Self-evaluation on ViLP
77AUROCVAUQ
Evaluation results
| Method | Links | |
|---|---|---|
| VAUQModel=LLaVA-1.5-7B2026.02 | 77 | |
| VAUQModel=LLaVA-1.5-13B2026.02 | 69.5 | |
| VL-UncertaintyBackbone=InternVL3.5-8B2026.02 | 67.4 | |
| VAUQBackbone=InternVL3.5-8B2026.02 | 65.2 | |
| Semantic EntropyBackbone=InternVL3.5-8B2026.02 | 64.2 | |
| VAUQBackbone=Qwen-2.5-VL-7B2026.02 | 64.1 | |
| Semantic EntropyModel=LLaVA-1.5-7B2026.02 | 63.7 | |
| Semantic EntropyModel=LLaVA-1.5-13B2026.02 | 63.5 | |
| EigenScoreModel=LLaVA-1.5-7B2026.02 | 63.2 | |
| Contextual LensBackbone=Qwen-2.5-VL-7B2026.02 | 61.3 | |
| EigenScoreModel=LLaVA-1.5-13B2026.02 | 61.1 | |
| Chain-of-EmbeddingsModel=LLaVA-1.5-13B2026.02 | 60.5 | |
| Contextual LensModel=LLaVA-1.5-13B2026.02 | 60.4 | |
| EigenScoreBackbone=InternVL3.5-8B2026.02 | 59.3 | |
| VL-UncertaintyModel=LLaVA-1.5-13B2026.02 | 58.6 | |
| VL-UncertaintyBackbone=Qwen-2.5-VL-7B2026.02 | 57.9 | |
| Contextual LensModel=LLaVA-1.5-7B2026.02 | 56.7 | |
| VerbalizedModel=LLaVA-1.5-7B2026.02 | 56.3 | |
| VL-UncertaintyModel=LLaVA-1.5-7B2026.02 | 55.6 | |
| VerbalizedBackbone=Qwen-2.5-VL-7B2026.02 | 55.3 | |
| PerplexityBackbone=InternVL3.5-8B2026.02 | 55.3 | |
| PerplexityBackbone=Qwen-2.5-VL-7B2026.02 | 55 | |
| PerplexityModel=LLaVA-1.5-7B2026.02 | 54.6 | |
| PerplexityModel=LLaVA-1.5-13B2026.02 | 54.2 | |
| EigenScoreBackbone=Qwen-2.5-VL-7B2026.02 | 53 | |
| VerbalizedModel=LLaVA-1.5-13B2026.02 | 52.4 | |
| Chain-of-EmbeddingsModel=LLaVA-1.5-7B2026.02 | 52 | |
| Semantic EntropyBackbone=Qwen-2.5-VL-7B2026.02 | 52 | |
| SVARBackbone=InternVL3.5-8B2026.02 | 51.6 | |
| Contextual LensBackbone=InternVL3.5-8B2026.02 | 51.3 | |
| SVARModel=LLaVA-1.5-7B2026.02 | 50.6 | |
| Chain-of-EmbeddingsBackbone=InternVL3.5-8B2026.02 | 50.1 | |
| SVARBackbone=Qwen-2.5-VL-7B2026.02 | 49.6 | |
| VerbalizedBackbone=InternVL3.5-8B2026.02 | 48.7 | |
| SVARModel=LLaVA-1.5-13B2026.02 | 48.3 | |
| Chain-of-EmbeddingsBackbone=Qwen-2.5-VL-7B2026.02 | 47.4 |