Hallucination Evaluation on CHAIR (test)
50.9CS ScoreGreedy
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| GreedyBase Model=InstructBLIP2026.05 | 50.9 | 13.2 | |
| GreedyBase Model=LLaVA-1.52026.05 | 48.3 | 14.1 | |
| VCDBase Model=Shikra2026.05 | 47.1 | 13.6 | |
| GreedyBase Model=Shikra2026.05 | 46.8 | 13.8 | |
| VCDBase Model=LLaVA-1.52026.05 | 45.5 | 12.8 | |
| OPERABase Model=InstructBLIP2026.05 | 45.5 | 12.3 | |
| OPERABase Model=Shikra2026.05 | 45.1 | 12.7 | |
| SIDBase Model=LLaVA-1.52026.05 | 44.5 | 12.7 | |
| OPERABase Model=LLaVA-1.52026.05 | 44.1 | 12.2 | |
| SIDBase Model=Shikra2026.05 | 43.9 | 12.5 | |
| SIDBase Model=InstructBLIP2026.05 | 43.5 | 12.1 | |
| VCDBase Model=InstructBLIP2026.05 | 43.2 | 13.9 | |
| ACEBase Model=InstructBLIP2026.05 | 41.7 | 10.8 | |
| ACEBase Model=Shikra2026.05 | 41.1 | 11.4 | |
| GreedyBase Model=LLaVA-NeXT2026.05 | 40.7 | 12.1 | |
| ACEBase Model=LLaVA-1.52026.05 | 40.4 | 9.6 | |
| VCDBase Model=LLaVA-NeXT2026.05 | 38.3 | 11.4 | |
| SIDBase Model=LLaVA-NeXT2026.05 | 37.9 | 11.5 | |
| OPERABase Model=LLaVA-NeXT2026.05 | 37.6 | 10.2 | |
| ACEBase Model=LLaVA-NeXT2026.05 | 35.4 | 10.6 |