| machine-reading-comprehension-on-dream | ASA + RoBERTa | #1 | Accuracy: 69.2 |
| machine-reading-comprehension-on-dream | ASA + BERT-base | #3 | Accuracy: 64.3 |
| named-entity-recognition-on-wnut-2017 | ASA + RoBERTa | #8 | F1: 57.3 |
| named-entity-recognition-on-wnut-2017 | ASA + BERT-base | #15 | F1: 49.8 |
| natural-language-inference-on-multinli | ASA + RoBERTa | #14 | Matched: 88 |
| natural-language-inference-on-multinli | ASA + BERT-base | #25 | Matched: 85 |
| natural-language-inference-on-qnli | ASA + RoBERTa | #17 | Accuracy: 93.6% |
| natural-language-inference-on-qnli | ASA + BERT-base | #23 | Accuracy: 91.4% |
| paraphrase-identification-on-quora-question | ASA + RoBERTa | #7 | F1: 73.7 |
| paraphrase-identification-on-quora-question | ASA + BERT-base | #10 | F1: 72.3 |
| semantic-textual-similarity-on-sts-benchmark | ASA + RoBERTa | #25 | Spearman Correlation: 0.892 |
| semantic-textual-similarity-on-sts-benchmark | ASA + BERT-base | #38 | Spearman Correlation: 0.865 |
| sentiment-analysis-on-sst-2-binary | ASA + RoBERTa | #17 | Accuracy: 96.3 |
| sentiment-analysis-on-sst-2-binary | ASA + BERT-base | #32 | Accuracy: 94.1 |