OpenAI Develops CriticGPT Model Based on GPT-4 to Catch Errors in ChatGPT's Code Output; Begins Integrating CriticGPT-Like Models Into RLHF Labeling Pipeline to Assist Trainers; Plans to Further Scale and Apply RLHF to GPT-4
Portfolio Pulse from Benzinga Newsdesk
OpenAI has developed a new model called CriticGPT, based on GPT-4, to identify errors in ChatGPT's code output. The company is integrating CriticGPT-like models into its Reinforcement Learning from Human Feedback (RLHF) labeling pipeline to assist trainers and plans to further scale and apply RLHF to GPT-4.

June 27, 2024 | 5:32 pm
News sentiment analysis
Sort by:
Ascending
POSITIVE IMPACT
Microsoft, a major investor in OpenAI, may benefit from the development of CriticGPT as it enhances the reliability and accuracy of AI models like GPT-4, which could be integrated into Microsoft's products and services.
Microsoft's significant investment in OpenAI means advancements like CriticGPT can enhance the AI capabilities of Microsoft's offerings, potentially leading to improved products and services.
CONFIDENCE 90
IMPORTANCE 70
RELEVANCE 50