{"paper":{"title":"External Large Foundation Model: How to Efficiently Serve Trillions of Parameters for Online Ads Recommendation","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.IR","authors_text":"Alex Gong, Alireza Vahdatpour, Amit Anand Amlesahwaram, Angie Huang, Benjamin Au, Bin Gao, Bo Long, Boyang Liu, Buyun Zhang, Chi Zhang, Chonglin Sun, Chunzhi Yang, Edison Gao, Ellie Wen, Ernest Wang, Evelyn Lyu, Fan Yang, Feifan Gu, Gaoxiang Liu, Hang Yin, Hansey Chen, Huayu Li, Hua Zheng, Jack Hsueh, Jackie Xu, Jade Nie, Jae-Woo Choi, Jamey Zhang, Jared Yang, Jason Rudy, Jiayi Xu, Jie Zheng, Jingzheng Qin, Jiyan Yang, John Bocharov, Kai Wang, Laming Chen, Lei Zhang, Lexi Song, Liang Luo, Longhao Jin, Luoshang Pan, Marcus Chen, Mengyue Hang, Mingfu Liang, Musharaf Sultan, Nancy Yu, Nan Xiao, Peggy Yao, Peng Sun, Qiang Jin, Qianru Li, Qinghai Zhou, Qin Huang, Qiuling Suo, Rich Zhu, Rocky Liu, Rong Jin, Rui Zhang, Sagar Chordia, Santanu Kolay, Shali Jiang, Shiquan Wang, Shuaiwen Wang, Shuo Chang, Shuo Gu, Shupin Mao, Shuyu Xu, Sihan Zeng, Song Zhou, Ted Lee, Tongyi Tang, Toshinari Kureha, Weilin Zhang, Wenguang Mao, Wenjing Lu, Wenjun Wang, Wenlin Chen, Wen-yen Chen, Xian Chen, Xiaohan Wei, Xiaorui Gan, Xiaoyi Liu, Xiaozhen Xia, Xihuan Zeng, Xi Liu, Xingliang Huang, Xingyuan Wang, Xin Zhang, Yantao Yao, Yasmine Badr, Yavuz Yetim, Yi Meng, Yinbin Ma, Yiping Han, Yiqun Liu, Yuchen Hao, Yujie Zha, Yuxi Hu, Yuxin Chen, Yuzhen Huang, Zhehui Zhou, Zhengli Zhao, Zhengyu Zhang, Zhichen Zeng, Zhijing Li, Zhiyuan Li","submitted_at":"2025-02-20T22:35:52Z","abstract_excerpt":"Ads recommendation is a prominent service of online advertising systems and has been actively studied. Recent studies indicate that scaling-up and advanced design of the recommendation model can bring significant performance improvement. However, with a larger model scale, such prior studies have a significantly increasing gap from industry as they often neglect two fundamental challenges in industrial-scale applications. First, training and inference budgets are restricted for the model to be served, exceeding which may incur latency and impair user experience. Second, large-volume data arriv"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.17494","kind":"arxiv","version":7},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2502.17494/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}