Definitely, top-k is a very naive way to do RAG. I think people have experimented with using a cross encoder like approach or even letting the LLM choose the so

Definitely, top-k is a very naive way to do RAG. I think people have experimented with using a cross encoder like approach or even letting the LLM choose the sources. We will experiment with more approaches like this :)