AWS AI Practitioner (AIF-C01) · 28% practice allocation

AWS AIF-C01: Applications of Foundation Models Practice Questions

5 free sample questions from the Applications of Foundation Models domain, with answers and explanations. The full AIF-C01 bank has 60 questions across 5 domains. Questions are scenario-based and written from the official AWS exam guide.

1.

A workflow requires the model to take multiple steps and call external tools/APIs (e.g., look up an order, then process a return). Which Amazon Bedrock capability orchestrates such multi-step, tool-using tasks?

  • A. Amazon Bedrock Agents✓ Correct
  • B. AWS Cost Explorer
  • C. Amazon Bedrock Guardrails
  • D. Amazon S3 Lifecycle policies
Explanation

Amazon Bedrock Agents orchestrate multi-step tasks, breaking a request into steps and invoking APIs/tools to complete them. Guardrails (C) enforce safety, S3 Lifecycle (D) manages object tiers, and Cost Explorer (B) analyzes spend — none orchestrates agentic tasks.

2.

A prompt gives the model an instruction with NO examples, relying on the model's general capabilities to produce the answer. What is this called?

  • A. Retrieval-augmented generation
  • B. Few-shot prompting
  • C. Zero-shot prompting✓ Correct
  • D. Fine-tuning
Explanation

Prompting with only an instruction and no examples is zero-shot prompting. Few-shot (B) includes examples, fine-tuning (D) retrains the model, and RAG (A) adds retrieved context — none of which is 'instruction only, no examples.'

3.

Which inference parameter primarily controls the RANDOMNESS/creativity of a model's output — higher values producing more varied text, lower values more focused and deterministic?

  • A. Maximum context window
  • B. The AWS Region
  • C. Number of tokens billed
  • D. Temperature✓ Correct
Explanation

Temperature adjusts randomness: higher = more diverse/creative, lower = more focused/deterministic. The context window (A) is a size limit, token billing (C) is a cost measure, and Region (B) is a deployment location — none controls output randomness.

4.

Which parameter limits the LENGTH of a model's generated response?

  • A. Temperature
  • B. The IAM policy attached to the caller
  • C. Top-p only
  • D. Maximum tokens (max output length)✓ Correct
Explanation

The maximum-tokens parameter caps how many tokens the model may generate, controlling response length. Temperature (A) and top-p (C) shape randomness/diversity, not length. An IAM policy (B) governs permissions, not output length.

5.

A developer includes several worked examples in the prompt to show the model the desired input-output pattern before asking it to handle a new case. Which prompt-engineering technique is this?

  • A. Few-shot prompting✓ Correct
  • B. Continued pre-training
  • C. Fine-tuning
  • D. Zero-shot prompting
Explanation

Providing a few examples in the prompt to demonstrate the pattern is few-shot prompting. Zero-shot (D) gives no examples, just the instruction. Fine-tuning (C) and continued pre-training (B) modify the model's weights with additional training — not an in-prompt technique.

Test yourself across every AIF-C01 domain

The free AIF-C01 practice test mixes about 20 questions weighted like the exam. The full 60-question bank adds instant grading, flashcards and a timed mock exam.

Take the free AIF-C01 practice testPractice the full bank

Study this domain

Applications of Foundation Models study guide (AIF-C01) →

Other AIF-C01 domains

← All AWS AI Practitioner study resources