Fetching the paper…

Direct Preference Optimization: Your Language Model is Secretly a Reward Model · Around