CriticGPT, a model based on GPT-4, writes critiques of ChatGPT responses to help human trainers spot mistakes during RLHF We've trained a model, based on GPT-4, called CriticGPT to catch errors in ChatGPT's code output. We found that when people get help from CriticGPT to review ChatGPT code they outperform those without help 60% of the time. We are beginning the work to integrate CriticGPT-like m
![Finding GPT-4’s mistakes with GPT-4](https://cdn-ak-scissors.b.st-hatena.com/image/square/bd02461525a6ef51d3a2293006b99203514807cc/height=288;version=1;width=512/https%3A%2F%2Fimages.ctfassets.net%2Fkftzwdyauwt9%2F2vklrjATStHvRlA1pflP1h%2F4a22fe81fc36acedb9b59c66ee06e46d%2FBug_Finding_Research.jpg%3Fw%3D1600%26h%3D900%26fit%3Dfill)