# Excitation Dropout: Encouraging Plasticity in Deep Neural Networks

arXiv: https://arxiv.org/abs/1805.09092 · HTML: https://mlsanity.com/paper/1805.09092/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 23 May 2018.

## Built on by (most cited first, 7 of 7)

- [Beyond Explaining: Opportunities and Challenges of XAI-Based Model Improvement](https://mlsanity.com/paper/2203.08008/) (2022)
- [Brain-inspired Artificial Intelligence: A Comprehensive Review](https://mlsanity.com/paper/2408.14811/) (2024)
- [Neuroplasticity in Artificial Intelligence -- An Overview and Inspirations on Drop In & Out Learning](https://mlsanity.com/paper/2503.21419/) (2025)
- [Explaining Cross-Domain Recognition with Interpretable Deep Classifier](https://mlsanity.com/paper/2211.08249/) (2022)
- [Efficient and Flexible Neural Network Training through Layer-wise Feedback Propagation](https://mlsanity.com/paper/2308.12053/) (2023)
- [The Overfocusing Bias of Convolutional Neural Networks: A Saliency-Guided Regularization Approach](https://mlsanity.com/paper/2409.17370/) (2024)
- [IMPACTX: improving model performance by appropriately constraining the training with teacher explanations](https://mlsanity.com/paper/2502.12222/) (2025)

## Abstract (arXiv)

We propose a guided dropout regularizer for deep networks based on the evidence of a network prediction defined as the firing of neurons in specific paths. In this work, we utilize the evidence at each neuron to determine the probability of dropout, rather than dropping out neurons uniformly at random as in standard dropout. In essence, we dropout with higher probability those neurons which contribute more to decision making at training time. This approach penalizes high saliency neurons that are most relevant for model prediction, i.e. those having stronger evidence. By dropping such high-saliency neurons, the network is forced to learn alternative paths in order to maintain loss minimization, resulting in a plasticity-like behavior, a characteristic of human brains too. We demonstrate better generalization ability, an increased utilization of network neurons, and a higher resilience to network compression using several metrics over four image/video recognition benchmarks.