Mathematical Reasoning on Minerva (Avg@2)
57.7Avg@2DARL
Evaluation Results
| Method | Links | |
|---|---|---|
| DARLBase=Base, Verifier=None2026.01 | 57.7 | |
| RLPRBase=Base, Verifier=None2026.01 | 56.5 | |
| RLVRBase=Base, Verifier=Rule2026.01 | 54.9 | |
| TTRLBase=Base, Verifier=Rule2026.01 | 52.8 | |
| Oat-ZeroBase=Math, Verifier=Rule2026.01 | 52.1 | |
| General ReasonerBase=Base, Verifier=Model2026.01 | 51.7 | |
| SimpleRL-ZooBase=Math, Verifier=Rule2026.01 | 51 | |
| Qwen2.5-7B-InstBase=-, Verifier=None2026.01 | 49.4 | |
| SimpleRL-ZooBase=Base, Verifier=Rule2026.01 | 49.2 | |
| VeriFreeBase=Base, Verifier=None2026.01 | 49 | |
| PRIMEBase=Math, Verifier=Rule2026.01 | 45.5 | |
| RLPRBase=Inst, Verifier=None2026.01 | 39 | |
| DARLBase=Inst, Verifier=None2026.01 | 37.9 | |
| Qwen2.5-7BBase=-, Verifier=None2026.01 | 37.6 | |
| RLVRBase=Inst, Verifier=Rule2026.01 | 35.2 | |
| Llama3.1-8B-InstBase=-, Verifier=None2026.01 | 32.7 |