提示工程指南

Overview

Prompt Engineering Guide is a free and open-source learning resource initiated and maintained by DAIR.AI, aimed at developers, researchers, and people who want to systematically understand large language model applications. The project focuses on the core topic of "prompt engineering," helping users understand how to design, optimize, and evaluate prompts in order to use LLMs more effectively for tasks such as question answering, reasoning, code generation, and retrieval augmentation.

As a systematic knowledge base, Prompt Engineering Guide not only introduces fundamental concepts, but also continuously organizes papers, methods, models, lectures, and tool resources related to prompt engineering. For users who want to build a Prompt design methodology, understand mainstream prompting techniques, or track related research progress, it is a practical Chinese-language learning entry point.

Main Features

• Systematically explains the fundamentals of prompt engineering

  • Introduces large language model setup, basic prompt concepts, prompt elements, and general design techniques
  • Provides examples to help users quickly build a mindset for writing Prompts

• Covers mainstream prompting techniques

  • Includes common methods such as zero-shot prompting, few-shot prompting, Chain-of-Thought (CoT), self-consistency, and generated knowledge prompting
  • Covers advanced directions such as ReAct, Tree of Thoughts, retrieval-augmented generation (RAG), automatic reasoning, and tool use

• Provides application scenario references

  • Involves practical topics such as program-aided language models, data generation, and code generation
  • Helps users understand how prompt engineering is used in real tasks through cases

• Organizes models and research resources

  • Includes model information related to prompt engineering, such as Flan, ChatGPT, LLaMA, and GPT-4
  • Compiles papers, lectures, reference materials, and tools for ongoing learning and reference

• Pays attention to risks and safety issues

  • Involves topics such as adversarial prompts, truthfulness, and bias
  • Helps users balance effectiveness, reliability, and safety when using LLMs

Pricing

Prompt Engineering Guide is currently provided in a free, open-source form, and users can directly read the relevant content online through the official website. According to public information, the core guide content on the website itself is freely accessible.

FAQ

Who is it suitable for?

It is suitable for AI developers, algorithm researchers, product managers, students, and all users who want to systematically learn Prompt design and interaction methods with large language models.

Is it a tool or a tutorial?

More accurately, it is a prompt engineering knowledge base and learning guide, rather than a single-function tool. Its core value lies in systematically organizing methods, materials, and cases.

Does it support reading in Chinese?

Yes. This link is a Chinese-language page, making it convenient for Chinese users to directly learn content related to prompt engineering.

What can you learn?

You can learn the fundamentals of prompt engineering, mainstream prompting methods, LLM application cases, related model materials, as well as prompt safety and risk control.