Exam AI-102 Topic 3 Question 249 Discussion

Explanation:

High degree of certainty - No.
The code uses default generation parameters for chat completions (e.g., default temperature # 1.0), which means responses are probabilistic and not guaranteed to be highly certain or consistent. Azure OpenAI guidance explains that parameters like temperature and top_p control randomness; to increase determinism
/certainty you would lower temperature and/or adjust top_p, which this snippet does not do.
Clarifying the user question - Yes.
Prompt engineering guidance emphasizes that clear, specific instructions lead to better outputs. Rephrasing the user message from "What is an LLM?" to "What is an LLM in the context of AI models?" adds context and intent, improving the chance of getting the exact explanatory answer you want.
Constrain with a stronger system message - Yes.
In chat completions, the system message sets behavior and has the highest priority. Tightening it to "You must answer only within the context of AI language models." narrows scope and reduces topic drift, increasing the likelihood that the model's response stays on-task (definitions of AI terms, here specifically LLMs).
Microsoft References (Azure AI / Azure OpenAI Service)
Azure OpenAI Service - Chat Completions: roles (system, user), message ordering, and usage.
Azure OpenAI Service - Prompt engineering: write clear, specific prompts; add constraints to guide responses.
Azure OpenAI Service - Sampling parameters: temperature and top_p control variability/creativity vs.
determinism.
Azure OpenAI Service - Best practices for grounding, controllability, and reducing model drift with stronger system instructions.