AI 뉴스로 돌아가기

Finding GPT-4’s mistakes with GPT-4

CriticGPT, a model based on GPT-4, writes critiques of ChatGPT responses to help human trainers spot mistakes during RLHF

2024.06.2711원문 보기
CriticGPT, a model based on GPT-4, writes critiques of ChatGPT responses to help human trainers spot mistakes during RLHF