Mott Transition and Volume Law Entanglement with Neural Quantum States
Résumé
The interplay between delocalisation and repulsive interactions can cause electronic systems to undergo a Mott transition between a metal and an insulator. Here we use neural network hidden fermion determinantal states (HFDS) to uncover this transition in the disordered, fully-connected Hubbard model. Whilst dynamical mean-field theory (DMFT) provides exact solutions to physical observables of the model in the thermodynamic limit, our method allows us to directly access the wavefunction for finite system sizes well beyond the reach of exact diagonalisation. We directly benchmark our results against state-of-the-art calculations obtained using a Matrix Product State (MPS) ansatz. We demonstrate how HFDS is able to obtain more accurate results in the metallic regime and in the vicinity of the transition, with the volume law of entanglement exhibited by the system being prohibitive to the MPS ansatz. We use the HFDS method to calculate the amplitudes of the wavefunction, the energy and double occupancy, the quasi-particle weight and the energy gap, hence providing novel insights into this model and the nature of the transition. Our work paves the way for the study of strongly correlated electron systems with neural quantum states.