MADE MoG

Reported on 6 benchmarks across 1 task · 2 papers · 6 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Methodology6 results

Density EstimationonBSDS300
Log-likelihood· 2017-05-19
153.71
best: 159.8 (TAN)
SOTA
Masked Autoregressive Flow for Density Estimation arXiv:1705.07057
Density EstimationonUCI HEPMASS
Log-likelihood· 2017-05-19
-15.15
best: -11.3 (DDE)
SOTA
Masked Autoregressive Flow for Density Estimation arXiv:1705.07057
Density EstimationonUCI MINIBOONE
Log-likelihood· 2017-05-19
-12.27
best: -6.94 (DDE)
SOTA
Masked Autoregressive Flow for Density Estimation arXiv:1705.07057
Density EstimationonMNIST
Log-likelihood (nats)· 2017-05-19
-1038.5
SOTA
Masked Autoregressive Flow for Density Estimation arXiv:1705.07057
Density EstimationonUCI POWER
Log-likelihood· 2017-05-19
0.4
best: 1.78 (nMDMA)
SOTA
Masked Autoregressive Flow for Density Estimation arXiv:1705.07057
Density EstimationonUCI GAS
Log-likelihood· 2015-02-12
8.47
best: 12.06 (B-NAF)
SOTA
MADE: Masked Autoencoder for Distribution Estimation arXiv:1502.03509