This preprint studies whether expected-energy improvement guarantees improved constraint feasibility in a shallow QAOA setting. A frozen six-point, 24-variable, K=5, p=1 simulation panel is audited across all 15 pairwise comparisons. Five pairs show opposite directional changes between expected energy and P(K=5), while ten are aligned. The 5/15 count is explicitly descriptive because the pairs are dependent. A separate high-contrast transition is retained only as an illustration. The paper distinguishes finite-shot simulator evidence from exact statevector and physical-QPU evidence, removes a non-preregistered illustrative threshold from the primary analysis, and makes no quantum-advantage or universal-QAOA claim.
This preprint studies whether expected-energy improvement guarantees improved constraint feasibility in a shallow QAOA setting. A frozen six-point, 24-variable, K=5, p=1 simulation panel is audited across all 15 pairwise comparisons. Five pairs show opposite directional changes between expected energy and P(K=5), while ten are aligned. The 5/15 count is explicitly descriptive because the pairs are dependent. A separate high-contrast transition is retained only as an illustration. The paper distinguishes finite-shot simulator evidence from exact statevector and physical-QPU evidence, removes a non-preregistered illustrative threshold from the primary analysis, and makes no quantum-advantage or universal-QAOA claim.
Abstract Purpose : Poor feasible sampling in full-space Penalty-X quantum approximate optimization algorithm (QAOA) can reflect representation-induced dilution, classical optimizer inadequacy, objective misalignment, or limited ansatz capacity. We separate these explanations on a controlled routing benchmark. Methods : Exact-statevector experiments use 140 resource-constrained shortest-path tasks and a scale-controlled Hamiltonian. Exact depth embeddings diagnose optimizer inadequacy. A frozen discovery/held-out design compares mean-energy training (O0), an exact-feasibility capacity control (O2), and conditional value at risk (CVaR)-0.10 (O3), with post-hoc depth, optimizer, tail-fraction, and finite-shot checks. Results : O2 showed that the fixed ansatz could reach feasible mass not induced by O0. On 84 held-out tasks from 15 graphs, O3 improved graph-level log-feasibility gain over O0 by +0.3547 decades and was non-inferior to O2: the O3–O2 effect was −0.0086, with lower bound −0.0360 above the frozen −0.10 margin. Certified nested failures occurred under COBYLA, Nelder–Mead, and SLSQP. Fixed-endpoint sampling was stable, whereas both finite-shot retraining intervals crossed zero. Conclusion : Optimizer inadequacy and objective misalignment are distinct under the tested diagnostics, and CVaR approaches the mechanistic O2 capacity control on held-out graphs. The classically tractable study supports no quantum-advantage or extrapolative scaling claim and does not test hardware noise.
Abstract This paper studies how much Quantum Processing Unit (QPU) work can be avoided while preserving optimization quality in the Quantum Approximate Optimization Algorithm (QAOA). Using exact statevector simulation of QAOA on six graph optimization problems across seven graph families, three baselines are compared: a naive QPU-centric workflow with 30 quantum evaluations, a best-practice workflow that separates classical parameter optimization from a single QPU evaluation, and a classical-only solver. Six QPUminimization strategies are evaluated: preprocessing, pretraining, surrogate modeling, adaptive stopping, adaptive shots, and measurement reuse. Quality is measured using normalized expected value rather than the saturated best-of-b metric. The key findings are: the best-practice baseline achieves 510- to 2070-fold shot reduction across 8 to 16 qubits with at most 0.4 percent quality loss relative to naive QAOA; quality improves with QAOA depth from 0.82 to 0.91 while the gap between naive and best-practice remains negligible; four of six strategies achieve 100 percent quality retention at depth 3; quality is stable across graph families ranging from 0.77 to 0.87; and a structural predictor achieves perfect prediction on held-out graph families. These results support the thesis that the required QPU budget is a structured, predictable function of the problem instance and workflow policy, rather than a fixed algorithmic constant.
On a seven-compound drug-response model, warm-start quantum approximate optimization (QAOA) on IonQ Forte-1 returned valid assignments more often than random bitstrings, but this alone did not show effective optimization. Ideal QAOA raised optimum probability above uniform feasible sampling in only four of twelve reference circuits. Hardware often fell below its own noiseless circuits, while greedy search solved all hardware models within 200 objective evaluations. Expanded simulations showed a gain over feasible sampling in 28 of 35 CAMA-1 panels and none of four 647-V panels. Annealing solved all these panels in every seed. A Grover mixer preserved feasibility and improved optimum probability over feasible sampling in all ten tested models. We analyzed 25 completed tasks containing 5,300 shots from 18 circuits and 14 instances. The encodings use 6-35 qubits and at most 4,900 feasible assignments, which we enumerated to establish exact optima. Noiseless references now cover both the original six circuits and six wider circuits. At 35 qubits, with 4,900 feasible assignments, ideal feasibility was 35.85%, compared with 7 of 200 valid hardware outputs. Its ideal optimum probability was below both sampling controls. The circuits sample assignments in a model built from measured single-agent and pairwise responses.
This paper evaluates whether feasible-subspace-preserving variational circuit families can compete with the frozen QAOA+ baseline from Peters (2026) for Constraint-Level QUBO (CL-QUBO) fault tree minimal cut set identification. Six alternative constructions—an independently parameterized XX+YY generator family, a cost-phase-augmented variant, a transpiled-matched variant, a structure-ordered variant, an n = 6 restricted screen, and a warm-start variant—were tested against the QAOA+ comparator on a frozen pilot cohort of 14 reactor-scale fault tree subtrees (10 n = 6, 4 n = 8a instances) under noiseless statevector simulation with deterministic multi-restart optimization. The directly materialized family constructions passed feasible-subspace closure audits, and the screen-only variants inherited closure-qualified source families. These audits confirmed that the excitation-number-conserving XX+YY construction remained inside the intended fixed-Hamming-weight sector for the tested CL-QUBO instances. The failure mechanism is identified through per-instance analysis: the alternative families concentrate essentially all statevector probability (≈99.99%) on a single computational basis state that does not correspond to a minimal cut set, while QAOA+ distributes probability across the feasible subspace with sufficient structure to place MCS among the highest-ranked outcomes. The independently parameterized generator family lacks the problem-specific cost–mixer alternation that gives QAOA+ its concentration advantage, and adding a single cost phase layer does not rescue this deficit. The result is a disciplined negative finding: within the tested family definitions and frozen screening protocol, no approved alternative produced evidence strong enough to justify hardware execution. The contribution is a careful narrowing of the viable variational design space for CL-QUBO fault tree optimization, establishing that feasible-subspace preservation alone is insufficient for competitive MCS enrichment. Complete experimental artifacts with SHA-256 verification are provided.
Constraint-preserving mixers keep QAOA within a feasible subspace, but feasibility and global mixer connectivity do not determine which feasible configurations are available to a particular shallow circuit. We study constrained QAOA at the level of the circuit actually executed: a specified feasible initial state, finite depth, and finite schedule of mixer interactions. We define the finite-depth accessible set Rp(M,x0) and show that the computational-basis support of the ideal QAOA circuit is contained within it. This gives a parameter-independent objective ceiling fR⋆=maxx∈Rpf(x) and a normalized reachable-quality diagnostic QR, separating accessible volume from the objective quality contained within it.
We evaluate this framework using resource-matched block-local XY mixer schedules for constrained portfolio optimization and synthetic block-constrained quadratic problems. Across 207 favorable-tail CVaR QAOA configurations, QR shows the strongest association with realized shallow-QAOA performance among the structural diagnostics considered, with Pearson and Spearman correlations of 0.7003 and 0.8057; the relationship remains positive under controlled and dependence-aware analyses. The distinction persists across changes in feasible initialization, portfolio size, and objective family.
A compilation-only study further shows that, for the dense portfolio Hamiltonians considered here, phase-separator routing dominates the compiled two-qubit resource scale, while mixer schedules with comparable compiled two-qubit cost can exhibit markedly different finite-depth accessible quality. These results motivate objective-relevant finite-depth accessibility as a complementary diagnostic for resource-bounded constrained QAOA.
The Quantum Approximate Optimization Algorithm (QAOA) is a leading candidate for near-term quantum advantage in combinatorial optimization. A central design parameter is circuit depth p, with theoretical analyses suggesting higher p monotonically improves solution quality. This paper presents empirical evidence that this theoretical relationship does not hold on current noisy intermediate-scale quantum (NISQ) hardware for fault tree analysis applications. Across 120 QAOA executions on 60 QUBO-encoded fault tree instances (N ∈ {12, 16, 20, 24, 28, 32}) on IBM quantum hardware, circuit depths p = 1 and p = 2 were directly compared. The mean change in optimal state sampling probability from doubling circuit depth was Δmax_prob = 1.22 × 10⁻⁵ with a median Δ of zero, making the difference statistically indistinguishable from noise across all evaluated metrics. Output distribution entropy approaches the theoretical maximum (mean = 12.74 bits), indicating hardware noise dominates algorithmic signal. These findings establish that for fault tree minimal cut set identification on current NISQ devices, computational resources are more effectively allocated to increased shot counts than to circuit depth. Complete experimental artifacts with SHA-256 verification are provided, enabling full reproducibility.