Skip to main content

Quiz: Module 8 - Debugging and Optimizing Prompts

Test your knowledge with 30 multiple-choice questions.

Questions 1-10: Common Failures​

Q1: What causes hallucinations?​

  • A) Slow internet
  • B) Model inventing facts it doesn't know
  • C) Too many tokens
  • D) Wrong language

Answer: B - Hallucinations occur when models confidently state incorrect information.

Q2-10 Answers​

Q2: C (Ask for uncertainty levels), Q3: A (Be more specific for off-topic), Q4: B (Provide format templates), Q5: C (Add word limits for verbosity), Q6: A (Specify tone explicitly), Q7: B (Use checklists for completeness), Q8: C (Diagnostic questions help), Q9: A (Identify failure type first), Q10: B (Within model capabilities)

Questions 11-20: Iterative Refinement​

Q11: What is the iteration cycle?​

  • A) Write once, use forever
  • B) Write β†’ Test β†’ Analyze β†’ Modify β†’ Repeat
  • C) Only test at the end
  • D) Skip testing

Answer: B - The cycle is Write β†’ Test β†’ Analyze β†’ Identify issues β†’ Modify β†’ Repeat.

Q12-20 Answers​

Q12: C (A/B test significant changes), Q13: A (Version control prompts), Q14: B (One change at a time), Q15: C (Document what works), Q16: A (Measure success objectively), Q17: B (Track changes over versions), Q18: C (Compare systematically), Q19: A (Incremental improvements), Q20: B (Keep prompt history)

Questions 21-30: Testing and Evaluation​

Q21: What should be automated in testing?​

  • A) All evaluation
  • B) Format/structure checks
  • C) Tone assessment only
  • D) Nothing

Answer: B - Automate format, structure, and pattern checks; human review for nuance.

Q22-30 Answers​

Q22: C (Define metrics first), Q23: A (Include edge cases), Q24: B (Rubrics for consistency), Q25: C (Human review for nuance), Q26: A (Test suite with diverse cases), Q27: B (Scoring 1-5 common), Q28: C (Pass/fail for format), Q29: A (Reviewer guidelines help), Q30: B (Continuous improvement)

← Back to Module 8 | Proceed to Module 9 β†’