Cancelled: Data-Driven Mechanism Design: Jointly Eliciting Preferences and Information and an Application to LLM Auctions

Lunch

Lunch will be served at 11:45 AM.

Abstract 

We consider a mechanism design setting where agents with quasi-linear utilities possess private information about their preferences and a common payoff-
relevant state.  A mechanism jointly elicits preferences and signals.  Given the multidimensional nature of types and interdependent values, we first show that there are instances for which no standards message-driven mechanism can implement the efficient allocation in posterior equilibrium.

We propose a data-driven mechanism that conditions transfer on additional information, in particular an estimator of the state.  With the full resolution of
uncertainty or an unbiased estimator and linearity of utilities in the state, the data-driven VCG mechanism achieves implementation in posterior equilibrium. With a consistent estimator and regularity conditions, the mechanism yields implementation in approximate posterior equilibrium, with the error diminishing as the estimator converges in probability.  We apply these results to incentive alignment in digital advertising setting, including sponsored search auction and emerging formats with Large Language Models, where additional information is naturally revealed through user engagement, such as clicks on sponsored ads.

Speaker Bio

Dirk Bergemann is the Douglass and Marion Campbell Professor of Economics at Yale University. He has secondary appointments as a Professor of Computer Science in the School of Engineering and Professor of Finance in the School of Management.  He received his Ph.D. in Economics from the University of Pennsylvania in 1994. He joined Yale in 1995 as an assistant professor, having previously served as a faculty member at Princeton University. He is a member of the Cowles Foundation for Research in Economics and a fellow of the American Academy of Arts and Sciences and of the Econometric Society.
Dirk currently is Co-Editor of American Economic Review: Insights and is a Member of the Executive Committee of the Econometric Society. Dirk Bergemann was Chair of the Department of Economics from 2013-2019 and Co-Editor of Econometrica from 2014-2018.
His research is in the area of game theory, contract theory, venture capital, and market design. His most recent work is in the area of auction and information 
design. His research is supported by grants from the National Science Foundation, the Sloan Foundation, the Knight Foundation, the Omidyar Network, and Google.