Prompt engineering: Worked Example — Experimentation (NVIDIA-Certified Associate: Generative AI LLM)

Prompt Engineering: A Worked Example for NVIDIA-Certified Associate Generative AI LLM

Prompt engineering is a critical skill within the Experimentation domain of the NVIDIA-Certified Associate: Generative AI LLM exam. It involves designing and refining input prompts to effectively guide large language models (LLMs) to produce desired outputs. This worked example demonstrates a step-by-step approach to prompt engineering in a realistic scenario, illustrating the reasoning and concrete actions involved.

Scenario

You are developing a customer support chatbot using an LLM. The goal is to generate concise, polite, and accurate responses to common customer inquiries about order status.

Step 1: Define the Task and Desired Output

Clearly specify what the prompt should achieve. Here, the LLM must:

• Understand the customer query about order status.
• Respond politely and concisely.
• Include relevant order information if provided.

Step 2: Create an Initial Prompt

Start with a simple instruction prompt to test the model’s baseline behavior.

Example Prompt 1

"You are a helpful customer support assistant. Answer the following question politely and briefly: 'Where is my order?'

Run this prompt through the LLM and observe the output.

Step 3: Analyze the Output

Suppose the model replies:

"Your order is being processed and should arrive soon."

This is polite and concise but lacks personalization and actionable details.

Step 4: Refine the Prompt with Context

Incorporate placeholders for dynamic data (e.g., order number, status) to improve relevance.

Example Prompt 2

"You are a helpful customer support assistant. A customer asks: 'Where is my order?' The order number is 12345, and the current status is 'shipped'. Provide a polite and concise response including this information."

Step 5: Test and Evaluate the New Output

Expected model response:

"Your order #12345 has been shipped and is on its way. Thank you for your patience!"

This output is more informative and maintains politeness.

Step 6: Iterate with Additional Instructions

To ensure consistent tone and length, add explicit instructions.

Example Prompt 3

"You are a helpful customer support assistant. Respond to the question 'Where is my order?' for order #12345 with status 'shipped'. Keep the response polite, concise (under 30 words), and include the order number and status."

Step 7: Final Output and Validation

Model output might be:

"Your order #12345 has shipped and will arrive soon. Thank you for choosing us!"

This meets all criteria: polite, concise, informative, and personalized.

Summary of Prompt Engineering Process

• Start simple: Test baseline model behavior with a straightforward prompt.
• Analyze output: Identify gaps in relevance, tone, or completeness.
• Add context: Include dynamic data and specific instructions.
• Iterate: Refine prompt to control length, style, and content.
• Validate: Ensure output meets task requirements consistently.

This systematic approach exemplifies experimentation in prompt engineering, a key competency for the NVIDIA-Certified Associate: Generative AI LLM certification.

More in this topic

Data augmentation techniques — Experimentation (NVIDIA-Certified Associate: Generative AI LLM)Data augmentation techniques: Practice Questions — Experimentation (NVIDIA-Certified Associate: Generative AI LLM)Experiment design and execution: Common Mistakes — Experimentation (NVIDIA-Certified Associate: Generative AI LLM)Comparing model outputs: Worked Example — Experimentation (NVIDIA-Certified Associate: Generative AI LLM)Experiment design and execution: Quick Reference — Experimentation (NVIDIA-Certified Associate: Generative AI LLM)Testing model performance across tasks: Common Mistakes — Experimentation (NVIDIA-Certified Associate: Generative AI LLM)Experiment design and execution: Practice Questions — Experimentation (NVIDIA-Certified Associate: Generative AI LLM)Comparing model outputs: Practice Questions — Experimentation (NVIDIA-Certified Associate: Generative AI LLM)Testing model performance across tasks: Quick Reference — Experimentation (NVIDIA-Certified Associate: Generative AI LLM)Experimentation — NVIDIA-Certified Associate: Generative AI LLMPrompt engineering — Experimentation (NVIDIA-Certified Associate: Generative AI LLM)Prompt engineering: Practice Questions — Experimentation (NVIDIA-Certified Associate: Generative AI LLM)Comparing model outputs — Experimentation (NVIDIA-Certified Associate: Generative AI LLM)Testing model performance across tasks: Worked Example — Experimentation (NVIDIA-Certified Associate: Generative AI LLM)Data augmentation techniques: Worked Example — Experimentation (NVIDIA-Certified Associate: Generative AI LLM)Prompt engineering: Quick Reference — Experimentation (NVIDIA-Certified Associate: Generative AI LLM)Data augmentation techniques: Quick Reference — Experimentation (NVIDIA-Certified Associate: Generative AI LLM)Prompt engineering: Common Mistakes — Experimentation (NVIDIA-Certified Associate: Generative AI LLM)Data augmentation techniques: Common Mistakes — Experimentation (NVIDIA-Certified Associate: Generative AI LLM)Comparing model outputs: Quick Reference — Experimentation (NVIDIA-Certified Associate: Generative AI LLM)Testing model performance across tasks: Practice Questions — Experimentation (NVIDIA-Certified Associate: Generative AI LLM)Experiment design and execution — Experimentation (NVIDIA-Certified Associate: Generative AI LLM)Testing model performance across tasks — Experimentation (NVIDIA-Certified Associate: Generative AI LLM)Comparing model outputs: Common Mistakes — Experimentation (NVIDIA-Certified Associate: Generative AI LLM)Experiment design and execution: Worked Example — Experimentation (NVIDIA-Certified Associate: Generative AI LLM)