Fetching the paper…

Beyond Reverse KL: Generalizing Direct Preference Optimization with Diverse Divergence Constraints · Around