Causal Spatial Reasoning on CausalSpatial
84.49Overall Average ScoreHuman
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Human2026.01 | 84.49 | — | |
| GPT52026.01 | 54.17 | 1 | |
| Gemini2.5 Pro2026.01 | 50.55 | 2 | |
| Qwen3-VL-30B-A3B-T.2026.01 | 50.5 | 3 | |
| Gemini2.5 Flash2026.01 | 48.82 | 4 | |
| Sonnet 3.72026.01 | 48.07 | 5 | |
| Qwen3-VL-30B-A3B-I.2026.01 | 45.8 | 6 | |
| GPT5-mini2026.01 | 45.44 | 7 | |
| Qwen3-VL-4B-I.2026.01 | 44.76 | 8 | |
| Qwen3-VL-8B-I.2026.01 | 43.53 | 9 | |
| Qwen3-VL-8B-T.2026.01 | 42.65 | 10 | |
| Qwen3-VL-2B-I.2026.01 | 40.6 | 11 | |
| Spatial-MLLM2026.01 | 40 | 12 | |
| Qwen2.5-VL-7B-I.2026.01 | 39.61 | 13 | |
| LLaVAOneVision1.5-8B-I.2026.01 | 34.62 | 14 | |
| Spatial-VLM2026.01 | 33.5 | 15 | |
| Random2026.01 | 28.6 | — |