RLHF Fine-Tuning Quiz

5 questions • Pass: 70% • +25 pts

Quiz covering Advanced Training

RLHF Fine-Tuning Quiz

5 questions | Pass: 70% | Earn 25 points

Questions in this quiz

A preview of the 5 questions covered. Start the quiz above to answer them, check your score, and read the explanations.

  1. 1

    What is the primary purpose of Reinforcement Learning from Human Feedback (RLHF) in training Large Language Models?

  2. 2

    In the RLHF pipeline, what is the role of the Reward Model?

  3. 3

    Which algorithm is most commonly used to update the policy model during the RLHF phase to prevent the model from deviating too far from the original supervised fine-tuned model?

  4. 4

    Why is a KL-divergence penalty typically included in the PPO objective function during RLHF?

  5. 5

    When training a Reward Model, why is it common practice to use pairwise comparisons (A is better than B) rather than absolute scalar ratings (1-5 stars)?