Guide-linked quiz · Medium Level
Guided Beam Search with Constraints Quiz
Covers beam search, lexical and grammar constraints, dynamic beam allocation, and the distinction between syntax and semantic validity.
Question 1 of 8
What does plain beam search keep at each generation step?
Preview every question8 questions
Answers and explanations appear as you take the quiz.
- What does plain beam search keep at each generation step?
- The top-k most probable partial sequences
- Only the single most probable token
- All possible sequences
- A random sample of sequences
- How do lexically constrained methods like Grid Beam Search organize beams?
- By output length only
- By how many required constraints each beam has satisfied
- Alphabetically
- By the model's confidence alone
- In grammar-constrained decoding, how is invalid output prevented?
- By deleting bad tokens after generation
- By retraining the model on valid examples
- By masking token probabilities so only grammar-valid tokens are allowed
- By lowering the temperature
- What problem does Dynamic Beam Allocation solve?
- It removes the need for constraints
- It makes the model larger
- It eliminates beam search entirely
- It efficiently shares beam slots across levels of constraint progress
- When can grammar-constrained decoding produce syntactically valid JSON?
- When the grammar constraint is enforced through completion and the output is not truncated
- Whenever JSON is mentioned in the prompt
- Even if the runtime stops before required closing tokens
- Regardless of whether the active grammar matches the intended schema
- Why can hard constraints crowd out good sequences without special handling?
- They make all beams identical
- High-probability beams that ignore constraints could dominate the top-k
- They slow the GPU
- They increase the vocabulary size
- A lexical constraint requires a translation to include “bridge” and “river.” What does constrained beam search enforce?
- The output has exactly two words
- The model uses only one beam
- The completed output includes the required terms
- The translation is factually accurate
- A grammar-constrained generator returns JSON that parses successfully. What does that not establish by itself?
- That the output follows the grammar
- That required fields can be present
- That the syntax is valid
- That the values are factually correct or appropriate for the application
Keep testing yourself
More quizzes picked for your level.