Regression on 1-D GP Synthetic (OOD)
-0.405Log-likelihoodRelaxedConvCNP
Evaluation results
| Method | Links | |
|---|---|---|
| RelaxedConvCNPfixed_inputs=tilde{T}, approximate_equivariance=true2024.06 | -0.405 | |
| ConvCNPfixed_inputs=tilde{T}, approximate_equivariance=true2024.06 | -0.412 | |
| EquivCNPfixed_inputs=tilde{E}, approximate_equivariance=true2024.06 | -0.413 | |
| TNPfixed_inputs=tilde{T}, approximate_equivariance=true2024.06 | -0.424 | |
| TNPfixed_inputs=hat{I}2024.06 | -0.43 | |
| ConvCNPfixed_inputs=T, equivariance=Strictly equivariant2024.06 | -0.442 | |
| EquivCNPfixed_inputs=E, equivariance=Strictly equivariant2024.06 | -0.443 | |
| TNPtype=Standard2024.06 | -1.3734 |