OpenAI has introduced CriticGPT, an AI model that detects and highlights errors in code developed by other AI services like ChatGPT.
The model is based on the GPT-4 family and uses “Reinforcement Learning from Human Feedback” (RLHF), where it collects comparisons in which AI trainers rate different ChatGPT responses against each other. The company claims that the model can identify subtle mistakes and spot errors in data previously marked flawless by human annotators. It can also identify standard and inserted bugs in coding.
By using this site, you agree to allow SPEEDA Edge and our partners to use cookies for analytics and personalization. Visit our privacy policy for more information about our data collection practices.