Stiennon, Ouyang, Wu, Ziegler, Lowe, Voss, Radford, Amodei, and Christiano. Organization: OpenAI.
Neural Information Processing Systems ( NeurIPS 2020 ). qsUa0c4PEVK2XcJiGiow.zip
This paper introduced a method to train models (like GPT-3) to summarize text by using Reinforcement Learning from Human Feedback (RLHF) . 📂 What is in the ZIP? Stiennon, Ouyang, Wu, Ziegler, Lowe, Voss, Radford, Amodei,