Embedding
Also called vector embedding
A list of numbers standing for the meaning of some text or an image, so a computer can measure how alike 2 things are.
Computers can't compare meanings directly, but they can compare numbers. An embedding turns a word, a sentence, a document or a photo into a long list of numbers, chosen by a model so that things with similar meanings get similar lists.
The distance between 2 lists then measures how related the 2 things are. That makes it possible to search by meaning instead of exact words, group similar complaints together, recommend the next video, or find the passages in your documents that relate to a question before a chatbot answers it. The numbers themselves mean nothing to a person; only the distances between them matter.
For example
"My card was charged twice" and "double payment on my account" share almost no words, but their embeddings sit close together, so a help centre can match them.
Put it to use
Where you'll meet embedding in ChatGPT, Gemini and Claude, how to use it there, and the mix-up most people make.
Daily Bytes Plus is in the Android app.