Expert AI Coding Quiz – Hard Level

A challenging set of questions designed for seasoned developers working with intelligent systems.

hard quizdeep learningmachine learningartificial intelligenceneural networksalgorithm designAI ethicsprogramming puzzlesmodel tuning
Difficulty:HARD

Quiz Details

Questions9
CategoryTechnology & Programming
DifficultyHARD
Start Quiz
Progress
0/0
0%

Quiz Questions

Answer all questions below and test your knowledge.

  1. 1

    In the original Transformer architecture, what is the computational complexity of the self‑attention mechanism with respect to the sequence length n?

    Question 1
  2. 2

    Which component of Reinforcement Learning from Human Feedback directly converts human preference data into a reward model?

    Question 2
  3. 3

    When training large language models, which clipping technique preserves the direction of gradients while limiting their norm?

    Question 3
  4. 4

    Which formal verification method is most commonly applied to prove properties of neural network controllers?

    Question 4
  5. 5

    Which optimizer introduces a decoupled weight‑decay term separate from the gradient update?

    Question 5
  6. 6

    According to empirical scaling laws, how does model performance typically change when dataset size is increased by a factor of eight while keeping model size constant?

    Question 6
  7. 7

    What is the primary advantage of 4‑bit integer quantization for inference of transformer models?

    Question 7
  8. 8

    In diffusion probabilistic models, which term in the loss function corresponds to predicting the added noise at each timestep?

    Question 8
  9. 9

    Which technique most reliably mitigates token‑leakage when chaining multiple LLM calls?

    Question 9

Never miss a quiz!

Daily challenges on Telegram

Join Now