AIF-C01 notes
Optimizing foundation models

Model Evaluation

9 exam-style questions on this lesson.

  1. Question 1 of 9A team evaluates AI-generated summaries and wants to know how much of the important information from the reference summary was captured. Which metric fits?
  2. Question 2 of 9Which metric focuses on precision and includes a brevity penalty for overly short output?
  3. Question 3 of 9A model writes "The automobile is quick" where the reference says "The car is fast". Which metric is most likely to recognize these as equivalent?
  4. Question 4 of 9Which ROUGE variant measures the longest common subsequence, reflecting coherence and order?
  5. Question 5 of 9What does ROUGE-2 measure?
  6. Question 6 of 9What is a limitation of ROUGE and BLEU?
  7. Question 7 of 9. Select two.Which TWO statements about BERTScore are correct? (Select TWO.)
    0 of 2 selected
  8. Question 8 of 9. Select two.A company is evaluating machine translation output against human reference translations. Which TWO metrics can it use? (Select TWO.)
    0 of 2 selected
  9. Question 9 of 9In the fashion retail case, why did the company care about ROUGE, BLEU, and BERTScore?