What do embeddings represent?
Why C is the answer
Text as numeric vectors, so that similar meanings sit close together
Embeddings capture meaning as numbers, which is what makes semantic search possible.
In detail
- Embeddings turn text into numeric vectors.
- Similar meanings sit close together.
- Embeddings power semantic search.
Use case
An illustrative scenario, not a real incident.
A help site finds "reset my password" when a user types "can't log in". Embeddings place the two close together, because their meanings are similar.
See it in different ways
stateDiagram-v2 [*] --> Tokens: prompt becomes tokens Tokens --> Predict Predict --> Append: pick the next token Append --> Predict: not finished Append --> [*]: stop or token limit
flowchart LR S0["Text"] --> S1["Embedding model"] --> S2["Vector"] --> S3["Compare by distance"]
Key point: Embeddings turn meaning into numbers.
Why the other options are wrong
- A. A list of the keywords that are banned by a content filter
- They are not filter lists.
- B. The final price of each model request
- They are not prices.
- D. Images that have been placed inside a document as attachments
- They are not image attachments.