yongliang-wu's tracked open-source repos, sorted by stars.
[ICLR 2026] On the Generalization of SFT: A Reinforcement Learning Perspective with Reward Rectification.
