Geometry reasoning on Geometry3K
66.7Geometry ScoreGPT-4o + Sketchpad
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| GPT-4o + SketchpadVisual Input=true, Sketchpad=true2024.06 | 66.7 | — | |
| GPT-4oVisual Input=true, Sketchpad=false2024.06 | 62.5 | — | |
| Claude 3 OPUSVisual Input=false2024.06 | 56.3 | — | |
| GPT-4 Turbo + SketchpadVisual Input=true, Sketchpad=true2024.06 | 45.8 | — | |
| GPT-4 TurboVisual Input=true, Sketchpad=false2024.06 | 37.5 | — | |
| SpatialMath-SXIntegration=With SpatialMath-SX, Base Model=LLaVA-NeXT-34B, Samples Size=6012026.01 | 14.81 | 5.99 | |
| LLaVA-NeXT-34BIntegration=Default, Base Model=LLaVA-NeXT-34B, Samples Size=6012026.01 | 8.82 | — |