Self-Play Q-learners Can Provably Collude in the Iterated Prisoner's Dilemma
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2023
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911328391659520 |
|---|---|
| author | Bertrand, Quentin Duque, Juan Calvano, Emilio Gidel, Gauthier |
| author_facet | Bertrand, Quentin Duque, Juan Calvano, Emilio Gidel, Gauthier |
| contents | A growing body of computational studies shows that simple machine learning agents converge to cooperative behaviors in social dilemmas, such as collusive price-setting in oligopoly markets, raising questions about what drives this outcome. In this work, we provide theoretical foundations for this phenomenon in the context of self-play multi-agent Q-learners in the iterated prisoner's dilemma. We characterize broad conditions under which such agents provably learn the cooperative Pavlov (win-stay, lose-shift) policy rather than the Pareto-dominated "always defect" policy. We validate our theoretical results through additional experiments, demonstrating their robustness across a broader class of deep learning algorithms. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2312_08484 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | Self-Play Q-learners Can Provably Collude in the Iterated Prisoner's Dilemma Bertrand, Quentin Duque, Juan Calvano, Emilio Gidel, Gauthier Computer Science and Game Theory A growing body of computational studies shows that simple machine learning agents converge to cooperative behaviors in social dilemmas, such as collusive price-setting in oligopoly markets, raising questions about what drives this outcome. In this work, we provide theoretical foundations for this phenomenon in the context of self-play multi-agent Q-learners in the iterated prisoner's dilemma. We characterize broad conditions under which such agents provably learn the cooperative Pavlov (win-stay, lose-shift) policy rather than the Pareto-dominated "always defect" policy. We validate our theoretical results through additional experiments, demonstrating their robustness across a broader class of deep learning algorithms. |
| title | Self-Play Q-learners Can Provably Collude in the Iterated Prisoner's Dilemma |
| topic | Computer Science and Game Theory |
| url | https://arxiv.org/abs/2312.08484 |