Robotic Failure Analysis VQA on ViFailback-Bench Hard
40AccuracyGPT-4o
Evaluation results
| Method | Links | |
|---|---|---|
| GPT-4oModel Category=General Closed-Source Models2025.12 | 40 | |
| Qwen2.5-VL-72B-InstructModel Category=General Open-Source Models2025.12 | 36.56 | |
| Qwen3-VL-32B-InstructModel Category=General Open-Source Models2025.12 | 35.23 | |
| Qwen3-VL-4B-InstructModel Category=General Open-Source Models2025.12 | 33.37 | |
| Qwen3-VL-8B-InstructModel Category=General Open-Source Models2025.12 | 33.04 | |
| Qwen2.5-VL-32B-InstructModel Category=General Open-Source Models2025.12 | 32.5 | |
| Gemini-2.5-ProModel Category=General Closed-Source Models2025.12 | 32.45 | |
| InternVL3-78BModel Category=General Open-Source Models2025.12 | 30.77 | |
| InternVL3-8BModel Category=General Open-Source Models2025.12 | 29.82 | |
| RoboBrain2.0-32BModel Category=Embodied Models2025.12 | 29.22 | |
| Cosmos-Reason1-7BModel Category=Embodied Models2025.12 | 28.6 | |
| Qwen2.5-VL-3B-InstructModel Category=General Open-Source Models2025.12 | 22.1 | |
| RoboBrain2.0-3BModel Category=Embodied Models2025.12 | 21.21 | |
| Qwen3-VL-2B-InstructModel Category=General Open-Source Models2025.12 | 20.28 | |
| Qwen2.5-VL-7B-InstructModel Category=General Open-Source Models2025.12 | 19.26 | |
| RoboBrain2.0-7BModel Category=Embodied Models2025.12 | 19.15 |