arXiv:2610.03114
Circuit growth and subspace read-out in quantum-selected configuration interaction: Separating the Hartree-Fock determinant
Ayaka Fujimoto, Norifumi Matsumoto, Shota Kanasugi, Kazunori Maruyama, Hirotaka Oshima
Abstract
Predicting many molecular properties requires accurate ground-state energies of molecular Hamiltonians. Quantum-selected configuration interaction (QSCI) obtains such energies by measuring an input state to select important electronic configurations and classically diagonalizing the Hamiltonian in the subspace they span. ADAPT-QSCI, an adaptive extension of QSCI, improves the input state through iterative circuit growth. For single-reference molecules, the sampling distribution is dominated by the Hartree-Fock (HF) configuration, so most measurements reproduce a configuration already known. We reduce this redundancy by placing the HF configuration in the diagonalization subspace classically, before any measurement. This invalidates the operator-selection criterion of ADAPT-QSCI, and we derive a consistent replacement with energy and gradient in closed form, adding no query to the quantum device. At the per-iteration shot budgets previously used for ADAPT-QSCI, the proposed scheme reaches chemical accuracy with approximately $63\%$ and $43\%$ of the CNOT gates for $\mathrm{H}_4$ and $\mathrm{H}_6$, and $44\%$ and $40\%$ of the total shots. The advantage depends on the budget, however: in the infinite-shot limit the $\mathrm{H}_4$ CNOT count increases to $125\%$ of that of ADAPT-QSCI. These results identify the per-iteration shot budget as a key design variable governing resource savings in QSCI-based algorithms.
AI Summary
Quantum-selected configuration interaction (QSCI) samples computational-basis configurations and classically diagonalizes the Hamiltonian in their span. This paper asks whether ADAPT-QSCI can avoid spending most measurements on the already-known Hartree–Fock (HF) determinant in single-reference molecules. The proposed variant supplies HF classically and initializes the quantum sampler on a superposition of two non-HF determinants selected using perturbative coefficients. For adaptive operator selection, it removes HF from the subspace eigenvector and uses the lowest generalized eigenvalue in the span of HF and the rotated remainder. The objective and its ranking gradient have closed forms evaluated classically, requiring no additional quantum queries. Noiseless-device simulations examine $\mathrm{H}_4$ and $\mathrm{H}_6$ at $1.0\,$Å in STO-3G, using 8 and 12 qubits. At previously used per-iteration budgets of $10^5$ and $5\times10^6$ shots, the scheme reaches chemical accuracy with approximately $63\%$ and $43\%$ of the baseline CNOT counts, and $44\%$ and $40\%$ of its total shots. Finite-budget comparisons use 20 trials. A separate fixed-state sampling test finds shot reductions of $21.2$ and $18.0$, consistent with the renormalization prediction $1/(1-p_{\mathrm{HF}})$. Savings depend on the shot budget: in the infinite-shot limit, the $\mathrm{H}_4$ CNOT count is $125\%$ of baseline. Limitations include only two small single-reference systems, no device noise, fixed shots per iteration, and no joint optimization of circuit construction and read-out. Several observed trajectory and distribution differences remain unexplained.
Central claim may not hold: The proposed Pauli-rotation updates do not enforce exclusion of the Hartree–Fock determinant from the measured state, so the ideal HF-deleted sampling experiment does not establish the defining operational claim. The reported H6 cost of 85 CNOTs, used to obtain the headline 43% ratio, is incompatible with both individual counts and conventional 20-trial medians under the paper's six-CNOT-per-rotation rule. The reported total-shot savings use samples from a renormalized physical-sector distribution without accounting for rejected measurements, although the circuit's operator pool need not preserve that sector.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
- Comments
- 14 pages, 4 figures