Adaptive Human-AI Coordination via Hierarchical Action Disentanglement

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Ahmad, Adnan, Nakisa, Bahareh, Rastgoo, Mohammad Naim
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916041437741056
author Ahmad, Adnan
Nakisa, Bahareh
Rastgoo, Mohammad Naim
author_facet Ahmad, Adnan
Nakisa, Bahareh
Rastgoo, Mohammad Naim
contents 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. We propose Intrinsic Action Disentanglement (IAD), a deep hierarchical reinforcement learning (DHRL) framework that learns distinct, partner-aware low-level action sequences conditioned on high-level latent skills. IAD introduces an intrinsic reward that explicitly encourages disentangled action distributions of the agent's low-level policy across skills, yielding an interpretable mapping between high-level decisions and partner-specific behavioral responses. By capturing temporally extended interaction patterns, IAD enables flexible adaptation to heterogeneous partner dynamics under distributional shift. We evaluate IAD in the Overcooked-AI domain across multiple layouts and diverse partner settings, including unseen simulated partners, a human-proxy model trained on human-human gameplay, and real human partners. Results show that IAD consistently outperforms strong baselines and achieves more reliable, adaptive coordination across all settings.
format Preprint
id arxiv_https___arxiv_org_abs_2605_24343
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Adaptive Human-AI Coordination via Hierarchical Action Disentanglement
Ahmad, Adnan
Nakisa, Bahareh
Rastgoo, Mohammad Naim
Artificial Intelligence
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. We propose Intrinsic Action Disentanglement (IAD), a deep hierarchical reinforcement learning (DHRL) framework that learns distinct, partner-aware low-level action sequences conditioned on high-level latent skills. IAD introduces an intrinsic reward that explicitly encourages disentangled action distributions of the agent's low-level policy across skills, yielding an interpretable mapping between high-level decisions and partner-specific behavioral responses. By capturing temporally extended interaction patterns, IAD enables flexible adaptation to heterogeneous partner dynamics under distributional shift. We evaluate IAD in the Overcooked-AI domain across multiple layouts and diverse partner settings, including unseen simulated partners, a human-proxy model trained on human-human gameplay, and real human partners. Results show that IAD consistently outperforms strong baselines and achieves more reliable, adaptive coordination across all settings.
title Adaptive Human-AI Coordination via Hierarchical Action Disentanglement
topic Artificial Intelligence
url https://arxiv.org/abs/2605.24343