NEW
Font size
WorksheetsModule 4 and 5 : Mock Quiz
Total questions: 10
Worksheet time: 8mins
Which of the following content types can Generative AI tools produce?
Only text and images
Only audio and music
Text, images, video, audio, music, and code
Only video and code
What is the main purpose of tokenization in natural language processing (NLP)?
To remove all stop words from a document
To convert text into numbers that machines can process
To train deep learning models from scratch
To improve hardware performance
Which Azure service helps deploy generative AI applications?
Azure Cosmos DB
Azure Function
Azure SQL Database
Azure AI Foundry
To make a general AI model better at answering legal contract questions, a law firm trains it further on 1,000 past contracts. What is this process called?
Grounding data
Fine-tuning
Retrieval augmentation
Security control implementation
Which element is required for an AI Agent to perform actions such as sending emails or updating a calendar?
An Action Tool
A multi-agent system
A Knowledge Tool
A specific User Prompt
Large Language Models (LLMs) generate text by predicting what?
The next token in a sequence
The fastest search result
The longest word in a sentence
A random vocabulary entry
A user wants their generative AI application to only respond in a professional and concise manner. Which type of prompt should be used to establish this constraint?
A User Prompt
A Conversation History Prompt
A System Prompt
A Grounding Prompt
What is the main purpose of using Retrieval Augmented Generation (RAG) in a generative AI application?
To train a new language model from scratch
To make the model generate faster responses
To reduce the size of the input prompt
To add relevant external information to the prompt so the answer is more accurate and grounded
What does Azure AI Foundry allow developers to do?
Customize and use language models in apps
Store files in the cloud
Create websites
Manage virtual machines
What is the purpose of risk and safety evaluators in generative AI?
To improve model speed
To expand the model’s vocabulary
To reduce training cost
To detect harmful or inappropriate content
