Evaluating the Cross-Lingual Effectiveness of Massively Multilingual Neural Machine Translation

  • Aditya Siddhant Google Research
  • Melvin Johnson Google Research
  • Henry Tsai Google Research
  • Naveen Ari Google Research
  • Jason Riesa Google Research
  • Ankur Bapna Google Research
  • Orhan Firat Google Research
  • Karthik Raman Google Research


The recently proposed massively multilingual neural machine translation (NMT) system has been shown to be capable of translating over 100 languages to and from English within a single model (Aharoni, Johnson, and Firat 2019). Its improved translation performance on low resource languages hints at potential cross-lingual transfer capability for downstream tasks. In this paper, we evaluate the cross-lingual effectiveness of representations from the encoder of a massively multilingual NMT model on 5 downstream classification and sequence labeling tasks covering a diverse set of over 50 languages. We compare against a strong baseline, multilingual BERT (mBERT) (Devlin et al. 2018), in different cross-lingual transfer learning scenarios and show gains in zero-shot transfer in 4 out of these 5 tasks.

AAAI Technical Track: Natural Language Processing