Ground state properties of quantum skyrmions described by neural network quantum states

We investigate the ground state properties of quantum skyrmions in a ferromagnet using variational Monte Carlo with neural network as ansatz. study states two-dimensional Heisenberg model presence Dzyaloshinskii-Moriya interaction (DMI) and show that accommodates skyrmion for large range parameters, especially at DMI. The spins these are weakly entangled, entanglement increases decreasing also find central spin is completely disentangled from rest lattice, establishing non-destructive way detecting this type by local magnetization measurements. While networks well suited to detect entangled DMI, they struggle describe small DMI regime due nearly degenerate strong entanglement. In paper, we propose method identify technique alleviate problem. Finally, analyze workings explore its limits pruning. Our work shows can be efficiently used magnetism systems exceeding size manageable exact diagonalization far.