# utkuozbulak/pytorch-cnn-visualizations

Pytorch implementation of convolutional neural network visualization techniques

Repository: https://github.com/utkuozbulak/pytorch-cnn-visualizations
Canonical: https://ross.abutalabs.com/products/pytorch-cnn-visualizations
Language: Python
License: MIT
License Family: permissive
Topics: grad-cam, cam, pytorch, guided-backpropagation, guided-grad-cam, gradient, saliency, segmentation, gradient-visualization, deep-dream, cnn-visualization, smooth-grad
Last push: 2025-01-01T10:31:10+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 3238, "days_push": 609, "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 8233, forks 1503 (observed 2026-08-28T04:10:19.752637+00:00)

## What it is
A PyTorch library implementing a wide range of convolutional neural network visualization and interpretability techniques, including Grad-CAM, guided backpropagation, saliency maps, Smooth Grad, Deep Dream, and integrated gradients. Each technique is provided as a standalone, well-commented Python file designed for learning and research.

## Use cases
- visualize what a CNN layer has learned
- generate grad-cam heatmaps for image classification
- compute saliency maps for model interpretability
- run deep dream on pretrained networks
- learn how cnn visualization techniques work
- explain pytorch model predictions with class activation maps

## When to choose
- you want reference implementations of many CNN visualization methods in one place
- you are studying or teaching interpretability techniques with readable per-technique code
- you need Grad-CAM, Smooth Grad, or integrated gradients on AlexNet/VGG-style models

## When to avoid
- you need a maintained library compatible with recent PyTorch versions (code targets torch 0.4.1)
- you need production-grade explainability tooling with broad model support
- you work with models lacking a features/classifier layer split without modifying code

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning, image-processing, data-visualization
- domain: deep-learning, computer-vision, machine-learning, education
- platform: python, cross-platform
- tags: pytorch, cnn-visualization, grad-cam, saliency-maps, explainability, deep-dream, smooth-grad, guided-backpropagation, interpretability

## Member repositories
- utkuozbulak/pytorch-cnn-visualizations (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:10:19.752637+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 2026-08-29T17:29:30.475519+00:00, confidence not recorded.
  - readme: https://github.com/utkuozbulak/pytorch-cnn-visualizations (fetched 2026-08-28T04:10:19.752637+00:00, sha 46d3338892a2)
- Data as of 2026-08-30T08:39:29.467469+00:00.