Prompt Wording
1
Introduction to Prompt Engineering
Module 003 –Prompt Wording
At the end of this module you are expected to:
1. Understand the different types of prompts and how to use them to
generate text.
2. Write effective prompts that will generate the desired output.
Prompt Wording
A prompt is a piece of text that guides the output of a large language model (LLM). Prompts can be as
simple as a few words or as complex as a paragraph. They can be written in natural language or in a more
technical format.
Specificity: The more specific the prompt, the more likely the LLM will generate the desired output. For
example, a prompt that says "Write a poem about a cat" is more likely to generate a poem about a cat
than a prompt that says "Write a poem."
The LLM is trained on a massive dataset of text. The more specific the prompt, the more likely the LLM is
to find a match in the dataset and generate text that is similar to the prompt.
Example:
"Write a poem about a cat" is a specific prompt because it specifies the topic of the poem.
"Write a poem" is a less specific prompt because it does not specify the topic of the poem.
Clarity: The prompt should be clear and easy to understand. Avoid using jargon or technical terms that
the LLM may not understand.
The LLM is not a human being. It does not have the same understanding of language as a human being. It
is important to use language that the LLM can understand.
Example:
"Write a poem about a feline" is a less clear prompt than "Write a poem about a cat" because the
word "feline" is a technical term that may not be understood by the LLM.
Course Module
Prompt Wording
2
Introduction to Prompt Engineering
Length: The length of the prompt will depend on the task that you are trying to achieve. For simple tasks,
a short prompt may be sufficient. For more complex tasks, a longer prompt may be necessary.
The length of the prompt will also depend on the complexity of the task. For simple tasks, the LLM may be
able to generate the desired output from a short prompt. For more complex tasks, the LLM may need
more information to generate the desired output.
Example:
"Write a poem about a cat" is a short prompt that can be used to generate a simple poem about a
cat.
"Write a poem about a cat that is both funny and sad" is a longer prompt that can be used to
generate a more complex poem about a cat.
Structure: The prompt should be well-structured and easy to read. Avoid using run-on sentences or
complex grammatical structures.
The LLM is a machine learning model. It is not able to understand the structure of a sentence. It is
important to write the prompt in a way that is easy for the LLM to understand.
Example:
"Write a poem about a cat who is sad because he lost his toy" is a well-structured prompt that is
easy to read.
"Write a poem about a sad cat who lost his toy" is a less well-structured prompt that is more
difficult to read.
Variety: Using a variety of prompts can help you to generate more creative and interesting text. Try using
different types of prompts, such as input/output prompts, zero-shot prompts, and few-shot prompts.
The LLM can learn from the different types of prompts that you use. This can help the LLM to generate
more creative and interesting text.
Example:
Try using a few-shot prompt to generate a poem about a cat.
Try using a zero-shot prompt to generate a story about a robot who falls in love with a human.
Try using an input/output prompt to generate a list of 10 things that you should do to improve
your writing skills.
Course Module
Prompt Wording
3
Introduction to Prompt Engineering
Experimentation: The best way to learn how to write effective prompts is to experiment. Try different
types of prompts and see what works best for you.
There is no one-size-fits-all approach to writing effective prompts. The best way to learn how to write
effective prompts is to experiment and see what works best for you.
For non-technical background
Prompts are used to tell the LLM what to do.
The type of prompt that you use will depend on the task that you are trying to achieve.
For example, if you are trying to generate text, you would use an input/output prompt.
If you are trying to translate languages, you would use a translation prompt.
If you are trying to answer questions, you would use a question-answering prompt.
For technical background
Prompts can be written in natural language or in a more technical format.
Natural language prompts are easier to write, but they can be less precise.
Technical prompts are more precise, but they can be more difficult to write.
The best type of prompt to use will depend on the task that you are trying to achieve and your
own level of expertise.
Example of a platform to use to be effective
Any platform that can generate text can be used to use prompts effectively.
However, some platforms are better suited for certain types of prompts.
For example, platforms with a large dataset of text are better suited for zero-shot prompts.
Platforms with a smaller dataset of text are better suited for few-shot prompts.
Technical differences:
The different types of prompts have different technical requirements.
Input/output prompts are the simplest type of prompt and have the fewest technical
requirements.
Zero-shot prompts require the LLM to learn new information, which can be time-consuming and
computationally expensive.
Few-shot prompts require the LLM to learn new information, but they do not require as much
learning as zero-shot prompts.
Adversarial prompts require the LLM to learn new information and to generate text that is not
similar to any of the text that it has seen before.
Course Module
Prompt Wording
4
Introduction to Prompt Engineering
References and Supplementary Materials
Books and Journals
1. https://www.researchgate.net/publication/360310862_Prompt_Engineering_for_Tex
t-Based_Generative_Art
2. https://arxiv.org/pdf/2107.13586.pdf
3. Oppenlaender, Jonas. (2022). Prompt Engineering for Text-Based Generative Art.
Online Supplementary Reading Materials
1. https://www.classcentral.com/course/chatgpt-for-developers-180241
2. https://www.flowrite.com/blog/introduction-to-prompt-engineering
3. https://docs.cohere.com/docs/prompt-engineering
4. https://solutions.yieldbook.com/content/dam/yieldbook/en_us/documents/publicat
ions/using-chatgpt-with-prompt-engineering.pdf
Online Instructional Videos
1. https://youtu.be/dOxUroR57xs?feature=shared
2. https://youtu.be/JTxsNm9IdYU?feature=shared
3. https://youtu.be/BP9fi_0XTlw?feature=shared
Course Module
