Hallucination Evaluation on Hallucination Evaluation Suite (Object HalBench, AMBER, HallusionBench)
5.84Avg. Hallucination ScoreViPSy
Evaluation Results
| Method | Links | |
|---|---|---|
| ViPSyPolicy Model=mPLUG-Owl3-7B2026.06 | 5.84 | |
| VanillaPolicy Model=mPLUG-Owl3-7B2026.06 | 6.66 | |
| ViPSyPolicy Model=InternVL3.5-8B2026.06 | 6.78 | |
| ViPSyPolicy Model=Qwen2.5-VL-7B2026.06 | 8.16 | |
| VanillaPolicy Model=InternVL3.5-8B2026.06 | 8.66 | |
| ViPSyPolicy Model=Kimi-VL-A3B2026.06 | 8.7 | |
| VanillaPolicy Model=Qwen2.5-VL-7B2026.06 | 9.48 | |
| VanillaPolicy Model=Kimi-VL-A3B2026.06 | 11.52 |