Mathematical Reasoning on MATH500 (Accuracy and Avg Gain)
82.63AccuracyMIND
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| MINDBackbone=Qwen2.5-7B, Fusion=with2026.01 | 82.63 | — | |
| MCCBackbone=Qwen2.5-7B, Fusion=none2026.01 | 82.2 | 2.71 | |
| MINDBackbone=Qwen2.5-7B, Fusion=without2026.01 | 80.6 | 2.31 | |
| EDITBackbone=Qwen2.5-7B, Fusion=none2026.01 | 79.5 | 1.53 | |
| MoDEBackbone=Qwen2.5-7B, Fusion=none2026.01 | 77.67 | 2.23 | |
| SbSBackbone=Qwen2.5-7B, Fusion=none2026.01 | 77.4 | 2.23 | |
| BaseBackbone=Qwen2.5-7B-Instruct, Fusion=none2026.01 | 77.2 | 3.99 | |
| MINDBackbone=Qwen2.5-1.5B, Fusion=with2026.01 | 54.09 | — | |
| MINDBackbone=Llama3.1-8B, Fusion=with2026.01 | 53.73 | — | |
| MCCBackbone=Qwen2.5-1.5B, Fusion=none2026.01 | 53.6 | 3.16 | |
| MCCBackbone=Llama3.1-8B, Fusion=none2026.01 | 53 | 2.63 | |
| EDITBackbone=Llama3.1-8B, Fusion=none2026.01 | 52.8 | 0.62 | |
| SbSBackbone=Llama3.1-8B, Fusion=none2026.01 | 52.6 | 0.67 | |
| MoDEBackbone=Qwen2.5-1.5B, Fusion=none2026.01 | 52.6 | 3.72 | |
| MINDBackbone=Qwen2.5-1.5B, Fusion=without2026.01 | 52 | 4.41 | |
| MINDBackbone=Llama3.1-8B, Fusion=without2026.01 | 51.6 | 3.78 | |
| EDITBackbone=Qwen2.5-1.5B, Fusion=none2026.01 | 51.2 | 2.71 | |
| MoDEBackbone=Llama3.1-8B, Fusion=none2026.01 | 50.41 | 4.28 | |
| BaseBackbone=Qwen2.5-1.5B-Instruct, Fusion=none2026.01 | 50.4 | 3.63 | |
| SbSBackbone=Qwen2.5-1.5B, Fusion=none2026.01 | 49.8 | 3.73 | |
| BaseBackbone=Llama3.1-8B-Instruct, Fusion=none2026.01 | 46 | 4.11 |