TY  - RPRT
TI  - Drift-Aware Multimodal User Representation Learning via Multi-Scale Temporal Modeling and Sparse Mixture-of-Experts
AU  - Ziqing Qian
AU  - Haohang Chen
AU  - Shengqi Dang
AU  - Yuhan Xiong
AU  - Canyu Shen
AU  - Jiaying Lei
AU  - Nan Cao
PY  - 2026
UR  - https://arxiv.org/abs/2608.25773
ID  - 2608.25773
ER  -