Natural Language Understanding on GLUE (SST-2, MNLI, MRPC, CoLA, QNLI, QQP, STS-B Scores)
85.78SST-2 AccuracyBoT
Evaluation results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| BoTArchitecture=BERT-B, Evaluation Protocol=Directly fine-tuned2026.03 | 85.78 | 77.33 | 74.02 | 26.34 | 84.26 | 88.53 | 76.4 | 73.24 | |
| BoTArchitecture=BERT-S, Evaluation Protocol=Directly fine-tuned2026.03 | 85.67 | 74.64 | 70.34 | 17.05 | 81.97 | 88.13 | 74.2 | 70.29 | |
| LiGOArchitecture=BERT-B, Evaluation Protocol=Directly fine-tuned2026.03 | 85.47 | 77.31 | 73.04 | 25.81 | 84.12 | 88.21 | 75.7 | 72.81 | |
| MangoArchitecture=BERT-B, Evaluation Protocol=Directly fine-tuned2026.03 | 85.24 | 77.3 | 73.04 | 25.85 | 84.16 | 88.24 | 75.92 | 72.82 | |
| WSArchitecture=BERT-S, Evaluation Protocol=Directly fine-tuned2026.03 | 81.77 | 74.26 | 69.61 | 10.21 | 77.72 | 86.43 | 26.33 | 60.9 | |
| ScratchArchitecture=BERT-S, Evaluation Protocol=Directly fine-tuned2026.03 | 81.19 | 63.45 | 69.85 | 8.02 | 59.09 | 81.08 | 20.12 | 54.69 | |
| bert2BERTArchitecture=BERT-B, Evaluation Protocol=Directly fine-tuned2026.03 | 81.19 | 71.44 | 69.61 | 6.09 | 75.27 | 85.84 | 23.44 | 58.98 | |
| ScratchArchitecture=BERT-B, Evaluation Protocol=Directly fine-tuned2026.03 | 80.85 | 63.31 | 68.38 | 2.06 | 61.19 | 79.2 | 11.44 | 52.35 |