Reinforcement Learning from LLM Feedback to Counteract Goal Misgeneralization

Re-assign community