adversarial-attacks · GitHub Topics

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

翻译 - 对抗性鲁棒性工具箱（ART）-用于机器学习安全性的Python库-规避，中毒，提取，推理

python attack adversarial-machine-learning poisoning trusted-ai artificial-intelligence extraction adversarial-attacks adversarial-examples evasion

Python4.89 k

3 天前

TextAttack

@QData

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

翻译 - TextAttack🐙是Python框架，用于NLP中的对抗性攻击，数据增强和模型训练

machine-learning security natural-language-processing nlp adversarial-machine-learning adversarial-attacks data-augmentation adversarial-examples

Python2.97 k

4 个月前

foolbox

@bethgelab

A Python toolbox to create adversarial examples that fool neural networks in PyTorch, TensorFlow, and JAX

翻译 - Python工具箱可创建对抗示例，这些示例会欺骗PyTorch，TensorFlow，Keras等神经网络。

adversarial-examples adversarial-attacks

Python2.71 k

8 个月前

promptbench

Microsoft@microsoft

#大语言模型# A unified evaluation framework for large language models

adversarial-attacks chatgpt evaluation large-language-models robustness prompt prompt-engineering benchmark

Python2.47 k

1 个月前

llm-guard

@protectai

#大语言模型# The Security Toolkit for LLM Interactions

large-language-models llm chatgpt security-tools adversarial-attacks adversarial-machine-learning prompt prompt-engineering prompt-injection llmops

Python1.25 k

1 个月前