Ad-Hoc Human-AI Coordination Challenge

Tin Dizdarević, Ravi Hammond, Tobias Gessler, Anisoara Calinescu, Jonathan Cook, Matteo Gallici, Andrei Lupu, Jakob Nicolaus Foerster
Proceedings of the 42nd International Conference on Machine Learning, PMLR 267:13900-13937, 2025.

Abstract

Achieving seamless coordination between AI agents and humans is crucial for real-world applications, yet it remains a significant open challenge. Hanabi is a cooperative card game featuring imperfect information, constrained communication, theory of mind requirements, and coordinated action – making it an ideal testbed for human-AI coordination. However, its use for human-AI interaction has been limited by the challenges of human evaluation. In this work, we introduce the Ad-Hoc Human-AI Coordination Challenge (AH2AC2) to overcome the constraints of costly and difficult-to-reproduce human evaluations. We develop human proxy agents on a large-scale human dataset that serve as robust, cheap, and reproducible human-like evaluation partners in AH2AC2. To encourage the development of data-efficient methods, we open-source a dataset of 3,079 games, deliberately limiting the amount of available human gameplay data. We present baseline results for both two- and three- player Hanabi scenarios. To ensure fair evaluation, we host the proxy agents through a controlled evaluation system rather than releasing them publicly. The code is available at https://github.com/FLAIROx/ah2ac2.

Cite this Paper

BibTeX
@InProceedings{pmlr-v267-dizdarevic25a, title = {Ad-Hoc Human-{AI} Coordination Challenge}, author = {Dizdarevi\'{c}, Tin and Hammond, Ravi and Gessler, Tobias and Calinescu, Anisoara and Cook, Jonathan and Gallici, Matteo and Lupu, Andrei and Foerster, Jakob Nicolaus}, booktitle = {Proceedings of the 42nd International Conference on Machine Learning}, pages = {13900--13937}, year = {2025}, editor = {Singh, Aarti and Fazel, Maryam and Hsu, Daniel and Lacoste-Julien, Simon and Berkenkamp, Felix and Maharaj, Tegan and Wagstaff, Kiri and Zhu, Jerry}, volume = {267}, series = {Proceedings of Machine Learning Research}, month = {13--19 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v267/main/assets/dizdarevic25a/dizdarevic25a.pdf}, url = {https://proceedings.mlr.press/v267/dizdarevic25a.html}, abstract = {Achieving seamless coordination between AI agents and humans is crucial for real-world applications, yet it remains a significant open challenge. Hanabi is a cooperative card game featuring imperfect information, constrained communication, theory of mind requirements, and coordinated action – making it an ideal testbed for human-AI coordination. However, its use for human-AI interaction has been limited by the challenges of human evaluation. In this work, we introduce the Ad-Hoc Human-AI Coordination Challenge (AH2AC2) to overcome the constraints of costly and difficult-to-reproduce human evaluations. We develop human proxy agents on a large-scale human dataset that serve as robust, cheap, and reproducible human-like evaluation partners in AH2AC2. To encourage the development of data-efficient methods, we open-source a dataset of 3,079 games, deliberately limiting the amount of available human gameplay data. We present baseline results for both two- and three- player Hanabi scenarios. To ensure fair evaluation, we host the proxy agents through a controlled evaluation system rather than releasing them publicly. The code is available at https://github.com/FLAIROx/ah2ac2.} }
Endnote
%0 Conference Paper %T Ad-Hoc Human-AI Coordination Challenge %A Tin Dizdarević %A Ravi Hammond %A Tobias Gessler %A Anisoara Calinescu %A Jonathan Cook %A Matteo Gallici %A Andrei Lupu %A Jakob Nicolaus Foerster %B Proceedings of the 42nd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2025 %E Aarti Singh %E Maryam Fazel %E Daniel Hsu %E Simon Lacoste-Julien %E Felix Berkenkamp %E Tegan Maharaj %E Kiri Wagstaff %E Jerry Zhu %F pmlr-v267-dizdarevic25a %I PMLR %P 13900--13937 %U https://proceedings.mlr.press/v267/dizdarevic25a.html %V 267 %X Achieving seamless coordination between AI agents and humans is crucial for real-world applications, yet it remains a significant open challenge. Hanabi is a cooperative card game featuring imperfect information, constrained communication, theory of mind requirements, and coordinated action – making it an ideal testbed for human-AI coordination. However, its use for human-AI interaction has been limited by the challenges of human evaluation. In this work, we introduce the Ad-Hoc Human-AI Coordination Challenge (AH2AC2) to overcome the constraints of costly and difficult-to-reproduce human evaluations. We develop human proxy agents on a large-scale human dataset that serve as robust, cheap, and reproducible human-like evaluation partners in AH2AC2. To encourage the development of data-efficient methods, we open-source a dataset of 3,079 games, deliberately limiting the amount of available human gameplay data. We present baseline results for both two- and three- player Hanabi scenarios. To ensure fair evaluation, we host the proxy agents through a controlled evaluation system rather than releasing them publicly. The code is available at https://github.com/FLAIROx/ah2ac2.
APA
Dizdarević, T., Hammond, R., Gessler, T., Calinescu, A., Cook, J., Gallici, M., Lupu, A. & Foerster, J.N.. (2025). Ad-Hoc Human-AI Coordination Challenge. Proceedings of the 42nd International Conference on Machine Learning, in Proceedings of Machine Learning Research 267:13900-13937 Available from https://proceedings.mlr.press/v267/dizdarevic25a.html.

Related Material