Math and Knowledge Reasoning on AIME 2024 (Avg)
86.3Average ScoreRL initialization model
Evaluation Results
| Method | Links | |
|---|---|---|
| RL initialization modelSampling protocol=@322026.05 | 86.3 | |
| RDPOSampling protocol=@322026.05 | 85.73 | |
| GRPOSampling protocol=@322026.05 | 85.56 |