Fetching the paper…

Learning Optimal Advantage from Preferences and Mistaking it for Reward · Around