Prompt Engineering and Augmentation GENERATIVE-AI-ENGINEER-ASSOCIATE Practice Questions (Page 5)

Questions 21–25

1. 21application · medium

   A developer is using an LLM to generate SQL queries from natural language. The first prompt 'Show me all customers' returns 'SELECT * FROM customers;' but the team needs only active customers. The developer updates the prompt to 'Show me all active customers.' The model still returns all customers. What should the developer do next?

   • AUse the same prompt but run it multiple times until the model returns the correct query.
   • BAdd an explicit instruction and an example: 'Return only rows where status = ''active''. Example: Show me active customers -> SELECT * FROM customers WHERE status = ''active'';'
   • CRemove the word 'active' from the prompt to avoid confusion.
   • DSwitch to a different model without changing the prompt.
   Reveal Solution Discussion Select an answer first
2. 22foundation · easy

   A user types: "Book a flight from Seattle to Denver on March 15th for 2 passengers." Which fields would be most important to extract as key fields for augmentation?

   • AFlight number, seat preference, and meal choice
   • BOrigin (Seattle), destination (Denver), date (March 15th), and number of passengers (2)
   • CAirline preference, ticket class, and baggage allowance
   • DUser's name, email address, and phone number
   Reveal Solution Discussion Select an answer first
4. 23application · medium

   A team is building a recipe assistant. The baseline LLM response to 'How do I make a chocolate cake?' is a long, unstructured paragraph. The team wants a concise, step-by-step list. Which prompt adjustment is most effective?

   • AAdd the constraint 'Keep it short' without any other changes.
   • BChange the prompt to 'Give me a recipe' and hope the model infers the desired format.
   • CAppend 'Please be detailed' to the original prompt.
   • DAdd the instruction 'Respond in a numbered list of steps' and provide an example of a formatted recipe.
   Reveal Solution Discussion Select an answer first
5. 24foundation · easy

   After extracting the intent, key fields, and terms from a user's request, what is the next step in the augmentation process?

   • ASend the extracted elements to a separate model for validation.
   • BStore the extracted elements in a database for future use.
   • CIntegrate these extracted elements into the prompt to provide relevant context to the LLM.
   • DDelete the original user input to avoid confusion.
   Reveal Solution Discussion Select an answer first
6. 25application · medium

   A developer is building a prompt for a movie recommendation system. The user says: 'I like action movies with a strong female lead, similar to Mad Max: Fury Road.' The developer wants to augment the prompt with context. Which integration is most effective?

   • AExtract only 'Mad Max: Fury Road' and ignore the rest.
   • BUse the raw user input as the context without any extraction.
   • CExtract the terms 'action movies', 'strong female lead', and 'Mad Max: Fury Road' and integrate them as a structured list of preferences.
   • DExtract 'I like' and 'similar to' as the key terms.
   Reveal Solution Discussion Select an answer first