ypeiyu/attribution_recalibration
[ICLR 2023 Spotlight] Re-calibrating Feature Attributions for Model Interpretation
This tool helps machine learning researchers and practitioners better understand why their image classification models make certain predictions. It takes an existing deep learning model and an image as input, then re-calibrates and enhances standard 'attribution maps' that highlight the most important parts of the image influencing the model's decision. The output is a more accurate visual explanation, making it easier to interpret model behavior.
Use this if you need more reliable and theoretically sound visual explanations for your deep learning model's predictions, especially when using path integration-based attribution methods.
Not ideal if you are looking for new attribution methods rather than improving existing ones, or if your primary need is for model debugging rather than interpretation.
Stars
47
Forks
2
Language
Python
License
—
Last pushed
Jan 21, 2026
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/ypeiyu/attribution_recalibration"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
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...