A decoder is a model component that converts an internal, dense, or processed representation into a more external or task-specific output representation. In sequence-to-sequence and transformer systems, a decoder often uses encoded state or prior tokens to generate the next part of an output sequence.
In general, any ML system that converts from a processed, dense, or internal representation to a more raw, sparse, or external representation. Decoders are often a component of a larger model, where they are frequently paired with an encoder. In sequence-to-sequence tasks, a decoder starts with the internal state generated by the encoder to predict the next sequence. Refer to Transformer for the definition of a decoder within the Transformer architecture. See Large language models in Machine Learning Crash Course for more information.
In general, any ML system that converts from a processed, dense, or internal representation to a more raw, sparse, or external representation. Decoders are often a component of a larger model, where they are frequently paired with an encoder. In sequence-to-sequence tasks, a decoder starts with the internal state generated by the encoder to predict the next sequence.