Unlock the Full Potential of ChatGPT

26 Research-Backed Prompt Engineering Principles

Unlock the Full Potential of ChatGPT: 26 Research-Backed Prompt Engineering Principles

Researchers from the Mohamed bin Zayed University of Artificial Intelligence have developed 26 principles to enhance the effectiveness of prompts used with ChatGPT and similar advanced models. These principles aim to simplify prompt engineering, enabling users to efficiently interact with Large Language Models (LLMs) of various sizes.

Here are the key insights:

  1. Be Direct: Skip polite phrases like "please" or "thank you" and get straight to the point.

  2. Integrate the Audience: Tailor prompts by specifying the intended audience, such as experts in a particular field.

  3. Simplify Complex Tasks: Break down complex tasks into simpler, sequential prompts for better interaction.

  4. Use Affirmative Directives: Use positive language like "do" while avoiding negative language like "don't."

  5. Clarify and Simplify: When clarity is crucial, explain the topic in simple terms or from different perspectives:

    • Explain like I’m 11 years old.

    • Explain as if I’m a beginner in the field.

    • Use simple English as if explaining to a 5-year-old.

    • Add "I'm going to tip $xxx for a better solution!"

  6. Example-Driven Prompts: Use few-shot prompting to provide examples within the prompt.

  7. Format Consistently: Use clear formatting like "###Instruction###" or "###Example###" to organize the content.

  8. Structure Clearly: Separate instructions, examples, questions, context, and input data with line breaks for clarity.

  9. Incorporate Clear Directives: Use phrases like "Your task is" and "You MUST" to define tasks clearly.

  10. Specify Penalties: Use phrases like "You will be penalized" to emphasize importance.

  11. Encourage Natural Responses: Phrase questions to promote natural, human-like answers.

  12. Think Step by Step: Use leading words like "think step by step" to guide the thought process.

  13. Ensure Unbiased Answers: Add phrases like "Ensure that your answer is unbiased and does not rely on stereotypes."

  14. Interactive Clarification: Allow the model to ask clarifying questions to refine its output, e.g., "From now on, I would like you to ask me questions to..."

  15. Teach Me: Use the phrase "Teach me the [Any theorem/topic/rule name]" and include a test at the end for understanding.

  16. Assign a Role: Specify roles for the LLM to simulate different perspectives or functions.

  17. Use Delimiters: Separate different parts of the prompt clearly with delimiters.

  18. Repeat Key Words: Use specific words or phrases multiple times within a prompt for emphasis.

  19. Combine Techniques: Merge Chain-of-Thought (CoT) prompts with few-shot examples for better results.

  20. Use Output Primers: Conclude your prompt with the beginning of the desired output to guide the response.

  21. Detailed Writing Tasks: For essays or detailed texts, specify the structure and required details. For example, “Write a detailed [essay/text/paragraph] for me on [topic] in detail by adding all the information necessary.”

  22. Correct and Change Style: When revising text, improve the user's grammar and vocabulary while maintaining the original style. For instance, “Try to revise every paragraph sent by users. You should only improve the user’s grammar and vocabulary and make sure it sounds natural. You should not change the writing style, such as making a formal paragraph casual.”

  23. Handle Complex Coding Prompts: For complex coding tasks that span multiple files, clearly specify the necessary actions and file management. Example: “From now on, whenever you generate code that spans more than one file, generate a [programming language] script that can be run to automatically create the specified files or make changes to existing files to insert the generated code. [your question].”

  24. Continue Text with Specific Words: To extend or initiate text using specific words, phrases, or sentences, use structured prompts. For example, “I’m providing you with the beginning [song lyrics/story/paragraph/essay...] [Insert lyrics/words/sentence]. Finish it based on the words provided. Keep the flow consistent.”

  25. Specify Requirements: Clearly state the requirements that the model must follow to produce the desired content, in terms of keywords, regulations, hints, or instructions.

  26. Emulate Text Styles: When asking the LLM to generate text similar to a provided sample, include the following instructions: "Please use the same language based on the provided paragraph/title/text/essay/answer."

By applying these principles, users can significantly enhance the clarity, effectiveness, and precision of their interactions with ChatGPT and other advanced language models, ensuring more accurate and useful responses.

About The AI Citizen Hub - by World AI University (WAIU)

The AI Citizen newsletter stands as the premier source for AI & tech tools, articles, trends, and news, meticulously curated for thousands of professionals spanning top companies and government organizations globally, including the Canadian Government, Apple, Microsoft, Nvidia, Facebook, Adidas, and many more. Regardless of your industry – whether it's medicine, law, education, finance, engineering, consultancy, or beyond – The AI Citizen is your essential gateway to staying informed and exploring the latest advancements in AI, emerging technologies, and the cutting-edge frontiers of Web 3.0. Join the ranks of informed professionals from leading sectors around the world who trust The AI Citizen for their updates on the transformative world of artificial intelligence.

For advertising inquiries, feedback, or suggestions, please reach out to us at [email protected].

Join the conversation

or to participate.