Implicitly Aligning Humans and Autonomous Agents Through Shared Task Abstractions
Abstract
In collaborative tasks, autonomous agents fall short of humans in their capability to quickly adapt to new and unfamiliar teammates. We posit that a limiting factor for zero-shot coordination is the lack of shared task abstractions, a mechanism humans rely on to implicitly align with teammates. To address this gap, we introduce HA^2: Hierarchical Ad Hoc Agents, a framework leveraging hierarchical reinforcement learning to mimic the structured approach humans use in collaboration. We evaluate HA^2 in the Overcooked environment, demonstrating statistically significant improvement over existing baselines when paired with both unseen agents and humans, providing better resilience to environmental shifts, and outperforming all state-of-the-art methods.
Cite
Text
Aroca-Ouellette et al. "Implicitly Aligning Humans and Autonomous Agents Through Shared Task Abstractions." International Joint Conference on Artificial Intelligence, 2025. doi:10.24963/IJCAI.2025/457Markdown
[Aroca-Ouellette et al. "Implicitly Aligning Humans and Autonomous Agents Through Shared Task Abstractions." International Joint Conference on Artificial Intelligence, 2025.](https://mlanthology.org/ijcai/2025/arocaouellette2025ijcai-implicitly/) doi:10.24963/IJCAI.2025/457BibTeX
@inproceedings{arocaouellette2025ijcai-implicitly,
title = {{Implicitly Aligning Humans and Autonomous Agents Through Shared Task Abstractions}},
author = {Aroca-Ouellette, Stéphane and Aroca-Ouellette, Miguel and von der Wense, Katharina and Roncone, Alessandro},
booktitle = {International Joint Conference on Artificial Intelligence},
year = {2025},
pages = {4101-4109},
doi = {10.24963/IJCAI.2025/457},
url = {https://mlanthology.org/ijcai/2025/arocaouellette2025ijcai-implicitly/}
}