Papers › Cross-lingual Retrieval for Iterative Self-Supervised Training

Cross-lingual Retrieval for Iterative Self-Supervised Training

16 Jun 2020NeurIPS 2020 12arXiv:2006.09526archive 2025-07-28

Chau Tran, Yuqing Tang, Xi-An Li, Jiatao Gu

Recent studies have demonstrated the cross-lingual alignment ability of multilingual pretrained language models. In this work, we found that the cross-lingual alignment can be further improved by training seq2seq models on sentence pairs mined using their own encoder outputs. We utilized these findings to develop a new approach -- cross-lingual retrieval for iterative self-supervised training (CRISS), where mining and training processes are applied iteratively, improving cross-lingual alignment and translation ability at the same time. Using this method, we achieved state-of-the-art unsupervised machine translation results on 9 language directions with an average improvement of 2.4 BLEU, and on the Tatoeba sentence retrieval task in the XTREME benchmark on 16 languages with an average improvement of 21.5% in absolute accuracy. Furthermore, CRISS also brings an additional 1.8 BLEU improvement on average compared to mBART, when finetuned on supervised machine translation downstream tasks.

PaperPDFConference PDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

pytorch/fairseq officialpytorch report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Machine TranslationRetrievalSentenceSentence RetrievalTranslationUnsupervised Machine Translation

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Methods

Introduced by this paper: CRISS

Absolute Position EncodingsAdamAttentionBPECRISSDense ConnectionsDropoutLabel SmoothingLayer NormalizationLinear LayerMulti-Head AttentionPosition-Wise Feed-Forward LayerResidual ConnectionSeq2SeqSoftmaxTransformermBART

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections