Abstention Classification on HotPotQA (test)
1Abs(0)Ternary
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Ternary2025.10 | 1 | 0 | |
| OTC2025.10 | 0.953 | 0.414 | |
| MASH w/ OTC2025.10 | 0.948 | 0.523 | |
| MASH w/ EXP2025.10 | 0.945 | 0.527 | |
| MASH w/ OTC-ST2025.10 | 0.912 | 0.603 | |
| AFHmode=Multisample2025.10 | 0.892 | 0.576 | |
| DPO2025.10 | 0.859 | 0.735 | |
| Promptingshots=5-shot2025.10 | 0.605 | 0.269 | |
| AFHmode=Absolute2025.10 | 0.504 | 0.354 |