EconPapers    
Economics at your fingertips  
 

Simultaneous Gradient Learning in First-Price Auctions

Janik Bürgermeister, Julius Durmann, Martin Bichler (), Mete Şeref Ahunbay (), Bary Pradelski () and Marco Scarsini ()
Additional contact information
Janik Bürgermeister: TUM - Technische Universität Munchen = Technical University Munich = Université Technique de Munich
Julius Durmann: TUM - Technische Universität Munchen = Technical University Munich = Université Technique de Munich
Martin Bichler: TUM - Technische Universität Munchen = Technical University Munich = Université Technique de Munich
Mete Şeref Ahunbay: Department of Computer of Science [University of Oxford] - University of Oxford, GHOST - Jeux, optimisation mathématique et systèmes stochastiques - Centre Inria de l'Université Grenoble Alpes - Inria - Institut National de Recherche en Informatique et en Automatique - LIG - Laboratoire d'Informatique de Grenoble - Inria - Institut National de Recherche en Informatique et en Automatique - CNRS - Centre National de la Recherche Scientifique - UGA - Université Grenoble Alpes - Grenoble INP - Institut polytechnique de Grenoble - Grenoble Institute of Technology - UGA - Université Grenoble Alpes
Bary Pradelski: MFO - Maison Française d'Oxford - MEAE - Ministère de l'Europe et des Affaires étrangères - CNRS - Centre National de la Recherche Scientifique, GHOST - Jeux, optimisation mathématique et systèmes stochastiques - Centre Inria de l'Université Grenoble Alpes - Inria - Institut National de Recherche en Informatique et en Automatique - LIG - Laboratoire d'Informatique de Grenoble - Inria - Institut National de Recherche en Informatique et en Automatique - CNRS - Centre National de la Recherche Scientifique - UGA - Université Grenoble Alpes - Grenoble INP - Institut polytechnique de Grenoble - Grenoble Institute of Technology - UGA - Université Grenoble Alpes
Marco Scarsini: LUISS - Libera Università Internazionale degli Studi Sociali Guido Carli [Roma]

Post-Print from HAL

Abstract: We study equilibrium learning in discretized Bayesian games, focusing on first-price auctions as a testbed for gradient-based learning dynamics under private information. While the symmetric independent private values model admits a unique Bayesian correlated equilibrium (BCE) in the continuous limit, standard convergence guarantees for gradient-based learners do not apply: classical no-regret theory predicts convergence only to Bayesian coarse correlated equilibria (BCCE), which can remain large and far from the competitive equilibrium even under fine discretizations. To bridge this gap, we introduce the notion of Bayesian semicoarse correlated equilibrium (BSCCE), an equilibrium concept that refines BCCE by restricting deviations to probability-preserving transformations consistent with projected gradient dynamics. We provide a linear programming characterization of BSCCE with polynomially many variables and constraints and establish the inclusion hierarchy BCE ⊆ BSCCE ⊆ BCCE. Our main finding is that the BSCCE set contracts in many cases as the discretization of the Bayesian game is refined. Even with priors for which BCCE fails to concentrate, the diameter of the BSCCE polytope shrinks toward zero, indicating convergence to the unique continuous-game Bayes-Nash equilibrium. These results provide a principled explanation for the observed robustness of gradient-based learning in discretized Bayesian games, showing that these dynamics select a smaller equilibrium set than standard no-regret theory predicts.

Keywords: Correlated equilibrium; Gradient dynamics; Equilibrium learning; Bayesian games; First-price auctions (search for similar items in EconPapers)
Date: 2026-12-09
References: Add references at CitEc
Citations:

Published in NeurIPS 2026 - 40th Conference on Neural Information Processing Systems, Dec 2026, Paris, France

There are no downloads for this item, see the EconPapers FAQ for hints about obtaining it.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:hal:journl:hal-05805515

Access Statistics for this paper

More papers in Post-Print from HAL
Bibliographic data for series maintained by CCSD ().

 
Page updated 2026-10-27
Handle: RePEc:hal:journl:hal-05805515