Decoder Deep Learning, NLP
noun
Definition: A model component that generates or reconstructs outputs from encoded representations, commonly used in sequence-to-sequence models and autoencoders [Robertson 2017].
Example in context: “The essential components, like feed forward, residual add, and layer norm, are almost identical in an encoder and a decoder transformer block.” [Yu et al. 2023]
“For instance, GraphIE (Qian et al., 2018) discovers non-local relationships between textual units, generates local context-aware latent reconstructions of words or sentences, and utilizes a decoder to label words at the word level.” [Chikwendu et al. 2023]
Related terms: sequence decoder, generator module, encoder-decoder architecture