Mathematical Problem Solving on MATH (Vanilla/Attack)
57.41Vanilla AccuracyAutoRAS
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| AutoRASEvaluation Protocol=Vanilla, Cost ($)=3.512026.06 | 57.41 | — | |
| OWLEvaluation Protocol=Vanilla, Cost ($)=3.232026.06 | 50.41 | — | |
| CAMELEvaluation Protocol=Vanilla, Cost ($)=2.042026.06 | 46.7 | — | |
| Magentic-OneEvaluation Protocol=Vanilla, Cost ($)=1.212026.06 | 45.88 | — | |
| AutoRASEvaluation Protocol=Attack, Cost ($)=3.482026.06 | — | 54.94 | |
| CAMELEvaluation Protocol=Attack, Cost ($)=2.452026.06 | — | 41.98 | |
| Magentic-OneEvaluation Protocol=Attack, Cost ($)=1.622026.06 | — | 23.25 | |
| OWLEvaluation Protocol=Attack, Cost ($)=3.312026.06 | — | 45.68 |