NEW
Font size
WorksheetsEmbeddings & RAG Quiz - Answer Key
Total questions: 15
Worksheet time: 8mins
What do embeddings represent?
Numerical form of meaning
Fixed-length token ids
Raw text sequences
One-hot label vectors
Similar vectors indicate:
Similar meaning
Identical tokens
Equal magnitudes
Same document IDs
Which method is used to measure similarity between vectors?
Cosine similarity
Edit distance
Hamming parity
Levenshtein score
Which system stores and searches embeddings efficiently?
Vector database
Relational spreadsheet
Full-text indexer
Blob object store
Which of the following is NOT a vector database?
Oracle MySQL
Pinecone service
FAISS index
Milvus engine
Word2Vec and GloVe are examples of:
Word-level embeddings
Sentence-level prompts
Character hash features
Topic probability models
Embedding enables:
Semantic similarity
Exact keyword match
Syntactic parsing only
Lossless compression
Metadata in vector databases is used for:
Source mapping
Gradient clipping
Model checkpointing
GPU memory paging
Which tool is commonly used for PDF/DOCX ingestion in RAG pipelines?
LangChain loaders
POS taggers
HTML minifiers
OpenGL shaders
Ideal text chunk length for embedding pipelines is:
200–500 tokens
10–30 tokens
900–1500 tokens
1–5 tokens
Which step converts text chunks into embeddings?
Embedding generation
Vector indexing
Document splitting
Prompt templating
Indexing in a vector database helps in:
Fast similarity search
Data encryption only
Version control diffs
Schema normalization
Retrieval during querying means:
Fetching top‑k similar chunks
Merging all sources blindly
Generating new training data
Compiling query to bytecode
Vector search vs keyword search primarily differs because:
Vector equals semantic similarity
Vector equals exact string match
Vector equals binary hashing
Vector equals regex filtering
Vector databases are essential for:
RAG pipelines
Image rendering
Network routing
Operating system boot
