# Retrieval Augmented Generation
{pause up}
# The Retriever
{pause up}
{pause up}
{pause up}
{#rag}
{pause up}
some of the snippets retrieved from the vector database are not always relevant or
of good quality, e.g. the contents pages of a book
{pause .block #solution}
> **Solution?**
>
> {pause focus-at-unpause=solution}
> Add another **LLM** of course!
{pause up}
# Reranker
{pause up}
A common reranking model is the cross encoder:
We plug this reranking model into the rag pipeline...
{pause up-at-unpause=rag}