Quantum

The sign network expresses the Marshall sign rule for small J2 and the stripe sign rule for large J2, which greatly improves the accuracy of neural network quantum states on the square lattice.

A two-layer autoregressive network is used for solving the ground state of 1D transverse field Ising model and 1D hard core Bose-Hubbard model. The insets show the relative error of energy.

The stochastic series expansion (SSE) is used to study the phase diagram of the 1D Bose-Hubbard model. In the animations, the x-axis shows different atoms in the chain and the y-axis shows the slices in SSE.