2 papers
cs.LG2026
Proximal Supervised Fine-Tuning
Wenhong Zhu, Ruobing Xie, Rui Wang +3
Supervised fine-tuning (SFT) of foundation models often leads to poor generalization, where prior capabilities deteriorate after tuning on new tasks or domains. Inspired by trust-r…
cs.CL2025
Weak-to-Strong Preference Optimization: Stealing Reward from Weak Aligned Model
Wenhong Zhu, Zhiwei He, Xiaofeng Wang +2
Aligning language models (LMs) with human preferences has become a key area of research, enabling these models to meet diverse user needs better. Inspired by weak-to-strong general…