AI and LLMs · Junior · Question 19 of 75
Why are documents split into chunks for RAG?
- 1Because models can't read documents longer than one page
- 2To make the documents easier for people to read
- 3To retrieve only the relevant parts and fit them in the context
Show the answer
Correct answer: To retrieve only the relevant parts and fit them in the context
Explanation
Chunk size is a trade-off: too small loses context, too large brings in noise and uses tokens.