Benchmark
InterGPS
math
spatial_reasoning
text
Interpretable Geometry Problem Solver (Inter-GPS) with Geometry3K dataset of 3,002 geometry problems with dense annotation in formal language using theorem knowledge and symbolic reasoning
语言EN
满分1
参评模型2
模型排名
| 名次 | 模型 | 机构 | 分数 | 来源 |
|---|---|---|---|---|
| 1 | Phi-4-multimodal-instruct | Microsoft | 48.6 | 来源 ↗ |
| 2 | Phi-3.5-vision-instruct | Microsoft | 36.3 | 来源 ↗ |