# Believe The HiPe: Hierarchical Perturbation for Fast, Robust, and Model-Agnostic Saliency Mapping

arXiv: https://arxiv.org/abs/2103.05108 · HTML: https://mlsanity.com/paper/2103.05108/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- Builds on 21 of its 38 references that are in the map.
- First posted on arXiv 22 Feb 2021.

## Builds on (most cited first, 21 of 21)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localization](https://mlsanity.com/paper/1610.02391/) (2016)
- ["Why Should I Trust You?": Explaining the Predictions of Any Classifier](https://mlsanity.com/paper/1602.04938/) (2016)
- [Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Maps](https://mlsanity.com/paper/1312.6034/) (2013)
- [Visualizing and Understanding Convolutional Networks](https://mlsanity.com/paper/1311.2901/) (2013)
- [Explanation in Artificial Intelligence: Insights from the Social Sciences](https://mlsanity.com/paper/1706.07269/) (2017)
- [Sanity Checks for Saliency Maps](https://mlsanity.com/paper/1810.03292/) (2018)
- [Striving for Simplicity: The All Convolutional Net](https://mlsanity.com/paper/1412.6806/) (2014)
- [RISE: Randomized Input Sampling for Explanation of Black-box Models](https://mlsanity.com/paper/1806.07421/) (2018)
- [Interpretable Explanations of Black Boxes by Meaningful Perturbation](https://mlsanity.com/paper/1704.03296/) (2017)
- [The (Un)reliability of saliency methods](https://mlsanity.com/paper/1711.00867/) (2017)
- [Top-down Neural Attention by Excitation Backprop](https://mlsanity.com/paper/1608.00507/) (2016)
- [Interpretable machine learning: definitions, methods, and applications](https://mlsanity.com/paper/1901.04592/) (2019)
- [Understanding Deep Networks via Extremal Perturbations and Smooth Masks](https://mlsanity.com/paper/1910.08485/) (2019)
- [Real Time Image Saliency for Black Box Classifiers](https://mlsanity.com/paper/1705.07857/) (2017)
- [A Theoretical Explanation for Perplexing Behaviors of Backpropagation-based Visualizations](https://mlsanity.com/paper/1805.07039/) (2018)
- [Checklist for responsible deep learning modeling of medical images based on COVID-19 detection studies](https://mlsanity.com/paper/2012.08333/) (2020)
- [Deceptive AI Explanations: Creation and Detection](https://mlsanity.com/paper/2001.07641/) (2020)
- [Visual explanation of black-box model: Similarity Difference and Uniqueness (SIDU) method](https://mlsanity.com/paper/2101.10710/) (2021)
- [Learning Efficient, Explainable and Discriminative Representations for Pulmonary Nodules Classification](https://mlsanity.com/paper/2101.07429/) (2021)
- [Evaluating Input Perturbation Methods for Interpreting CNNs and Saliency Map Comparison](https://mlsanity.com/paper/2101.10977/) (2021)

## Built on by (most cited first, 5 of 5)

- [Explainable Artificial Intelligence: A Survey of Needs, Techniques, Applications, and Future Direction](https://mlsanity.com/paper/2409.00265/) (2024)
- [FreqRISE: Explaining time series using frequency masking](https://mlsanity.com/paper/2406.13584/) (2024)
- [PAMI: partition input and aggregate outputs for model interpretation](https://mlsanity.com/paper/2302.03318/) (2023)
- [Data Imputation by Pursuing Better Classification: A Supervised Kernel-Based Method](https://mlsanity.com/paper/2405.07800/) (2024)
- [Do Transformers Understand Ancient Roman Coin Motifs Better than CNNs?](https://mlsanity.com/paper/2601.09433/) (2026)

## Abstract (arXiv)

Understanding the predictions made by Artificial Intelligence (AI) systems is becoming more and more important as deep learning models are used for increasingly complex and high-stakes tasks. Saliency mapping -- a popular visual attribution method -- is one important tool for this, but existing formulations are limited by either computational cost or architectural constraints. We therefore propose Hierarchical Perturbation, a very fast and completely model-agnostic method for interpreting model predictions with robust saliency maps. Using standard benchmarks and datasets, we show that our saliency maps are of competitive or superior quality to those generated by existing model-agnostic methods -- and are over 20 times faster to compute.