Vector, in AI and machine learning, is a list of numbers that represents data in a form a model can process. Vectors are the fundamental way information — words, images, features — is fed to and handled by AI systems.
What it means in plain English
Computers work with numbers, so any data a model uses must first be turned into numbers arranged in a list: a vector. A vector might represent the features of a house, the meaning of a word (an embedding), or the pixels of an image. Representing data as vectors lets models perform the mathematical operations that underlie learning and prediction. Much of machine learning is, under the hood, mathematics on vectors.
Vectors also let systems measure similarity — how “close” two pieces of data are in meaning or properties.
A simple example
The word “cat” might be represented as a long vector of numbers (an embedding). Because “kitten” has a similar vector, a system can tell the two words are related by comparing their vectors mathematically.
Why it matters
Vectors are the basic language in which AI represents and manipulates information. Understanding that data becomes vectors clarifies how concepts like embeddings, similarity search, and neural network computation actually work.
Related terms
- Embedding — a vector that captures meaning.
- Vector Database — stores and searches vectors.
- Parameter — the values models learn, also stored as vectors.
Frequently asked questions
What is a vector in AI?
A vector is an ordered list of numbers that represents data in a form models can process — for example, an embedding that captures the meaning of a word or image.
Why are vectors central to AI?
Representing data as vectors lets models measure similarity mathematically, which underpins embeddings, semantic search, and much of how neural networks work.
