Mathematical Reasoning on FlowVerse
70.4AccuracyMathFlow
Evaluation Results
| Method | Links | |
|---|---|---|
| MathFlowBase LLM=Gemini 2.5-pro, Perception Model=MathFlow-P-7B2025.03 | 70.4 | |
| Gemini 2.5-pro2025.03 | 68.9 | |
| MathFlowBase LLM=Claude-sonnet-3.5, Perception Model=MathFlow-P-7B2025.03 | 68.1 | |
| Claude-sonnet-3.52025.03 | 64 | |
| MathFlowBase LLM=InternVL-2.5-78B, Perception Model=MathFlow-P-7B2025.03 | 60.1 | |
| MathFlowBase LLM=GPT-4V, Perception Model=MathFlow-P-7B2025.03 | 59.3 | |
| MathFlowBase LLM=Qwen2-VL-72B, Perception Model=MathFlow-P-7B2025.03 | 58.3 | |
| GPT-4V2025.03 | 56.2 | |
| InternVL-2.5-78B2025.03 | 54.7 | |
| Qwen2-VL-72B2025.03 | 52.3 |