Screening Gender Transfer in Neural Machine Translation
This addresses the issue of understanding information flow in machine translation for researchers, but it is incremental as it focuses on a specific case without broad new methods.
The paper tackled the problem of identifying how gender information flows in neural machine translation systems, specifically from French to English, and found that gender information is present in all token representations of the encoder and decoder, indicating multiple pathways for gender transfer.
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.