Partner Capability Estimation for Task-Agnostic Adaptation in Ad-Hoc Teamwork

2026-07-29

AI Summary

This work addresses the challenges of ad hoc human-agent collaboration in multi-task settings without pre-training, where partners exhibit unknown, suboptimal, and diverse behaviors. The problem is formulated as joint planning and decentralized execution under hidden capabilities. The authors propose a task-agnostic capability estimation method based on approximate Bayesian inference, which learns a partner’s capability vector online from minimal interactions and integrates it into a Contextual Multiagent Markov Decision Process (C-MMDP) for adaptive planning. A novel CE-CM-Div mechanism is introduced to evaluate capability hypotheses using diverse planning trajectories, significantly enhancing robustness to human behavioral diversity. Experiments demonstrate that the approach rapidly recovers hidden capabilities and reduces ineffective action assignments in simulation, and substantially outperforms baselines in an offline human study with 15 participants, validating the effectiveness and interpretability of the proposed capability modeling framework.

Abstract

Effective collaboration with novel and diverse partners is a crucial skill for autonomous agents. Most current ad-hoc teamwork (AHT) approaches assume that agents will collaborate on a single, fixed task and that the partner's capabilities, their ability to successfully execute the desired action, are already known. In reality, a partner's true capabilities are often hidden, and human collaborators may act sub-optimally on tasks with multiple valid strategies. To address these limitations, we extend ad-hoc teamwork into a multi-task setting by re-framing it as a problem of joint planning with decentralised execution under hidden partner capabilities. We introduce CE-CM (Capability Estimation via Contextual Models), an approximate Bayesian method that infers task-invariant capability vectors. By using simulation-based sampling, the agent estimates capabilities and induces a contextual Multi-agent Markov Decision Processes for planning. This approach requires no population pre-training and refines its beliefs online from just a few tasks. To account for human unpredictability, we propose CE-CM-Div, an extension that evaluates capability hypotheses against diverse planner rollouts rather than a single optimal trajectory. Simulated experiments demonstrate that CE-CM rapidly recovers hidden capabilities, reduces infeasible action assignments, and adapts to changes over time. Furthermore, in an offline human study of 225 trajectories from 15 participants, CE-CM-Div substantially improved capability estimates over the baseline CE-CM method. Our results suggest capability-based modelling is a promising interpretable, task-agnostic representation in the studied settings, demonstrating that accounting for behavioural diversity is essential for robust human-AI teaming.

Problem

• ad-hoc teamwork
• partner capability estimation
• task-agnostic adaptation
• hidden capabilities
• human-AI teaming

Innovation

• Ad-Hoc Teamwork
• Capability Estimation
• Task-Agnostic Adaptation
• Bayesian Inference
• Behavioral Diversity