Normalization

Randomly zeroing activations during training to reduce co-adaptation and overfitting.

An RNN variant using gates to mitigate vanishing gradients and capture longer context.

Injects sequence order into Transformers, since attention alone is permutation-invariant.

Converting text into discrete units (tokens) for modeling; subword tokenizers balance vocabulary size and coverage.

Activation max(0, x); improves gradient flow and training speed in deep nets.

Gradients shrink through layers, slowing learning in early layers; mitigated by ReLU, residuals, normalization.

Gradients grow too large, causing divergence; mitigated by clipping, normalization, careful init.

Methods to set starting weights to preserve signal/gradient scales across layers.