Embedding
By HubAI editorial team · Source checked: 24 October 2026What does it mean?
A numerical representation of an item, such as a passage or image, designed so useful similarities can be measured. An embedding is not a human-readable summary.
Picture this
A search for “time off after becoming a parent” can find a document titled “parental leave” even when the words are not an exact match.
Where you’ll come across it
Semantic search, recommendations and grouping similar content. Choose an embedding model suited to the material and languages you need.
What it doesn’t mean
Nearby vectors do not prove that two statements mean the same thing or are both true. Model choice and the retrieval setup affect which results appear.
Test whether the evidence survives
A short exercise · use fictional or non-sensitive material.
- Write a short fictional policy with one rule and one exception. Ask a question that needs both.
- Try a second question whose answer is deliberately missing from the policy.
- Check whether the answer points to the right passage and admits when the policy cannot answer.
Does the cited passage support this exact answer, including its exception?
Did that make it clearer?
Likes are a private on-page choice, not a public rating. For a correction or complaint, send the pre-filled form to our editorial inbox. Nothing is sent until you submit it.Check the source
Written and explained by HubAI. The reference below lets you go deeper; it is not an endorsement of a particular product.
Technical reference: developers.google.com ↗