# Towards Robust Explanations for Deep Neural Networks

arXiv: https://arxiv.org/abs/2012.10425 · HTML: https://mlsanity.com/paper/2012.10425/

## Key facts

- Cited by 43 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 18 Dec 2020.

## Built on by (most cited first, 30 of 43)

- [AttnLRP: Attention-Aware Layer-Wise Relevance Propagation for Transformers](https://mlsanity.com/paper/2402.05602/) (2024)
- [Adversarial attacks and defenses in explainable artificial intelligence: A survey](https://mlsanity.com/paper/2306.06123/) (2023)
- [A Survey on the Robustness of Feature Importance and Counterfactual Explanations](https://mlsanity.com/paper/2111.00358/) (2021)
- [MambaLRP: Explaining Selective State Space Sequence Models](https://mlsanity.com/paper/2406.07592/) (2024)
- [Diffeomorphic Counterfactuals with Generative Models](https://mlsanity.com/paper/2206.05075/) (2022)
- [Investigating the fidelity of explainable artificial intelligence methods for applications of convolutional neural networks in geoscience](https://mlsanity.com/paper/2202.03407/) (2022)
- [SAFARI: Versatile and Efficient Evaluations for Robustness of Interpretability](https://mlsanity.com/paper/2208.09418/) (2022)
- [Decoupling Pixel Flipping and Occlusion Strategy for Consistent XAI Benchmarks](https://mlsanity.com/paper/2401.06654/) (2024)
- [Disentangled Explanations of Neural Network Predictions by Finding Relevant Subspaces](https://mlsanity.com/paper/2212.14855/) (2022)
- [Robust Explanation Constraints for Neural Networks](https://mlsanity.com/paper/2212.08507/) (2022)
- [What Makes a Good Explanation?: A Harmonized View of Properties of Explanations](https://mlsanity.com/paper/2211.05667/) (2022)
- [Efficiently Training Low-Curvature Neural Networks](https://mlsanity.com/paper/2206.07144/) (2022)
- [On the Robustness of Removal-Based Feature Attributions](https://mlsanity.com/paper/2306.07462/) (2023)
- [Neural Network Attribution Methods for Problems in Geoscience: A Novel Synthetic Benchmark Dataset](https://mlsanity.com/paper/2103.10005/) (2021)
- [Explaining Bayesian Neural Networks](https://mlsanity.com/paper/2108.10346/) (2021)
- [Scrutinizing XAI using linear ground-truth data with suppressor variables](https://mlsanity.com/paper/2111.07473/) (2021)
- [Navigating the Maze of Explainable AI: A Systematic Approach to Evaluating Methods and Metrics](https://mlsanity.com/paper/2409.16756/) (2024)
- [Visual explanation of black-box model: Similarity Difference and Uniqueness (SIDU) method](https://mlsanity.com/paper/2101.10710/) (2021)
- [Regulating Model Reliance on Non-Robust Features by Smoothing Input Marginal Density](https://mlsanity.com/paper/2407.04370/) (2024)
- [A.I. Robustness: a Human-Centered Perspective on Technological Challenges and Opportunities](https://mlsanity.com/paper/2210.08906/) (2022)
- [SPES: Spectrogram Perturbation for Explainable Speech-to-Text Generation](https://mlsanity.com/paper/2411.01710/) (2024)
- [Self-learn to Explain Siamese Networks Robustly](https://mlsanity.com/paper/2109.07371/) (2021)
- [On the Robustness of Explanations of Deep Neural Network Models: A Survey](https://mlsanity.com/paper/2211.04780/) (2022)
- [Unveiling the Hessian's Connection to the Decision Boundary](https://mlsanity.com/paper/2306.07104/) (2023)
- [SmoothHess: ReLU Network Feature Interactions via Stein's Lemma](https://mlsanity.com/paper/2311.00858/) (2023)
- [Don't Get Me Wrong: How to Apply Deep Visual Interpretations to Time Series](https://mlsanity.com/paper/2203.07861/) (2022)
- [Towards More Robust Interpretation via Local Gradient Alignment](https://mlsanity.com/paper/2211.15900/) (2022)
- [Pixel-level Certified Explanations via Randomized Smoothing](https://mlsanity.com/paper/2506.15499/) (2025)
- [Robustness of Explanation Methods for NLP Models](https://mlsanity.com/paper/2206.12284/) (2022)
- [Verified Training for Counterfactual Explanation Robustness under Data Shift](https://mlsanity.com/paper/2403.03773/) (2024)

## Abstract (arXiv)

Explanation methods shed light on the decision process of black-box classifiers such as deep neural networks. But their usefulness can be compromised because they are susceptible to manipulations. With this work, we aim to enhance the resilience of explanations. We develop a unified theoretical framework for deriving bounds on the maximal manipulability of a model. Based on these theoretical insights, we present three different techniques to boost robustness against manipulation: training with weight decay, smoothing activation functions, and minimizing the Hessian of the network. Our experimental results confirm the effectiveness of these approaches.