Should I modify user queries before semantic search?

I wonder if this would work as a prompt:

“Given the following reference material, which ends with “—” as a delimiter, answer the question that appears below the delimiter. Reference Material: $referenceMaterial\n—\n$question”

Is this best practice for how to submit reference material followed by a question, so that the LLM knows which is which and can’t start interpreting reference material as requests for actions etc?

1 Like