How to obtain the Jacobian Matrix with respect to the inputs of a keras model neural network?

Question

I recently started learning and using automatic differentiation to determine the gradients and jacobian matrix of a neural network with respect to a given input. The method