3.3 Identify Grounding Techniques and Their Use Cases. GENERATIVE-AI-LEADER Practice Questions (Page 5)

Questions 21–25

1. 21foundation · easy

   What is the purpose of safety settings in a generative AI model?

   • ATo increase the model's creativity
   • BTo filter harmful or inappropriate content from model outputs
   • CTo limit the length of generated responses
   • DTo improve the model's factual accuracy
   Reveal Solution Discussion Select an answer first
2. 22foundation · easy

   Which type of data does grounding with Google Search use?

   • AWorld data
   • BSynthetic data
   • CThird-party proprietary data
   • DFirst-party enterprise data
   Reveal Solution Discussion Select an answer first
4. 23expert · hard

   A research institute wants to build a question-answering system over a proprietary corpus of scientific papers. The team needs to control the exact chunking strategy, the retrieval scoring function, and the prompt template. They also need to log which documents were retrieved for each query for audit purposes. Which approach is most appropriate?

   • AUse the RAG APIs to build a custom workflow with controlled chunking, retrieval, and prompt construction.
   • BUse Agent Search with default settings.
   • CFine-tune the model on the corpus and use default sampling.
   • DGround with Google Search to retrieve the scientific papers.
   Reveal Solution Discussion Select an answer first
5. 24foundation · easy

   What is a key benefit of using RAG for a question-answering system?

   • AIt guarantees the model will never produce incorrect answers
   • BIt eliminates the need for any external data sources
   • CIt ensures the model's responses are based on the latest available information
   • DIt reduces the computational cost of model inference
   Reveal Solution Discussion Select an answer first
6. 25foundation · easy

   What does top-p (nucleus sampling) control in text generation?

   • AThe number of tokens to generate before stopping
   • BThe safety filters applied to the output
   • CThe maximum number of tokens in the output
   • DThe set of tokens considered for sampling based on cumulative probability
   Reveal Solution Discussion Select an answer first