Evaluating and refining prompts · Question 48 of 60
The error analysis shows that boundary values are missing in 70% of the faulty test cases for a travel expense app. Which refinement is most effective?
1 answer is correct. Hard Created with AI
Show answer
Answer
- B Explicitly request boundary values and include a short example of them in the prompt
Explanation
The most frequent type of error shows where the prompt needs to be sharpened. A clear instruction with an example steers the model specifically towards the gap.