On the Correspondence between Compositionality and Imitation in Emergent Neural Communication - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2023

On the Correspondence between Compositionality and Imitation in Emergent Neural Communication

Résumé

Compositionality is a hallmark of human language that not only enables linguistic generalization, but also potentially facilitates acquisition. When simulating language emergence with neural networks, compositionality has been shown to improve communication performance; however, its impact on imitation learning has yet to be investigated. Our work explores the link between compositionality and imitation in a Lewis game played by deep neural agents. Our contributions are twofold: first, we show that the learning algorithm used to imitate is crucial: supervised learning tends to produce more average languages, while reinforcement learning introduces a selection pressure toward more compositional languages. Second, our study reveals that compositional languages are easier to imitate, which may induce the pressure toward compositional languages in RL imitation settings.
Fichier principal
Vignette du fichier
2023.findings-acl.787.pdf (2.45 Mo) Télécharger le fichier
Origine : Fichiers éditeurs autorisés sur une archive ouverte

Dates et versions

hal-04242002 , version 1 (14-10-2023)

Licence

Paternité - Pas de modifications

Identifiants

Citer

Emily Cheng, Mathieu Rita, Thierry Poibeau. On the Correspondence between Compositionality and Imitation in Emergent Neural Communication. ACL 2023 - 61st Annual Meeting of the Association for Computational Linguistics, ACL, Jul 2023, Toronto, Canada. pp.12432-12447, ⟨10.18653/v1/2023.findings-acl.787⟩. ⟨hal-04242002⟩
12 Consultations
13 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More