arXiv:2610.29919
Resource Analysis of Quantum Optimization Algorithms via Variational Methods
Vitor Dallabona, Lucas Friedrich, Jonas Maziero
Abstract
Variational Quantum Algorithms (VQAs) are promising approaches for solving combinatorial optimization problems on near-term quantum devices. However, the role played by quantum resources such as entanglement and coherence during the optimization process remains poorly understood. In this work, we investigate the evolution of these resources in two VQAs--the Quantum Approximate Optimization Algorithm (QAOA) and the Feedback-based Algorithm for Quantum Optimization (FALQON)--applied to the Minimum Vertex Cover problem. We employ a variational resource-quantification framework based on the Bures metric and Uhlmann's theorem, in which an auxiliary quantum circuit is optimized to identify the closest resource-free states. This approach allows us to quantify Bures entanglement and Bures coherence of the algorithmic states without requiring full quantum state tomography. For the considered problem instance, QAOA exhibits a pronounced accumulation of entanglement in its initial layers, followed by a progressive reduction as the algorithm converges toward the optimal solution. Its coherence likewise decreases overall with increasing circuit depth, although a finite residual value persists due to the degeneracy of the optimal-solution subspace. In contrast, FALQON exhibits a substantially lower entanglement peak, reached at larger circuit depth, followed by a decay as convergence proceeds. Its coherence decreases smoothly and monotonically together with the optimization error. These results reveal distinct quantum-resource trajectories associated with the two optimization strategies and illustrate how entanglement and coherence evolve as variational quantum states approach solutions of a classical combinatorial optimization problem.
AI Summary
The paper examines how entanglement and computational-basis coherence vary as quantum optimization algorithms approach solutions, and as molecular ground states become more correlated. It applies an existing variational resource-quantification framework to QAOA, FALQON, and FALQON+ on one four-vertex minimum vertex cover instance, and to four-qubit VQE simulations of $H_2$ in the STO-3G basis. The framework estimates Bures-based resource measures by optimizing auxiliary circuits representing purifications of separable or incoherent states. Uhlmann’s theorem converts the calculation into fidelity optimization; overlaps can be measured using a SWAP test without full state tomography. Resource estimates are evaluated after optimizing each circuit depth or molecular geometry. For the vertex-cover instance, QAOA’s estimated entanglement peaks at depth $p=1$ and subsequently decreases. FALQON has a lower peak near $p=16$, followed by decay. FALQON+ peaks near $0.35$ at $p=2$ and approaches zero for $p\geq5$. Coherence generally decreases alongside the reported probability-distribution error. Residual coherence in QAOA and FALQON+ is interpreted as compatible with superpositions of the two optimal bitstrings, $1010$ and $0110$. For $H_2$, entanglement and coherence are low near equilibrium, $R\approx1.40\,a_0$, and increase upon bond stretching. The paper explicitly treats these trajectories as instance-, algorithm-, and quantification-dependent rather than universal. It notes that correlations do not establish causation, degeneracy does not guarantee residual coherence, and low resource estimates alone establish neither efficient classical simulation nor reliable execution on noisy hardware.
Error found: The paper defines MSE against a single optimal bitstring but interprets vanishing MSE as compatible with finite pure-state coherence from a superposition of two optimal bitstrings; symmetry-preserving QAOA instead has an MSE floor of 1/32 against either bitstring when entirely within this optimal subspace. The paper reports bipartite Bures entanglement without identifying the qubit bipartition, leaving the entanglement trajectories incompletely specified.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
- Comments
- 12 pages, 9 figures