Answer Verification on Omni-MATH
0.8591AUROCMSV 16
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| MSV 16Sequences=16, Evaluation Protocol=Streaming2026.03 | 0.8591 | 0.11 | |
| MSV 64Sequences=64, Evaluation Protocol=Streaming2026.03 | 0.8352 | 0.1255 | |
| MSV 4Sequences=4, Evaluation Protocol=Streaming2026.03 | 0.8327 | 0.116 | |
| ProbeEvaluation Protocol=Streaming2026.03 | 0.8229 | 0.1527 | |
| Probe + WV 4Sequences=4, Evaluation Protocol=Weighted Voting2026.03 | 0.8167 | 0.1246 | |
| Probe + WV 16Sequences=16, Evaluation Protocol=Weighted Voting2026.03 | 0.8167 | 0.1246 | |
| Probe + WV 64Sequences=64, Evaluation Protocol=Weighted Voting2026.03 | 0.8167 | 0.1246 | |
| MSV 1Sequences=1, Evaluation Protocol=Streaming2026.03 | 0.7951 | 0.1381 | |
| MSV 1 + WV 4Sequences=4, Evaluation Protocol=MSV 1 + Weighted Voting2026.03 | 0.7813 | 0.1293 | |
| MSV 1 + WV 16Sequences=16, Evaluation Protocol=MSV 1 + Weighted Voting2026.03 | 0.7813 | 0.1293 | |
| MSV 1 + WV 64Sequences=64, Evaluation Protocol=MSV 1 + Weighted Voting2026.03 | 0.7813 | 0.1293 |