Thanks Sharon. Yes, I'm planning to add this feature. Will show the chunks that were retrieved and used by the llm for generating the query. I'm also planning to add some evaluation metrics.