ResearchBenchmarksSynthetic Regression on Synthetic GP Matérn-5/2 kernel, InterpolationSynthetic Regression on Synthetic GP Matérn-5/2 kernel, InterpolationFollow0.01KL Divergence (dx=2)ConvGNP-0.00520.09740.20.3026Jun 19, 2023Evaluation resultsMethodMethodLinksKL Divergence (dx=2)KL Divergence (dx=3)KL Divergence (dx=5)ConvGNP2023.060.01——RGNPkernel_type=stationarykernel_type=stationary2023.060.030.050.11GNP2023.060.070.110.19FullRGNPkernel_type=isotropickernel_type=isotropic2023.060.090.160.21ConvCNP2023.060.26——RCNPkernel_type=stationarykernel_type=stationary2023.060.30.390.35FullRCNPkernel_type=isotropickernel_type=isotropic2023.060.30.320.29CNP2023.060.390.460.47