Sparsity in Continuous-Depth Neural Networks