Interpretable and Editable Programmatic ...
Type de document :
Communication dans un congrès avec actes
Titre :
Interpretable and Editable Programmatic Tree Policies for Reinforcement Learning
Auteur(s) :
Kohler, Hector [Auteur]
Université de Lille
Centrale Lille
Centre de Recherche en Informatique, Signal et Automatique de Lille - UMR 9189 [CRIStAL]
Scool [Scool]
Delfosse, Quentin [Auteur]
Technische Universität Darmstadt - Technical University of Darmstadt [TU Darmstadt]
Akrour, Riad [Auteur]
Université de Lille
Centrale Lille
Centre de Recherche en Informatique, Signal et Automatique de Lille - UMR 9189 [CRIStAL]
Scool [Scool]
Kersting, Kristian [Auteur]
Technische Universität Darmstadt - Technical University of Darmstadt [TU Darmstadt]
Preux, Philippe [Auteur]
Université de Lille
Centrale Lille
Centre de Recherche en Informatique, Signal et Automatique de Lille - UMR 9189 [CRIStAL]
Scool [Scool]
Université de Lille
Centrale Lille
Centre de Recherche en Informatique, Signal et Automatique de Lille - UMR 9189 [CRIStAL]
Scool [Scool]
Delfosse, Quentin [Auteur]
Technische Universität Darmstadt - Technical University of Darmstadt [TU Darmstadt]
Akrour, Riad [Auteur]
Université de Lille
Centrale Lille
Centre de Recherche en Informatique, Signal et Automatique de Lille - UMR 9189 [CRIStAL]
Scool [Scool]
Kersting, Kristian [Auteur]
Technische Universität Darmstadt - Technical University of Darmstadt [TU Darmstadt]
Preux, Philippe [Auteur]
Université de Lille
Centrale Lille
Centre de Recherche en Informatique, Signal et Automatique de Lille - UMR 9189 [CRIStAL]
Scool [Scool]
Titre de la manifestation scientifique :
European Workshop on Reinforcement Learning
Ville :
Toulouse
Pays :
France
Date de début de la manifestation scientifique :
2024-10-28
Date de publication :
2024
Mot(s)-clé(s) en anglais :
Interpretable AI
Reinforcement Learning
Decision Tree
Reinforcement Learning
Decision Tree
Discipline(s) HAL :
Informatique [cs]/Intelligence artificielle [cs.AI]
Résumé en anglais : [en]
Deep reinforcement learning agents are prone to goal misalignments. The black-box nature of their policies hinders the detection and correction of such misalignments, and the trust necessary for real-world deployment. So ...
Lire la suite >Deep reinforcement learning agents are prone to goal misalignments. The black-box nature of their policies hinders the detection and correction of such misalignments, and the trust necessary for real-world deployment. So far, solutions learning interpretable policies are inefficient or require many human priors. We propose INTERPRETER, a fast distillation method producing INTerpretable Editable tRee Programs for ReinforcEmenT lEaRning. We empirically demonstrate that INTERPRETER compact tree programs match oracles across a diverse set of sequential decision tasks and evaluate the impact of our design choices on interpretability and performances. We show that our policies can be interpreted and edited to correct misalignments on Atari games and to explain real farming strategies.Lire moins >
Lire la suite >Deep reinforcement learning agents are prone to goal misalignments. The black-box nature of their policies hinders the detection and correction of such misalignments, and the trust necessary for real-world deployment. So far, solutions learning interpretable policies are inefficient or require many human priors. We propose INTERPRETER, a fast distillation method producing INTerpretable Editable tRee Programs for ReinforcEmenT lEaRning. We empirically demonstrate that INTERPRETER compact tree programs match oracles across a diverse set of sequential decision tasks and evaluate the impact of our design choices on interpretability and performances. We show that our policies can be interpreted and edited to correct misalignments on Atari games and to explain real farming strategies.Lire moins >
Langue :
Anglais
Comité de lecture :
Oui
Audience :
Internationale
Vulgarisation :
Non
Collections :
Source :
Fichiers
- document
- Accès libre
- Accéder au document
- Interpretable_and_Editable_.pdf
- Accès libre
- Accéder au document