RAG & Search
Similarity score
A number indicating how closely two embeddings or items match.
Example
Search assigns a higher score to passages whose meaning is closer to the question.
Why it matters
It helps systems order possible matches and select which items deserve a closer look.
You'll hear it
“This is the closest match, but the score still looks weak.”
Can you control it?
Not directly
No, not directly. A similarity score describes a learned property, observed result, or broader concept rather than a normal user setting. You may influence it indirectly by changing the tool, data, prompt, or workflow.
Common questions
- Is the score a chance that the answer is correct?
- Usually not. It describes a particular comparison, not verified factual accuracy.
- Are scores from different systems directly comparable?
- Not necessarily. Their methods and scales can differ.
- Can the highest-scoring result still be irrelevant?
- Yes. The best available match may be poor when the collection lacks the needed information.