Figure 5From: Adaptive neural-domain refinement for solving time-dependent differential equationsVisualisation of grid-point distribution in a subdomain, (green/circle) 10 equidistant training points \(t_{i}\), (orange/cross) 12 equidistant verification points. Please note that, e.g. \(n_{{\mathrm{TP}}}=9\) refers to ten training points in Table 1Back to article page