1 paper · 1 filter
Yoav Kolumbus, Joe Halpern, Éva Tardos
In repeated games, such as auctions, players rely on autonomous learning agents to choose their actions. We study settings in which players have their agents make monetary transfer…