Temporal Coherence Verification on Fluid Video 1.0 (test)
70.1AccuracyGPT-4o
Evaluation Results
| Method | Links | |
|---|---|---|
| GPT-4oCategory=Closed-source, #Param=—, Stride=2, Zero-shot=true2026.03 | 70.1 | |
| Gemini-2.5-FlashCategory=Closed-source, #Param=—, Stride=4, Zero-shot=true2026.03 | 70.06 | |
| GPT-4oCategory=Closed-source, #Param=—, Stride=4, Zero-shot=true2026.03 | 69.91 | |
| Gemini-2.5-FlashCategory=Closed-source, #Param=—, Stride=2, Zero-shot=true2026.03 | 67 | |
| Qwen2.5-VLCategory=Open-source, #Param=7B, Stride=2, Zero-shot=true2026.03 | 57.8 | |
| Qwen2.5-VLCategory=Open-source, #Param=7B, Stride=4, Zero-shot=true2026.03 | 56.63 | |
| InternVL3Category=Open-source, #Param=8B, Stride=4, Zero-shot=true2026.03 | 54.94 | |
| GLM-4.1VCategory=Open-source, #Param=9B, Stride=4, Zero-shot=true2026.03 | 54.01 | |
| Qwen2-VLCategory=Open-source, #Param=7B, Stride=2, Zero-shot=true2026.03 | 54 | |
| Qwen2-VLCategory=Open-source, #Param=7B, Stride=4, Zero-shot=true2026.03 | 53.24 | |
| mPLUG-Owl3Category=Open-source, #Param=7B, Stride=4, Zero-shot=true2026.03 | 53.2 | |
| InternVL3Category=Open-source, #Param=8B, Stride=2, Zero-shot=true2026.03 | 53 | |
| InternVL2.5Category=Open-source, #Param=8B, Stride=2, Zero-shot=true2026.03 | 52.95 | |
| InternVL2.5Category=Open-source, #Param=8B, Stride=4, Zero-shot=true2026.03 | 52.31 | |
| mPLUG-Owl3Category=Open-source, #Param=7B, Stride=2, Zero-shot=true2026.03 | 52.19 | |
| GLM-4.1VCategory=Open-source, #Param=9B, Stride=2, Zero-shot=true2026.03 | 51.4 |