Fetching the paper…

Inverse Preference Learning: Preference-based RL without a Reward Function · Around