Your question is Measuring LLM Output Quality. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
How do you objectively measure the quality of LLM output?
Describe a practical evaluation framework that combines task-specific criteria, representative test data, automated metrics, human review, and error analysis. Explain how you would validate that the evaluation is reliable, aligned with user needs, and robust to nondeterministic model behavior.
