Design Retrieval Systems Using Advanced Chunking Strategies GENERATIVE-AI-ENGINEER-ASSOCIATE Practice Questions (Page 1)

Questions 1–5

1. 1foundation · easy

   Why is it important to evaluate different chunking strategies using retrieval metrics?

   • ATo ensure that all chunks have the same number of tokens for consistency.
   • BTo identify which strategy yields the best retrieval performance for the specific use case and data.
   • CTo reduce the number of chunks in the index to save storage space.
   • DTo determine which strategy produces the most aesthetically pleasing chunks.
   Reveal Solution Discussion Select an answer first
2. 2foundation · easy

   What is a potential drawback of semantic chunking compared to fixed-size chunking?

   • AIt requires additional computation to generate embeddings for similarity comparisons.
   • BIt cannot be used with documents that lack headings.
   • CIt produces chunks of uniform size, which may not fit the embedding model's token limit.
   • DIt always produces chunks that are too large for retrieval.
   Reveal Solution Discussion Select an answer first
4. 3application · medium

   A financial services firm is building a RAG system over annual reports (10-K filings). Each report has sections like 'Risk Factors', 'Management's Discussion', and 'Financial Statements'. Users often ask questions that require context from both a section heading and the content within that section. The team notices that when a chunk is retrieved, the model sometimes cannot tell which section it came from, leading to ambiguous answers. Which enhancement should they implement?

   • AUse a sliding window with a fixed window size over the raw text.
   • BAdd metadata to each chunk, such as the section heading and document title, and store it with the chunk.
   • CSwitch to semantic chunking to group related sentences across sections.
   • DIncrease the chunk size to include the entire section in one chunk.
   Reveal Solution Discussion Select an answer first
5. 4foundation · easy

   What is the purpose of adding metadata to chunks in a context-aware chunking strategy?

   • ATo reduce the storage size of the vector index.
   • BTo automatically split the chunk into smaller pieces if it is too long.
   • CTo make the chunk embeddings more similar to each other.
   • DTo provide additional context about the chunk's origin, such as the document title or section, which can be used to filter or enrich retrieval results.
   Reveal Solution Discussion Select an answer first
6. 5foundation · easy

   What is the main benefit of document-based chunking that uses headings and sections?

   • AIt eliminates the need for overlap between chunks.
   • BIt preserves the logical organization of the document, so each chunk represents a coherent topic.
   • CIt reduces the total number of chunks compared to other methods.
   • DIt ensures all chunks have the same number of tokens for consistent embedding.
   Reveal Solution Discussion Select an answer first