Learning from Explanations and Demonstrations: A Pilot Study - Sorbonne Université Accéder directement au contenu
Communication Dans Un Congrès Année : 2020

Learning from Explanations and Demonstrations: A Pilot Study

Résumé

We discuss the relationship between explainability and knowledge transfer in reinforcement learning. We argue that explainability methods, in particular methods that use counterfactuals, might help increasing sample efficiency. For this, we present a computational approach to optimize the learner's performance using explanations of another agent and discuss our results in light of effective natural language explanations for both agents and humans.
Fichier non déposé

Dates et versions

hal-03152179 , version 1 (09-03-2021)

Identifiants

  • HAL Id : hal-03152179 , version 1

Citer

Silvia Tulli, Sebastian Wallkötter, Ana Paiva, Francisco Melo, Mohamed Chetouani. Learning from Explanations and Demonstrations: A Pilot Study. 2nd Workshop on Interactive Natural Language Technology for Explainable Artificial Intelligence, Dec 2020, Dubin, Virtual, Ireland. pp.61-66. ⟨hal-03152179⟩
53 Consultations
1 Téléchargements

Partager

Gmail Facebook X LinkedIn More