RAG Basics MCQ Questions with Answers (Latest 2026)

Q1. Which option best describes retrieval-augmented generation?

A. Pure search only. B. Pure logs. C. Augment LLM with retrieved context. D. Pure parametric memory.

Select an answer to check.

Answer: Augment LLM with retrieved context.

Here, Augment LLM with retrieved context. is the right choice. Reduces hallucinations. It aligns directly with what the question asks about which option best describes retrieval-augmented generation. A quick elimination of partially true options helps confirm it.

Q2. What is the primary purpose of retrieval-augmented generation?

A. Pure logs. B. Pure search only. C. Augment LLM with retrieved context. D. Pure parametric memory.

Select an answer to check.

Answer: Augment LLM with retrieved context.

In this case, Augment LLM with retrieved context. is correct. Reduces hallucinations. It aligns directly with what the question asks about what is the primary purpose of retrieval-augmented generation. A quick elimination of partially true options helps confirm it.

Q3. Which statement about retrieval-augmented generation is most accurate?

A. Pure logs. B. Augment LLM with retrieved context. C. Pure search only. D. Pure parametric memory.

Select an answer to check.

Answer: Augment LLM with retrieved context.

The best option here is Augment LLM with retrieved context.. Reduces hallucinations. It aligns directly with what the question asks about which statement about retrieval-augmented generation is most accurate. A quick elimination of partially true options helps confirm it.

Q4. How is retrieval-augmented generation best characterized?

A. Pure logs. B. Augment LLM with retrieved context. C. Pure search only. D. Pure parametric memory.

Select an answer to check.

Answer: Augment LLM with retrieved context.

For this question, Augment LLM with retrieved context. is correct. Reduces hallucinations. It aligns directly with what the question asks about how is retrieval-augmented generation best characterized. A quick elimination of partially true options helps confirm it.

Q5. Which option best describes a vector database?

A. A pure log. B. A relational DB only. C. A pure cache. D. Stores embeddings for similarity search.

Select an answer to check.

Answer: Stores embeddings for similarity search.

Stores embeddings for similarity search. is the correct answer here. Pinecone, Weaviate, pgvector, etc. It aligns directly with what the question asks about which option best describes a vector database. A quick elimination of partially true options helps confirm it.

Q6. What is the primary purpose of a vector database?

A. A relational DB only. B. Stores embeddings for similarity search. C. A pure cache. D. A pure log.

Select an answer to check.

Answer: Stores embeddings for similarity search.

Here, Stores embeddings for similarity search. is the right choice. Pinecone, Weaviate, pgvector, etc. This matches the core idea being tested around what is the primary purpose of a vector. A quick elimination of partially true options helps confirm it.

Q7. Which statement about a vector database is most accurate?

A. Stores embeddings for similarity search. B. A pure cache. C. A relational DB only. D. A pure log.

Select an answer to check.

Answer: Stores embeddings for similarity search.

In this case, Stores embeddings for similarity search. is correct. Pinecone, Weaviate, pgvector, etc. This matches the core idea being tested around which statement about a vector database is most. A quick elimination of partially true options helps confirm it.

Q8. How is a vector database best characterized?

A. A pure log. B. A relational DB only. C. Stores embeddings for similarity search. D. A pure cache.

Select an answer to check.

Answer: Stores embeddings for similarity search.

The best option here is Stores embeddings for similarity search.. Pinecone, Weaviate, pgvector, etc. This matches the core idea being tested around how is a vector database best characterized. A quick elimination of partially true options helps confirm it.

Q9. Which option best describes embeddings?

A. Pure logs. B. Pure tokens. C. Dense vector representations of text. D. Sparse only.

Select an answer to check.

Answer: Dense vector representations of text.

For this question, Dense vector representations of text. is correct. Generated by embedding models. This matches the core idea being tested around which option best describes embeddings. A quick elimination of partially true options helps confirm it.

Q10. What is the primary purpose of embeddings?

A. Dense vector representations of text. B. Pure tokens. C. Pure logs. D. Sparse only.

Select an answer to check.

Answer: Dense vector representations of text.

Dense vector representations of text. is the correct answer here. Generated by embedding models. This matches the core idea being tested around what is the primary purpose of embeddings. A quick elimination of partially true options helps confirm it.

Q11. Which statement about embeddings is most accurate?

A. Pure logs. B. Dense vector representations of text. C. Pure tokens. D. Sparse only.

Select an answer to check.

Answer: Dense vector representations of text.

Here, Dense vector representations of text. is the right choice. Generated by embedding models. That is exactly the concept behind which statement about embeddings is most accurate in this context. A quick elimination of partially true options helps confirm it.

Q12. How is embeddings best characterized?

A. Pure tokens. B. Dense vector representations of text. C. Sparse only. D. Pure logs.

Select an answer to check.

Answer: Dense vector representations of text.

In this case, Dense vector representations of text. is correct. Generated by embedding models. That is exactly the concept behind how is embeddings best characterized in this context. A quick elimination of partially true options helps confirm it.

Q13. Which option best describes an embedding model?

A. Generates embeddings from text. B. A pure parser. C. A generative LLM only. D. A pure tokenizer.

Select an answer to check.

Answer: Generates embeddings from text.

The best option here is Generates embeddings from text.. E.g., text-embedding-3. That is exactly the concept behind which option best describes an embedding model in this context. A quick elimination of partially true options helps confirm it.

Q14. What is the primary purpose of an embedding model?

A. A pure parser. B. A generative LLM only. C. A pure tokenizer. D. Generates embeddings from text.

Select an answer to check.

Answer: Generates embeddings from text.

For this question, Generates embeddings from text. is correct. E.g., text-embedding-3. That is exactly the concept behind what is the primary purpose of an embedding in this context. A quick elimination of partially true options helps confirm it.

Q15. Which statement about an embedding model is most accurate?

A. A generative LLM only. B. A pure tokenizer. C. A pure parser. D. Generates embeddings from text.

Select an answer to check.

Answer: Generates embeddings from text.

Generates embeddings from text. is the correct answer here. E.g., text-embedding-3. That is exactly the concept behind which statement about an embedding model is most in this context. A quick elimination of partially true options helps confirm it.

Q16. How is an embedding model best characterized?

A. A pure parser. B. A generative LLM only. C. Generates embeddings from text. D. A pure tokenizer.

Select an answer to check.

Answer: Generates embeddings from text.

Here, Generates embeddings from text. is the right choice. E.g., text-embedding-3. It fits the requirement in the prompt about how is an embedding model best characterized. A quick elimination of partially true options helps confirm it.

Q17. Which option best describes approximate nearest neighbors?

A. Pure index only. B. Exact NN only. C. Pure full scan. D. Fast similarity search at scale.

Select an answer to check.

Answer: Fast similarity search at scale.

In this case, Fast similarity search at scale. is correct. HNSW, IVF. It fits the requirement in the prompt about which option best describes approximate nearest neighbors. A quick elimination of partially true options helps confirm it.

Q18. What is the primary purpose of approximate nearest neighbors?

A. Pure index only. B. Pure full scan. C. Exact NN only. D. Fast similarity search at scale.

Select an answer to check.

Answer: Fast similarity search at scale.

The best option here is Fast similarity search at scale.. HNSW, IVF. It fits the requirement in the prompt about what is the primary purpose of approximate nearest. A quick elimination of partially true options helps confirm it.

Q19. Which statement about approximate nearest neighbors is most accurate?

A. Pure index only. B. Exact NN only. C. Pure full scan. D. Fast similarity search at scale.

Select an answer to check.

Answer: Fast similarity search at scale.

For this question, Fast similarity search at scale. is correct. HNSW, IVF. It fits the requirement in the prompt about which statement about approximate nearest neighbors is most. A quick elimination of partially true options helps confirm it.

Q20. How is approximate nearest neighbors best characterized?

A. Pure index only. B. Exact NN only. C. Pure full scan. D. Fast similarity search at scale.

Select an answer to check.

Answer: Fast similarity search at scale.

Fast similarity search at scale. is the correct answer here. HNSW, IVF. It fits the requirement in the prompt about how is approximate nearest neighbors best characterized. A quick elimination of partially true options helps confirm it.

Q21. Which option best describes HNSW?

A. FAISS Flat only. B. Hierarchical Navigable Small World ANN index. C. Pure linear scan. D. IVF only.

Select an answer to check.

Answer: Hierarchical Navigable Small World ANN index.

Here, Hierarchical Navigable Small World ANN index. is the right choice. Strong recall and speed. This is the most accurate statement for which option best describes hnsw. A quick elimination of partially true options helps confirm it.

Q22. What is the primary purpose of HNSW?

A. Hierarchical Navigable Small World ANN index. B. Pure linear scan. C. IVF only. D. FAISS Flat only.

Select an answer to check.

Answer: Hierarchical Navigable Small World ANN index.

In this case, Hierarchical Navigable Small World ANN index. is correct. Strong recall and speed. This is the most accurate statement for what is the primary purpose of hnsw. A quick elimination of partially true options helps confirm it.

Q23. Which statement about HNSW is most accurate?

A. IVF only. B. FAISS Flat only. C. Hierarchical Navigable Small World ANN index. D. Pure linear scan.

Select an answer to check.

Answer: Hierarchical Navigable Small World ANN index.

The best option here is Hierarchical Navigable Small World ANN index.. Strong recall and speed. This is the most accurate statement for which statement about hnsw is most accurate. A quick elimination of partially true options helps confirm it.

Q24. How is HNSW best characterized?

A. Hierarchical Navigable Small World ANN index. B. IVF only. C. Pure linear scan. D. FAISS Flat only.

Select an answer to check.

Answer: Hierarchical Navigable Small World ANN index.

For this question, Hierarchical Navigable Small World ANN index. is correct. Strong recall and speed. This is the most accurate statement for how is hnsw best characterized. A quick elimination of partially true options helps confirm it.

Q25. Which option best describes IVF?

A. HNSW only. B. Pure linear scan. C. Pure brute force. D. Inverted File partitioning for ANN.

Select an answer to check.

Answer: Inverted File partitioning for ANN.

Inverted File partitioning for ANN. is the correct answer here. Used in FAISS. This is the most accurate statement for which option best describes ivf. A quick elimination of partially true options helps confirm it.

Q26. What is the primary purpose of IVF?

A. Pure brute force. B. HNSW only. C. Inverted File partitioning for ANN. D. Pure linear scan.

Select an answer to check.

Answer: Inverted File partitioning for ANN.

Here, Inverted File partitioning for ANN. is the right choice. Used in FAISS. It aligns directly with what the question asks about what is the primary purpose of ivf. The other options are either incomplete or contextually incorrect.

Q27. Which statement about IVF is most accurate?

A. HNSW only. B. Pure linear scan. C. Inverted File partitioning for ANN. D. Pure brute force.

Select an answer to check.

Answer: Inverted File partitioning for ANN.

In this case, Inverted File partitioning for ANN. is correct. Used in FAISS. It aligns directly with what the question asks about which statement about ivf is most accurate. The other options are either incomplete or contextually incorrect.

Q28. How is IVF best characterized?

A. Pure linear scan. B. Inverted File partitioning for ANN. C. Pure brute force. D. HNSW only.

Select an answer to check.

Answer: Inverted File partitioning for ANN.

The best option here is Inverted File partitioning for ANN.. Used in FAISS. It aligns directly with what the question asks about how is ivf best characterized. The other options are either incomplete or contextually incorrect.

Q29. Which option best describes recall?

A. Fraction of true neighbors retrieved. B. Precision only. C. Pure latency. D. Pure throughput.

Select an answer to check.

Answer: Fraction of true neighbors retrieved.

For this question, Fraction of true neighbors retrieved. is correct. Key ANN quality metric. It aligns directly with what the question asks about which option best describes recall. The other options are either incomplete or contextually incorrect.

Q30. What is the primary purpose of recall?

A. Precision only. B. Fraction of true neighbors retrieved. C. Pure latency. D. Pure throughput.

Select an answer to check.

Answer: Fraction of true neighbors retrieved.

Fraction of true neighbors retrieved. is the correct answer here. Key ANN quality metric. It aligns directly with what the question asks about what is the primary purpose of recall. The other options are either incomplete or contextually incorrect.

Q31. Which statement about recall is most accurate?

A. Pure latency. B. Pure throughput. C. Precision only. D. Fraction of true neighbors retrieved.

Select an answer to check.

Answer: Fraction of true neighbors retrieved.

Here, Fraction of true neighbors retrieved. is the right choice. Key ANN quality metric. This matches the core idea being tested around which statement about recall is most accurate. The other options are either incomplete or contextually incorrect.

Q32. How is recall best characterized?

A. Pure latency. B. Fraction of true neighbors retrieved. C. Pure throughput. D. Precision only.

Select an answer to check.

Answer: Fraction of true neighbors retrieved.

In this case, Fraction of true neighbors retrieved. is correct. Key ANN quality metric. This matches the core idea being tested around how is recall best characterized. The other options are either incomplete or contextually incorrect.

Q33. Which option best describes chunking?

A. Pure full doc. B. Pure pure tokens only. C. Split documents into retrievable units. D. Pure pages only.

Select an answer to check.

Answer: Split documents into retrievable units.

The best option here is Split documents into retrievable units.. Affects retrieval quality. This matches the core idea being tested around which option best describes chunking. The other options are either incomplete or contextually incorrect.

Q34. What is the primary purpose of chunking?

A. Pure full doc. B. Pure pure tokens only. C. Split documents into retrievable units. D. Pure pages only.

Select an answer to check.

Answer: Split documents into retrievable units.

For this question, Split documents into retrievable units. is correct. Affects retrieval quality. This matches the core idea being tested around what is the primary purpose of chunking. The other options are either incomplete or contextually incorrect.

Q35. Which statement about chunking is most accurate?

A. Pure pure tokens only. B. Pure full doc. C. Pure pages only. D. Split documents into retrievable units.

Select an answer to check.

Answer: Split documents into retrievable units.

Split documents into retrievable units. is the correct answer here. Affects retrieval quality. This matches the core idea being tested around which statement about chunking is most accurate. The other options are either incomplete or contextually incorrect.

Q36. How is chunking best characterized?

A. Split documents into retrievable units. B. Pure pure tokens only. C. Pure pages only. D. Pure full doc.

Select an answer to check.

Answer: Split documents into retrievable units.

Here, Split documents into retrievable units. is the right choice. Affects retrieval quality. That is exactly the concept behind how is chunking best characterized in this context. The other options are either incomplete or contextually incorrect.

Q37. Which option best describes chunk size?

A. Pure full doc. B. Pure 1 token. C. Tokens per chunk; balances context vs precision. D. Pure logs.

Select an answer to check.

Answer: Tokens per chunk; balances context vs precision.

In this case, Tokens per chunk; balances context vs precision. is correct. Tune to data. That is exactly the concept behind which option best describes chunk size in this context. The other options are either incomplete or contextually incorrect.

Q38. What is the primary purpose of chunk size?

A. Pure logs. B. Tokens per chunk; balances context vs precision. C. Pure 1 token. D. Pure full doc.

Select an answer to check.

Answer: Tokens per chunk; balances context vs precision.

The best option here is Tokens per chunk; balances context vs precision.. Tune to data. That is exactly the concept behind what is the primary purpose of chunk size in this context. The other options are either incomplete or contextually incorrect.

Q39. Which statement about chunk size is most accurate?

A. Tokens per chunk; balances context vs precision. B. Pure 1 token. C. Pure full doc. D. Pure logs.

Select an answer to check.

Answer: Tokens per chunk; balances context vs precision.

For this question, Tokens per chunk; balances context vs precision. is correct. Tune to data. That is exactly the concept behind which statement about chunk size is most accurate in this context. The other options are either incomplete or contextually incorrect.

Q40. How is chunk size best characterized?

A. Pure 1 token. B. Tokens per chunk; balances context vs precision. C. Pure full doc. D. Pure logs.

Select an answer to check.

Answer: Tokens per chunk; balances context vs precision.

Tokens per chunk; balances context vs precision. is the correct answer here. Tune to data. That is exactly the concept behind how is chunk size best characterized in this context. The other options are either incomplete or contextually incorrect.

Q41. Which option best describes chunk overlap?

A. No overlap always. B. Pure full doc. C. Overlap to preserve context across chunks. D. Pure logs.

Select an answer to check.

Answer: Overlap to preserve context across chunks.

Here, Overlap to preserve context across chunks. is the right choice. Improves continuity. It fits the requirement in the prompt about which option best describes chunk overlap. The other options are either incomplete or contextually incorrect.

Q42. What is the primary purpose of chunk overlap?

A. Overlap to preserve context across chunks. B. No overlap always. C. Pure full doc. D. Pure logs.

Select an answer to check.

Answer: Overlap to preserve context across chunks.

In this case, Overlap to preserve context across chunks. is correct. Improves continuity. It fits the requirement in the prompt about what is the primary purpose of chunk overlap. The other options are either incomplete or contextually incorrect.

Q43. Which statement about chunk overlap is most accurate?

A. No overlap always. B. Pure logs. C. Pure full doc. D. Overlap to preserve context across chunks.

Select an answer to check.

Answer: Overlap to preserve context across chunks.

The best option here is Overlap to preserve context across chunks.. Improves continuity. It fits the requirement in the prompt about which statement about chunk overlap is most accurate. The other options are either incomplete or contextually incorrect.

Q44. How is chunk overlap best characterized?

A. Pure full doc. B. No overlap always. C. Overlap to preserve context across chunks. D. Pure logs.

Select an answer to check.

Answer: Overlap to preserve context across chunks.

For this question, Overlap to preserve context across chunks. is correct. Improves continuity. It fits the requirement in the prompt about how is chunk overlap best characterized. The other options are either incomplete or contextually incorrect.

Q45. Which option best describes hybrid retrieval?

A. Pure logs. B. Pure vector. C. Pure BM25. D. Combine vector + keyword (BM25) search.

Select an answer to check.

Answer: Combine vector + keyword (BM25) search.

Combine vector + keyword (BM25) search. is the correct answer here. Robust across queries. It fits the requirement in the prompt about which option best describes hybrid retrieval. The other options are either incomplete or contextually incorrect.

Q46. What is the primary purpose of hybrid retrieval?

A. Pure BM25. B. Combine vector + keyword (BM25) search. C. Pure vector. D. Pure logs.

Select an answer to check.

Answer: Combine vector + keyword (BM25) search.

Here, Combine vector + keyword (BM25) search. is the right choice. Robust across queries. This is the most accurate statement for what is the primary purpose of hybrid retrieval. The other options are either incomplete or contextually incorrect.

Q47. Which statement about hybrid retrieval is most accurate?

A. Combine vector + keyword (BM25) search. B. Pure vector. C. Pure logs. D. Pure BM25.

Select an answer to check.

Answer: Combine vector + keyword (BM25) search.

In this case, Combine vector + keyword (BM25) search. is correct. Robust across queries. This is the most accurate statement for which statement about hybrid retrieval is most accurate. The other options are either incomplete or contextually incorrect.

Q48. How is hybrid retrieval best characterized?

A. Pure BM25. B. Combine vector + keyword (BM25) search. C. Pure vector. D. Pure logs.

Select an answer to check.

Answer: Combine vector + keyword (BM25) search.

The best option here is Combine vector + keyword (BM25) search.. Robust across queries. This is the most accurate statement for how is hybrid retrieval best characterized. The other options are either incomplete or contextually incorrect.

Q49. Which option best describes BM25?

A. Pure embeddings. B. Pure ANN. C. Sparse, lexical relevance scoring. D. Pure cache.

Select an answer to check.

Answer: Sparse, lexical relevance scoring.

For this question, Sparse, lexical relevance scoring. is correct. Strong baseline. This is the most accurate statement for which option best describes bm25. The other options are either incomplete or contextually incorrect.

Q50. What is the primary purpose of BM25?

A. Sparse, lexical relevance scoring. B. Pure ANN. C. Pure embeddings. D. Pure cache.

Select an answer to check.

Answer: Sparse, lexical relevance scoring.

Sparse, lexical relevance scoring. is the correct answer here. Strong baseline. This is the most accurate statement for what is the primary purpose of bm25. The other options are either incomplete or contextually incorrect.