Debate on MATH
36.9AccuracyOPTIMA-iDPO SC
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| OPTIMA-iDPO SCoptimization_strategy=iDPO, self_consistency=true2024.10 | 36.9 | 2,743.1 | |
| SCn=82024.10 | 35.7 | 2,600.9 | |
| OPTIMA-iSFT-DPO SCoptimization_strategy=iSFT-DPO, self_consistency=true2024.10 | 34.8 | 2,788.5 | |
| OPTIMA-iSFT SCoptimization_strategy=iSFT, self_consistency=true2024.10 | 32.4 | 2,432.9 | |
| OPTIMA-iDPOoptimization_strategy=iDPO2024.10 | 30.4 | 272.8 | |
| OPTIMA-iSFToptimization_strategy=iSFT2024.10 | 30.1 | 830.3 | |
| MAD2024.10 | 29.8 | 1,517.6 | |
| OPTIMA-iSFT-DPOoptimization_strategy=iSFT-DPO2024.10 | 29.3 | 488.1 | |
| AutoForm2024.10 | 26.1 | 644.3 | |
| CoT2024.10 | 23.9 | 329.8 |