Visibility Reasoning on CrossPoint-Bench
96.15Object AccuracyHuman
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| Human2025.12 | 96.15 | 89.66 | 92.73 | |
| CroPond-7BParameters=7B2025.12 | 81.73 | 76.72 | 79.09 | |
| o32025.12 | 79.81 | 71.55 | 75.45 | |
| CroPond-3BParameters=3B2025.12 | 79.81 | 65.52 | 72.27 | |
| Gemini-2.5-Pro2025.12 | 68.27 | 66.38 | 67.27 | |
| Claude-Sonnet-42025.12 | 66.35 | 50.86 | 58.18 | |
| Qwen3-VL-30B-A3B-InstructParameters=30B, MoE=A3B, Mode=Instruct2025.12 | 66.35 | 52.59 | 59.09 | |
| Qwen3-VL-235B-A22B-ThinkingParameters=235B, MoE=A22B, Mode=Thinking2025.12 | 65.38 | 59.48 | 62.27 | |
| Qwen3-VL-235B-A22B-InstructParameters=235B, MoE=A22B, Mode=Instruct2025.12 | 59.62 | 56.9 | 58.18 | |
| Qwen2.5-VL-32B-InstructParameters=32B, Mode=Instruct2025.12 | 56.73 | 58.62 | 57.73 | |
| Qwen2.5-VL-3B-InstructParameters=3B, Mode=Instruct2025.12 | 50.96 | 43.97 | 47.27 | |
| Qwen2.5-VL-7B-InstructParameters=7B, Mode=Instruct2025.12 | 50 | 63.79 | 57.27 | |
| Qwen2.5-VL-72B-InstructParameters=72B, Mode=Instruct2025.12 | 49.04 | 58.62 | 54.09 | |
| Qwen3-VL-30B-A3B-ThinkingParameters=30B, MoE=A3B, Mode=Thinking2025.12 | 49.04 | 59.48 | 54.55 | |
| Molmo-7B-DParameters=7B2025.12 | 49.04 | 51.72 | 50.45 | |
| RoboBrain2.0-7BParameters=7B, reasoning_mode=no-thinking2025.12 | 46.15 | 55.17 | 50.91 | |
| RoboBrain2.0-7BParameters=7B, reasoning_mode=thinking2025.12 | 46.15 | 52.59 | 49.55 | |
| RoboRefer-8B-SFTParameters=8B, Mode=SFT2025.12 | 37.5 | 37.07 | 37.27 | |
| RoboPoint-v1-Vicuna-13BBackbone=Vicuna-13B2025.12 | 33.65 | 41.38 | 37.73 |