Table 3 Hyperparameter tuning of the weights for variance, invariance, and covariance loss terms.
From: Deep representation learning using layer-wise VICReg losses
Variance weight | Invariance weight | Covariance weight | Best DB | Avg. DB | SD of DB values across trials |
|---|---|---|---|---|---|
1 | 1 | 1 | 3.052 | 3.288 | 0.247 |
1 | 25 | 1 | 3.291 | 3.315 | 0.038 |
10 | 10 | 1 | 3.227 | 3.289 | 0.060 |
25 | 25 | 1 | 3.109 | 3.217 | 0.107 |
50 | 50 | 1 | 3.246 | 3.309 | 0.084 |
