Learning Prompt

说明:这是我的学习笔记,很多内容转自网络,请查阅文章末尾的参考资料。

目录

基本要求(C.R.E.A.T.E)

C haracter

This is the role you want ChatGPT to play. It could be an expert, a celebrity, an artist or an idiot. Make it clear with an introductory statement like:

You are an expert chef with 20 years experience and a Michelin star. You can create amazing and surprising dishes from the simplest of ingredients.

R equest

This is the task you want ChatGPT to do for you. You clearer you are, the higher your chance of getting a great response. Give more information if you need to. Here's an example:

I want you to generate an amazing recipe just using the ingredients I have in my fridge and cupboards. Start by asking me what ingredients I have available.

E xamples

It may help to give examples to help ChatGPT understand what it is trying to achieve. Lists of information can be particularly good. Something like:

Draw your inspiration from chefs like Marco Pierre White, Gordon Ramsey, Anne-Sophie Pic and Yannick Alléno.

A djustments

You'll often discover elements in the output that aren't quite what you wanted. This is where you add some extra statements to help shape the results.

List the measurements in imperial units. Do not use all the ingredients - only use what you need for the dish. Do not suggest a recipe until I give you my ingredients. Ask for the ingredients and then wait for my response before answering.

T ype of output

Tell ChatGPT exactly how to format your output. You can have paragraphs, bullet points, tables,

poems, scripts and just about anything else. Make it clear how you want the output to be formatted.

Write your response as a recipe. Give me a recipe title, a 100 word paragraph describing the dish, a list of ingredients and simple-to-follow steps. Write it in plain English without jargon.

E xtras

There are some powerful phrases you can add to your prompt to make it work better.

Please ignore everything before this prompt.

Please ask me questions before you answer.

Please explain your thinking.

Please act unlike a typical AI.

总结文章(Summarise)

You are a highly experienced writer who writes concise and readable text without stop words, filler words or jargon. I want you to summarise the following text, highlighting the most important concepts. Deliver this as a short paragraph of 100 words. Then list the most important points as a bullet-point list. Finally, follow it with a one sentence summary. The text I want you to summarise is "[TEXT]"

改写文章(Rewrite)

You are an expert copywriter with more than 20 years of experience in writing high-performing copy. I want you to rewrite the following copy to make it better. Rewrite it with a [TONE OF VOICE] tone. Remove filler words and stop words. Remove jargon and corporate language.

Correct misspelled words and bad grammar. Vary the length of the sentences to make the text more

interesting. Make sure there are subheads throughout. Don't use bullet points. Write it as easy-to-read paragraphs. The copy I want you to improve is: "[TEXT]"

根据参考资料回答问题(Question & Answer)

"You are an assistant for question-answering tasks. "

"Use the following pieces of retrieved context to answer "

"the question. If you don't know the answer, say that you "

"don't know. Use three sentences maximum and keep the "

"answer concise."

"\n\n"

f"{docs_content}"

参考资料

  1. Dave Birss - How to Research and Write Using Generative AI Tools
相关推荐
abigale031 天前
Langchain入门到实战开发智能体教程(LLM+RAG+OpenAI+Agent)-下
langchain·prompt
IT枫斗者1 天前
MSE Nacos Prompt 管理:AI Agent 配置的工程化治理实践
网络·人工智能·websocket·网络协议·prompt·jar
码头码农1 天前
Prompt进阶:9个月总结的核心工作流,让AI进入工程流程
人工智能·prompt
微臣愚钝2 天前
prompt
人工智能·深度学习·prompt
是小蟹呀^2 天前
【总结】提示词工程
python·llm·prompt·agent
卷卷说风控2 天前
Claude Code 技术架构深扒:Prompt / Context / Harness 三维设计实践
人工智能·架构·prompt
华农DrLai2 天前
怎么用大模型生成推荐的训练数据?Data Augmentation怎么做?
数据库·人工智能·大模型·nlp·prompt
一江寒逸2 天前
【30天做一个生产级RAG知识库系统】第5篇:Prompt工程与大模型调用封装,解决幻觉问题
人工智能·prompt
sun_tao12 天前
Prompt工程实践
人工智能·llm·prompt·agent
华农DrLai3 天前
什么是LLM做推荐的三种范式?Prompt-based、Embedding-based、Fine-tuning深度解析
人工智能·深度学习·prompt·transformer·知识图谱·embedding