Shikang(Danny) Wu   /   吴世康

Shikang Wu is a Senior Staff ML Scientist and Tech Lead in Douyin Search Ranking, where he leads the development of the next-generation Douyin Search Engine. His work focuses on Generative Search Foundation Models, unified retrieval-and-ranking modeling, and Large Recommendation Models.

Prior to joining Douyin, he was a Senior ML Scientist at Baidu Search. He received his M.S. and B.E. degrees from Beijing Jiaotong University, advised by Zhihao Wu and Youfang Lin.

He is working toward meaningful impact. Feel free to get in touch:

Email  /  CV  /  Google Scholar  /  Linkedin

profile photo
News
  • 2026.08: 🎉🎉 Our paper “LEMUR” is accepted by CIKM 2026, presenting the first large-scale multimodal recommender system trained end-to-end from raw data.
  • 2026.02: 🎉🎉 Our papers “MSN”, “MDL”, and “TokenMixer-Large” are accepted by KDD 2026. MSN and TokenMixer-Large present sparse activation scaling frameworks based on memory networks and Sparse MoE, while MDL unifies multi-scenario and multi-task learning through tokenization.
  • 2026.01: 🎉🎉 Our paper “HyFormer” is accepted by SIGIR 2026, presenting a unified framework of sequence modeling and feature interaction.
Research Interests

Shikang's research interests include recommender systems and information retrieval. More specifically, his research focuses on:

  • Large-scale recommendation models inspired by recent progress in LLMs and VLMs, emphasizing performance improvements through scaling model capacity.
  • The integration of multimodal signals and personalization in recommender systems.
  • Representation learning, particularly graph representation learning.
  • Generative recommendation models, including sequential and multimodal paradigms.
Experience and Education

Experience

ByteDance · Douyin Search Ranking Dec. 2023 - Present
Current responsibility: Tech Lead Jan. 2026 - Present
Welcome to join my team!
Sr. Staff ML Scientist Jul. 2026 - Present
Leading the direction of the Douyin Next-Gen Search Engine.
Staff ML Scientist Jan. 2026 - Jun. 2026
Built a next-generation search engine based on a Generative Search Foundation Model, a unified model combining retrieval and ranking capabilities, and a Large Recommendation Model scaling towards 4B dense parameters, 100T sparse features, and 100K sequence lengths.
Senior ML Scientist Dec. 2023 - Dec. 2025
Led the Large Recommendation Model (LRM) for Douyin Search, focusing on sparse activation, sequence, and model scaling; all-token and multi-distribution learning; multimodal co-training; and next-generation search tools.
Baidu · Baidu Search Jun. 2021 - Dec. 2023
Senior ML Scientist Jul. 2023 - Dec. 2023
Received the only cross-level promotion within the Business Group.
ML Scientist Jun. 2021 - Jun. 2023
Meituan Jun. 2020 - Sep. 2020
ML Scientist Intern

Education

Beijing Jiaotong University
M.E., School of Information Technology Sep. 2018 - Jun. 2021
B.E., School of Information Technology Sep. 2014 - Jun. 2018
Publications

Note: * indicates equal contributions, † indicates Corresponding authors.

MSN framework
MSN: A Memory-based Sparse Activation Scaling Framework for Large-scale Industrial Recommendation.
Shikang Wu*, Hui Lu*, Jinqiu Jin*, Zheng Chai*, Shiyong Hong, Junjie Zhang, Shanlei Mu, Kaiyuan Ma, Tianyi Liu, Yuchao Zheng, Zhe Wang†, Jingjian Lin
KDD 2026 Accepted
[arXiv]
MDL framework
MDL: A Unified Multi-Distribution Learner in Large-scale Industrial Recommendation through Tokenization.
Shanlei Mu*, Yuchen Jiang*, Shikang Wu*, Shiyong Hong, Tianmu Sha, Junjie Zhang, Jie Zhu, Zhe Chen, Zhe Wang†, Jingjian Lin
KDD 2026 Accepted
[arXiv]
TokenMixer-Large framework
TokenMixer-Large: Scaling Up Large Ranking Models in Industrial Recommenders.
Yuchen Jiang*, Jie Zhu*, Xintian Han*, Hui Lu*, Kunmin Bai*, Mingyu Yang*, Shikang Wu*, Ruihao Zhang*, Wenlin Zhao*, Shipeng Bai, Sijin Zhou, Huizhi Yang, Tianyi Liu, Wenda Liu, Ziyan Gong, Haoran Ding, Zheng Chai, Deping Xie, Zhe Chen, Yuchao Zheng†, Peng Xu
KDD 2026 Accepted
[arXiv]
HyFormer framework
HyFormer: Revisiting the Roles of Sequence Modeling and Feature Interaction in CTR Prediction.
Yunwen Huang*, Shiyong Hong*, Xijun Xiao*, Jinqiu Jin*, Xuanyuan Luo, Zhe Wang, Zheng Chai†, Shikang Wu†, Yuchao Zheng, Jingjian Lin
SIGIR 2026 Accepted
[arXiv]
LEMUR framework
LEMUR: Large scale End-to-end MUltimodal Recommendation.
Xintian Han*, Honggang Chen*, Quan Lin*, Jingyue Gao*, Xiangyuan Ren*, Lifei Zhu, Zhisheng Ye, Shikang Wu, XiongHang Xie, Xiaochu Gan, Bingzheng Wei, Peng Xu, Zhe Wang, Yuchao Zheng†, Jingjian Lin, Di Wu, Junfeng Ge
CIKM 2026 Accepted
[arXiv]

Last update: August 2026      Template