Mathematical Reasoning on MathVerse-Plus Vision Only 1.0 (test)
31AccuracySFT
Evaluation Results
| Method | Links | |
|---|---|---|
| SFTBase Model=Phi-42026.01 | 31 | |
| SpatialMathBase Model=Phi-42026.01 | 31 | |
| ICL (1-shot)Base Model=Phi-42026.01 | 24 | |
| CoT (1-shot)Base Model=Phi-42026.01 | 24 | |
| Zero-ShotBase Model=Phi-42026.01 | 20 | |
| SpatialMathBase Model=LLaVA-NeXT-34B2026.01 | 20 | |
| SFT + Data AugmentationBase Model=Phi-42026.01 | 19 | |
| SFT + Data AugmentationBase Model=LLaVA-NeXT-34B2026.01 | 14 | |
| Zero-ShotBase Model=LLaVA-NeXT-34B2026.01 | 13 | |
| SFTBase Model=LLaVA-NeXT-34B2026.01 | 11 | |
| ICL (1-shot)Base Model=LLaVA-NeXT-34B2026.01 | 9 | |
| CoT (1-shot)Base Model=LLaVA-NeXT-34B2026.01 | 9 |