Interactive POMDP Lite: Towards Practical Planning to Predict and Exploit Intentions for Interacting with Self-Interested Agents

TL;DR

This paper introduces an intention-aware planning framework for non-cooperative multi-agent systems that efficiently predicts and exploits other agents' intentions, improving robustness and performance in stochastic games.

Contribution

It presents a practical planning approach that integrates intention prediction with bounded performance loss, outperforming existing algorithms in multi-agent interactions.

Findings

Performance loss is linearly bounded by intention prediction error.

Policies achieve better robustness and effectiveness than state-of-the-art methods.

Empirical results demonstrate improved success in stochastic game scenarios.

Abstract

A key challenge in non-cooperative multi-agent systems is that of developing efficient planning algorithms for intelligent agents to interact and perform effectively among boundedly rational, self-interested agents (e.g., humans). The practicality of existing works addressing this challenge is being undermined due to either the restrictive assumptions of the other agents' behavior, the failure in accounting for their rationality, or the prohibitively expensive cost of modeling and predicting their intentions. To boost the practicality of research in this field, we investigate how intention prediction can be efficiently exploited and made practical in planning, thereby leading to efficient intention-aware planning frameworks capable of predicting the intentions of other agents and acting optimally with respect to their predicted intentions. We show that the performance losses incurred by the resulting planning policies are linearly bounded by the error of intention prediction. Empirical evaluations through a series of stochastic games demonstrate that our policies can achieve better and more robust performance than the state-of-the-art algorithms.