Adversarial Machine Learning Python Packages

llm-guard

The Security Toolkit for LLM Interactions

adversarial-robustness-toolbox

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

textattack

TextAttack 🐙 is a Python framework for adversarial attacks, data augmentation, and model training in NLP https://textattack.readthedocs.io/en/master/

robustbench

RobustBench: a standardized adversarial robustness benchmark [NeurIPS 2021 Benchmarks and Datasets Track]

auto-lirpa

auto_LiRPA: An Automatic Linear Relaxation based Perturbation Analysis Library for Neural Networks and General Computational Graphs

secml

A Python library for Secure and Explainable Machine Learning

ml-pentest

Robustness evaluation framework for ML-Based Windows malware detectors

advertorch

A Toolbox for Adversarial Robustness Research

adversarial-labeller

Adversarial labeller is a sklearn compatible instance labelling tool for model selection under data drift.

graphgallery

GraphGallery is a gallery for benchmarking Graph Neural Networks

fawkes

An utility to protect user privacy

secml-malware

Create adversarial attacks against machine learning Windows malware detectors

trojai

A repository to quickly generate synthetic data and associated trojaned deep learning models

augmented-pca

Python implementations of supervised and adversarial linear factor models.

aixploit

Engineered to help red teams and penetration testers exploit large language model AI solutions vulnerabilities.

layerssustainabilityanalysis

LSA : Layer Sustainability Analysis framework for the analysis of layer vulnerability in a given neural network. LSA can be a helpful toolkit to assess deep neural networks and to extend the adversarial training approaches towards improving the sustainability of model layers via layer monitoring and analysis.

advsecurenet

Machine Learning Security Library

robustness-framework

An efficient framework for establishing a baseline for standard and adversarial machine learning training projects

pro-gan-pth

Unofficial PyTorch implementation of the paper titled "Progressive growing of GANs for improved Quality, Stability, and Variation"

versatile-evasion-attacks

Security protocols for estimating adversarial robustness of machine learning models for both tabular and image datasets. This package implements a set of evasion attacks based on metaheuristic optimization algorithms, and complex cost functions to give reliable results for tabular problems.

niteshade

Library for simulating data poisoning attacks against online learning.

universal-evasion-attacks

Security protocols for estimating adversarial robustness of machine learning models for both tabular and image datasets. This package implements a set of evasion attacks based on metaheuristic optimization algorithms, and complex cost functions to give reliable results for tabular problems.

mia

A library for running membership inference attacks against ML models

scratchai-nightly

Scratch AI