Action-Conditioned Person Grounded Reasoning on Know-Show (test)
84T/Acc.Human
Evaluation Results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| HumanSubset=Know-Show subset2025.12 | 84 | 92 | 82 | 86 | 76 | — | — | |
| HumanSubset=Know-Show subset2025.12 | 81.3 | 94.7 | 74.7 | 94.7 | 76 | 88 | 76 | |
| Gemini-2.5-ProSubset=Know-Show subset2025.12 | 72 | 4 | 4 | 0 | 0 | — | — | |
| Ovis2.5-9B2025.12 | 51.6 | 34.8 | 4.5 | 12.8 | 0.6 | 9.4 | 1.6 | |
| MiniCPM-V 4.52025.12 | 50.4 | 17.1 | 0.4 | 2.6 | 0 | 4 | 0.9 | |
| Ovis2.5-9B2025.12 | 48.7 | 34.3 | 16.1 | 13.3 | 4.9 | — | — | |
| Gemini-2.5-FlashSubset=Know-Show subset2025.12 | 48 | 0 | 0 | 0 | 0 | 0 | 0 | |
| GRAM2025.12 | 46.7 | 66.7 | 12.7 | 27.4 | 2 | 31.8 | 10.3 | |
| MiniCPM-V 4.52025.12 | 45.4 | 8.2 | 2.7 | 1.6 | 0.6 | — | — | |
| GPT-4oSubset=Know-Show subset2025.12 | 44 | 16 | 0 | 12 | 0 | 0 | 0 | |
| GPT-4oSubset=Know-Show subset2025.12 | 40 | 32 | 8 | 8 | 0 | — | — | |
| Gemini-2.5-FlashSubset=Know-Show subset2025.12 | 40 | 0 | 0 | 0 | 0 | — | — | |
| Qwen2.5 VL 7B2025.12 | 26.8 | 42.4 | 2 | 10.1 | 0.4 | 9.1 | 2.2 | |
| GRAM2025.12 | 25.6 | 81.4 | 20.9 | 46 | 12.3 | — | — | |
| LLaVA-OneVision2025.12 | 25.1 | 37.8 | 9 | 6.3 | 1 | — | — | |
| LLaVA-OneVision2025.12 | 24 | 35.2 | 5.2 | 9.7 | 0.4 | 48 | 20.5 | |
| Qwen2.5 VL 7B2025.12 | 19.8 | 29.4 | 6.3 | 8 | 2 | — | — | |
| Gemini-2.5-ProSubset=Know-Show subset2025.12 | 16 | 0 | 0 | 0 | 0 | 0 | 0 | |
| VideoLLaVA2025.12 | 0.9 | 0 | 0 | 0 | 0 | 0 | 0 | |
| VideoLLaVA2025.12 | 0.6 | 62.8 | 0.2 | 24.9 | 0 | — | — |