| natural-language-inference-on-scitail | MT-DNN | #2 | Accuracy: 94.1 |
| natural-language-inference-on-snli | MT-DNN | #8 | % Test Accuracy: 91.6% Train Accuracy: 97.2Parameters: 330m |
| natural-language-inference-on-snli | Ntumpha | #9 | % Test Accuracy: 90.5% Train Accuracy: 99.1Parameters: 220 |
| paraphrase-identification-on-quora-question-pairs | MT-DNN | #10 | F1: 72.4Accuracy: 89.6 |
| linguistic-acceptability-on-cola | MT-DNN | #15 | Accuracy: 68.4% |
| natural-language-inference-on-multinli | MT-DNN | #21 | Matched: 86.7Mismatched: 86.0 |
| sentiment-analysis-on-sst-2-binary-classification | MT-DNN | #23 | Accuracy: 95.6 |