for what "it" refers to. For the current token, the model builds a query; each token provides a key compared to that query and a value containing the information to mix. The similarities become weights via softmax. The output is a weighted sum of the values. A large attention signals a strong influence in this calculation, not a full explanation of the decision.

Preview — the rest of the lesson is for enrolled readers.

Already enrolled with a code?

Your access is tied to your account, not to this link. Sign in with the same email you used in class: your course is waiting, no need to enter the code again.

Sign inNo account yet? Create one
This lesson is part of the “Language” module

The first modules of the course are open to everyone. For the rest you have three options: buy this course once and for all, subscribe, or enter the code handed out in class.

Are you a student on this course?

The code is tied to your account: sign in or create an account and it will be applied automatically when you come back.

No account yet? Create one