October 26, 2024 Model Rodin Gen-1 RLHF More stable through Reinforcement Learning from Human Feedback.