Relative Position Reasoning on Scene 3 Laboratory storage 1.0 (overall)
74Accrelgemini
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| geminiModel=Google Gemini2026.01 | 74 | 80 | 70 | 71 | |
| SpatialMem2026.01 | 74 | 81 | 62 | 72 | |
| internvl (Entry 1)Model=InternVL 2.5, Deployment=Local2026.01 | 72 | 78 | 68 | 69 | |
| llavaModel=LLaVA-Video, Deployment=Local2026.01 | 70 | 76 | 66 | 67 | |
| internvl (Entry 2)Model=InternVL 2.5, Deployment=Local2026.01 | 66 | 72 | 62 | 63 | |
| openaiModel=OpenAI GPT-4o2026.01 | 64 | 70 | 60 | 61 | |
| togetherModel=Qwen2.5-VL, Deployment=Cloud2026.01 | 58 | 66 | 54 | 55 | |
| qwen25vlModel=Qwen2.5-VL, Deployment=Local2026.01 | 50 | 59 | 46 | 47 |