Autoregressive Transformer
A type of neural network architecture (like those used in GPT models) that predicts text one token at a time, with each token prediction based on all previous tokens. 'Autoregressive' means it uses its own previous outputs as inputs; 'Transformer' refers to the specific attention-based architecture.
As defined in The Future of Knowledge Graphs in a World of Large Language Models
Where it comes up · 1
Most important claims · 1 of 1Search all claims for “Autoregressive Transformer” →