Attention
A mechanism that lets a model weigh which parts of the input matter most for each piece it's processing. Instead of treating every word equally, it focuses on the ones that are actually relevant.
Think of It Like This
Like reading with a highlighter — you weight the important words and skim the rest.
For each word it processes, an attention mechanism scores how relevant every other word is, then blends them together weighted by those scores. That's how a model links a pronoun back to the noun it refers to, even when they sit far apart in the sentence.
Think of It Like This
Picture a spotlight on a stage: as the model reads each word it swivels the light onto the handful of other words that actually matter, leaving the rest dim.