PositionalEncoding class

Injects information about the relative or absolute position of tokens in a sequence.

Since the Transformer architecture contains no recurrence, it has no inherent sense of word order. This layer adds a unique, non-trainable vector to each input embedding, allowing the model to learn from the sequence order.

It uses the standard sinusoidal formula from the "Attention Is All You Need" paper.

Inheritance

• Object
• Layer
• PositionalEncoding

Constructors

PositionalEncoding(int maxLength, int dModel)

Properties

dModel ↔ int

getter/setter pair

encodingMatrix ↔ Tensor<Matrix>

getter/setter pair

hashCode → int

The hash code for this object.

no setter, inherited

maxLength ↔ int

getter/setter pair

name ↔ String

A user-friendly name for the layer (e.g., 'dense', 'lstm').

getter/setter pair, override-getter

parameters → List<Tensor>

A list of all trainable tensors (weights and biases) in the layer.

no setter, override

runtimeType → Type

A representation of the runtime type of the object.

no setter, inherited

Methods

build(Tensor input) → void

Initializes the layer's parameters based on the shape of the first input.

override

call(Tensor input) → Tensor

The public, callable interface for the layer.

inherited

forward(Tensor input) → Tensor<Matrix>

The core logic of the layer's transformation.

override

noSuchMethod(Invocation invocation) → dynamic

Invoked when a nonexistent method or property is accessed.

inherited

toString() → String

A string representation of this object.

inherited

Operators

operator ==(Object other) → bool

The equality operator.

inherited