Mathematical Reasoning on Numina (accuracy)
48.1Accuracy (Numina)Prismatic Syn.
Evaluation Results
| Method | Links | |
|---|---|---|
| Prismatic Syn.Backbone=Qwen 3B, Training Distribution=MedMCQA, Method Category=Synthesis2026.05 | 48.1 | |
| ProximityBackbone=Qwen 3B, Training Distribution=MedMCQA, Method Category=AS2026.05 | 43.2 | |
| OriginalBackbone=Qwen 3B, Training Distribution=MedMCQA, Method Category=Synthesis2026.05 | 40 | |
| Answer VarianceBackbone=Qwen 3B, Training Distribution=MedMCQA, Method Category=AS2026.05 | 38.4 | |
| ConfidenceBackbone=Qwen 3B, Training Distribution=MedMCQA, Method Category=AS2026.05 | 38.1 | |
| DiversityBackbone=Qwen 3B, Training Distribution=MedMCQA, Method Category=AS2026.05 | 38.1 | |
| Gradient NormBackbone=Qwen 3B, Training Distribution=MedMCQA, Method Category=AS2026.05 | 36.4 | |
| DataEnvGymBackbone=Qwen 3B, Training Distribution=MedMCQA, Method Category=Synthesis2026.05 | 35.7 | |
| RandomBackbone=Qwen 3B, Training Distribution=MedMCQA, Method Category=Selection2026.05 | 34.3 | |
| FormatBackbone=Qwen 3B, Training Distribution=MedMCQA, Method Category=AS2026.05 | 34.2 | |
| FilteredBackbone=Qwen 3B, Training Distribution=MedMCQA, Method Category=Selection2026.05 | 29.5 | |
| Qwen 3BBackbone=Qwen 3B, Method Category=Untrained model2026.05 | 28.3 |