Mlpepr

Latest version: v0.1b6

Safety actively analyzes 682387 Python packages for vulnerabilities to keep your Python projects secure.

0.1b6

- Add filter function for outlier samples, based on the [gmia attack](https://hallojs.github.io/ml-pepr/gmia.html).
- Fix gmia threshold search.
- Fix mia report generation.

0.1b5

Add [ART](https://github.com/Trusted-AI/adversarial-robustness-toolbox)-Inference attacks.

0.1b4

Add [ART](https://github.com/Trusted-AI/adversarial-robustness-toolbox)-Extraction Attacks.

0.1b3

Add [ART](https://github.com/Trusted-AI/adversarial-robustness-toolbox)-Evasion Attacks Integration and improve [FoolBox](https://github.com/bethgelab/foolbox)-Integration.

0.1b2

Add [FoolBox](https://github.com/bethgelab/foolbox)-Intergration. Thanks mastercaution!

0.1b1

This first release contains two attacks and the basic foundation of our library.

Releases

Has known vulnerabilities

0.1b6
0.1b5
0.1b4
0.1b3
0.1b2
0.1b1