What is the primary purpose of text chunking in the RAG pipeline?	To break down large source documents into smaller, manageable pieces of text.
What is the function of a text embedding?	To convert text chunks into numerical vectors that capture the semantic meaning of the text.
What does the normalization step do to a generated embedding vector?	It scales the magnitude (length) of the vector to exactly 1.0.
What is a vector database optimized for?	Storing, comparing, and looking up long lists of numerical vectors (embeddings).
How is a user query prepared before it is searched against the vector database?	It is run through the exact same embedding model used for the source text chunks.
What mathematical calculation is used to determine which stored embedding is most similar to the user's query?	Cosine similarity.
What does a cosine similarity score close to 1 indicate?	The user's query and the retrieved embedding are highly similar.
How is cosine distance calculated, and what does a low value indicate?	It is calculated as 1 minus the cosine similarity; a low value indicates high similarity.
What is the final step in the RAG pipeline after the relevant text chunk is retrieved?	The user's original query and the retrieved chunk are combined into a prompt and sent to the LLM.
List the major stages of the complete RAG pipeline.	Chunking, generating embeddings, storing them in a vector database, querying/retrieval, and final generation.