Multi-classification on Yahoo
65.02AccuracyClean
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| CleanBackbone=BERT, Triggers=152023.05 | 65.02 | 9.95 | 6.17 | 0 | |
| POR-1Backbone=BERT, Triggers=152023.05 | 64.82 | 99.63 | 20 | 0.2 | |
| POR-2Backbone=BERT, Triggers=152023.05 | 64.7 | 97.18 | 36.7 | 0.2 | |
| NeuBABackbone=BERT, Triggers=152023.05 | 64.52 | 37.32 | 30.25 | 0 | |
| UORBackbone=BERT, Triggers=15, Gradient Search=false2023.05 | 63.58 | 95.07 | 74.36 | 0.6 | |
| UOR-GBackbone=BERT, Triggers=15, Gradient Search=true2023.05 | 63.54 | 92.25 | 77.77 | 0.8 |