Get ready for the GARP Risk and AI Exam with flashcards and multiple choice questions. Each question comes with hints and explanations. Prepare for success!

Multiple Choice

Which prompting style provides examples?

Providing examples in the prompt lets the model learn the task by imitation. In few-shot prompting, you include a small number of example input-output pairs before asking for a new input. Those demonstrations show the expected format, the kind of reasoning, and the typical answer length, so the model generalizes from them to produce a matching response for the new case. This is exactly what sets few-shot prompting apart: it leverages concrete demonstrations to guide the model. Zero-shot prompting provides no examples—it relies on instruction alone. Prompt engineering is a broader term about crafting prompts for better performance, not specifically about including demonstrations. Chain-of-thought prompting focuses on eliciting step-by-step reasoning in the model’s output, which can involve demonstrations but aims at reasoning traces rather than just showing how to respond.

Providing examples in the prompt lets the model learn the task by imitation. In few-shot prompting, you include a small number of example input-output pairs before asking for a new input. Those demonstrations show the expected format, the kind of reasoning, and the typical answer length, so the model generalizes from them to produce a matching response for the new case. This is exactly what sets few-shot prompting apart: it leverages concrete demonstrations to guide the model. Zero-shot prompting provides no examples—it relies on instruction alone. Prompt engineering is a broader term about crafting prompts for better performance, not specifically about including demonstrations. Chain-of-thought prompting focuses on eliciting step-by-step reasoning in the model’s output, which can involve demonstrations but aims at reasoning traces rather than just showing how to respond.