OSS Alternative - Discover Top Open Source Alternatives to Popular Software

CyberAlbSecOP/Awesome_GPT_Super_Prompting

A comprehensive collection of resources for ChatGPT jailbreaks, GPT prompt leaks, prompt injection techniques, and LLM prompt security.

Core Features

Extensive list of ChatGPT jailbreak techniques and resources.

Collection of leaked GPT agent system prompts.

Resources for understanding and mitigating prompt injection attacks.

Focus on LLM prompt security and adversarial machine learning.

Curated links to relevant GitHub repositories and communities.

Detailed Introduction

This project serves as a vital knowledge hub for anyone interested in the security and advanced usage of Large Language Models (LLMs), particularly GPT models. It meticulously compiles and categorizes a wide array of resources, including methods for bypassing model restrictions (jailbreaks), uncovering hidden system prompts (prompt leaks), and exploring vulnerabilities like prompt injection. Beyond offensive techniques, it also touches upon defensive strategies for LLM prompt security and the broader field of adversarial machine learning, making it an indispensable guide for researchers, security professionals, and AI enthusiasts.