Mutual Information Estimation on BMI benchmark
1.59MI Estimate Score (Mn-dense 5-5-0.5)KDE
Evaluation results
| Method | Links | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| KDEN=5000 samples per task, Average over=10 random seeds2026.05 | 1.59 | 2.87 | 2.43 | 2.36 | 1.17 | 2.23 | 2.94 | 1.23 | 2.04 | |
| KNIFEN=5000 samples per task, Average over=10 random seeds2026.05 | 0.93 | 0.1 | 0.66 | 0.5 | 0.07 | 0.92 | 0.05 | 0.65 | 48.4 | |
| MINEN=5000 samples per task, Average over=10 random seeds2026.05 | 0.6 | 1 | 0.53 | 0.21 | 1.03 | 1.06 | 1.63 | 0.39 | 25.9 | |
| MINE-5sN=5000 samples per task, Average over=10 random seeds2026.05 | 0.6 | 0.9 | 0.33 | 0.15 | 0.93 | 1.06 | 1.61 | 0.38 | 4.92 | |
| InfoAtlasN=5000 samples per task, Average over=10 random seeds2026.05 | 0.6 | 0.89 | 0.41 | 0.21 | 0.93 | 0.96 | 1.46 | 0.39 | 0.09 | |
| MINDEN=5000 samples per task, Average over=10 random seeds2026.05 | 0.58 | 0.92 | 0.43 | 0.36 | 0.89 | 1.01 | 1.42 | 0.5 | 34.2 | |
| InfoNCEN=5000 samples per task, Average over=10 random seeds2026.05 | 0.56 | 0.98 | 0.49 | 0.18 | 0.97 | 1.03 | 1.62 | 0.4 | 67.6 | |
| KSGN=5000 samples per task, Average over=10 random seeds2026.05 | 0.54 | 0.75 | 0.25 | 0.07 | 0.79 | 0.58 | 1.61 | 0.41 | 0.13 |