Video Generation Assessment on GenAI-Bench Video (test)
82.5AccuracyUnifiedReward-Flex
Evaluation Results
| Method | Links | |
|---|---|---|
| UnifiedReward-Flex2026.02 | 82.5 | |
| UnifiedReward-FlexDPO=False, Protocol=Both correct2026.02 | 80.6 | |
| UnifiedReward-Think2026.02 | 80.3 | |
| UnifiedReward-FlexDPO=False2026.02 | 79.4 | |
| UnifiedReward2026.02 | 76.8 | |
| VideoReward2026.02 | 73.1 | |
| VideoScore2026.02 | 70.6 | |
| LiFT2026.02 | 60.1 |