Grounding Language to Autonomously-Acquired Skills via Goal Generation - ENSTA Paris - École nationale supérieure de techniques avancées Paris Accéder directement au contenu
Communication Dans Un Congrès Année : 2021

Grounding Language to Autonomously-Acquired Skills via Goal Generation

Résumé

We are interested in the autonomous acquisition of repertoires of skills. Language-conditioned reinforcement learning (LC-RL) approaches are great tools in this quest, as they allow to express abstract goals as sets of constraints on the states. However, most LC-RL agents are not autonomous and cannot learn without external instructions and feedback. Besides, their direct language condition cannot account for the goal-directed behavior of pre-verbal infants and strongly limits the expression of behavioral diversity for a given language input. To resolve these issues, we propose a new conceptual approach to language-conditioned RL: the Language-Goal-Behavior architecture (LGB). LGB decouples skill learning and language grounding via an intermediate semantic representation of the world. To showcase the properties of LGB, we present a specific implementation called DECSTR. DECSTR is an intrinsically motivated learning agent endowed with an innate semantic representation describing spatial relations between physical objects. In a first stage (G -> B), it freely explores its environment and targets self-generated semantic configurations. In a second stage (L -> G), it trains a language-conditioned goal generator to generate semantic goals that match the constraints expressed in language-based inputs. We showcase the additional properties of LGB w.r.t. both an end-to-end LC-RL approach and a similar approach leveraging non-semantic, continuous intermediate representations. Intermediate semantic representations help satisfy language commands in a diversity of ways, enable strategy switching after a failure and facilitate language grounding.
Fichier principal
Vignette du fichier
2006.07185.pdf (4.16 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03121146 , version 1 (26-01-2021)

Identifiants

Citer

Ahmed Akakzia, Cédric Colas, Pierre-Yves Oudeyer, Mohamed Chetouani, Olivier Sigaud. Grounding Language to Autonomously-Acquired Skills via Goal Generation. ICLR 2021 - Ninth International Conference on Learning Representation, May 2021, Vienna / Virtual, Austria. ⟨hal-03121146⟩
307 Consultations
536 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More