Attention Mechanism

Level 3

Short Description

A neural network technique that lets a model weigh the importance of different parts of an input when producing each part of the output — the foundation of modern transformer models.

Friendly Description: Attention is what lets an AI focus on the parts of a sentence (or image, or anything) that matter most for what it's doing next. It's a lot like how, when you read a long email, your eyes naturally land on the most important sentences. Attention helps AI do the same thing, so its responses feel relevant and on-topic instead of scattered.

Example: If you ask an AI, "What did the author say about the dog in chapter three?" the attention mechanism helps the model zoom in on chapter three and on the parts that mention the dog, instead of treating every word in the book equally. That focused reading is what lets it answer accurately.