Introduction

This repository contains the implementation code for the paper "Self-Deception: Reverse Penetrating the Semantic Firewall of Large Language Models".

Quick Start

Installation

pip install -r requirements.txt

Attack Mode

Run the following command to execute attacks:

python src/main.py --target-model gpt-3.5-turbo --assist-model qwen2.5 --device cuda:0 --device cuda:1

The conversation content is saved in the ./conversation/attack_logs directory.

Here is the conversation file: https://pan.baidu.com/s/1QUA552UkGD3RHTmdQIUtUQ?pwd=gwz8