Image Classification on Imagenette (AUROC, AUARC, ECE)
76.9AUROCCov Eigenscore
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| Cov EigenscoreEvaluation=Averaged across LVLM architectures2025.12 | 76.9 | 57.6 | 9.4 | |
| SGPUEvaluation=Averaged across LVLM architectures2025.12 | 74.9 | 77.4 | 19.9 | |
| Cos EigenscoreEvaluation=Averaged across LVLM architectures2025.12 | 62.5 | 60.4 | 33.6 | |
| PEEvaluation=Averaged across LVLM architectures2025.12 | 62.4 | 64.1 | 26 | |
| UMPIREEvaluation=Averaged across LVLM architectures2025.12 | 60.9 | 60 | 33.8 | |
| DSEEvaluation=Averaged across LVLM architectures2025.12 | 60.5 | 61.4 | 28.5 | |
| SEEvaluation=Averaged across LVLM architectures2025.12 | 60.4 | 61.7 | 30.9 | |
| KLE-MaternEvaluation=Averaged across LVLM architectures2025.12 | 59 | 61.6 | 30.1 | |
| KLE-HeatEvaluation=Averaged across LVLM architectures2025.12 | 58.9 | 61.6 | 37.2 |