Alec Radford, Karthik Narasimhan, Tim Salimans, & Ilya Sutskever (2018)
OpenAI Technical Report.
URL: https://cdn.openai.com/research-covers/language-unsupervised/language_understanding_paper.pdf
Abstract. Introduces GPT (Generative Pre-trained Transformer), a decoder-only transformer trained with autoregressive language modelling then fine-tuned on downstream tasks. GPT established the decoder-only paradigm that underpins modern large language models.
Tags: transformer gpt language-models url-only