Multi-label Text Classification on AAPD
75.61Micro-F1Sim-Dissim
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Sim-DissimBackbone=Llama-3.1-8B2024.10 | 75.61 | 63.42 | |
| MulSupConBackbone=Llama-3.1-8B2024.10 | 74.72 | 61.58 | |
| Sim-DissimBackbone=RoBERTa-based2024.10 | 74.54 | 62.31 | |
| ALLBackbone=Llama-3.1-8B2024.10 | 74.32 | 60.47 | |
| MulSupConBackbone=RoBERTa-based2024.10 | 73.64 | 60.52 | |
| ANYBackbone=Llama-3.1-8B2024.10 | 73.41 | 59.63 | |
| ALLBackbone=RoBERTa-based2024.10 | 73.23 | 59.41 | |
| ANYBackbone=RoBERTa-based2024.10 | 72.31 | 58.55 | |
| BERT+AMNumber of labeled texts (Nl)=20,0002026.03 | 70.7 | 53.8 | |
| S2TC-BDDNumber of labeled texts (Nl)=20,0002026.03 | 70.7 | 55.3 | |
| BERT+BCENumber of labeled texts (Nl)=20,0002026.03 | 67.8 | 46.5 | |
| S2TC-BDDNumber of labeled texts (Nl)=2,0002026.03 | 65.1 | 48.3 | |
| BERT+AMNumber of labeled texts (Nl)=2,0002026.03 | 63.7 | 46.6 | |
| CAPNumber of labeled texts (Nl)=20,0002026.03 | 62.7 | 49.6 | |
| CAPNumber of labeled texts (Nl)=2,0002026.03 | 62.5 | 46.9 | |
| BERT+BCENumber of labeled texts (Nl)=2,0002026.03 | 62.3 | 42 | |
| MetaExpertNumber of labeled texts (Nl)=20,0002026.03 | 61 | 45.7 | |
| MetaExpertNumber of labeled texts (Nl)=2,0002026.03 | 60.1 | 44 | |
| S2TC-BDDNumber of labeled texts (Nl)=2002026.03 | 55.4 | 35.6 | |
| BERT+AMNumber of labeled texts (Nl)=2002026.03 | 54.3 | 34.2 | |
| MetaExpertNumber of labeled texts (Nl)=2002026.03 | 52.5 | 32.8 | |
| CAPNumber of labeled texts (Nl)=2002026.03 | 49.7 | 31.5 | |
| BERT+BCENumber of labeled texts (Nl)=2002026.03 | 45.6 | 21.2 | |
| SDRLNumber of labeled texts (Nl)=20,0002026.03 | 44 | 12.5 | |
| SDRLNumber of labeled texts (Nl)=2,0002026.03 | 41.6 | 10.2 | |
| SDRLNumber of labeled texts (Nl)=2002026.03 | 36.1 | 7 |