Diegetic representation of feedback in open games

Research output: Chapter in Book/Report/Conference proceedingConference contribution book

2 Citations (Scopus)
15 Downloads (Pure)

Abstract

We improve the framework of open games with agency by showing how the players' counterfactual analysis giving rise to Nash equilibria can be described in the dynamics of the game itself (hence diegetically), getting rid of devices such as equilibrium predicates. This new approach overlaps almost completely with the way gradient-based learners are specified and trained. Indeed, we show feedback propagation in games can be seen as a form of backpropagation, with a crucial difference explaining the distinctive character of the phenomenology of non-cooperative games. We outline a functorial construction of arena of games, show players form a subsystem over it, and prove that their 'fixpoint behaviours' are Nash equilibria.
Original languageEnglish
Title of host publicationProceedings Fifth International Conference on Applied Category Theory
EditorsJade Master, Martha Lewis
Place of PublicationWaterloo, NSW
PublisherOpen Publishing Association
Pages145-158
Number of pages14
Volume380
DOIs
Publication statusPublished - 7 Aug 2023
Event5th International Conference on Applied Category Theory - Glasgow, United Kingdom
Duration: 18 Jul 202222 Jul 2022

Publication series

NameElectronic Proceedings in Theoretical Computer Science
PublisherOpen Publishing Association
Volume380
ISSN (Print)2075-2180

Conference

Conference5th International Conference on Applied Category Theory
Country/TerritoryUnited Kingdom
CityGlasgow
Period18/07/2222/07/22

Keywords

  • open games
  • Nash equilibria
  • gradient-based learners

Fingerprint

Dive into the research topics of 'Diegetic representation of feedback in open games'. Together they form a unique fingerprint.

Cite this