LMG Hamiltonian has some symmetries, so we could exploit them for a new ansatz that may or may not improve TwoLocal performances. It's the case of changing only system ansatz since the ancilla one works fine in providing right probabilities (for beta \simeq 0 we have almost perfect fidelity, and that only depend on the entropy part of the Helmoltz energy, right?)
LMG Hamiltonian has some symmetries, so we could exploit them for a new ansatz that may or may not improve TwoLocal performances. It's the case of changing only system ansatz since the ancilla one works fine in providing right probabilities (for beta \simeq 0 we have almost perfect fidelity, and that only depend on the entropy part of the Helmoltz energy, right?)