Collaborative AI Teaming in Unknown Environments via Active Goal Deduction

TL;DR

This paper introduces a framework for AI agents to effectively team with unknown agents by actively deducing goals and adapting policies without prior reward knowledge, demonstrated in multi-agent and gaming environments.

Contribution

It presents a novel approach combining kernel density Bayesian inverse learning with pre-trained goal-conditioned policies for zero-shot adaptation in unknown teaming scenarios.

Findings

Significant improvement in teaming performance across diverse environments.

Effective goal deduction enables zero-shot policy adaptation.

Framework outperforms existing methods in collaborative tasks.

Abstract

With the advancements of artificial intelligence (AI), we're seeing more scenarios that require AI to work closely with other agents, whose goals and strategies might not be known beforehand. However, existing approaches for training collaborative agents often require defined and known reward signals and cannot address the problem of teaming with unknown agents that often have latent objectives/rewards. In response to this challenge, we propose teaming with unknown agents framework, which leverages kernel density Bayesian inverse learning method for active goal deduction and utilizes pre-trained, goal-conditioned policies to enable zero-shot policy adaptation. We prove that unbiased reward estimates in our framework are sufficient for optimal teaming with unknown agents. We further evaluate the framework of redesigned multi-agent particle and StarCraft II micromanagement environments with diverse unknown agents of different behaviors/rewards. Empirical results demonstrate that our framework significantly advances the teaming performance of AI and unknown agents in a wide range of collaborative scenarios.