Fetching the paper…

Reinforcement Learning from LLM Feedback to Counteract Goal Misgeneralization · Around