Global search helps fix traps in quantum algorithm tuning
When is global evolutionary search useful for variational quantum algorithms? A landscape-first study
Neural and Evolutionary Computing
Summary
Tuning quantum circuits to solve problems is hard because the process can get stuck in bad solutions. The authors studied when using a global search strategy helps find better settings compared to just trying many local tweaks. They discovered that certain problem features make local methods unreliable, but global search performs better, especially when parameters are reused or when the problem mixes certain types of cost components. They also developed a way to predict in advance if global search will be helpful, making the tuning process more efficient.
What this means in practice
- •For quantum algorithm developers: Improve tuning strategies by identifying when global optimization beats local search for quantum circuit parameters, leading to better performance on hard problem instances.
- •For quantum hardware testing teams: Use a predictive landscape score to determine whether investing in global search optimizers will enhance quantum algorithm calibration efficiency on new problems.
Authors
Vojtěch Novák, Ivan Zelinka
Abstract
Variational quantum algorithms recast state preparation as classical nonconvex optimization, but it is often unclear when multistart local search suffices and when population-based global search justifies its evaluation cost. Using a landscape-first design, controlled QAOA experiments identify two mechanisms making local search unreliable: parameter reuse across circuit layers and competition between 2-local and 3-local cost terms. Increasing QAOA depth alone does not replicate this effect. We test these mechanisms across eight unseen spin glasses, fresh $N=10$ and $N=12$ instances, and standard MaxCut, transverse-field Ising, and Heisenberg models. On all eight unseen spin glasses, adaptive differential evolution (DE) achieves lower median error than both multistart BFGS and multistart Powell on the two difficult constructions, whereas the independent-depth control does not. The parameter-reuse effect transfers to MaxCut at both sizes, while standard VQE models remain favorable to local search. A pre-benchmark landscape score combining random-start local outcomes and 1D parameter slices is fixed from confirmation data before evaluating new optimizers. It predicts whether adaptive DE outperforms multistart BFGS by over one spectral-range percentage point on 50 new quantum objectives with 80--86% condition-level accuracy. The clearest indicator of global-search utility is that local search frequently traps in inferior basins, rather than circuit depth or curvature anisotropy. The findings motivate whether quantum circuit complexity can be traded for harder classical optimization when robust global search is available.