3.2 Describe Prompt Engineering Techniques and How They Drive Better Results. GENERATIVE-AI-LEADER Practice Questions (Page 3)

Questions 11–15

1. 11foundation · easy

   What is the primary purpose of prompt engineering in the context of large language models?

   • ATo train the model on new data
   • BTo craft inputs that guide the model toward more accurate, relevant, and useful outputs
   • CTo replace the need for model fine-tuning
   • DTo design the model's architecture
   Reveal Solution Discussion Select an answer first
2. 12application · medium

   A data analytics team wants to use an LLM to generate a monthly report that summarizes sales data, identifies trends, and provides recommendations. The raw data is large and unstructured. The team wants to ensure the final report is accurate and well-structured. Which approach should they take?

   • AUse zero-shot prompting to ask for the report directly.
   • BUse role prompting to ask the model to act as a data analyst.
   • CAsk the model to generate the entire report in a single prompt with all the data.
   • DUse prompt chaining to first summarize the data, then identify trends, and finally generate recommendations.
   Reveal Solution Discussion Select an answer first
4. 13foundation · easy

   What impact does prompt engineering have on task performance with LLMs?

   • AIt only affects the speed of the model's response
   • BIt has no impact because LLMs are already perfect at understanding any input
   • CIt can significantly improve the accuracy, relevance, and usefulness of the model's responses
   • DIt can improve output quality but has no effect on task accuracy
   Reveal Solution Discussion Select an answer first
5. 14expert · hard

   A legal tech company is building an LLM application that reviews contracts and extracts key clauses, then generates a summary and a risk assessment. The contract text is long and complex, and the company has strict requirements for accuracy and traceability. They also need to minimize token usage to control costs. Which approach best balances accuracy, traceability, and cost?

   • AUse a single prompt with the full contract and ask for the summary and risk assessment.
   • BUse few-shot prompting with examples of contract summaries.
   • CUse chain-of-thought prompting to reason through the entire contract in one prompt.
   • DUse prompt chaining: first extract clauses, then summarize, then assess risks.
   Reveal Solution Discussion Select an answer first
6. 15foundation · easy

   What is role prompting?

   • AAssigning a specific persona or role to the model
   • BEncouraging the model to reason step-by-step
   • CProviding examples of the desired output
   • DBreaking a task into sequential prompts
   Reveal Solution Discussion Select an answer first