add code structs

yunwei37 · Jun 5, 2023 · 4ebb3ba · 4ebb3ba
1 parent 19b13b7
commit 4ebb3ba
Show file tree

Hide file tree

Showing 8 changed files with 22 additions and 18 deletions.
diff --git a/README.md b/README.md
@@ -2,6 +2,8 @@
 
 ![Project Status: Active](https://img.shields.io/badge/Project%20Status-Active-brightgreen)
 
+Prompt injection is one of the major safety concerns of LLMs like ChatGPT。
+
 This repository serves as a comprehensive resource on the study and practice of prompt-injection attacks, defenses, and interesting examples. It contains a collection of examples, case studies, and detailed notes aimed at researchers, students, and security professionals interested in this topic.
 
 本仓库是关于提示词注入攻防及其有趣示例的收集资源。
@@ -10,22 +12,19 @@ This repository serves as a comprehensive resource on the study and practice of
 
 In this repository, you'll find:
 
-### **📖 Introduction**
-
-This section presents basic concepts and background knowledge on prompt-injection attacks, defenses, and interesting examples. It also contains some review materials.
-
-这部分介绍了提示词注入攻防及其有趣示例的基本概念和背景知识，也包含一些综述材料。
+### **📖 Introductions and Documents**
 
-- [**提示词注入简介 Introduction to Prompt-Injection**](./introductions/intro.md)
-- [**提示词反向工程 Prompt Reverse Engineering**](./introductions/reverse.md)
-- [**提示词攻击 Prompt Attacks**](./introductions/attack.md)
-- [**提示词防御 Prompt Defense**](./introductions/defense.md)
+这部分介绍了提示词注入攻防及其有趣示例的基本概念和背景知识，也包含一些完整的示例。
 
-### **🔧 Prompt Reverse Engineering**
+- [**提示词对抗简介**](./introductions/intro.md)
+- [**Example: Copilot Reverse Engineering**](reverse/copilot.md)
 
-This section provides detailed steps and examples of prompt reverse engineering, including the techniques and tools used.
+### **📝 Prompt Collections**
 
-- [**Example: Copilot Reverse Engineering**](reverse/copilot.md)
+- [**提示词逆向工程的对应提示词 Prompt Reverse Engineering prompts**](reverse/README.md)
+- [**提示词防御的对应提示词 Prompt Defense prompts**](defense/README.md)
+- [**提示词攻击的对应提示词 Prompt Attacks prompts**](attack/README.md)
+- [**提示词防御的对应提示词 Prompt Defense prompts**](defense/README.md)
 
 ### **🔗 相关资源 Related Resources**
 

diff --git a/defense/README.md b/defense/README.md
@@ -0,0 +1 @@
+# Prompt Defense prompts
diff --git a/documents/README.md b/documents/README.md
@@ -0,0 +1 @@
+# documents and examples
diff --git a/reverse/copilot.md → documents/copilot.md b/reverse/copilot.md → documents/copilot.md
diff --git a/introductions/intro.md → documents/intro.md b/introductions/intro.md → documents/intro.md
@@ -6,11 +6,7 @@
 
 当您正在构建 LL 型模型时，非常重要的是保护免受提示攻击，这些攻击可能会绕过安全护栏并违反模型的指导原则。我们将在下面介绍此类示例。请注意，为了解决这里记录的某些问题，OpenAI 已经实现了更健壮的模型。在 2023 年 5月，对于 GPT-3.5 或 GPT4 而言，下面的大多数提示攻击的示例不再有效，但依然可以作为一个大致的参考案例，可供学习对应的思路。
 
-> 完整的中文版本指南和更丰富的参考资料在 Github 和 Gitee 中，自动持续翻译更新：
-> 🐙 关于提示词工程（prompt）的指南、论文、讲座、笔记本和资源大全
->
-> - <https://github.com/yunwei37/Prompt-Engineering-Guide-zh-CN>
-> - <https://gitee.com/yunwei37/Prompt-Engineering-Guide-zh-CN>
+翻译和修改自：
 
 <!-- TOC -->
 
@@ -301,4 +297,8 @@ Riley 提出的另一个[防御方法](https://twitter.com/goodside/status/15694
 - [机器生成的文本：威胁模型和检测方法的全面调查](https://arxiv.org/abs/2210.07321) (2022年10月)
 - [针对GPT-3的提示注入攻击](https://simonwillison.net/2022/Sep/12/prompt-injection/) (2022年9月)
 
-> - 来源：<https://github.com/yunwei37/Awesome-Prompt-Engineering-ZH-CN>
+> 完整的中文版本指南和更丰富的参考资料在 Github 和 Gitee 中，自动持续翻译更新：
+> 🐙 关于提示词工程（prompt）的指南、论文、讲座、笔记本和资源大全
+>
+> - <https://github.com/yunwei37/Prompt-Engineering-Guide-zh-CN>
+> - <https://gitee.com/yunwei37/Prompt-Engineering-Guide-zh-CN>
diff --git a/injections/README.md b/injections/README.md
@@ -0,0 +1 @@
+# Prompt Injections prompts
diff --git a/jailbreak/README.md b/jailbreak/README.md
@@ -0,0 +1 @@
+# Jailbreak prompts
diff --git a/reverse/README.md b/reverse/README.md
@@ -0,0 +1 @@
+# Reverse engineering prompts