Mn-Cu-Ni-Fe damping alloys, such as M2052, are key structural materials for vibration suppression in large-scale machinery due to their high specific strength, stiffness, and exceptional damping capacity. This damping mechanism stems from the migration of twin boundaries within the face-centered tetragonal (FCT) structure. However, accurately describing atomic-level behavior in such multi-component systems is challenging with conventional empirical potentials such as the Embedded Atom Method (EAM). To address this, we previously constructed a Neural Network Potential (NNP) for Mn-Cu-Ni-Fe alloys using DeePMD-kit, enabling high-precision Molecular Dynamics (MD) simulations. In this study, we investigated the mechanical behavior of twin boundaries in the M2052 alloy using these MD simulations. We constructed periodic single-crystal models with varying twin spacings and evaluated the Critical Resolved Shear Stress (CRSS), defined as the stress required to initiate twin boundary movement. Based on these results, we discuss how differences in twin spacing affect the frictional resistance to interface migration.