Andreas Herzig
CNRS Research Director
Logic, Interaction, Language, and Computation Group (LILaC)
IRIT, Université Paul Sabatier, France
Date: April 29, 2026 (Wednesday)
Time: 15:00–17:00
Venue: Zhixin Building, Room A1618
Location: Shandong University, Central Campus
Abstract
In multi-agent planning tasks, agents may have false beliefs about the world and about other agents that they may have to revise when they learn about the occurrence of actions. We introduce a simple specification language that is based on a lightweight fragment of an epistemic-doxastic logic and that includes ontic, observation, and communication actions. It accounts for well-known false-belief tasks such as the Sally-Anne Task. It leads to an epistemic planning formalism that is expressive enough to model second-order false-belief tasks, yet has the same computational complexity as classical planning.
About the Speaker
Andreas Herzig is a CNRS Research Director (Directeur de Recherches CNRS) in the Logic, Interaction, Language, and Computation Group (LILaC) of IRIT at Université Paul Sabatier, France. He served as Head of IRIT’s AI Department from 2009 to 2020 and as Head of the LILaC team from 2000 to 2005. He has been a member of the Scientific Board of CNRS GDR RADIA (formerly GDR IA) since 2016 and of the Executive Board of the CIMI Excellence Laboratory since 2024. He also served on the Scientific Board of the CNRS Institute for Information Sciences and Technologies (INS2I) from 2014 to 2018. He is a Fellow of the European Association for Artificial Intelligence (EurAI), Editor-in-Chief of the Journal of Applied Non-Classical Logics, Associate Editor of Artificial Intelligence, and a member of the Editorial Board of the Journal of Philosophical Logic. He has served on numerous international conference program committees.