Layer Normalisation is a core component of modern Transformer architectures. This article explains normalization fundamentals, internal covariate shift, why batch normalization fails in self-attention, and how layer normalization works mathematically inside Transformers—step by step with clear examples.