OpenAI lance CriticGPT, un modèle basé sur GPT-4

Odaily Daily News reports that OpenAI has recently launched a model based on GPT-4 called CriticGPT, designed to capture errors in the output of ChatGPT code. This model can provide critical comments on ChatGPT response results, helping AI trainers to identify errors discovered during the 'Reinforcement Learning based on Human Feedback (RLHF)' process and providing clear AI assistance for AI trainers. (Financial Association)