HOTSPOT (Drag and Drop is not supported)Select the answer that correctly completes the sentence.Hot Area:
Answer(s): A
Box: API endpoint An API endpoint URL is what applications call to send requests to a deployed generative AI model.Key Components of Deployed Model Requests API Endpoint: The specific web address hosted by a cloud provider or server.HTTP Method: Typically a POST request used to send data to the model.Authentication: A secure API key passed in the request header to verify identity.Payload: A JSON body containing the prompt, token limits, and temperature settings.
https://help.sap.com/docs/sap-ai-core/generative-ai/consume-models
HOTSPOT (Drag and Drop is not supported)For each of the following statements, select Yes if the statement is true. Otherwise, select No.Note: Each correct selection is worth one point.Hot Area:
Box 1: Yes Yes, generative AI models like large language models operate precisely by predicting the next token in a sequence using statistical patterns learned during trainingHow Next-Token Prediction Works Tokenization: Input text splits into small chunks called tokens, which can be whole words, parts of words, or characters.Context Processing: The model reads the existing sequence of tokens to evaluate the surrounding context.Probability Distribution: It calculates statistical likelihoods for every possible next token in its vocabulary.Iterative Loop: A single token gets chosen and added to the text, then the model repeats the entire process to generate the next one.Box 2: Yes Yes, a system prompt is used to influence the behavior, tone, and boundaries of a generative AI model. It acts as a foundational, high-level instruction set provided before user inputs are processed.Box 3: No Increasing the temperature makes generative AI responses more random, creative, and varied, while lowering the temperature makes them more deterministic, predictable, and consistent.Low Temperature (near 0): The model picks the most likely words every time. This creates consistent and factual answers.High Temperature (closer to 1 or higher): The model takes more risks and picks less common words. This creates unique and surprising answers, but can cause mistakes or off-topic text.
https://www.ucf.edu/artificial-intelligence/how-do-generative-ai-tools-like-chatgpt-work/
Your company provides users with clear information about an AI system’s purpose, functionality, and limitations.Which Microsoft responsible AI principle is this an example of?
Transparency is the Microsoft responsible AI principle that provides clear information about an AI system's purpose, how it works, and its limitations.System Purpose: Explains what the AI is designed to do and how it helps users.Functionality: Shows how the model makes decisions or generates content.Limitations: Warns users about potential errors, blind spots, or times when the AI might be wrong.User Awareness: Ensures people know they are interacting with an AI system.
https://www.microsoft.com/en-us/ai/principles-and-approach
Box 1: No Transparency under Microsoft's responsible AI principles focuses on making sure people understand how the system works and why it makes decisions, rather than increasing performance or throughput.Box 2: Yes Yes, transparency is actively supported and defined under Microsoft’s Responsible AI framework when users and stakeholders are clearly informed about an AI system's purpose, intended use, capabilities, and limitations.Box 3: Yes Yes, providing meaningful explanations of how an AI system generates its results directly supports the principle of transparency under Microsoft's responsible AI framework.Builds trust: Users feel safe when they understand system actions.Aids understanding: People see the data and logic behind choices.Shares limitations: Users know what the AI cannot do.
https://medium.com/accredian/responsible-ai-the-path-to-ethical-and-inclusive-technology-1b9ed83a0884
You are building an AI solution that processes customer information.Which task is an example of the Microsoft responsible AI principle of privacy and security?
Encrypting stored customer data and restricting system access using role-based controls are exact and correct examples of the Microsoft responsible AI principle of privacy and security.Key Goals of Privacy and Security Protect data: Keep user and customer data safe from unauthorized eyes.Control access: Ensure only the right people use the system.Build trust: Make users feel secure when they use AI tools.
Post your Comments and Discuss Microsoft AI-901 exam prep with other Community members:
Our website is free, but we have to fight against AI bots and content theft. We're sorry for the inconvenience caused by these security measures. You can access the rest of the AI-901 content, but please register or login to continue.