Question Answering on HotPotQA (test) (Metrics: Acc, Prec)
20.7AccuracyAFH
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| AFHmode=Absolute2025.10 | 20.7 | 34.2 | |
| DPO2025.10 | 19.9 | 53.1 | |
| MASH w/ OTC-ST2025.10 | 17.3 | 59.9 | |
| MASH w/ OTC2025.10 | 15.6 | 66.5 | |
| MASH w/ EXP2025.10 | 15.2 | 64 | |
| Promptingshots=5-shot2025.10 | 14.7 | 31.5 | |
| AFHmode=Multisample2025.10 | 12.9 | 53.8 | |
| OTC2025.10 | 12.6 | 64.5 | |
| Ternary2025.10 | 0 | — |