Transformer

Intermediate

Architecture based on self-attention and feedforward layers; foundation of modern LLMs and many multimodal models.

Full Definition

Architecture based on self-attention and feedforward layers; foundation of modern LLMs and many multimodal models.

Keywords

attention sequence

Domains

Transformers & LLMs

Related Terms

Attention used in

Mechanism that computes context-aware mixtures of representations; scales well and captures long-range dependencies.

 Neural Network narrower than

A parameterized function composed of interconnected units organized in layers with nonlinear activations.

 Large Language Model narrower than

A high-capacity language model trained on massive corpora, exhibiting broad generalization and emergent behaviors.

 Speech Recognition used in

Converting audio speech into text, often using encoder-decoder or transducer architectures.

 Attention related to

Mechanism that computes context-aware mixtures of representations; scales well and captures long-range dependencies.

 Self-Attention related to

Attention where queries/keys/values come from the same sequence, enabling token-to-token interactions.

 Vocabulary related to

The set of tokens a model can represent; impacts efficiency, multilinguality, and handling of rare strings.

 Context Window related to

Maximum number of tokens the model can attend to in one forward pass; constrains long-document reasoning.

Concept Map

See how Transformer connects to other concepts.

Open Knowledge Graph