Question Answering on HotPotQA (test) (Metrics: Acc, TC, TP)
56.34AccuracySearch-R1
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Search-R1inference mode=w/ search tools2025.10 | 56.34 | 3 | 14.09 | |
| MASH w/ OTCinference mode=w/ search tools2025.10 | 55.42 | 1.14 | 32.91 | |
| MASH w/ EXPinference mode=w/ search tools2025.10 | 53.78 | 1.07 | 32.09 | |
| MASH w/ OTC-STinference mode=w/ search tools2025.10 | 53.32 | 1.1 | 32.55 | |
| OTCinference mode=w/ search tools2025.10 | 44.76 | 0.81 | 28.64 | |
| R1inference mode=w/ search tools2025.10 | 26.53 | 0 | 26.53 |