Hallucination Detection on MMBench
60.8AUROCSCoOP
Evaluation Results
| Method | Links | |
|---|---|---|
| SCoOP#VLMs=4-VLM2026.03 | 60.8 | |
| SCoOP#VLMs=3-VLM2026.03 | 58.7 | |
| SCoOP#VLMs=2-VLM2026.03 | 58.6 | |
| Naive Selection#VLMs=2-VLM2026.03 | 55.5 | |
| Majority Voting#VLMs=2-VLM2026.03 | 55.3 | |
| Majority Voting#VLMs=3-VLM2026.03 | 54.9 | |
| Naive Selection#VLMs=3-VLM2026.03 | 54.5 | |
| Naive Selection#VLMs=4-VLM2026.03 | 53.3 | |
| Majority Voting#VLMs=4-VLM2026.03 | 53.2 |