Abstract:Dropout is a widely used regularization technique in deep learning, but its effects are typically realized through stochastic masking rather than explicit optimization objectives. We propose a deterministic formulation that expresses dropout as an additive regularizer directly incorporated into the training loss. The framework derives explicit regularization terms for Transformer architectures, covering attention query, key, value, and feed-forward components with independently controllable strengths. This formulation removes reliance on stochastic perturbations while providing clearer and fine-grained control over regularization strength. Experiments across image classification, temporal action detection, and audio classification show that explicit dropout matches or outperforms conventional implicit methods, with consistent gains when applied to attention and feed-forward network layers. Ablation studies demonstrate stable performance and controllable regularization through regularization coefficients and dropout rates. Overall, explicit dropout offers a practical and interpretable alternative to stochastic regularization while maintaining architectural flexibility across diverse tasks.
| Subjects: | Machine Learning (cs.LG) |
| MSC classes: | 68T07 |
| ACM classes: | I.2.6; I.5.1 |
| Cite as: | arXiv:2604.20505 [cs.LG] |
| (or arXiv:2604.20505v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2604.20505 arXiv-issued DOI via DataCite (pending registration) |
From: Vidhi Agrawal [view email]
[v1]
Wed, 22 Apr 2026 12:45:51 UTC (537 KB)
此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。