Tag

#interpretability

1 post

• May 9, 2023 · Essay · 1 min

  GPT-4 Explains Neurons in GPT-2

  In a new study, OpenAI's GPT-4 is used to explain the behavior of neurons in GPT-2.