ACL·2026

Trident: Self-Supervised Preference Alignment via Triplet Regularization

Yingnan Guo, Kejia Chen, Xiaofeng Zhang, Zifei Wu, Yu Zhang