Embedding
A way of representing words, images, or other items as lists of numbers (vectors) so a model can work with them. Items with similar meaning end up with similar vectors.
Think of It Like This
It's a map where things with related meanings sit close together.
A model can't do math on the raw word "cat," so it maps every token to a vector — often a few hundred numbers — learned during training. Tokens that show up in similar contexts drift toward nearby vectors, which is what lets a model treat "car" and "automobile" as close cousins.
Think of It Like This
The vectors capture relationships, not just closeness: start at "king," subtract "man," add "woman," and you land right next to "queen."