Skip to content

Generative AI & LLMs

Attention

A mechanism that lets a model weigh which parts of the input matter most for the current prediction.

Example

While translating a pronoun, the model gives extra importance to the earlier noun it refers to.

Why people use it

Knowing how “Attention” works helps teams choose and configure generative systems more effectively.

What you'll hear

“How does Attention affect the answer the model gives the user?”

Related terms