Visual Reasoning on Vis. Corr.
80.8AccuracyGPT-4o
Evaluation results
| Method | Links | |
|---|---|---|
| GPT-4oVisual Sketchpad=Enabled2024.06 | 80.8 | |
| GPT-4oVisual Sketchpad=Disabled2024.06 | 73.3 | |
| GPT-4 TurboVisual Sketchpad=Enabled2024.06 | 52.3 | |
| GPT-4 TurboVisual Sketchpad=Disabled2024.06 | 48.8 | |
| Gemini-Pro2024.06 | 42.4 | |
| SOTA [42]2024.06 | 42.4 | |
| Claude 3 OPUS2024.06 | 36.6 | |
| GPT-4V-preview2024.06 | 33.7 | |
| LLaVA-NeXT-34B2024.06 | 30.8 | |
| LLaVA-1.5-13B2024.06 | 29.1 | |
| LLaVA-1.5-7B2024.06 | 25.6 |