# Dropout Reduces Underfitting

arXiv: https://arxiv.org/abs/2303.01500 · HTML: https://mlsanity.com/paper/2303.01500/

## Key facts

- Cited by 21 papers in the map (data through 14 Oct 2026).
- Builds on 60 of its 84 references that are in the map.
- First posted on arXiv 2 Mar 2023.

## Builds on (most cited first, 30 of 60)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [Very Deep Convolutional Networks for Large-Scale Image Recognition](https://mlsanity.com/paper/1409.1556/) (2014)
- [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://mlsanity.com/paper/2103.14030/) (2021)
- [Masked Autoencoders Are Scalable Vision Learners](https://mlsanity.com/paper/2111.06377/) (2021)
- [Mask R-CNN](https://mlsanity.com/paper/1703.06870/) (2017)
- [Hierarchical Text-Conditional Image Generation with CLIP Latents](https://mlsanity.com/paper/2204.06125/) (2022)
- [Rethinking the Inception Architecture for Computer Vision](https://mlsanity.com/paper/1512.00567/) (2015)
- [mixup: Beyond Empirical Risk Minimization](https://mlsanity.com/paper/1710.09412/) (2017)
- [Training data-efficient image transformers & distillation through attention](https://mlsanity.com/paper/2012.12877/) (2020)
- [Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning](https://mlsanity.com/paper/1506.02142/) (2015)
- [Going Deeper with Convolutions](https://mlsanity.com/paper/1409.4842/) (2014)
- [A ConvNet for the 2020s](https://mlsanity.com/paper/2201.03545/) (2022)
- [CutMix: Regularization Strategy to Train Strong Classifiers with Localizable Features](https://mlsanity.com/paper/1905.04899/) (2019)
- [Benchmarking Neural Network Robustness to Common Corruptions and Perturbations](https://mlsanity.com/paper/1903.12261/) (2019)
- [RandAugment: Practical automated data augmentation with a reduced search space](https://mlsanity.com/paper/1909.13719/) (2019)
- [The Many Faces of Robustness: A Critical Analysis of Out-of-Distribution Generalization](https://mlsanity.com/paper/2006.16241/) (2020)
- [Improved Regularization of Convolutional Neural Networks with Cutout](https://mlsanity.com/paper/1708.04552/) (2017)
- [An Empirical Study of Training Self-Supervised Vision Transformers](https://mlsanity.com/paper/2104.02057/) (2021)
- [Do ImageNet Classifiers Generalize to ImageNet?](https://mlsanity.com/paper/1902.10811/) (2019)
- [Accurate, Large Minibatch SGD: Training ImageNet in 1 Hour](https://mlsanity.com/paper/1706.02677/) (2017)
- [MLP-Mixer: An all-MLP Architecture for Vision](https://mlsanity.com/paper/2105.01601/) (2021)
- [ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness](https://mlsanity.com/paper/1811.12231/) (2018)
- [Natural Adversarial Examples](https://mlsanity.com/paper/1907.07174/) (2019)
- [MMDetection: Open MMLab Detection Toolbox and Benchmark](https://mlsanity.com/paper/1906.07155/) (2019)
- [Unified Perceptual Parsing for Scene Understanding](https://mlsanity.com/paper/1807.10221/) (2018)

## Built on by (most cited first, 21 of 21)

- [LayerSkip: Enabling Early Exit Inference and Self-Speculative Decoding](https://mlsanity.com/paper/2404.16710/) (2024)
- [SlimPajama-DC: Understanding Data Combinations for LLM Training](https://mlsanity.com/paper/2309.10818/) (2023)
- [Initializing Models with Larger Ones](https://mlsanity.com/paper/2311.18823/) (2023)
- [Deep Transfer Learning for Automatic Speech Recognition: Towards Better Generalization](https://mlsanity.com/paper/2304.14535/) (2023)
- [Efficient Stagewise Pretraining via Progressive Subnetworks](https://mlsanity.com/paper/2402.05913/) (2024)
- [A Coefficient Makes SVRG Effective](https://mlsanity.com/paper/2311.05589/) (2023)
- [Super Tiny Language Models](https://mlsanity.com/paper/2405.14159/) (2024)
- [How Effective Can Dropout Be in Multiple Instance Learning ?](https://mlsanity.com/paper/2504.14783/) (2025)
- [Keeping Deep Learning Models in Check: A History-Based Approach to Mitigate Overfitting](https://mlsanity.com/paper/2401.10359/) (2024)
- [Weak-to-Strong Generalization via Bregman Bias-Variance Decomposition](https://mlsanity.com/paper/2505.24313/) (2025)
- [Fine-tuning with Very Large Dropout](https://mlsanity.com/paper/2403.00946/) (2024)
- [Adapting LLaMA Decoder to Vision Transformer](https://mlsanity.com/paper/2404.06773/) (2024)
- [Phase Diagram of Dropout for Two-Layer Neural Networks in the Mean-Field Regime](https://mlsanity.com/paper/2510.07554/) (2025)
- [Frustratingly Easy Model Generalization by Dummy Risk Minimization](https://mlsanity.com/paper/2308.02287/) (2023)
- [GoalNet: Goal Areas Oriented Pedestrian Trajectory Prediction](https://mlsanity.com/paper/2402.19002/) (2024)
- [A robust deep learning-based damage identification approach for SHM considering missing data](https://mlsanity.com/paper/2304.00040/) (2023)
- [Revisiting Implicit Models: Sparsity Trade-offs Capability in Weight-tied Model for Vision Tasks](https://mlsanity.com/paper/2307.08013/) (2023)
- [SAMSA: Efficient Transformer for Many Data Modalities](https://mlsanity.com/paper/2408.05391/) (2024)
- [Reducing the Cost of Dropout in Flash-Attention by Hiding RNG with GEMM](https://mlsanity.com/paper/2410.07531/) (2024)
- [Dropout and Random Gradient Masking Are Asymptotically Equivalent in Large ResNets](https://mlsanity.com/paper/2607.16761/) (2026)
- [SpecDrop: Parameter-Free Category-Conditioned Routing for Modular Specialization](https://mlsanity.com/paper/2608.04084/) (2026)

## Abstract (arXiv)

Introduced by Hinton et al. in 2012, dropout has stood the test of time as a regularizer for preventing overfitting in neural networks. In this study, we demonstrate that dropout can also mitigate underfitting when used at the start of training. During the early phase, we find dropout reduces the directional variance of gradients across mini-batches and helps align the mini-batch gradients with the entire dataset's gradient. This helps counteract the stochasticity of SGD and limit the influence of individual batches on model training. Our findings lead us to a solution for improving performance in underfitting models - early dropout: dropout is applied only during the initial phases of training, and turned off afterwards. Models equipped with early dropout achieve lower final training loss compared to their counterparts without dropout. Additionally, we explore a symmetric technique for regularizing overfitting models - late dropout, where dropout is not used in the early iterations and is only activated later in training. Experiments on ImageNet and various vision tasks demonstrate that our methods consistently improve generalization accuracy. Our results encourage more research on understanding regularization in deep learning and our methods can be useful tools for future neural network training, especially in the era of large data. Code is available at https://github.com/facebookresearch/dropout.