Hi, It is known that the number of spin-up and spin-down electrons are conserved quantities in the Hubbard model. I just noticed that in the mps_optim paper (Computer Physics Communications 185 (2014) 3430-3440) that the example for the hubbard ladder has Nup_total and Ndown_total are the same even if the system is away from half-filling. Is this a general property of the model's ground state? If so, I would like to know references pertaining to this property. Related to that is, if I used the alternative fermionic hubbard model, is the property mentioned above the same as setting Sz_total = 0? Thanks, Robertson C. Esperanza