String-level response similarity on RA-QA Global, Discriminative tasks
0.9BERTScoreRAMoEA-QA
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| RAMoEA-QA2026.03 | 0.9 | 88.38 | |
| CareAQA-operaGTBackbone=OPERA-GT2026.03 | 0.89 | 87.05 | |
| CareAQA-operaCTBackbone=OPERA-CT2026.03 | 0.87 | 84.89 | |
| CareAQA-operaGTBackbone=OPERA-GT2026.03 | 0.87 | 83.22 | |
| CareAQA-operaCTBackbone=OPERA-CT2026.03 | 0.86 | 83.15 | |
| RAMoEA-QA2026.03 | 0.83 | 81.02 | |
| PENGI2026.03 | -0.01 | 2.61 | |
| PENGI2026.03 | -0.09 | 0 |