Johannes HEIDECKE

Johannes Heidecke is a researcher at OpenAI, focusing on AI safety and reinforcement learning. He has co-authored works on rule-based rewards for language model safety and training models to prioritize privileged instructions, contributing to the development of more secure AI systems.