Visual Reasoning on V* cross-domain (test)
79.06AccuracyVIRC-7B
Evaluation Results
| Method | Links | |
|---|---|---|
| VIRC-7B#Params=7B2025.12 | 79.06 | |
| Hint-GRPO-Qwen2-VL-7B#Params=7B2025.12 | 75.92 | |
| Qwen2.5-VL-7B-Instruct#Params=7B2025.12 | 75.5 | |
| InternVL2.5-8B-MPO#Params=8B2025.12 | 72.25 | |
| VIRC-3B#Params=3B2025.12 | 72.25 | |
| LLaVA-OV-Qwen2-7b-ov#Params=7B2025.12 | 70.9 | |
| InternVL2.5-8B#Params=8B2025.12 | 68.59 | |
| Hint-GRPO-Qwen2.5-VL-3B#Params=3B2025.12 | 68.06 | |
| Qwen2.5-VL-3B-Instruct#Params=3B2025.12 | 68.06 | |
| DeepSeek-VL2#Params=4.5B2025.12 | 66.49 | |
| GPT-4o2025.12 | 65.2 | |
| MM-Eureka#Params=7B2025.12 | 59.16 | |
| R1-VL-7B#Params=7B2025.12 | 45.55 | |
| MINT-CoT-7B#Params=7B2025.12 | 43.46 | |
| Qwen3-VL-8B-Instructw/ tool#Params=8B2025.12 | 41.88 |