1: \begin{abstract}

2: 

3: In this paper, we present a novel approach for detecting the discontinuity interfaces of a discontinuous function. This approach leverages Graph-Informed Neural Networks (GINNs) and sparse grids to address discontinuity detection also in domains of dimension larger than 3. GINNs, trained to identify troubled points on sparse grids, exploit graph structures built on the grids to achieve efficient and accurate discontinuity detection performances. We also introduce a recursive algorithm for general sparse grid-based detectors, characterized by convergence properties and easy applicability.

4: 

5: Numerical experiments on functions with dimensions $n = 2$ and $n = 4$ demonstrate the efficiency and robust generalization of GINNs in detecting discontinuity interfaces. Notably, the trained GINNs offer portability and versatility, allowing integration into various algorithms and sharing among users.

6: 

7: \quad

8: 

9: \textbf{Keywords:} Discontinuous functions; Sparse grids; Deep learning; Graph neural networks; Discontinuity interface detection.

10: 

11: \textbf{MSC:} 68T07, % Artificial neural networks and deep learning

12: 03D32, % Algorithmic randomness and dimension

13: 65D40 % Numerical approximation of high-dimensional functions; sparse grids

14: 

15: \end{abstract}

16: