Image Generation Assessment on GenAI-Bench Image (test)
73.4AccuracyUnifiedReward-Flex
Evaluation Results
| Method | Links | |
|---|---|---|
| UnifiedReward-Flex2026.02 | 73.4 | |
| UnifiedReward-Think2026.02 | 72.3 | |
| UnifiedReward-FlexDPO=False, Protocol=Both correct2026.02 | 72 | |
| UnifiedReward2026.02 | 71.5 | |
| UnifiedReward-FlexDPO=False2026.02 | 71.5 | |
| HPSv32026.02 | 70.9 | |
| PickScore2026.02 | 70 | |
| HPSv22026.02 | 68.8 |