Natural Language Understanding on GLUE (7-Task Performance Subset)
76.98CoLADTS-T
Evaluation results
| Method | Links | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| DTS-TBackbone=GPT-22025.12 | 76.98 | 81.72 | 79.94 | 87.99 | 81.66 | 70 | 56.73 | 50.45 | 59.06 | |
| IndividualBackbone=GPT-22025.12 | 76.8 | 81.99 | 80.39 | 88.27 | 81.86 | 89.64 | 65.34 | 91.17 | 82.05 | |
| DTS-T*Backbone=GPT-22025.12 | 76.72 | 81.84 | 79.94 | 88.15 | 81.66 | 70.11 | 56.49 | 51.45 | 59.35 | |
| DTS-DBackbone=GPT-22025.12 | 76.69 | 81.58 | 80.33 | 88.64 | 81.81 | 70.03 | 57.65 | 51.03 | 59.57 | |
| DTS-D*Backbone=GPT-22025.12 | 76.51 | 81.62 | 80.26 | 88.41 | 81.7 | 71.12 | 56.93 | 52.08 | 60.04 | |
| Twin-MergingBackbone=GPT-22025.12 | 76.27 | 80.03 | 79.65 | 87.53 | 80.87 | 69.73 | 55.29 | 51.14 | 58.72 | |
| T-SwitchBackbone=GPT-22025.12 | 76.27 | 81.87 | 79.9 | 88.57 | 81.65 | 63.16 | 52.7 | 50.91 | 55.59 | |
| SVDBackbone=GPT-22025.12 | 76.22 | 81.26 | 79.9 | 87.84 | 81.31 | 63.16 | 52.7 | 50.91 | 55.59 | |
| TALL-MaskBackbone=GPT-22025.12 | 74.78 | 78.59 | 78.43 | 88.15 | 79.98 | 67.12 | 43.68 | 50.71 | 53.84 | |
| EMR-MERGINGBackbone=GPT-22025.12 | 73.63 | 81.74 | 80.14 | 87.04 | 80.64 | 69.94 | 54.87 | 51.12 | 58.64 | |
| WEMOEBackbone=GPT-22025.12 | 72.3 | 80.24 | 77.59 | 83.06 | 78.3 | 69.16 | 56.93 | 51.02 | 58.37 | |
| Task-ArithmeticBackbone=GPT-22025.12 | 68.26 | 68.44 | 72.54 | 60.69 | 67.48 | 67.12 | 43.68 | 50.71 | 53.84 | |
| DAREBackbone=GPT-22025.12 | 63.37 | 65.51 | 69.11 | 57 | 63.75 | 68.07 | 42.96 | 51.56 | 54.2 | |
| Ties-MergingBackbone=GPT-22025.12 | 63.08 | 79.93 | 34.06 | 71.16 | 62.06 | 70.36 | 57.03 | 51.26 | 59.55 |