PE-MTHRFCKR
  • Prompt Engineering Guide
    • Introduction
    • LLM Settings
    • Basics of Prompting
    • Elements of a Prompt
    • General Tips for Designing Prompts
    • Examples of Prompts
    • Prompting Techniques
    • Zero-Shot Prompting
    • Few-Shot Prompting
    • Chain-of-Thought Prompting
    • Self-Consistency
    • Generated Knowledge Prompting
    • Automatic Prompt Engineer (APE)
    • Active-Prompt
    • Directional Stimulus Prompting
    • ReAct Prompting
    • Multimodal CoT Prompting
    • GraphPrompts
    • Prompting Applications
    • PAL (Program-Aided Language Models)
    • Generating Data
    • Graduate Job Classification Case Study
    • Models
    • Scaling Instruction-Finetuned Language Models
    • ChatGPT Prompt Engineering
    • Llama
    • GPT-4
    • Model Collection
    • Risks & Misuses
    • Adversarial Prompting
    • Factuality
    • Biases
    • Papers
    • Tools & Libraries
    • Prompt Engineering Notebooks
    • Datasets
    • Additional Readings
  • Complex ChatGPT Prompts
    • C–Context
    • H–Heritage
    • A–Aim
    • T–Tone
    • GPT
    • Example of ChatGPT-method
  • The Art of ChatGPT Prompting
    • Preface
    • Introduction
    • What is ChatGPT and How Does it Work?
    • The Role of Prompts in ChatGPT Conversations
    • Principles of Clear Communication
    • Constructing Effective Prompts
    • Common Mistakes to Avoid When Crafting ChatGPT Prompts
    • Troubleshooting
    • Technical Issues
    • Case Studies
    • Conclusion
  • Prompt Engineering 101
    • Include direct instructions in prompts
    • Give examples in prompts to get the best response
    • Align prompt instructions with the task’s end goal
    • Use personas to get more specific voices
    • Include acceptable responses in prompts for consistency
    • Try different prompts to find what works best
    • Summary
  • Mastering ChatGPT
    • I. Introduction
    • II. Understanding Prompts
    • III. Tips for Crafting Effective Prompts
    • IV. Advanced Hacks and Techniques
    • V. Avoiding Common Pitfalls
    • VI. Real-life ChatGPT Examples and Case Studies
    • VII. Conclusion
Powered by GitBook
On this page

Complex ChatGPT Prompts

PreviousAdditional ReadingsNextC–Context

Last updated 2 years ago

Making your own prompts can sometimes be hard, but using our method it should be easy. You do it step by step, and enginerring your prompt with a plan. It’s easy to remember by it’s name CHATGPT.

  • C–Context

  • H–Heritage

  • A–Aim

  • T–Tone

  • GPT

  • Example of ChatGPT-method

Let’s look at what the different letters means:

  • C–Context: What is the situation or background information relevant to the task? Goal: Provide information on a topic of your choice.

  • H–Heritage: What knowledge, skills, or attributes should the model possess to carry out the task effectively? Objective: Up to date and knowledgeable about the subject.

  • A–Aim: What is the goal or objective of the task? Goal: Provide clear and accurate information.

  • T–Tone: What tone or voice should the model adopt when carrying out the task? Goal: Make ChatGPT reply in a professional and informative tone, or friendly for a email?

  • GPT: Make sure it behaves less like it self, this part should be in all your prompts.

Try it out yourself and make the prompts better for your needs.