Rag And Retrieval easy
Your model has a very large context window. The prompt template and chat history use 3,000 tokens, and you retrieve 8 chunks of about 600 tokens each. A PM suggests filling the rest of the window with 150 more chunks 'since the space is free.' What is the best response?
- A Agree; a larger context always improves grounding, so filling the window is ideal.
- B Agree, as long as the chunks are deduplicated so that no passage appears twice.
- C Disagree, because long-context models silently truncate anything past the first ten chunks.
- D Disagree; extra chunks add cost, latency, and distractors, so keep a tight budget.
Answer: D. Disagree; extra chunks add cost, latency, and distractors, so keep a tight budget.
**Context space is not free**: input tokens are billed and add latency, and many low-relevance chunks make it harder for the model to find the right evidence. A tight budget of a few well-ranked chunks usually beats a full window. Option A is the tempting misconception; a large budget makes sense only for tasks that genuinely need broad coverage, such as summarizing an entire document set.
One system design question a day, with spaced repetition so the ones you miss come back.
Get the appRelated questions
- You must index 100M chunks using a 3,072-dimension embedding in fp32, served from RAM via HNSW. The raw vectors alone take about 1.2 TB before graph overhead. You need to cut memory several-fold while losing at most a point or two of recall@10. Which approach fits?
- A multi-tenant SaaS app stores all tenants in one vector index with a tenant_id field. Queries fetch the top 20 chunks by similarity and then drop chunks from other tenants. Small tenants often get 0-2 results. Why does this happen, and what is the fix?
- Runbooks are split into fixed 300-token chunks with no overlap. When a procedure step starts near the end of one chunk and finishes in the next, answers come back incomplete. What is the lowest-cost fix?
- Your offline eval shows recall@5 of 92% on a labeled set, so the gold passage is in the context for nearly every question. Yet human graders mark 35% of answers wrong, and the wrong answers often ignore or contradict the gold passage. Where should the team invest next?
- Your product requires every RAG answer to include clickable source citations that users can verify. How should those citations be produced?
- A runbook RAG system uses fixed 500-token chunks with no overlap. Users who ask for a multi-step recovery procedure get answers that stop at step 4 of 9, and retrieval logs show the top hit ends mid-procedure. What is the most targeted fix?