rachtibat/zennit-crp
An eXplainable AI toolkit with Concept Relevance Propagation and Relevance Maximization
This project helps researchers and practitioners understand *why* an AI model makes a particular decision. Instead of just showing where in an image the model looked, it identifies the specific visual characteristics, like "large irides" or "heavy wrinkles," that influenced the prediction. This means you input an image and the model's prediction, and you get out a clear explanation of the underlying concepts the model used.
141 stars. Available on PyPI.
Use this if you need to go beyond simple heatmaps and pinpoint the exact semantic features your AI model is using to make its classifications.
Not ideal if you are looking for a simple 'black box' explanation that doesn't delve into specific learned concepts, or if you need to explain models that are not image-based.
Stars
141
Forks
27
Language
Jupyter Notebook
License
—
Last pushed
Jan 14, 2026
Commits (30d)
0
Dependencies
4
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/rachtibat/zennit-crp"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Related frameworks
obss/sahi
Framework agnostic sliced/tiled inference + interactive ui + error analysis plots
tensorflow/tcav
Code for the TCAV ML interpretability project
MAIF/shapash
🔅 Shapash: User-friendly Explainability and Interpretability to Develop Reliable and Transparent...
TeamHG-Memex/eli5
A library for debugging/inspecting machine learning classifiers and explaining their predictions
csinva/imodels
Interpretable ML package 🔍 for concise, transparent, and accurate predictive modeling...