yangarbiter/rare-spurious-correlation

Understanding Rare Spurious Correlations in Neural Network

28
/ 100
Experimental

This project helps machine learning researchers and practitioners understand how neural networks can unexpectedly learn from very few, subtle data patterns that shouldn't influence their decisions. It takes in training data, potentially with these 'rare spurious correlations,' and outputs insights into the model's sensitivity, accuracy, and privacy implications. The target user is anyone building or analyzing neural network models who needs to ensure their models are robust and private.

No commits in the last 6 months.

Use this if you are concerned about your neural networks picking up on hidden, privacy-leaking patterns from just a handful of training examples, or if you want to test and mitigate such vulnerabilities.

Not ideal if you are looking for a general-purpose machine learning library or a tool for routine model deployment and monitoring, as this focuses specifically on research into rare spurious correlations.

machine-learning-research neural-network-robustness model-privacy data-bias AI-ethics
No License Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 5 / 25
Maturity 8 / 25
Community 15 / 25

How are scores calculated?

Stars

12

Forks

4

Language

Jupyter Notebook

License

Last pushed

Jun 05, 2022

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/yangarbiter/rare-spurious-correlation"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.