Generative AI Concepts Quiz - Competitive Exams AWS_AI_Practitioner

Free practice quiz: Generative AI Concepts (Competitive Exams AWS_AI_Practitioner). 73 multiple-choice questions with answers and explanations. Play online, no sign-up needed.

This Competitive Exams AWS_AI_Practitioner practice quiz has 73 multiple-choice questions on Generative AI Concepts. Try the 5 sample questions below (tap an option to check yourself), then play the full quiz online. No sign-up needed.

Sample questions

1. What is the primary purpose of a foundation model (FM) in generative AI?

   1. To store relational data for enterprise applications
   2. To perform only one specific task like translation
   3. To replace all traditional machine learning algorithms
   4. To serve as a large, pre-trained model that can be adapted to many downstream tasks
   Show answer

   Answer: D. To serve as a large, pre-trained model that can be adapted to many downstream tasks

2. What is 'context window' in a large language model?

   1. The time limit for model training
   2. The number of users the model can serve concurrently
   3. The maximum amount of text (tokens) the model can consider at once
   4. The screen size of the user interface
   Show answer

   Answer: C. The maximum amount of text (tokens) the model can consider at once

3. Which term describes a model that has been trained on a large, broad dataset and can be adapted to many downstream tasks?

   1. Rule-based system
   2. Expert system
   3. Narrow AI model
   4. Foundation model
   Show answer

   Answer: D. Foundation model

4. What does 'in-context learning' refer to in large language models?

   1. The model's ability to learn from user feedback
   2. The model's ability to learn from examples provided in the prompt without weight updates
   3. The model's ability to learn from its own generated text
   4. The model's ability to learn from external databases
   Show answer

   Answer: B. The model's ability to learn from examples provided in the prompt without weight updates

5. What is the key function of a cross-encoder reranker in a retrieval-augmented generation system?

   1. To fine-tune the generator on user feedback
   2. To generate the initial vector embeddings for all documents
   3. To rescore retrieved passages against the query for better relevance ranking
   4. To compress the model for edge deployment
   Show answer

   Answer: C. To rescore retrieved passages against the query for better relevance ranking

More questions in this quiz

Answers and explanations are in the full quiz.

1. Which technique involves further training a pre-trained foundation model on a smaller, task-specific dataset?
2. In the context of large language models, what does 'inference' refer to?
3. What is a 'token' in the context of generative AI language models?
4. Which of the following best describes 'prompt engineering'?
5. What is a key characteristic of few-shot prompting?
6. Which approach involves training a model to learn from human preferences to align outputs with desired behavior?
7. What is 'hallucination' in the context of generative AI?
8. What is the main purpose of retrieval-augmented generation (RAG)?
9. Which statement best describes 'zero-shot' prompting?
10. What is 'temperature' in the context of generative AI model sampling?
11. Which of the following is an example of a multimodal generative AI model?
12. What is the primary role of an embedding in generative AI applications?
13. Which AWS service is specifically designed to provide access to foundation models for generative AI applications?
14. Which technique helps reduce the cost and latency of serving large foundation models by using smaller, distilled models?
15. What is the key difference between fine-tuning and prompt engineering?
16. Which of the following best describes 'chain-of-thought' prompting?
17. What is a potential risk of using generative AI models that produce content based on biased training data?
18. Which metric is commonly used to evaluate the quality and relevance of text generated by a language model?
19. What is 'prompt injection' in the context of generative AI security?
20. Which of the following is a benefit of using retrieval-augmented generation (RAG) over fine-tuning for frequently changing knowledge?
21. What does 'model quantization' typically achieve in generative AI deployment?
22. Which concept refers to the ability of a generative AI model to perform tasks it was not explicitly trained on?
23. What is the main purpose of guardrails in generative AI applications?
24. Which component of a generative AI system is responsible for converting raw text into a numerical vector representation that captures semantic meaning?
25. In the context of generative AI, what does the term 'latent space' refer to?
26. Which type of generative model uses a generator and a discriminator trained in opposition to each other?
27. What is the primary function of the self-attention mechanism in a transformer-based generative model?
28. What does 'top-p sampling' (nucleus sampling) control during text generation?
29. Which of the following best describes a 'diffusion model' in generative AI?
30. What is the role of a 'decoder' in an encoder-decoder generative architecture?
31. Which of the following is a common use case for generative AI in the field of software development?
32. What does 'beam search' refer to in the context of generative AI text generation?
33. Which of the following best describes 'transfer learning' in generative AI?
34. What is a 'diffusion model' primarily used for in generative AI?
35. Which term describes the phenomenon where a generative model produces the same output for different but semantically similar prompts?
36. What is the purpose of 'positional encoding' in transformer models?
37. Which of the following is a key advantage of using a generative AI model over a traditional rule-based system for content creation?
38. What does 'fine-tuning' typically involve in the context of generative AI?
39. What is 'model distillation' in the context of generative AI?
40. Which of the following is a common application of generative AI in healthcare?
41. Which of the following best describes a 'diffusion model' compared to a GAN?
42. What is the primary purpose of 'regularization' in training generative AI models?
43. Which of the following is a potential ethical concern with generative AI?
44. What does 'autoregressive' mean in the context of generative language models?
45. Which of the following best describes 'model drift' in generative AI?
46. In a generative AI application, what is the main role of a vector database?
47. Which sampling parameter directly controls how many of the most probable next tokens are considered before random selection?
48. What is the primary purpose of a system prompt in a conversational generative AI assistant?
49. Which approach best describes 'instruction tuning' of a foundation model?
50. What does the term 'grounding' mean when building a generative AI application?
51. Which of the following is a characteristic of a Mixture of Experts (MoE) architecture in large language models?
52. What is the main purpose of a tokenizer in a generative language model pipeline?
53. Which technique involves adding human-written examples of desired behavior directly into the prompt to guide the model?
54. What does the 'context window' limitation mean for a RAG pipeline?
55. Which of the following best describes 'multi-agent' generative AI systems?
56. What is the primary role of a reward model in reinforcement learning from human feedback (RLHF)?
57. Which of the following is a common technique for reducing latency in generative AI inference without retraining the model?
58. What is a 'LoRA' (Low-Rank Adaptation) used for in generative AI?
59. What does the 'BLEU' score primarily evaluate in generative AI?
60. Which concept describes a generative AI model's tendency to produce plausible-sounding but factually incorrect information?

More quizzes

• Responsible AI▶ Play
• AI/ML Fundamentals▶ Play
• AWS AI Services▶ Play

Browse all AWS_AI_Practitioner quizzes

More to play