A Simple Geometric Method for Cross-Lingual Linguistic Transformations with Pre-trained Autoencoders
Maarten De Raedt, Fréderic Godin, Pieter Buteneers, Chris Develder, Thomas Demeester
Abstract
Powerful sentence encoders trained for multiple languages are on the rise. These systems are capable of embedding a wide range of linguistic properties into vector representations. While explicit probing tasks can be used to verify the presence of specific linguistic properties, it is unclear whether the vector representations can be manipulated to indirectly steer such properties. For efficient learning, we investigate the use of a geometric mapping in embedding space to transform linguistic properties, without any tuning of the pre-trained sentence encoder or decoder. We validate our approach on three linguistic properties using a pre-trained multilingual autoencoder and analyze the results in both monolingual and cross-lingual settings.
BibTeX
@inproceedings{de-raedt-etal-2021-simple,
title = "A Simple Geometric Method for Cross-Lingual Linguistic Transformations with Pre-trained Autoencoders",
author = "De Raedt, Maarten and
Godin, Fr{\'e}deric and
Buteneers, Pieter and
Develder, Chris and
Demeester, Thomas",
editor = "Moens, Marie-Francine and
Huang, Xuanjing and
Specia, Lucia and
Yih, Scott Wen-tau",
booktitle = "Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing",
month = nov,
year = "2021",
address = "Online and Punta Cana, Dominican Republic",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2021.emnlp-main.792/",
doi = "10.18653/v1/2021.emnlp-main.792",
pages = "10108--10114"
}