gp_kernels: Gaussian process' kernels
In citiususc/voila: Variational Inference for Langevin Equations

Available kernels for creating gaussian processes

gp_kernel Abstract class representing a generic GP kernel. Final users should not use it.
exp_kernel An squared-exponential kernel with fixed amplitude.
rq_kernel A rational-quadratic kernel with fixed amplitude.
exp_const_kernel A sum of a squared-exponential kernel and a constant value. The sum of the constant value and the amplitude of the exponential kernel are tied so that they always sum the same value.
sum_exp_kernels A sum of two different squared-exponential kernels: The sum of the amplitudes of the exponential kernels are tied so that they always sum the same value.
clamped_exp_lin_kernel A sum of a squared-exponential kernel and a linear kernel. The sum of the amplitudes of the two terms are tied so that they always sum the same value. Note that this kernel is non-stationary.