arXiv:2610.07173
Connectivity Controls Variational Accessibility in Symmetry-Preserving Quantum Circuits
Sahinur Reja
Abstract
Designing expressive quantum circuits for the variational quantum eigensolver (VQE) is a central challenge in simulating interacting many-body systems. Even symmetry-preserving (SP) circuits need not access the full many-body symmetry sector. We demonstrate this explicitly for the spin-$1/2$ XXZ and Hubbard models, where Hamiltonian-inspired nearest-neighbor (NN) SP circuits generate restricted dynamical Lie algebras (DLAs). While these restricted manifolds can accurately represent the ground state near the free-fermion limits, $J_z=0$ for XXZ and $U=0$ for Hubbard, interactions expose inaccessible many-body directions and severely limit VQE performance even at large circuit depth. Remarkably, a minimal modification of the generator connectivity can reconstruct the full orthogonal DLA within all relevant symmetry sectors studied in this work. For XXZ model, the addition of {\it only} one SP next-to-NN single-excitation connection triggers this dramatic reconstruction, whereas the Hubbard case requires a SP correlated double-excitation;--showing that the relevant connectivity can extend from the spatial lattice to many-body configuration space. This reconstruction enlarges the tangent space and accessibility of the energy-lowering Hamiltonian direction, leading to near-unit ground-state fidelity in VQE simulations. Our results establish a direct link between generator connectivity and variational accessibility and suggest a broader circuit-design principle: identify minimal symmetry-preserving generators that enlarge the accessible many-body space rather than increasing circuit depth alone. This perspective may have broader relevance for quantum many-body simulation at the interface of condensed-matter physics and quantum computing.
AI Summary
The paper asks when symmetry-preserving variational circuits remain unable to reach interacting ground states despite increasing depth, and whether a minimal change to their generators removes that restriction. It studies nearest-neighbor single-excitation circuits for periodic spin-$1/2$ XXZ and Hubbard chains, combining numerical dynamical Lie-algebra (DLA) closure, tangent-space diagnostics, and state-vector VQE simulations. For XXZ, the nearest-neighbor generators realize $\mathfrak{so}(N)$, of dimension $N(N-1)/2$, rather than the full orthogonal algebra on the fixed-excitation sector of dimension $D=\binom{N}{k}$. Adding one next-nearest-neighbor spin-exchange generator produces $\mathfrak{so}(D)$ in every tested sector: at $N=8$, $k=4$, the dimension increases from $28$ to $2415$. For the four-site, half-filled Hubbard model, the reported DLA dimensions are $12$ with nearest-neighbor single excitations, $30$ with additional longer-range single excitations, and the full $630$ after adding one correlated double excitation. Twelve-qubit VQE simulations measure tangent-space rank, projection of the energy-lowering direction onto that space, energy error, and ground-state fidelity. Restricted circuits perform well near the free-fermion limits but saturate away from them. The augmented circuits achieve near-unit fidelity; XXZ improvements appear around 16 layers, with interaction sweeps performed at 32 layers. The full-algebra reconstruction is established numerically only for the finite systems and sectors examined, not by a general-size theorem. The accessible algebra is orthogonal, reflecting real circuit generators, rather than arbitrary complex-unitary control. Performance evidence is limited to small, noiseless simulations, including additional magnetization sectors.
Error found: For the Hubbard fermionic single-excitation generators explicitly defined in the supplement, adding next-nearest-neighbor connections cannot increase the dynamical Lie-algebra dimension from 12 to 30 as reported, because the connected nearest-neighbor generators already generate all same-spin real quadratic hopping generators.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
- Comments
- 5 pages, 3 figures