| paraphrase-identification-on-quora-question-pairs | ALICE | #1 | F1: 90.7 |
| paraphrase-identification-on-quora-question-pairs | FreeLB | #24 | Accuracy: 74.8Dev Accuracy: 92.6 |
| paraphrase-identification-on-quora-question-pairs | SMART-BERT | #26 | Dev Accuracy: 91.5Dev F1: 88.5 |
| semantic-textual-similarity-on-mrpc | MT-DNN-SMART | #1 | Accuracy: 93.7%F1: 91.7 |
| semantic-textual-similarity-on-mrpc | SMART | #5 | Accuracy: 91.3% |
| semantic-textual-similarity-on-sts-benchmark | MT-DNN-SMART | #1 | Pearson Correlation: 0.929Spearman Correlation: 0.925 |
| semantic-textual-similarity-on-sts-benchmark | SMARTRoBERTa | #58 | Dev Pearson Correlation: 92.8Dev Spearman Correlation: 92.6 |
| semantic-textual-similarity-on-sts-benchmark | SMART-BERT | #59 | Dev Pearson Correlation: 90.0Dev Spearman Correlation: 89.4 |
| natural-language-inference-on-multinli | T5 | #2 | Matched: 92.0Mismatched: 91.7 |
| natural-language-inference-on-multinli | MT-DNN-SMARTv0 | #52 | Accuracy: 85.7 |
| natural-language-inference-on-multinli | MT-DNN-SMART | #53 | Accuracy: 85.7 |
| natural-language-inference-on-multinli | SMART+BERT-BASE | #54 | Accuracy: 85.6 |
| natural-language-inference-on-multinli | SMARTRoBERTa | #56 | Dev Matched: 91.1Dev Mismatched: 91.3 |
| natural-language-inference-on-multinli | SMART-BERT | #57 | Dev Matched: 85.6Dev Mismatched: 86.0 |
| natural-language-inference-on-qnli | ALICE | #2 | Accuracy: 99.2% |
| natural-language-inference-on-qnli | MT-DNN-SMART | #3 | Accuracy: 99.2% |
| sentiment-analysis-on-sst-2-binary-classification | MT-DNN-SMART | #2 | Accuracy: 97.5 |
| sentiment-analysis-on-sst-2-binary-classification | MT-DNN | #37 | Accuracy: 93.6 |
| sentiment-analysis-on-sst-2-binary-classification | SMART+BERT-BASE | #42 | Accuracy: 93 |
| sentiment-analysis-on-sst-2-binary-classification | SMARTRoBERTa | #80 | Dev Accuracy: 96.9 |
| sentiment-analysis-on-sst-2-binary-classification | SMART-MT-DNN | #81 | Dev Accuracy: 96.1 |
| sentiment-analysis-on-sst-2-binary-classification | SMART-BERT | #82 | Dev Accuracy: 93.0 |
| natural-language-inference-on-rte | T5-XXL 11B | #7 | Accuracy: 92.5% |
| natural-language-inference-on-rte | SMARTRoBERTa | #10 | Accuracy: 92.0% |
| natural-language-inference-on-rte | SMART-BERT | #46 | Accuracy: 71.2% |
| natural-language-inference-on-rte | SMART | #47 | Accuracy: 71.2% |
| natural-language-inference-on-scitail | MT-DNN-SMART_100%ofTrainingData | #7 | Dev Accuracy: 96.1 |
| natural-language-inference-on-scitail | MT-DNN-SMART_10%ofTrainingData | #8 | Dev Accuracy: 91.3 |
| natural-language-inference-on-scitail | MT-DNN-SMART_1%ofTrainingData | #9 | Dev Accuracy: 88.6 |
| natural-language-inference-on-scitail | MT-DNN-SMART_0.1%ofTrainingData | #10 | Dev Accuracy: 82.3 |
| natural-language-inference-on-scitail | MT-DNN-SMARTLARGEv0 | #11 | % Dev Accuracy: 96.6% Test Accuracy: 95.2 |
| natural-language-inference-on-snli | MT-DNN-SMARTLARGEv0 | #7 | % Test Accuracy: 91.7% Dev Accuracy: 92.6 |
| natural-language-inference-on-snli | MT-DNN-SMART_100%ofTrainingData | #69 | Dev Accuracy: 91.6 |
| natural-language-inference-on-snli | MT-DNN-SMART_10%ofTrainingData | #70 | Dev Accuracy: 88.7 |
| natural-language-inference-on-snli | MT-DNN-SMART_1%ofTrainingData | #71 | Dev Accuracy: 86 |
| natural-language-inference-on-snli | MT-DNN-SMART_0.1%ofTrainingData | #72 | Dev Accuracy: 82.7 |