The Prompt Engineering Guide is a project initiated by DAIR.AI, aiming to help people in R&D and industry understand prompt engineering. With the goal of disseminating AI technology and research results, DAIR.AI's vision is to empower a new generation of innovators in the AI field. This free and open-source prompt engineering project has surpassed 30,000 stars on GitHub, and includes all the latest papers, learning guides, lectures, reference materials, and tools related to LLM prompt engineering.
Table of Contents of the Prompt Engineering Guide
1. Introduction to Prompt Engineering
- Large Language Model Settings
- Basic Concepts
- Elements of a Prompt
- General Tips for Designing Prompts
- Prompt Examples
2. Prompting Techniques
- Zero-Shot Prompting
- Few-Shot Prompting
- Chain-of-Thought (CoT) Prompting
- Self-Consistency
- Generated Knowledge Prompting
- Tree of Thoughts
- Retrieval-Augmented Generation
- Automatic Reasoning and Tool Use
- Automatic Prompt Engineer
- Active-Prompt
- Directional Stimulus Prompting
- ReAct Framework
- Multimodal Chain-of-Thought Prompting
- Graph-Based Prompting
3. Prompt Applications
- Program-Aided Language Models
- Generating Data
- Generating Code
- Graduate Job Classification Case Study
- Prompt Function
4. Models
- Flan
- ChatGPT
- LLaMA
- GPT-4
- Model Collection
5. Risks and Misuse
- Adversarial Prompting
- Truthfulness
- Bias
