Multiple Instance Learning classification on Aggregate 12 MIL Benchmarks
84.17Average AUROCICMIL
Evaluation results
| Method | Links | ||
|---|---|---|---|
| ICMIL2026.06 | 84.17 | 4.08 | |
| Mixed2026.06 | 83.38 | 4.67 | |
| Fact. (cont, MLP)2026.06 | 82.76 | 5.5 | |
| Joint2026.06 | 82.56 | 4.92 | |
| MeanLogReg2026.06 | 82.37 | 6.33 | |
| Fact. (cont, tree)2026.06 | 81.55 | 7.5 | |
| TabPFN-Cluster2026.06 | 81.21 | 5.83 | |
| Fact. (disc, lookup)2026.06 | 80.98 | 8 | |
| ABMIL2026.06 | 79.97 | 7.75 | |
| SVM-Summ2026.06 | 79.38 | 6.17 | |
| Fact. (disc, MLP)2026.06 | 75.12 | 11.33 | |
| TabPFN-Subsample2026.06 | 74.82 | 9.17 | |
| TabPFN-Concat2026.06 | 74.16 | 9.83 | |
| Fact. (disc, tree)2026.06 | 51.27 | 13.92 |