Task Statement 2.1: Explain the Basic Concepts of Generative AI (GenAI). AIF-C01 Practice Questions (Page 4)

Questions 16–20

1. 16foundation · easy

   How does token-based pricing typically affect the cost of running inference on a large language model?

   • AThe cost is determined by the number of model parameters, not the tokens processed
   • BThe cost is a flat fee per API call, regardless of the length of the prompt or response
   • CThe cost is only incurred for the output tokens, not the input tokens
   • DThe cost is based on the number of tokens in the input prompt and the generated output
   Reveal Solution Discussion Select an answer first
2. 17application · medium

   A company is building a customer-support chatbot using an Amazon Bedrock foundation model. They want to reduce token consumption and cost while maintaining answer quality for frequently asked questions about shipping policies. The team has a curated set of 50 policy documents. Which approach best meets these requirements?

   • ARetrieve only the most relevant policy sections using embeddings and include them in the prompt
   • BUse a larger foundation model to better understand the policies without retrieval
   • CInclude all 50 policy documents in the system prompt for every request
   • DIncrease the model's max output tokens to allow more detailed responses
   Reveal Solution Discussion Select an answer first
4. 18expert · hard

   A company is building an agentic AI system to handle customer refunds. The system must check order status, verify return eligibility, and issue refunds. The team wants to use multiple specialized agents. Which pattern best supports this workflow?

   • AOne agent that performs all three tasks sequentially without delegation
   • BThree independent agents that each call the same API without coordination
   • CA single agent that uses a large language model to decide all actions without tool calls
   • DA single orchestrator agent that delegates tasks to specialized agents for order lookup, eligibility, and refund processing
   Reveal Solution Discussion Select an answer first
5. 19foundation · easy

   A company wants to automatically generate realistic product images from text descriptions for its e-commerce catalog. Which type of generative AI model is best suited for this task?

   • AAn autoencoder
   • BA recurrent neural network (RNN)
   • CA transformer-based language model
   • DA diffusion model
   Reveal Solution Discussion Select an answer first
6. 20foundation · easy

   In the foundation model lifecycle, what is the purpose of the fine-tuning stage?

   • ATo deploy the model to a production environment for inference
   • BTo evaluate the model's performance on a held-out test set
   • CTo adapt a pre-trained model to a specific domain or task using a smaller, labeled dataset
   • DTo train a model from scratch on a large, diverse corpus of internet text
   Reveal Solution Discussion Select an answer first