athena
adversarial-robustness-toolbox
athena | adversarial-robustness-toolbox | |
---|---|---|
1 | 8 | |
42 | 4,483 | |
- | 1.7% | |
0.0 | 9.7 | |
over 2 years ago | 3 days ago | |
Python | Python | |
MIT License | MIT License |
Stars - the number of stars that a project has on GitHub. Growth - month over month growth in stars.
Activity is a relative number indicating how actively a project is being developed. Recent commits have higher weight than older ones.
For example, an activity of 9.0 indicates that a project is amongst the top 10% of the most actively developed projects that we are tracking.
athena
-
How to Put Out Democracy’s Dumpster Fire: Our democratic habits have been killed off by an internet kleptocracy that profits from disinformation, polarization, and rage. Here’s how to fix that.
While users could bookmark algorithms for use anywhere on reddit, the default sorting mode for a subreddit would be established by an ensemble of the algorithms, weighted by the usage of the those algorithms on that subreddit. Such a system could be robust against bot attacks, as an adversary must defeat not one algorithm, but the majority of algorithms used (see Athena: "A Framework for Defending Machine Learning Systems Against Adversarial Attacks").
adversarial-robustness-toolbox
- [D] Couldn't devs of major GPTs have added an invisible but detectable watermark in the models?
- [D] ML Researchers/Engineers in Industry: Why don't companies use open source models more often?
- [D]: How safe is it to just use a strangers Model?
-
[D] Does anyone care about adversarial attacks anymore?
Check out this project https://github.com/Trusted-AI/adversarial-robustness-toolbox
- adversarial-robustness-toolbox: Adversarial Robustness Toolbox (ART) - Python Library for Machine Learning Security - Evasion, Poisoning, Extraction, Inference - Red and Blue Teams
- Library for Machine Learning Security Evasion, Poisoning, Extraction, Inference
-
Introduction to Adversarial Machine Learning
Adversarial Robustness Toolbox (ART) is a Python library for Machine Learning Security. ART provides tools that enable developers and researchers to defend and evaluate Machine Learning models and applications against the adversarial threats of Evasion, Poisoning, Extraction, and Inference.
-
[D] Testing a model's robustness to adversarial attacks
Depending on what attacks you want I've found both https://github.com/cleverhans-lab/cleverhans and https://github.com/Trusted-AI/adversarial-robustness-toolbox to be useful.
What are some alternatives?
fawkes - Fawkes, privacy preserving tool against facial recognition systems. More info at https://sandlab.cs.uchicago.edu/fawkes
DeepRobust - A pytorch adversarial library for attack and defense methods on images and graphs
TextAttack - TextAttack 🐙 is a Python framework for adversarial attacks, data augmentation, and model training in NLP https://textattack.readthedocs.io/en/master/
auto-attack - Code relative to "Reliable evaluation of adversarial robustness with an ensemble of diverse parameter-free attacks"
faceswap - Deepfakes Software For All
alpha-zero-boosted - A "build to learn" Alpha Zero implementation using Gradient Boosted Decision Trees (LightGBM)
m2cgen - Transform ML models into a native code (Java, C, Python, Go, JavaScript, Visual Basic, C#, R, PowerShell, PHP, Dart, Haskell, Ruby, F#, Rust) with zero dependencies
waf-bypass - Check your WAF before an attacker does
Differential-Privacy-Guide - Differential Privacy Guide
gretel-synthetics - Synthetic data generators for structured and unstructured text, featuring differentially private learning.
unrpa - A program to extract files from the RPA archive format.