Anticipatory thinking in multi-agent environments: The role of theory of mind

Irina Rabkina, Constantine Nakos, Kenneth D. Forbus

Research output: Contribution to journalConference article

Abstract

Multi-agent environments pose unique challenges for the agents that interact with them. The complex behaviors of other agents can lead to novel modes of failure. Anticipating and mitigating failures due to other agents requires reasoning about their goals and beliefs, called theory of mind (ToM) reasoning. In this paper, we outline types of multiagent scenarios that require ToM reasoning and discuss the complexity of mental models required to address them.

Original languageEnglish (US)
JournalCEUR Workshop Proceedings
Volume2558
StatePublished - Jan 1 2020
Event2019 Workshop on Cognitive Systems for Anticipatory Thinking, COGSAT 2019 - Arlington, United States
Duration: Nov 7 2019Nov 9 2019

ASJC Scopus subject areas

  • Computer Science(all)

Cite this