GPT-3: Language Models Are Few-Shot Learners — Understanding the Paper from First Principles

I recently started reading the research paper “Language Models are Few-Shot Learners”, the paper that introduced GPT-3. Instead of jumping directly into the results, I first wanted to understand the c