Emergent Language Generalization and Acquisition Speed are not tied to Compositionality

Eugene Kharitonov, Marco Baroni


Abstract
Studies of discrete languages emerging when neural agents communicate to solve a joint task often look for evidence of compositional structure. This stems for the expectation that such a structure would allow languages to be acquired faster by the agents and enable them to generalize better. We argue that these beneficial properties are only loosely connected to compositionality. In two experiments, we demonstrate that, depending on the task, non-compositional languages might show equal, or better, generalization performance and acquisition speed than compositional ones. Further research in the area should be clearer about what benefits are expected from compositionality, and how the latter would lead to them.
Anthology ID:
2020.blackboxnlp-1.2
Volume:
Proceedings of the Third BlackboxNLP Workshop on Analyzing and Interpreting Neural Networks for NLP
Month:
November
Year:
2020
Address:
Online
Editors:
Afra Alishahi, Yonatan Belinkov, Grzegorz Chrupała, Dieuwke Hupkes, Yuval Pinter, Hassan Sajjad
Venue:
BlackboxNLP
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
11–15
Language:
URL:
https://aclanthology.org/2020.blackboxnlp-1.2
DOI:
10.18653/v1/2020.blackboxnlp-1.2
Bibkey:
Cite (ACL):
Eugene Kharitonov and Marco Baroni. 2020. Emergent Language Generalization and Acquisition Speed are not tied to Compositionality. In Proceedings of the Third BlackboxNLP Workshop on Analyzing and Interpreting Neural Networks for NLP, pages 11–15, Online. Association for Computational Linguistics.
Cite (Informal):
Emergent Language Generalization and Acquisition Speed are not tied to Compositionality (Kharitonov & Baroni, BlackboxNLP 2020)
Copy Citation:
PDF:
https://aclanthology.org/2020.blackboxnlp-1.2.pdf
Code
 facebookresearch/EGG