Screening Gender Transfer in Neural Machine Translation
Guillaume Wisniewski, Lichao Zhu, Nicolas Bailler, François Yvon
Abstract
This paper aims at identifying the information flow in state-of-the-art machine translation systems, taking as example the transfer of gender when translating from French into English. Using a controlled set of examples, we experiment several ways to investigate how gender information circulates in a encoder-decoder architecture considering both probing techniques as well as interventions on the internal representations used in the MT system. Our results show that gender information can be found in all token representations built by the encoder and the decoder and lead us to conclude that there are multiple pathways for gender transfer.- Anthology ID:
- 2021.blackboxnlp-1.24
- Volume:
- Proceedings of the Fourth BlackboxNLP Workshop on Analyzing and Interpreting Neural Networks for NLP
- Month:
- November
- Year:
- 2021
- Address:
- Punta Cana, Dominican Republic
- Venue:
- BlackboxNLP
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 311–321
- Language:
- URL:
- https://aclanthology.org/2021.blackboxnlp-1.24
- DOI:
- 10.18653/v1/2021.blackboxnlp-1.24
- Cite (ACL):
- Guillaume Wisniewski, Lichao Zhu, Nicolas Bailler, and François Yvon. 2021. Screening Gender Transfer in Neural Machine Translation. In Proceedings of the Fourth BlackboxNLP Workshop on Analyzing and Interpreting Neural Networks for NLP, pages 311–321, Punta Cana, Dominican Republic. Association for Computational Linguistics.
- Cite (Informal):
- Screening Gender Transfer in Neural Machine Translation (Wisniewski et al., BlackboxNLP 2021)
- PDF:
- https://preview.aclanthology.org/ingestion-script-update/2021.blackboxnlp-1.24.pdf