Lookahead Decoding Quiz

Answers and explanations appear as you take the quiz.

1. What distinguishes lookahead decoding from standard speculative decoding?
   • It requires two separately trained models
   • It uses no separate draft model — only the target model itself
   • It only works for translation tasks
   • It permanently quantizes the model
2. Which numerical method underpins lookahead decoding?
   • Gradient descent
   • Fourier transform
   • K-means clustering
   • Jacobi (fixed-point) iteration
3. What are the two parallel branches that run at each step?
   • A training branch and a testing branch
   • An encoder and a decoder
   • A lookahead branch and a verification branch
   • A draft model and a target model
4. What is stored in the 'n-gram pool'?
   • Plausible multi-token sequences seen during the parallel iterations
   • The user's chat history
   • Model weights
   • Attention masks
5. What supports correctness when lookahead decoding commits a guessed n-gram prefix?
   • Using the same model makes every unverified guess correct
   • Every candidate in the n-gram pool must be committed
   • Verification against the target's greedy predictions or an appropriate distribution-preserving sampling procedure
   • A larger window proves the generated statements are true
6. What resource tradeoff is important when deploying lookahead decoding on already busy hardware?
   • Candidate generation uses no compute or memory
   • Additional parallel work may offset the benefit of fewer sequential passes
   • Every larger window must lower wall-clock latency
   • The target weights are automatically compressed to zero size
7. Why should a lookahead candidate remain provisional until target verification?
   • Parallel iteration guesses can differ from the target's valid continuation
   • Every cached n-gram has already been proven correct by storage alone
   • The lookahead branch permanently changes the target weights
   • Verification is needed only to determine the output filename
8. Does a standard training-free lookahead iteration update the target model's learned weights?
   • Yes; each candidate applies a gradient update to every parameter
   • Yes; verified n-grams become a new training dataset immediately
   • Only the tokenizer's vocabulary is retrained at every step
   • No; it changes candidate token states and verifies outputs while keeping the model weights fixed