Improving Language Understanding by Generative Pre-Training (GPT-1)

作者： Alec Radford, Karthik Narasimhan, Tim Salimans, Ilya Sutskever (2018)

TLDR（中文）

OpenAI 首次提出 decoder-only + 自回归预训练 + 任务微调；为后来 GPT-2/3/4 系列奠定路线。当时不如 BERT 火，几年后被证明是真正的赢家方向。

TLDR (English)

OpenAI's first proposal of decoder-only + autoregressive pre-training + task fine-tuning, establishing the foundation for GPT-2/3/4 series. Less popular than BERT initially, but proven to be the winning direction years later.