| Benchmark | Model | Rank | Results |
|---|---|---|---|
| cross-lingual-natural-language-inference-on-xnli | Decoupled | #2 | Accuracy: 71.3 |
| cross-lingual-natural-language-inference-on-xnli | Coupled | #3 | Accuracy: 70.7 |
| cross-lingual-question-answering-on-tydiqa-goldp | Decoupled | #6 | EM: 42.8F1: 58.1 |
| cross-lingual-question-answering-on-xquad | Decoupled | #2 | EM: 46.9F1: 63.8 |
| cross-lingual-question-answering-on-xquad | Coupled | #3 | EM: 46.2F1: 63.2 |