This will be an in-person event only with attendance in Levine 307.
Autonomous AI agents are deployed in increasingly complex and uncertain environments where they must account for the presence of other agents while trying to achieve their own objectives. Moreover, such agents may require assistance from other agents to efficiently accomplish their assigned task or even be able to complete it at all.
This work aims to develop theoretical foundations for agents to learn and adopt cooperative behaviors by introducing Value of Assistance (VOA) – a novel measure of the potential performance improvement achieved by assistive actions. The benefit of using VOA will be demonstrated in multi-robot navigation and collaborative manipulation settings. In addition, adaptations of VOA to other stochastic and partially observable multi-agent settings will be discussed, including our ongoing work on AI-enabled energy systems.