Reproduce “AdCoEvo: A Co-Evolving Multi-Agent Advertising Delivery System” 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
Tianzhu Hu, Hang Yu, Zeyuan Zhao, Zeyu Pan, Huan Luo, Chun Gan, Mingzhe Cao, Liubin Wang and 4 more
AdCoEvo coordinates advertising delivery through Planner, Optimizer, and Inspector policies that allocate budgets, adjust unit-level target cost per acquisition, and review settled outcomes. It updates one role’s policy per period using delayed feedback. In five matched 50-period AuctionNet Sparse simulations, it outperformed a frozen multi-role version: ADV proxy rose 3.96% and CPA fell 1.24%; these are not live-market results.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.