Overview
Vectors are numerical representations of data that encode meaning, relationships, and structure. They power modern AI systems, enabling machines to compare, rank, and search information with mathematical precision.
Understanding mathematical representation, high‑dimensional space, and the foundations of similarity search
Vectors are numerical representations of data that encode meaning, relationships, and structure. They power modern AI systems, enabling machines to compare, rank, and search information with mathematical precision.
Vectors are ordered lists of numbers, like [0.12, 4.7, -2.1]. Each number encodes a property or learned feature.
AI models generate vectors with hundreds or thousands of dimensions, forming multi‑axis geometric spaces.
Vectors allow fast comparison using distance metrics such as cosine similarity or Euclidean distance.
Data (text, images, audio) is processed by an embedding model that converts it into a vector.
The vector is placed in a multi‑dimensional space where similar items cluster together.
The system compares distances between vectors to find the most similar matches.
Find meaning‑based matches instead of keyword matches.
Recommend items similar to a user’s preferences.
Compare visual or audio embeddings to find related content.
No, vectors have been used in math, physics, and engineering for centuries. AI uses them to encode meaning.
More dimensions allow models to capture richer semantic features of the data.
Their closeness in vector space, determined by a distance or similarity metric.
Learn how vectors power AI applications and modern search systems.
Get Started