CompassDPO: Dynamics-Controlled Direct Preference Optimization for Robust Safety Alignment

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Liu, Jilong, Yang, Yonghui, Shao, Pengyang, Tao, Wenjian, Zhan, Hao, Ma, Haokai, Qin, Wei, Hong, Richang
Format:	Preprint
Published:	2026
Subjects:	Machine Learning
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Internet

https://arxiv.org/abs/2603.07211