Reproduce “DTN: Deep Multiple Task-specific Feature Interactions Network for Multi-Task Recommendation” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Yaowen Bi, Yuteng Lian, Jie Cui, Jun Liu, Peijian Wang, Guanghui Li, Xuejun Chen, Jinglin Zhao and 9 more
DTN lets each task select among shared and task-specific feature-interaction modules, combining their outputs through task-sensitive gates. For sequential behaviors, its Task-Sensitive Network passes an earlier task’s prediction into a later task’s representation. Shopee offline tests report highest AUC across CTR, ATC, and CVR; a seven-day online test against PLE reports increases in clicks, orders, and GMV.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.