STEM Reasoning on TheoremQA (Avg@2)
55.4Avg@2RLPR
Evaluation Results
| Method | Links | |
|---|---|---|
| RLPRBase=Base, Verifier=None2026.01 | 55.4 | |
| DARLBase=Base, Verifier=None2026.01 | 55.2 | |
| Oat-ZeroBase=Math, Verifier=Rule2026.01 | 53.3 | |
| RLVRBase=Base, Verifier=Rule2026.01 | 52.2 | |
| General ReasonerBase=Base, Verifier=Model2026.01 | 52.1 | |
| SimpleRL-ZooBase=Math, Verifier=Rule2026.01 | 51.1 | |
| SimpleRL-ZooBase=Base, Verifier=Rule2026.01 | 49.5 | |
| TTRLBase=Base, Verifier=Rule2026.01 | 48.8 | |
| PRIMEBase=Math, Verifier=Rule2026.01 | 47.7 | |
| VeriFreeBase=Base, Verifier=None2026.01 | 47.6 | |
| Qwen2.5-7B-InstBase=-, Verifier=None2026.01 | 47.3 | |
| Qwen2.5-7BBase=-, Verifier=None2026.01 | 41.4 | |
| DARLBase=Inst, Verifier=None2026.01 | 39.4 | |
| RLPRBase=Inst, Verifier=None2026.01 | 36.5 | |
| RLVRBase=Inst, Verifier=Rule2026.01 | 32 | |
| Llama3.1-8B-InstBase=-, Verifier=None2026.01 | 31.3 |