A clock-tower help desk keeps a binder of winding procedures. Those pages are turned into a searchable store, the closest pages are stuffed into an enlarged prompt, and the language model’s weights are never changed. Which approach is that?
Select an answer to reveal the explanation.
Short Explanation
A binder of winding procedures turned into a searchable store, closest pages stuffed into an enlarged prompt, weights never changed: that is retrieval-augmented generation. RAG retrieves and augments the prompt. Fine-tuning is the contrast, because stuffing pages does not change the weights.
Full Explanation
RAG retrieves supporting passages and leaves the weights alone. Fine-tuning changes weights. From-scratch training and Chapter 6 techniques are different work.