Embeddings Explained

Relax. We'll talk through this in plain words — no textbook voice.

What you'll walk away with

Understand Embeddings Explained without any of the intimidation
Get hands-on practice trying it yourself
Learn to spot — and smile past — the easy-to-make mistakes

Learn how text meaning gets turned into number vectors and how that lets you search by semantic similarity.

Let's think about it this way for a second

An embedding represents the meaning of a piece of text, an image, or other data as a vector of numbers. Things with similar meaning end up close together in vector space, so you can search by meaning instead of just matching keywords.

Embeddings Explained lesson illustration — RAG and Vector Search — Embeddings Explained

Let's connect this to everyday life

"Car repair shop" and "auto mechanic shop" don't share the same words, but embedding search can tell they're closely related. This is useful for search, clustering, recommendations, and duplicate detection.

Turn text into a vector
Embed the query with the same model
Use a metric like cosine similarity
Check the quality of the top results

Let's try it hands-on together

javascript

import OpenAI from "openai";
const client = new OpenAI();

const result = await client.embeddings.create({
  model: "text-embedding-3-small",
  input: "မြန်မာစာဖြင့် AI လေ့လာခြင်း",
});

console.log(result.data[0].embedding.length);

You should see

You'll get an embedding vector for a line of text.

5-minute try-it

Embed three sentences and guess which two are closest in meaning. Then check your guess with similarity scores.

A quick word of caution

Don't treat AI output as the final word. Have a human review anything important — including code and user data — before it's actually used.

OpenAI — Embeddings — OpenAI

Easy traps

Mixing vectors from different embedding models
Assuming that a close vector means the fact is correct

Exercise

Embed three sentences and guess which two are closest in meaning. Then check your guess with similarity scores.

You'll know it worked when: You'll get an embedding vector for a line of text.