[2204.05862] Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback September 9, 2026 by jlamprecht https://arxiv.org/abs/2204.05862
