Towards Robust Alignment of Language Models: Distributionally Robustifying Direct Preference Optimization
Primary LanguagePython