Expected flow networks in stochastic environments and two-player zero-sum games

Marco Jiralerspong, Bilun Sun, Danilo Vucetic, Tianyu Zhang, Y. Bengio +2 more
2/12/2026

No official code found. Best implementation: GFNOrg/AdversarialFlowNetworks (9 stars, last commit 2024-01-31).

Abstract

Generative flow networks (GFlowNets) are sequential sampling models trained to match a given distribution. GFlowNets have been successfully applied to various structured object generation tasks, sampling a diverse set of high-reward objects quickly. We propose expected flow networks (EFlowNets), which extend GFlowNets to stochastic environments. We show that EFlowNets outperform other GFlowNet formulations in stochastic tasks such as protein design. We then extend the concept of EFlowNets to adversarial environments, proposing adversarial flow networks (AFlowNets) for two-player zero-sum games. We show that AFlowNets learn to find above 80% of optimal moves in Connect-4 via self-play and outperform AlphaZero in tournaments.

View on arXivView PDF

Code Implementations(1)

GFNOrg/AdversarialFlowNetworksCommunity100%
95PythonDec 5, 20232 years ago

Similar Papers

Stochastic Generative Flow Networks

87% match1 repo142023

Pessimistic Backward Policy for GFlowNets

84% match1 repo52024

Pre-Training and Fine-Tuning Generative Flow Networks

84% match2 repos5492026

Generative Flow Networks as Entropy-Regularized RL

84% match1 repo412023

Towards Understanding and Improving GFlowNet Training

84% match1 repo202023

Distributional GFlowNets with Quantile Flows

83% match1 repo132023

Cite this paper

@article{jiralerspong2026expected,
  title  = {Expected flow networks in stochastic environments and two-player zero-sum games},
  author = {Marco Jiralerspong and Bilun Sun and Danilo Vucetic and Tianyu Zhang and Y. Bengio and G. Gidel and Nikolay Malkin},
  year   = {2026},
  eprint = {2310.02779},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/pdf/2310.02779},
  journal = {ICLR 2023 2023}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2310.02779.svg)](https://codeofpaper.com/paper/2310.02779)

Discussion