3.2 Describe Prompt Engineering Techniques and How They Drive Better Results. GENERATIVE-AI-LEADER Practice Questions (Page 4)

Questions 16–20

1. 16foundation · easy

   Why is prompt engineering critical for effective interaction with large language models?

   • ABecause the quality of the output is heavily influenced by how the prompt is structured
   • BBecause prompt engineering is the only way to control an LLM's behavior
   • CBecause LLMs cannot generate text without a prompt
   • DBecause LLMs require prompts to be written in a specific programming language
   Reveal Solution Discussion Select an answer first
2. 17application · medium

   A product team is evaluating how to improve the quality of an LLM-based feature that generates email responses. They notice that the responses are often too verbose and miss key points. They want to understand the most effective way to guide the model's output without changing the underlying model. What should they focus on?

   • ARefining the prompts to be more specific and include constraints.
   • BIncreasing the model's temperature setting to make outputs more creative.
   • CSwitching to a larger model with more parameters.
   • DAdding more training data to the model.
   Reveal Solution Discussion Select an answer first
4. 18foundation · easy

   What is zero-shot prompting?

   • AProviding the model with a single example to guide its response
   • BAsking the model to perform a task without providing any examples
   • CProviding the model with a role to adopt before the task
   • DProviding the model with multiple examples to guide its response
   Reveal Solution Discussion Select an answer first
5. 19foundation · easy

   When should chain-of-thought prompting be used?

   • AWhen the task is simple and can be answered directly
   • BWhen the model needs to use external tools
   • CWhen the model needs to adopt a specific persona
   • DWhen the task requires multi-step reasoning or arithmetic
   Reveal Solution Discussion Select an answer first
6. 20expert · hard

   A healthcare startup is using an LLM to classify patient feedback into categories like 'appointment scheduling', 'billing', and 'clinical care'. They have a small labeled dataset of 10 examples per category. They want to maximize accuracy but also keep the prompt length short to reduce latency and cost. Which approach should they take?

   • AUse all 30 examples in the prompt.
   • BUse one example per category in the prompt.
   • CUse zero-shot prompting with a detailed description of each category.
   • DUse a few carefully selected examples per category, balancing accuracy and prompt length.
   Reveal Solution Discussion Select an answer first