Key principles of prompt engineering

ChatGPT Prompt Engineering for Developers

Fouad Trad

Machine Learning Engineer

Clear and precise prompts

A visual diagram showing a bakery and a house with two possible routes between them. One if simple and short (effective), and the other is complicated and long (ineffective).

Key principles

🎬 Appropriate action verbs
💬 Detailed and precise instructions
❗ Well-structured delimited prompts

Using action verbs

Guide model what to do

Table for the verbs to use including write, complete, explain, describe, or evaluate.

Avoid ambiguous verbs

Confuse model's understanding

Table for the verbs to use including write, complete, explain, describe, or evaluate, and the verbs to avoid using such as understand, think, feel, try, and know.

Effective prompt with verbs

prompt = "Think about the issue of 
deforestation."
response = get_response(prompt)
print(response)

Deforestation is a significant 
environmental issue involving 
permanently removing
or destroying forests and woodlands. 
It has far-reaching impacts on the 
environment, ecosystems, wildlife, 
and human communities. [...]

prompt = "Propose strategies to 
reduce deforestation."
response = get_response(prompt)
print(response)

Reducing deforestation requires 
a comprehensive and multi-dimensional 
approach involving various stakeholders. 
Here are several strategies 
that can help address the issue:
Strengthen Forest Governance
Promote Sustainable Land Use [...]

Formulating detailed instructions

Provide specific, descriptive, and detailed instructions regarding:

📖 Context
📏 Output length
🖌 Format and style
👥👥 Audience

Effective prompt with instructions

Ineffective prompt: ~~"Tell me about dogs."~~

Effective prompt

prompt = "Write a descriptive paragraph about the behavior and characteristics of 
Golden Retrievers, highlighting their friendly nature, intelligence, 
and suitability as family pets."
print(get_response(prompt))

Golden Retrievers are beloved worldwide for their exceptional behavior, remarkable 
characteristics, and friendly nature. They are highly intelligent, trainable, and 
adaptable, making them great companions for families, including those with children.

Limiting output length

max_tokens:

Limit on number of tokens
Output cannot bypass it
Might lead to incomplete or cut responses

Prompt:

Limit on words, sentences, or paragraphs
Output may bypass it
Complete responses

Prompt components

Instructions and input data to operate on
Example: text summarization
- Instruction: summarize the given text
- Input data: text to summarize

An icon representing the text summarization task

Crafting a well-structured prompt with delimiters

Start prompt with instructions
Use delimiters (parentheses, brackets, backticks, etc.) to specify input parts
Mention which delimiters are used

prompt = """Summarize the text delimited by triple backticks into bullet points.
           ```TEXT GOES HERE```"""
response = get_response(prompt)

Using formatted strings (f-strings)

Include defined string into another string

text = "This is a sample text to summarize"

prompt = f"""Summarize the text delimited by triple backticks into bullet points.
           ```{text}```"""

print(prompt)

Summarize the text delimited by triple backticks into bullet points.
```This is a sample text to summarize```

Let's practice!

ChatGPT Prompt Engineering for Developers