| Benchmark | Model | Rank | Results |
|---|---|---|---|
| multimodal-reasoning-on-rebus | GPT-4V | #1 | Accuracy: 24.0 |
| multimodal-reasoning-on-rebus | Gemini Pro | #2 | Accuracy: 13.2 |
| multimodal-reasoning-on-rebus | LLaVa-1.5-13B | #3 | Accuracy: 1.8 |
| multimodal-reasoning-on-rebus | LLaVa-1.5-7B | #4 | Accuracy: 1.5 |
| multimodal-reasoning-on-rebus | BLIP2-FLAN-T5-XXL | #5 | Accuracy: 0.9 |
| multimodal-reasoning-on-rebus | CogVLM | #6 | Accuracy: 0.9 |
| multimodal-reasoning-on-rebus | QWEN | #7 | Accuracy: 0.9 |
| multimodal-reasoning-on-rebus | InstructBLIP | #8 | Accuracy: 0.6 |