Table of Contents
Question
You’re trying to get better at prompt engineering, so you decided to try a new technique. You say, “Write a 500-word essay on large language models and hallucinations from the perspective of a computer science graduate student at a university.” What technique are you using here?
A. You are using a compression technique by limiting the results to 500 words.
B. You are taking an adversarial approach to get both sides of the story.
C. You’re brainstorming with the system about large language models and hallucinations.
D. You are using role-playing to get more accurate responses.
Answer
D. You are using role-playing to get more accurate responses.
Explanation
The correct answer is D. You are using role-playing to get more accurate responses.
Role-playing is a technique that involves specifying a persona or a scenario for the system to adopt when generating a response. By providing details such as the perspective, the background, the goal, or the tone of the response, you can guide the system to produce more accurate and relevant content. Role-playing can also help to avoid generic or vague responses that do not match the context or the purpose of the prompt .
In this case, you are asking the system to write an essay from the perspective of a computer science graduate student at a university. This implies that the system should use academic language, cite relevant sources, and demonstrate some knowledge and understanding of large language models and hallucinations. By role-playing as a graduate student, the system can generate a more realistic and informative essay than if it were to write from a different or unspecified perspective.
The other options are incorrect because they do not describe the technique you are using here.
- A compression technique is a technique that involves reducing the length or size of the output without losing important information or meaning. By limiting the results to 500 words, you are not compressing an existing output, but rather setting a constraint on the desired output.
- An adversarial approach is a technique that involves challenging or testing the system with difficult or controversial prompts, or evaluating the system’s responses with critical feedback. By asking for an essay on large language models and hallucinations, you are not necessarily taking an adversarial approach, unless you intend to critique or dispute the system’s arguments or claims.
- Brainstorming with the system is a technique that involves generating multiple ideas or suggestions for a given topic or problem, without judging or filtering them. By asking for an essay on large language models and hallucinations, you are not brainstorming with the system, but rather requesting a structured and coherent piece of writing.
Reference
- Prompt engineering techniques with Azure OpenAI – Azure OpenAI Service | Microsoft Learn
- Prompt engineering overview | Microsoft Learn
- Prompting Techniques | Prompt Engineering Guide (promptingguide.ai)
- Large language models and the perils of their hallucinations | Critical Care | Full Text (biomedcentral.com)
- A Gentle Introduction to Hallucinations in Large Language Models – MachineLearningMastery.com
- [2305.15852] Self-contradictory Hallucinations of Large Language Models: Evaluation, Detection and Mitigation (arxiv.org)
The latest Generative AI Skills Initiative certificate program actual real practice exam question and answer (Q&A) dumps are available free, helpful to pass the Generative AI Skills Initiative certificate exam and earn Generative AI Skills Initiative certification.