# Layer Normalization

> Source: https://sukruyusufkaya.com/en/glossary/layer-normalization
> Updated: 2026-05-13T19:58:42.531Z
> Type: glossary
> Category: derin-ogrenme
**TLDR:** A technique that normalizes activations at the sample level and provides more stable training especially in sequence models.

<p>Layer normalization normalizes activations within each sample rather than across the batch. This makes it more suitable than batch normalization in RNNs, Transformers, and small-batch training scenarios. It improves training stability and can help gradient behavior in deep architectures. It has become one of the core building blocks of modern Transformer design.</p>