Generative AI & LLMs
RAG (Retrieval-Augmented Generation)
8 practice questions. Free questions open a full answer guide; the rest unlock with Pro.
- A teammate proposes fixing RAG quality by switching to a long-context model and just stuffing the top 50 retrieved chunks into the prompt. What's your view on that approach, and how do you decide how much retrieved context to actually pass?
- Your RAG support assistant works well when users phrase questions the way the docs do, but fails on paraphrases and synonyms — a user asks about 'login loops' and the chunk says 'authentication redirect errors,' and retrieval misses it. How do you reason about and fix this vocabulary-mismatch problem?Go Pro
- In a RAG system, what is a reranking step, and why might you add one after your initial vector search instead of just using the top vector hits?Go Pro
- In a RAG system, why might pure vector (semantic) search fail to retrieve the right passage, and what can you do about it?Go Pro
- Your RAG system retrieves passages but the answers are still frequently wrong or unsupported. How would you diagnose whether the problem is retrieval or generation, and how would you fix the retrieval side?Go Pro
- When building a RAG system, how do you decide how to chunk your documents before embedding them, and why does chunk size matter?Go Pro
- Your RAG system retrieves relevant-looking passages but the answers are still missing information that's clearly in the corpus. How do you reason about retrieval quality, and what would you change to raise it?Go Pro
- In a RAG pipeline, how do you decide how to chunk your documents, and what goes wrong if the chunks are too big or too small?Go Pro
Want questions matched to your role? Paste a job title, job description, or CV for a personalized set, or go Pro to unlock the full bank.