Vector
A list of numbers that stands in for a thing, so a computer can do math on meaning.
Think of it like
GPS coordinates for an idea — a handful of numbers that pin it to a spot you can measure distances from.
Example
The word “dog” might become [0.2, -1.1, 0.7, …], and “puppy” lands nearby in the same number-space.
How it actually works
Almost everything inside a model is a vector: tokens, hidden states, embeddings. Represent things as vectors and “similar” becomes “close”, “combine” becomes “add”. The dimensions rarely mean anything human-readable, but distances and directions carry real structure.
For product teams
It’s the common currency that lets a model compare, search, and relate anything to anything.
For engineers
An ordered array of floats in R^d; the basic unit of representation throughout the network.
Related
- Embeddings — Meaningful ones are embeddings.
- Latent space — Live in a latent space.
Read anything AI without the jargon
Look up any term in plain English, or save terms as you read with the free Chrome extension.
Open DecoderAdd to Chrome