MESH: Scaling Up Retrieval with Heterogeneous Content Unification

Jiaxing Qu, Yilin Chen, Junpeng Hou +4

Optimizing large-scale retrieval hinges on the ability to efficiently surface candidates across diverse content tiers. However, to capture segments such as fresh and long-tail cont…