Auto-sync: 2026-04-20 00:02
This commit is contained in:
32
wiki/concepts/Prompt-Engineering.md
Normal file
32
wiki/concepts/Prompt-Engineering.md
Normal file
@@ -0,0 +1,32 @@
|
||||
---
|
||||
title: "Prompt Engineering"
|
||||
type: concept
|
||||
tags:
|
||||
- AI
|
||||
- Prompt-Engineering
|
||||
- LLM
|
||||
date: 2024-11-12
|
||||
---
|
||||
|
||||
## Summary
|
||||
提示词工程(Prompt Engineering)是创建、设计和优化提示词以引导大语言模型(LLM)响应的过程,确保输出的准确性和相关性。
|
||||
|
||||
## Definition
|
||||
Prompt Engineering(提示词工程)是指通过精心设计提示词来提升 AI 输出质量的技术,涵盖需求拆解、结构化表达、场景共情、迭代优化四个维度。
|
||||
|
||||
## Key Components
|
||||
提示词由以下组件构成:
|
||||
- **指令(Instructions)**:明确告诉模型要做什么
|
||||
- **上下文(Context)**:提供背景信息
|
||||
- **用户输入(User Input)**:具体问题或任务
|
||||
- **输出指示器(Output Indicator)**:提示模型输出格式
|
||||
|
||||
## Basic Techniques
|
||||
- **One-shot Prompting**:提供一个示例
|
||||
- **Few-shot Prompting**:提供多个示例
|
||||
- **Chain of Thoughts**:提供逐步思考过程
|
||||
|
||||
## Connections
|
||||
- [[LLM]] ← guides ← Prompt Engineering:提示词工程引导大语言模型
|
||||
- [[Generative AI]] ← uses ← Prompt Engineering:生成式 AI 应用提示词工程
|
||||
- [[Claude Skills]] ← relies_on ← Prompt Engineering:Claude Skills 基于提示词工程技术
|
||||
Reference in New Issue
Block a user