跳到正文
arXiv · AI· Hua Zheng, Shali Jiang, Boyang Liu, Laming Chen, Kenny Lov, Chuanqi Xu, Lisang Ding, Qinghai Zhou, Can Cui, Xiaolong Liu, Xiaoyi Liu, Yasmine Badr, Xin Xu, Mingfu Liang, Jiyan Yang, Ellie Dingqiao Wen, Gerard Jonathan Mugisha Akkerhuis, Jason Rudy, Xi Liu, Chenxiao Guan, Rong Jin, Ruichao Qiu, Xian Chen, Zhehui Zhou, Ping Chen, Rui Yang, Haicheng Chen, Meet Raval, Song Zhou, Dharak Kharod, Shuyu Xu, Xingyuan Wang, Liang Tao, Qiang Jin, Qiao Yang, Wankun Zhu, Qin Huang, Yuzhen Huang, Darren Liu, Parish Aggarwal, Hui Zhou, Erzhuo Wang, Shuo Chang, Xiaorui Gan, Wenlin Chen, Santanu Kolay, Huayu Li·· 2026-10-08AI 评分40

LoopFM:利用基础模型历史表征进行推荐

LoopFM: Learning frOm HistOrical RePresentations of Foundation Model for Recommendation

AI 导读

LoopFM 提出利用基础模型(FM)中间嵌入作为输入特征,替代传统知识蒸馏的单标量传输,在 TaobaoAd 基准上实现 6% 以上 AUC 提升。 在数十亿样本及万亿参数 FM 的工业级系统中,LoopFM 在 KD 基础上使知识传输比例翻倍,分别带来 +0.5%、+1.03% 和 +1.22% 的转化率提升。

来源:arXiv · AI · arxiv.org