Bi-Factorial Preference Optimization: Balancing Safety-Helpfulness in Language Models.

Wenxuan Zhang, Philip Torr, Mohamed Elhoseiny, Adel Bibi

Browse the full ICLR paper archive.