Perspective-shift reasoning on CAMCUE-DATA
80.12Overall Accuracy (Avg)CamCue
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| CamCueBackbone=Qwen2.5-VL-7B2026.02 | 80.12 | 92 | 88.24 | 83.52 | 78.52 | 60 | |
| CamCueBackbone=InternVL-2.5-8B2026.02 | 77.36 | 94 | 80.39 | 89.01 | 72.38 | 54.12 | |
| CamCueBackbone=Qwen2.5-VL-3B2026.02 | 75.92 | 94.12 | 82.35 | 80.43 | 67.97 | 63.53 | |
| MindJourneyBackbone=InternVL-2.5-8B2026.02 | 74.21 | 93 | 82.35 | 85.71 | 64.64 | 55.29 | |
| MindJourneyBackbone=Qwen2.5-VL-7B2026.02 | 72.83 | 92 | 84.31 | 80.22 | 65.75 | 50.59 | |
| Qwen2.5-VL-7BBackbone=Qwen2.5-VL-7B2026.02 | 71.06 | 93 | 84.31 | 71.43 | 59.73 | 55.29 | |
| MindJourneyBackbone=Qwen2.5-VL-3B2026.02 | 70.28 | 95 | 76.47 | 69.23 | 64.64 | 50.59 | |
| InternVL-2.5-8BBackbone=InternVL-2.5-8B2026.02 | 68.11 | 89 | 76.47 | 80.22 | 58.56 | 52.94 | |
| Qwen2.5-VL-3BBackbone=Qwen2.5-VL-3B2026.02 | 67.52 | 97 | 80.39 | 62.64 | 57.05 | 50.59 |