# What is Residual Connection? Definition & Technical Mechanics

> **Source:** [SPIDITS AI AI Glossary](https://spidits.com/ai-glossary)  
> **Category:** Neural Architectures  
> **Canonical Citation:** [https://spidits.com/ai-glossary/residual-connection](https://spidits.com/ai-glossary/residual-connection)

## Definition
A Residual Connection (or skip connection) is an architectural feature in deep neural networks that passes the input of a layer directly to its output, bypassing one or more intermediate layers by adding them together.

## Architectural Mechanics & Deep Dive
A residual connection (or skip connection) is a neural network component that passes input signals directly to a later layer, skipping intermediate non-linear transformations. ResNet architectures use residual connections to mitigate the vanishing gradient problem, enabling the training of neural networks with hundreds of layers.

## Industry Applications & Real-World Use Cases
ResNet architectures in computer vision, Transformer layer stacks, and ultra-deep neural networks.

## Frequently Asked Questions

### Why do residual connections prevent performance degradation in deep networks?
They create an uninterrupted highway for gradients to flow backward during backpropagation, effectively bypassing vanishing gradient bottlenecks in intermediate layers.

### What is the mathematical formulation of a residual block?
If the layers perform a mapping F(x), the output of the residual block is H(x) = F(x) + x, where x is the original input.

## Related Terminology
- [neural-network](https://spidits.com/ai-glossary/neural-network)
- [transformer](https://spidits.com/ai-glossary/transformer)
- [vanishing-gradient-problem](https://spidits.com/ai-glossary/vanishing-gradient-problem)

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