Hallucination Benchmark on POPE A
87.6Accuracy (POPE A)V-STAR
Evaluation Results
| Method | Links | |
|---|---|---|
| V-STARData Size=40k2026.04 | 87.6 | |
| ThinkLite-VLData Size=11k2026.04 | 87.1 | |
| VL-CogitoData Size=80k2026.04 | 84.1 | |
| Vision-R1Data Size=210k2026.04 | 84 | |
| Qwen2.5VL2026.04 | 83.1 | |
| VL-RethinkerData Size=39k2026.04 | 82.8 | |
| R1-OnevisionData Size=155k2026.04 | 82.5 | |
| OpenVLThinkerData Size=59.2k2026.04 | 79.1 |