On Incentivized Exploration beyond Bayesianism and Full-Information
It addresses the problem of designing incentive-compatible exploration mechanisms for non-Bayesian agents with private information, which is a foundational issue in online learning and mechanism design.
The paper extends incentivized exploration beyond Bayesian full-information settings, introducing new definitions for agents with external information or multiple priors, and showing that existing approaches fail in these contexts.
We extend Incentive Compatible Exploration beyond the Bayesian full-information setting of Kremer et al. [2014]. We consider agents that may possess external information unknown to the principal. We show such settings require new notions of incentivized exploration, as well as going beyond a Bayesian perspective, and we introduce a definition where agents choose any reasonable (undominated) action. Furthermore, our framework provides for a more robust treatment of ties, and extends to settings where agents lack a single common prior and instead only know that reward distributions belong to a collection of potential priors.