Adaptive Human-AI Coordination via Hierarchical Action Disentanglement
The paper presents a new framework for adaptive human-AI coordination called Intrinsic Action Disentanglement (IAD). This framework utilizes deep hierarchical reinforcement learning to create partner-aware action sequences that improve collaboration with diverse partners. The results demonstrate that IAD outperforms existing methods in various settings, including interactions with unseen partners and real humans.
- ▪IAD is designed to adapt to different partner behaviors and skill levels.
- ▪The framework introduces an intrinsic reward to encourage distinct action distributions.
- ▪IAD was evaluated in the Overcooked-AI domain and showed improved coordination across multiple scenarios.
Opening excerpt (first ~120 words) tap to expand
Computer Science > Artificial Intelligence arXiv:2605.24343 (cs) [Submitted on 23 May 2026] Title:Adaptive Human-AI Coordination via Hierarchical Action Disentanglement Authors:Adnan Ahmad, Bahareh Nakisa, Mohammad Naim Rastgoo View a PDF of the paper titled Adaptive Human-AI Coordination via Hierarchical Action Disentanglement, by Adnan Ahmad and 1 other authors View PDF HTML (experimental) Abstract:Human-AI collaboration requires agents that can adapt to diverse partner behaviors and skill levels while remaining robust to unseen partners. Existing methods often collapse to a single dominant behavior or learn poorly aligned skills, limiting effective coordination.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at arXiv cs.AI.
Discussion
0 commentsMore from arXiv cs.AI
