Human agreement correlation on VideoPhy-2 (test)
0.484Spearman Correlation (ρ)Gemini-3-Flash-Thinking
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Gemini-3-Flash-ThinkingTaxonomy=Refine, Thinking mode=true2026.06 | 0.484 | 0.138 | |
| Gemini-3-FlashTaxonomy=Refine, Thinking mode=false2026.06 | 0.429 | 0.136 | |
| Seed-2.0-Lite-ThinkingTaxonomy=Refine, Thinking mode=true2026.06 | 0.422 | 0.112 | |
| MiMo-v2.5-ThinkingTaxonomy=Refine, Thinking mode=true2026.06 | 0.406 | 0.159 | |
| Gemini-3-FlashTaxonomy=Seed, Thinking mode=false2026.06 | 0.4 | 0.107 | |
| Seed-2.0-Lite-ThinkingTaxonomy=Seed, Thinking mode=true2026.06 | 0.4 | 0.091 | |
| Qwen2.5-VL-7BTaxonomy=Refine, Thinking mode=false2026.06 | 0.383 | 0.066 | |
| Qwen3-VL-8B-InstructTaxonomy=Refine, Thinking mode=false2026.06 | 0.375 | 0.068 | |
| Gemini-3-Flash-ThinkingTaxonomy=Seed, Thinking mode=true2026.06 | 0.372 | 0.026 | |
| Gemma-4-31BTaxonomy=Refine, Thinking mode=false2026.06 | 0.371 | 0.086 | |
| MiMo-v2.5-ThinkingTaxonomy=Seed, Thinking mode=true2026.06 | 0.356 | 0.109 | |
| Gemma-4-31BTaxonomy=Seed, Thinking mode=false2026.06 | 0.355 | 0.07 | |
| Gemini-3-Flash-ThinkingTaxonomy=PhyGen, Thinking mode=true2026.06 | 0.346 | — | |
| MiMo-v2-OmniTaxonomy=Seed, Thinking mode=false2026.06 | 0.341 | 0.046 | |
| MiMo-v2.5Taxonomy=Refine, Thinking mode=false2026.06 | 0.33 | 0.012 | |
| Qwen3-VL-8B-InstructTaxonomy=Seed, Thinking mode=false2026.06 | 0.328 | 0.021 | |
| MiMo-v2-OmniTaxonomy=Refine, Thinking mode=false2026.06 | 0.319 | 0.025 | |
| MiMo-v2.5Taxonomy=PhyGen, Thinking mode=false2026.06 | 0.317 | — | |
| Qwen2.5-VL-7BTaxonomy=PhyGen, Thinking mode=false2026.06 | 0.317 | — | |
| Seed-2.0-Lite-ThinkingTaxonomy=PhyGen, Thinking mode=true2026.06 | 0.31 | — | |
| Seed-2.0-LiteTaxonomy=Refine, Thinking mode=false2026.06 | 0.308 | 0.047 | |
| Qwen3-VL-8B-InstructTaxonomy=PhyGen, Thinking mode=false2026.06 | 0.307 | — | |
| MiMo-v2.5Taxonomy=Seed, Thinking mode=false2026.06 | 0.305 | -0.012 | |
| MiMo-v2-OmniTaxonomy=PhyGen, Thinking mode=false2026.06 | 0.294 | — | |
| Gemini-3-FlashTaxonomy=PhyGen, Thinking mode=false2026.06 | 0.293 | — | |
| MiMo-v2-Omni-ThinkingTaxonomy=Refine, Thinking mode=true2026.06 | 0.291 | 0.057 | |
| Seed-2.0-LiteTaxonomy=Seed, Thinking mode=false2026.06 | 0.286 | 0.025 | |
| MiMo-v2-Omni-ThinkingTaxonomy=Seed, Thinking mode=true2026.06 | 0.285 | 0.051 | |
| Gemma-4-31BTaxonomy=PhyGen, Thinking mode=false2026.06 | 0.285 | — | |
| Qwen2.5-VL-7BTaxonomy=Seed, Thinking mode=false2026.06 | 0.276 | -0.041 | |
| Nova-2-LiteTaxonomy=Refine, Thinking mode=false2026.06 | 0.262 | 0.047 | |
| Seed-2.0-LiteTaxonomy=PhyGen, Thinking mode=false2026.06 | 0.261 | — | |
| MiMo-v2.5-ThinkingTaxonomy=PhyGen, Thinking mode=true2026.06 | 0.247 | — | |
| MiMo-v2-Omni-ThinkingTaxonomy=PhyGen, Thinking mode=true2026.06 | 0.234 | — | |
| Nova-2-LiteTaxonomy=PhyGen, Thinking mode=false2026.06 | 0.215 | — | |
| GLM-5V-TurboTaxonomy=Refine, Thinking mode=false2026.06 | 0.207 | 0.026 | |
| GLM-5V-TurboTaxonomy=Seed, Thinking mode=false2026.06 | 0.194 | 0.013 | |
| GLM-5V-TurboTaxonomy=PhyGen, Thinking mode=false2026.06 | 0.181 | — | |
| Nova-2-LiteTaxonomy=Seed, Thinking mode=false2026.06 | 0.174 | -0.041 | |
| GLM-4.6V-ThinkingTaxonomy=Refine, Thinking mode=true2026.06 | 0.163 | 0.126 | |
| Reka-EdgeTaxonomy=Refine, Thinking mode=false2026.06 | 0.151 | 0.081 | |
| GLM-4.6VTaxonomy=Refine, Thinking mode=false2026.06 | 0.138 | 0.022 | |
| GLM-4.6V-ThinkingTaxonomy=Seed, Thinking mode=true2026.06 | 0.136 | 0.1 | |
| GLM-4.6VTaxonomy=Seed, Thinking mode=false2026.06 | 0.132 | 0.016 | |
| GLM-4.6VTaxonomy=PhyGen, Thinking mode=false2026.06 | 0.116 | — | |
| Reka-EdgeTaxonomy=PhyGen, Thinking mode=false2026.06 | 0.07 | — | |
| Reka-EdgeTaxonomy=Seed, Thinking mode=false2026.06 | 0.043 | -0.028 | |
| GLM-4.6V-ThinkingTaxonomy=PhyGen, Thinking mode=true2026.06 | 0.037 | — |