TA on Curated Prompt Set (OOD)
0.6052RM AccuracyVideoPhy
Evaluation Results
| Method | Links | |
|---|---|---|
| VideoPhy2025.12 | 0.6052 | |
| SoliRewardBackbone=InternVL32025.12 | 0.6025 | |
| VideoScore2025.12 | 0.5405 | |
| VideoScore-v1.12025.12 | 0.5336 | |
| VideoAlign2025.12 | 0.4914 | |
| VisionRewardDegenerate Score Distribution=true2025.12 | 0.2224 | |
| UnifiedRewardDegenerate Score Distribution=true2025.12 | 0.1063 | |
| LiFTDegenerate Score Distribution=true2025.12 | 0.074 |