Toulouse Workshop on Explanation and Epistemic Reasoning

IRIT, Toulouse

5-6 December 2022

Sponsors:

  • CNRS project LEXIA ("The Logic of Explanation: From Explainable to Explaining Legal AI Systems")

  • ERC project CompuLaw ("Governance of Computational Entities Through An Integrated Legal and Technical Framework")

  • Artificial and Natural Intelligence Toulouse Institute (ANITI)

  • EU project TAILOR ("A Network for Trustworthy AI")

  • LIANDA project ("Logic and Artificial Intelligence: New Challenges and Adequacy")

Programme

Day 1 (Auditorium J. Herbrand, IRIT)

  • 9:00-9:15: Introduction

  • 9:15-9:45: "Explanations, counterfactuals, causality and statistics", Nicholas Asher

  • 9:45-10:15: Missing title, Leila Amgoud

  • 10:15-10:45: "Non-Classical Logics for Explanations in AI Systems", Emiliano Lorini

  • 10:45-11:15: Coffee break

  • 11:15-11:45: "Argumentation and Explanation in the Law", Antonino Rotolo

  • 11:45-12:15: "Fairness standards and algorithmic decision making. What about COMPAS?", Francesca Lagioia and Giovanni Sartor

  • 12:45-14:30: Lunch break

  • 14:30-15:00: "Combining deep learning based NLP with PROLEG (PROlog based LEGal reasoning support system)", Ken Satoh [online talk]

  • 15:00-15:30: "Modelling and Explaining Legal Cased-based Reasoners through Classifiers", Xinghan Liu

  • 15:30-16:00: "Data Science for Justice: Evidence from a Randomized Judicial Reform in Kenya", Daniel Chen

  • 16:00-16:30: Coffee break

  • 16:30-17:00: "From Explainability to Fairness in AI: A Study of Automated Loan Lending", Thomas Souverain

  • 17:00-17:30: "Explanations in team formation scenario", Athina Georgara

Day 2 (Salle des thèses, IRIT)

  • 9:30-10:00: "Cumulative Consequence in Neural Networks", Hannes Leitgeb [online talk]

  • 10:00-10:30: "Attention! A Dynamic Epistemic Logic Model for Inattentive Agents", Gaia Belardinelli & Thomas Bolander

  • 10:30-11:00: "Logic-based Multi-Agent Ethical Planning", Timothy Parker

  • 11:00-11:30: Coffee break

  • 11:30-12:00: "Deontic Explanations", Leon van der Torre

  • 12:00-12:30: "Awareness Neighbourhood Model for representing Explicit Aware Knowledge", Claudia Fernández-Fernández

  • 12:30-14:15: Lunch break

  • 14:15-14:45: "Trust and Logic", Andreas Herzig

  • 14:45:15:15: "A Dynamic Procedure for Information Evaluation", Benjamin Icard

Workshop organizers

Andreas Herzig and Emiliano Lorini

Registration

Both online and on site attendance are free. In both cases, please register by sending an email to herzig@irit.fr by specifying whether you will attend the workshop online or on site.