# What is Dropout? Definition & Technical Mechanics

> **Source:** [SPIDITS AI AI Glossary](https://spidits.com/ai-glossary)  
> **Category:** Model Training  
> **Canonical Citation:** [https://spidits.com/ai-glossary/dropout](https://spidits.com/ai-glossary/dropout)

## Definition
Dropout is a regularization technique used in neural networks during training where a fraction of network nodes are randomly deactivated (dropped out) in each forward pass, preventing co-adaptation of features.

## Architectural Mechanics & Deep Dive
Dropout is a popular regularization technique used to prevent overfitting in deep neural networks. During training, dropout randomly deactivates (zeroes out) a specified fraction of neurons in a layer during each forward pass. This forces the network to learn redundant representations and prevents individual neurons from co-adapting too closely, resulting in a more robust model that generalizes better to unseen testing data.

## Industry Applications & Real-World Use Cases
Neural network training, overfitting prevention, and model regularization.

## Frequently Asked Questions

### How does dropout prevent overfitting?
By turning off random neurons, it forces the remaining network to learn redundant representations and robust features rather than depending on specific pathways.

### Is dropout active during inference?
No, dropout is only active during training. During inference, all weights are utilized but scaled to match the training probability distribution.

## Related Terminology
- [overfitting](https://spidits.com/ai-glossary/overfitting)
- [neural-network](https://spidits.com/ai-glossary/neural-network)
- [bias-variance-tradeoff](https://spidits.com/ai-glossary/bias-variance-tradeoff)

---
*From the SPIDITS AI Knowledge Base (336+ terms indexed). Explore full technical definitions: [https://spidits.com/ai-glossary](https://spidits.com/ai-glossary)*