# jacobgil/pytorch-grad-cam

Advanced AI Explainability for computer vision.  Support for CNNs, Vision Transformers, Classification, Object detection, Segmentation, Image similarity and more.

Repository: https://github.com/jacobgil/pytorch-grad-cam
Canonical: https://ross.abutalabs.com/products/pytorch-grad-cam
Homepage: https://jacobgil.github.io/pytorch-gradcam-book
Language: Python
License: MIT
License Family: permissive
Topics: deep-learning, pytorch, grad-cam, visualizations, interpretability, interpretable-ai, interpretable-deep-learning, score-cam, class-activation-maps, vision-transformers, explainable-ai, xai, image-classification, machine-learning, object-detection, computer-vision, explainable-ml
Last push: 2026-08-13T07:15:26+00:00

## Health v2 (maintenance only)
Score: 76/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 97, release rhythm 35, longevity 100
- inputs: {"age_days": 3381, "days_push": 20, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 12958, forks 1707 (observed 2026-08-28T04:11:00.246507+00:00)

## What it is
A PyTorch library providing state-of-the-art pixel attribution (saliency) methods like GradCAM, ScoreCAM, and AblationCAM for explainable AI in computer vision. It supports CNNs and Vision Transformers across classification, object detection, segmentation, and embedding-similarity tasks, with metrics for evaluating explanation trustworthiness.

## Use cases
- visualize which image regions a CNN uses for its predictions
- explain vision transformer predictions with saliency maps
- debug why my image classifier misclassifies images
- generate class activation maps for object detection models
- benchmark new explainability methods for computer vision
- check if I can trust my model's explanations with faithfulness metrics

## When to choose
- you need pixel attribution/saliency visualizations for PyTorch vision models
- you want a comprehensive, well-tested collection of CAM variants including for Vision Transformers
- you need explanation quality metrics and batch processing for research or production diagnostics

## When to avoid
- you use TensorFlow or JAX instead of PyTorch
- you need explainability for NLP or tabular models rather than vision
- you want a GUI tool rather than a Python library

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, deep-learning, data-visualization, image-processing, computer-vision
- domain: deep-learning, computer-vision, machine-learning
- platform: python, cross-platform
- tags: explainable-ai, grad-cam, interpretability, pytorch, saliency-maps, vision-transformers, xai, class-activation-maps, gpu

## Member repositories
- jacobgil/pytorch-grad-cam (main) score 76

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:11:00.246507+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:13:39.458562+00:00, confidence not recorded.
  - readme: https://github.com/jacobgil/pytorch-grad-cam (fetched 2026-08-28T04:11:00.246507+00:00, sha 6ac59f4029de)
  - homepage: https://jacobgil.github.io/pytorch-gradcam-book (fetched 2026-08-29T08:09:32.821304+00:00, sha 02134641d6c2)
- Data as of 2026-08-30T08:39:29.467469+00:00.