Abstract | ||
---|---|---|
In this paper, we introduce ELECTRA-style tasks (Clark et al., 2020b) to cross-lingual language model pre-training. Specifically, we present two pre-training tasks, namely multilingual replaced token detection, and translation replaced token detection. Besides, we pretrain the model, named as XLM-E, on both multilingual and parallel corpora. Our model outperforms the baseline models on various cross-lingual understanding tasks with much less computation cost. Moreover, analysis shows that XLM-E tends to obtain better cross-lingual transferability. |
Year | DOI | Venue |
---|---|---|
2022 | 10.18653/v1/2022.acl-long.427 | PROCEEDINGS OF THE 60TH ANNUAL MEETING OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS (ACL 2022), VOL 1: (LONG PAPERS) |
DocType | Volume | Citations |
Conference | Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) | 2 |
PageRank | References | Authors |
0.36 | 0 | 11 |
Name | Order | Citations | PageRank |
---|---|---|---|
Zewen Chi | 1 | 2 | 2.39 |
Shaohan Huang | 2 | 57 | 10.29 |
Li Dong | 3 | 582 | 31.86 |
Shuming Ma | 4 | 83 | 15.92 |
Bo Zheng | 5 | 12 | 10.73 |
Saksham Singhal | 6 | 2 | 1.71 |
Payal Bajaj | 7 | 6 | 1.44 |
Xia Song | 8 | 30 | 3.19 |
Xian-Ling Mao | 9 | 99 | 25.19 |
Heyan Huang | 10 | 173 | 61.47 |
Furu Wei | 11 | 1956 | 107.57 |