Screening Gender Transfer in Neural Machine Translation

Fran\c{c}ois Yvon; Guillaume Wisniewski; Lichao Zhu; Nicolas Ballier

RECORD

Screening Gender Transfer in Neural Machine Translation

Guillaume Wisniewski, Lichao Zhu, Nicolas Ballier, François Yvon

Computation and Language · Submitted 25 Feb 2022

abstract compiled · 2026-09-23 · no language-model review

The authors' main claim 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.

how it is stated · asserted 0.94 read the record →

Comment (0) Read on arXiv ↗

Cite ▾

arXiv ID2202.12568 (version 1)

Pith IDHVQNRZPW, Pith's permanent record of this paper and its review

CitationFran\c{c}ois Yvon, Guillaume Wisniewski, Lichao Zhu, Nicolas Ballier. "Screening Gender Transfer in Neural Machine Translation." arXiv:2202.12568 (2022). https://pith.science/paper/HVQNRZPW

Claim paper

The compiled record

The paper's own sentences, arranged. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. Numbers in the tables are the authors', checked by code. No sentence in this record was written by a model.

Open the compiled record

Compiled by classifier-1.13.0 via TypeSafe API, question bank abstract-p1-20260923, 2026-09-23: 1 request, 2,976 input tokens.

What the abstract itself states, as probabilities: claims first 0.03 · states limitation 0.08 · human participants 0.04 · controlled comparison 0.21 · randomized assignment 0.02 · states quantitative result 0.04 · compares with prior results 0.09 · mentions code or data availability 0.02.

The abstract, sentence by sentence

arXiv listing abstract v1. Results and the main claim are underlined; show every role.

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.

background contribution method result interpretation main claim

What the classifier reports for each sentence. The number is the probability it gave the label.
#	sentence	role	the authors' own work	states the main conclusion	strength
s0	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.	contribution 0.82	0.26	0.03	asserted 0.54
s1	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.	method 0.99	0.91	0.02	unclear 0.67
s2	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.	result 0.80	0.96	0.92	asserted 0.94

What kind of paper the classifier reports this is: empirical study 1.00 · evidence from observation or experiment 0.99.

What was examined, and what was not

sources read

arXiv listing abstract v1

not examined

full text
figures and tables
references and citing sentences
relations to other papers

levels run

P abstract

not run

S sections
T body sentences
C citations
R cross-paper relations

Every quoted sentence on this record is the authors'. Every label is from Pith's closed vocabulary, with the probability shown beside it. Every number in the measurements table is the authors', checked by code. Record as JSON

The paper

Original 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.

Human review

How Pith checks papers → · Full audit of this paper →

How the literature uses it

What this paper does with the arXiv papers it cites

Each citing sentence is taken from the citing paper's LaTeX source and typed by Pith's classifier. A label appears at probability 0.8 or more (0.95 for support), a contest at 0.7 or more; the quoted words are the citing authors'. How each label was checked · JSON

What it cites · 5 sentences without a confident label

Builds on 1 paper

We extended the scope of the investigation of the locus of gender transfer beyond the determiner/noun analysis of [ref] and questioned the role of predicates and epicene determiners for French.

Gender Bias in Multilingual Neural Machine Translation: The Architecture Matters · Discussion and Conclusion · 1.00

Cites as background 2 papers

Analyzing misclassified occupations in terms of gender, [ref] investigated the architectural bias for the translation of occupational nouns, suggesting that using language-specific encoders and decoder yields less bias than a shared encoder-decoder architecture.

Gender Bias in Multilingual Neural Machine Translation: The Architecture Matters · Related Work · 0.93
[ref] synthesizes the studies and datasets on gender bias for translation.

Gender Bias in Machine Translation · Related Work · 0.93

Discussion

Know something about this paper? Leave a comment, or highlight any passage above to comment on it.