Few-shot text classification on BinaryClfs
0.698AUC-ROCICT
Evaluation Results
| Method | Links | |
|---|---|---|
| ICTBackbone=GPT2-Large, K-shot=5-shot, Instruction presence=With instructions2021.10 | 0.698 | |
| InsT + FTBackbone=GPT2-Large, K-shot=5-shot, Instruction presence=With instructions2021.10 | 0.694 | |
| ICTBackbone=GPT2-Medium, K-shot=5-shot, Instruction presence=With instructions2021.10 | 0.674 | |
| InsT + FTBackbone=GPT2-Medium, K-shot=5-shot, Instruction presence=With instructions2021.10 | 0.67 | |
| ICTBackbone=GPT2-Large, K-shot=0-shot, Instruction presence=With instructions2021.10 | 0.663 | |
| ICTBackbone=GPT2-Medium, K-shot=0-shot, Instruction presence=With instructions2021.10 | 0.629 | |
| Raw IC-LBackbone=GPT2-Large, K-shot=5-shot, Instruction presence=With instructions2021.10 | 0.583 | |
| Raw IC-LBackbone=GPT2-Medium, K-shot=5-shot, Instruction presence=With instructions2021.10 | 0.578 | |
| Raw IC-LBackbone=GPT2-Large, K-shot=0-shot, Instruction presence=With instructions2021.10 | 0.51 | |
| Raw IC-LBackbone=GPT2-Medium, K-shot=0-shot, Instruction presence=With instructions2021.10 | 0.505 |