Popular GitHub Repositories for Attack

Adversarial Robustness Toolbox (ART) - Python Library for Machine Learning Security - Evasion, Poisoning, Extraction, Inference - Red and Blue Teams

FedML - The Research and Production Integrated Federated Learning Library: https://fedml.ai

Master Federated Learning in 2 Hours—Run It on Your PC!

Must-read Papers on Textual Adversarial Attack and Defense

Keras implementation of "One pixel attack for fooling deep neural networks" using differential evolution on Cifar10 and ImageNet

It's not AI that takes away your job, but the people who master the use of AI tools. The most deadly attack is a dimension-reducing strike: destroying you has nothing to do with you - from "The Three-Body Problem". 中文说明： 抢走你工作的不是AI，而是掌握使用AI工具的人。 降维打击最为致命：毁灭你，与你何干《三体》

A pytorch adversarial library for attack and defense methods on images and graphs

Chinese safety prompts for evaluating and improving the safety of LLMs. 中文安全prompts，用于评估和提升大模型的安全性。

A curated list of adversarial attacks and defenses papers on graph-structured data.

An Open-Source Package for Textual Adversarial Attack.

Privacy Meter: An open-source library to audit data privacy in statistical and machine learning algorithms.

A Model for Natural Language Attack on Text Classification and Inference

Model Merging in LLMs, MLLMs, and Beyond: Methods, Theories, Applications and Opportunities. arXiv:2408.07666.

A script for credentials-based attack surface enumeration and general reconnaissance of massive networks

Security and Privacy Risk Simulator for Machine Learning (arXiv:2312.17667)

TransferAttack is a pytorch framework to boost the adversarial transferability for image classification.

A curated list of trustworthy deep learning papers. Daily updating...

Pytorch implementation of convolutional neural network adversarial attack techniques

WifiForge is a tool developed by Black Hills InfoSec to help train Pentesters on different Wi-Fi attack vectors and Wireless capabilities.

A unified, easily extensible repository for LLM unlearning benchmarks (TOFU, MUSE) - enabling new datasets, evaluations, methods, and tasks.