arXiv:2602.04878 presents thermal-state simulation in two operator bases: Pauli and Majorana. qbm/pauli_prop.py implements only the Pauli side. The Majorana version is the natural route to fermionic problems (the paper demonstrates the Fermi-Hubbard model on a triangular lattice), and the update rule is structurally identical.
What to do
A Majorana analogue of qbm/pauli_prop.py: monomials M_b = i^{r_b} m_1^{b_1} ... m_2N^{b_2N} indexed by a binary vector, with the same imaginary-time branching rule and a length-based truncation replacing Pauli-weight truncation.
Acceptance
- validated against the dense Gibbs state of a small fermionic Hamiltonian via Jordan-Wigner
- exact for quadratic (free-fermion) Hamiltonians, where an independent covariance-matrix reference exists
Getting oriented
Weeks 20-21 of the codebase tour; Section II and Appendix I of the paper.
arXiv:2602.04878 presents thermal-state simulation in two operator bases: Pauli and Majorana.
qbm/pauli_prop.pyimplements only the Pauli side. The Majorana version is the natural route to fermionic problems (the paper demonstrates the Fermi-Hubbard model on a triangular lattice), and the update rule is structurally identical.What to do
A Majorana analogue of
qbm/pauli_prop.py: monomialsM_b = i^{r_b} m_1^{b_1} ... m_2N^{b_2N}indexed by a binary vector, with the same imaginary-time branching rule and a length-based truncation replacing Pauli-weight truncation.Acceptance
Getting oriented
Weeks 20-21 of the codebase tour; Section II and Appendix I of the paper.