AI and LLMs · Junior · Question 18 of 75
What is retrieval-augmented generation (RAG)?
- 1Fetching relevant documents and adding them to the prompt
- 2Training the model again every time a new document is added
- 3Generating answers and then hiding the sources from the user
Show the answer
Correct answer: Fetching relevant documents and adding them to the prompt
Explanation
RAG grounds answers in your data, such as docs or tickets, without retraining. Quality depends heavily on what gets retrieved.