← Library · Core concept

Embeddings

Embeddings are numerical representations of discrete data (like words, images, or users) in a continuous vector space where semantically similar items are located closer together. These high-dimensional vectors capture contextual meaning and relationships, allowing algorithms to process and understand complex data types effectively. They translate abstract concepts into a mathematical format that models can use.

In plain terms

Think of it as converting every unique item in a language or image into a precise coordinate on a multi-dimensional map, where similar items are clustered near each other.

Why it matters

Embeddings enable AI models to process and understand complex, non-numerical data like text and images by turning them into meaningful numerical representations.

Learn one new AI thing every day.

Daily Deck sends you seven plain-English cards like this every morning. Free.

Start free