Intent Classification on ToDs benchmark GPT-2 backbone (test)
0.875AccuracyMULTI
Evaluation Results
| Method | Links | |
|---|---|---|
| MULTIMemory size (Mem)=Full data, Evaluation type=Modularized, Role=Upper bound2024.05 | 0.875 | |
| HESITMemory size (Mem)=t|M|, Evaluation type=Modularized, Perspective=Performance-perspective2024.05 | 0.8271 | |
| GSSMemory size (Mem)=t|M|, Evaluation type=Modularized, Perspective=Model-perspective2024.05 | 0.8136 | |
| UNIFORMMemory size (Mem)=t|M|, Evaluation type=Modularized, Perspective=Data-perspective2024.05 | 0.811 | |
| RANDOMMemory size (Mem)=t|M|, Evaluation type=Modularized, Perspective=Data-perspective2024.05 | 0.7992 | |
| ARPERMemory size (Mem)=t|M|, Evaluation type=Modularized, Perspective=Model-perspective2024.05 | 0.7963 | |
| A-GEMMemory size (Mem)=t|M|, Evaluation type=Modularized, Approach=Rehearsal-based2024.05 | 0.1057 | |
| EWCMemory size (Mem)=0, Evaluation type=Modularized, Approach=Regularization-based2024.05 | 0.0372 | |
| L2Memory size (Mem)=0, Evaluation type=Modularized, Approach=Regularization-based2024.05 | 0.0359 | |
| VANILLAMemory size (Mem)=0, Evaluation type=Modularized2024.05 | 0.0306 | |
| LAMOLMemory size (Mem)=0, Evaluation type=Modularized, Approach=Rehearsal-based2024.05 | 0.0272 |