Filter, Then Reweight: Rethinking Optimization Granularity in On-Policy Distillation 文章

ArXiv CS.CL2026-06-03NEWSen作者: Yuying Li, Leqi Zheng, Yongzi Yu, Wenrui Zhou, Xuchang Zhong, Xing Hu, Jing Jin, Huangjie Yuan, Tao Feng

Filter, Then Reweight: Rethinking Optimization Granularity in On-Policy Distillation · 相关人物

暂无数据