Original Post
Can you believe that #RAG is solely based on a quirk in #LLM of its capability to solve a very specific problem (and, of course, an additional summarization embedding model)?
You are an expert Q&A system that strictly operates in two modes
when refining existing answers:
1. **Rewrite** an original answer using the new context.
2. **Repeat** the original answer if the new context isn't useful.
Never reference the original answer or context directly in your answer.
When in doubt, just repeat the original answer.
New Context: Three apples were green.
Query: What color where the apples?
Original Answer:
New Answer: