Pith. sign in

REVIEW 4 major objections 4 minor 67 references

Principles of Quantum Optimization for Constrained Problems

T0 review · 4 major / 4 minor · reviewed 2026-08-02 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The paper argues that computational slowdown in constrained quantum optimization is driven by entanglement restructuring, not by small spectral gaps, and that tiny avoided crossings can be jumped to speed up evolution.

desk verdict A plausible unification of constraint-aware quantum optimization, but the central fast-jump theorem has an unproved eigenvector alignment step. read the letter →

arxiv 2607.14227 v1 pith:5G4K6G6L submitted 2026-07-15 quant-ph math.OC

Einar Gabbassov , Gurpahul Singh , Achim Kempf This is my paper
classification quant-phmath.OC MSC 81P6890C1081P40 PACS 03.67.-a03.67.Lx
keywords quantumoptimizationconstrainedcombinatorialentanglementrestructuringspectralgapslevelcrossingseigenvectorswapadiabaticevolutionpenalty-freemethods
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper tries to establish that the real cost of a constrained quantum optimization run is entanglement restructuring—the forced creation, redistribution, or destruction of entanglement among qubits as the state evolves. It argues that constraints induce sequences of narrow or exactly closed spectral gaps, and that the difficulty of a problem is set by how much and how quickly entanglement must change, not by the minimum gap alone. If this is right, small gaps are not intrinsic bottlenecks: a fast-evolving system can jump across a narrow avoided crossing and keep the eigenvector it needs, arriving at the optimal feasible solution without restructuring. The paper also unifies penalty-free and penalty-based methods as spectral duals, and shows that constraint-aware dynamics—which keep restructuring minimal while keeping the optimum reachable—should outperform generic penalty-based encodings. A sympathetic reader would care because this gives a common physical explanation for when quantum optimization fails or succeeds, plus a concrete design principle.

What carries the argument

The central object is the ϵ-avoided level crossing: an avoided crossing whose gap closes with ϵ and is bounded below at order ϵ^2. The workhorse is the eigenvector-swap theorem for a rank-one perturbation nearly aligned with one eigenvector: at the crossing, the two adjacent energy levels exchange their eigenvectors over a short interval. In the fast-jump regime the system deliberately jumps during that exchange, so it lands on the neighboring level still carrying its original eigenvector, thereby avoiding entanglement restructuring. The other load-bearing piece is the relaxed mixer H_ϵ = (Π_F + ϵΠ_Q)H_init(Π_F + ϵΠ_Q), with Π_F projecting onto feasible states and Π_Q onto infeasible states;

What would settle it

Simulate a small constrained instance under the fast-jump schedule with ϵ small and measure the final overlap with |x*>. If a third energy level approaches within order ϵ of a crossing that the system is supposed to jump, and the overlap does not approach 1 as ϵ→0, the two-level assumption fails and the claimed jump-and-swap behavior is refuted.

Watch

Extended reading notes

Core claim

The paper's central formal claim is Theorem 2: for sufficiently small leakage ϵ, a system that starts in the ground state of a slightly relaxed feasibility-preserving mixer and evolves fast enough will jump upward at each ϵ-avoided level crossing—a narrow gap that closes like ϵ^2—rather than slowly following the instantaneous eigenstate. Because the dynamical jump and the eigenvector swap happen at the same crossing, the state preserves its eigenvector and entanglement structure while moving to the next energy level. It therefore reaches the level E_k(T)=f(x*) whose eigenvector is the optimal feasible solution |x*>, in time T much shorter than adiabatic following would require. The paper fur

Load-bearing premise

The paper's central result rests on the assumption that every narrow avoided crossing the system jumps is a clean two-level crossing with all other levels far away, and that there is a continuous feasible energy path to the optimal solution—true only for certain mixers and not for fragmented feasible spaces (Section VII.A, items 3–4).

Editorial extensions

If this is right

  • A sequence of narrow avoided crossings induced by constraints can be jumped rather than slowly followed; evolution times of order T≪ϵ^{-2} suffice, so small gaps can reduce rather than increase runtime.
  • An exactly closed global gap caused by crossing between orthogonal invariant subspaces does not by itself imply a bottleneck; only the gap of the feasible-restricted Hamiltonian matters for the evolution that stays in the feasible subspace.
  • Penalty-free and penalty-based formulations are spectral duals: the useful trajectory ascends to an excited level of the problem Hamiltonian in one and descends to the ground level in the other, which clarifies when starting in the mixer ground state is the right choice.
  • Transitions inside the feasible subspace are mediated by infeasible states with opposite weighting rules: penalty-based methods favor mildly violating mediators, while relaxed penalty-free methods favor mediators whose objective value is close to the instantaneous scaled energy.
  • Constraints should be built into the evolution rather than absorbed into penalties; the guiding design target is minimal entanglement restructuring with the optimum kept dynamically accessible.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The framework suggests a practical diagnostic: locate bottlenecks by computing where the rate of change of entanglement entropy peaks, not where the spectral gap is smallest; this extends the paper's own example.
  • It also suggests a design lever the paper does not pursue: deliberately engineering an infeasible mediator state whose objective value is near-resonant could create virtual shortcuts that accelerate penalty-free search, at the cost of controlled leakage.
  • The spectral-duality view implies that initialization in an excited mixer eigenstate within an invariant subspace is a general resource, and the Hamming-weight example indicates this can shrink the effective search dimension from exponential to polynomial.
  • If small gaps are exploitable, then noisy or randomized schedules that create many weak avoided crossings might be harnessed deliberately; a testable extension would be to compare stochastic schedules against smooth adiabatic schedules on the same constraint instances.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 4 minor

Summary. The paper develops a spectral theory of constrained quantum optimization, arguing that computational slowdown is governed by entanglement restructuring rather than by the minimum spectral gap alone. It introduces a general mixer H^ε_init = (Π_F + εΠ_Q)H_init(Π_F + εΠ_Q) that interpolates between feasibility-preserving and unconstrained mixers, and shows in small examples that penalty-based methods induce abrupt entanglement restructuring while certain constraint-aware mixers avoid it. The central technical result, Theorem 2 (Section VII.A), claims that for sufficiently small leakage ε, a fast evolution jumps upward at ε-avoided level crossings, preserves its instantaneous eigenvector and entanglement structure through each jump, and terminates in the optimal feasible state |x*⟩ at energy f(x*). The paper also proposes a spectral duality between penalty-free and penalty-based methods (Section VIII) and derives effective second-order transition weights for both (Section IX).

Significance. If the central claim were fully supported, the paper would provide a valuable conceptual shift: narrow avoided crossings could be used as computational shortcuts rather than bottlenecks, and constraint structure could be exploited to reduce entanglement restructuring. The paper also gives concrete illustrative examples, a clear graph-theoretic interpretation of mixer-induced transitions, and an explicit derivation of effective transition weights that could guide mixer design. However, the proof of Theorem 2 rests on a key unproved assertion — the eigenvector alignment in Eq. (41) — and on several assumptions that are stated but not established. The framework is promising, but the central theorem is not yet rigorous enough for publication as stated.

major comments (4)
  1. [Section VII.A, Eq. (41)] The proof of Theorem 2 reduces each ε-avoided crossing to the rank-one setting of Theorem 1 by asserting that D_init(t′) has an eigenvector |d_j⟩ with |⟨z|d_j⟩|² = 1−O(ε²). This is not proved. At ε=0 all infeasible states are degenerate zero eigenstates of H^0_init; the ε² Π_Q H_init Π_Q term can mix |z⟩ with other infeasible states, and the ε Π_F H_init Π_Q term can mix it with feasible states. The diagonal differences μ_x(t′)−μ_z(t′) can be O(ε) or smaller, so the eigenvector near |z⟩ need not have overlap 1−O(ε²) with |z⟩. Since this alignment is the basis for applying Theorem 1, the 'jump and swap occur together' mechanism and the conclusion E_k(T)=f(x*), |E_k(T)⟩=|x*⟩ are unsupported. Please provide a proof of Eq. (41) or state and verify explicit conditions on the spectrum of D_init(t′).
  2. [Section VII.A, item 4] The assumption that each ε-avoided crossing is locally a two-level crossing is load-bearing but not justified. The proof requires that no third level participates; otherwise the eigenvector-swap statement of Theorem 1 does not apply. The manuscript also does not establish that the coupling matrix elements that open the crossings are nonzero for every relevant infeasible level. A concrete condition in terms of the mixer's transition graph and the coefficients of H_init is needed.
  3. [Theorem 2 proof (Section VII.A) and Corollary 2.1] The proof is local and only explicitly treats infeasible levels with f(z)<f(x⋆). Crossings with infeasible levels that terminate above f(x⋆) but cross the feasible trajectory earlier are not discussed. More importantly, no global error bound is given for the accumulation of errors over repeated diabatic jumps. Corollary 2.1 gives T≪ε^{-2} as a per-crossing sufficient condition, but the final fidelity after all jumps is not estimated. Even if each jump has small error, the product over many crossings could be large; a rigorous statement needs a global error estimate.
  4. [Section IX, Eqs. (54)–(60)] The Schur-complement derivation of the effective Hamiltonian H^η_eff uses a Taylor expansion in η=ε² or 1/λ. For penalty-free methods, the weight W_z^ε = ε²/(f(z)−E/s) has a singular denominator at resonance, and the regime where f(z)≈E/s is precisely the avoided-crossing regime central to Theorem 2. The expansion therefore requires a regularization or a validity condition. This does not invalidate Theorem 2, but it weakens the classification of second-order transitions as stated.
minor comments (4)
  1. [Definition 1 vs. Corollary 2.1] Definition 1 says ε-avoided gaps are 'bounded below at the order ε²', while Corollary 2.1 assumes the gap is 'bounded above at order ε'. Please harmonize these statements and clarify whether the upper bound is meant to hold for all relevant crossings.
  2. [Section VII.A, item 3] The phrase 'This assumption holds for mixers that connect all feasible basis states through nonpositive off-diagonal matrix elements' is unclear. Perron–Frobenius arguments give a positive ground state, but they do not by themselves guarantee a continuous eigenvalue trajectory ending at |x⋆⟩. Please state the precise graph-theoretic condition and cite the relevant result.
  3. [Eq. (41)] The phrase 'one can show' should be replaced by an explicit lemma with proof. As written, the assertion is a black box in the central theorem.
  4. [Figure 9] Please specify the value of ε used in panel (b) and state whether the spectra are schematic or numerically computed. This would help readers assess the scale of the avoided gaps.

Circularity Check

0 steps flagged · score 2.0 of 10

No construction-level circularity; main theorem is conditional and contains a load-bearing unproved assertion (Eq. 41), with an author-overlapping theorem as the imported mechanism.

full rationale

The paper's derivation does not reduce any prediction to fitted inputs by construction. Theorem 2 is a conditional statement: for ε=0 it assumes a feasible eigenvalue trajectory ending at f(x*) (Section VII.A, item 3), and for ε>0 it assumes each crossing is locally two-level (item 4). The proof then applies the eigenvector-swap mechanism. The swap mechanism itself is imported from [45] (Theorem 1 and Corollaries 1.1/1.2), which is author-overlapping self-citation; however, [45] is a parameter-free theorem with stated assumptions, so under the review rules it counts as independent support and does not by itself constitute circularity. The main rigor gap is in the proof of Theorem 2: after Eq. (41) the paper asserts 'Eigenvector |d_j> exists by the construction of the mixer H^ε_init. That is, one can show that |d_j>=α|v>+β|ζ> ... with |α|^2=1−O(ε^2).' This is unproved and load-bearing: without it the local reduction to the rank-one Hamiltonian of Theorem 1 is unsupported, because the ε H_FQ coupling can mix |z> with the feasible level that is only O(ε) away at the crossing. This is a correctness/rigor concern, not a circular reduction. No fitted parameters are called predictions, and no external benchmarks are used. Accordingly, circularity is minimal: score 2.

Assumptions & free parameters 3 free parameters · 6 assumptions · 0 invented entities

The central claims rest on standard adiabatic theorems, the invariant-subspace structure of feasibility-preserving mixers, and several domain-specific assumptions about spectral connectivity and two-level isolation. No new physical entities are introduced beyond the generalized mixer operator H^ϵ_init, which is a mathematical construction rather than a new physical object.

free parameters (3)
  • epsilon (leakage strength ϵ) = sufficiently small, ϵ << Δ; T << ϵ^{-2}
    Controls coupling between feasible/infeasible subspaces. The fast-jump theorem and Corollary 2.1 require a gap bounded at order ϵ, but no constructive bound is given for generic problems.
  • penalty multiplier λ (penalty-based methods) = large (λ ≫ 1)
    Defines the large-penalty regime used for the second-order transition weight W^λ_z = 1/(λ g(z,w)). The duality is qualitative and does not depend on a specific value.
  • minimum protected gap Δ = must satisfy Δ^{-2} << T << ϵ^{-2}
    Separates crossings that must be jumped from gaps that must be adiabatically followed. No estimate of Δ is supplied for any problem class.
assumptions (6)
  • standard math Adiabatic theorem / Landau-Zener passage governs survival or transition at avoided crossings
    Used for the T ≫ gap^{-2} and T ≪ gap^{-2} conditions in Corollary 2.1.
  • domain assumption For ϵ=0, H_feas_init annihilates infeasible states and feasible/infeasible subspaces are invariant
    Equations (16)-(17) and (50); basis of the level-crossing arguments in Proposition 1 and Theorem 2.
  • domain assumption The restricted feasible Hamiltonian H_FF(t) has a continuous eigenvalue trajectory from the initial feasible eigenstate to the optimal state |x*>
    Section VII.A, item 3. Asserted to hold for mixers connecting all feasible states via nonpositive off-diagonal elements, following [17]; not proven here.
  • domain assumption Each ϵ-avoided level crossing is locally a two-level crossing; all other levels are asymptotically farther away
    Section VII.A, item 4. Needed for the local rank-one reduction to Theorem 1; not justified for dense spectra.
  • ad hoc to paper Eigenvector alignment |d_j> ≈ |v> with O(ϵ²) error holds by construction of H^ϵ_init (Eq. 41)
    Stated without proof in the proof of Theorem 2; this is the bridge that lets Corollary 1.2 apply to optimization Hamiltonians.
  • domain assumption For generic 0-1 IPs, infeasible solutions tend to have lower objective values than feasible optima (Observation 1)
    Motivates the spectral-duality picture; not universally true.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Principles of Quantum Optimization for Constrained Problems." pith.science (2026). https://pith.science/paper/5G4K6G6L

@misc{pith2026260714227,
  author       = {Pith},
  title        = {Pith review of: Principles of Quantum Optimization for Constrained Problems},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5G4K6G6L}},
  note         = {Machine review of arXiv:2607.14227}
}
read the original abstract

Constrained combinatorial optimization underlies many industrial and technological decision problems. We develop a spectral theory that unifies many quantum optimization algorithms. We show that computational slowdown is driven by entanglement restructuring: the creation, redistribution, and destruction of entanglement during system evolution. The severity of the slowdown depends on how much entanglement must be changed. We show that algebraic properties of constraints induce such restructuring, and that constraint-aware dynamics reduce the associated slowdown by avoiding unnecessary restructuring. This framework explains why constraint-aware quantum methods can outperform generic penalty-based approaches. The theory connects constrained optimization, computational complexity, entanglement dynamics, and Hamiltonian spectral structure across continuous-time and circuit-based quantum optimization paradigms.

Figures

Figures reproduced from arXiv: 2607.14227 by the authors.

Figure 1
Figure 1. FIG. 1. (a) A three-dimensional bounded polyhedron [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2. Conceptual illustration of iterative tightening in LP [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3. Transition graph induced by a feasibility-preserving [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (9 more)
Figure 4
Figure 4. Figure 4: both the entanglement entropy and its rate of [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]
Figure 5
Figure 5. Figure 5: FIG. 5. Transition graphs for three quantum optimization [PITH_FULL_IMAGE:figures/full_fig_p009_5.png]
Figure 4
Figure 4. Figure 4: FIG. 4. Bipartite entanglement entropy of the reduced sys [PITH_FULL_IMAGE:figures/full_fig_p009_4.png]
Figure 6
Figure 6. Figure 6: FIG. 6. Schematic illustration of an eigenvector swap at a [PITH_FULL_IMAGE:figures/full_fig_p010_6.png]
Figure 7
Figure 7. Figure 7: FIG. 7. The cascade of eigenvector swaps described in Corol [PITH_FULL_IMAGE:figures/full_fig_p012_7.png]
Figure 8
Figure 8. Figure 8: FIG. 8. Relevant eigenvalue trajectories of the total Hamiltonian [PITH_FULL_IMAGE:figures/full_fig_p013_8.png]
Figure 9
Figure 9. Figure 9: FIG. 9. Relevant eigenvalue trajectories of the total Hamiltonian [PITH_FULL_IMAGE:figures/full_fig_p017_9.png]
Figure 10
Figure 10. Figure 10: FIG. 10. A [PITH_FULL_IMAGE:figures/full_fig_p018_10.png]
Figure 9
Figure 9. Figure 9: These consist of an instance of the Maximum [PITH_FULL_IMAGE:figures/full_fig_p021_9.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

67 extracted references · 9 linked inside Pith

  1. [1]

    Penalty-based setup Let us consider the penalty-based approach first. The objective function is quadratic, and it is given by: q(x) = 4X i=1 cixi +λ(x 3 +x 4 −1) 2 (23) The infeasible solutions of the form (x 1, x2,1,1) and (x1, x2,0,0) are penalized by the energetic penaltyλ, and the solutions of the form (x 1, x2,1,0) and (x 1, x2,0,1) have no penalty. ...

  2. [2]

    For this, we consider the penalty-free approach

    Penalty-free setup It is possible to eliminate entanglement restructuring entirely, thereby achieving a significantly faster evolution time. For this, we consider the penalty-free approach. Then, the problem Hamiltonian is 1-local, Hf =− 1 2 4X i=1 ciZi.(26) First, consider the simplest feasibility-preserving mixer, obtained from the transverse-field Hami...

  3. [3]

    As noted in Observation 1, the lowest values off(x) are typically attained by in- feasible bit strings

    The problem HamiltonianH p encodes the linear objectivef(x). As noted in Observation 1, the lowest values off(x) are typically attained by in- feasible bit strings. Hence, the ground eigenstate and many lower-energy eigenstates ofH p lie inH Q, 13 FIG. 8. Relevant eigenvalue trajectories of the total HamiltonianH(t) for a 0–1 IP [see Section B 2]. The blu...

  4. [4]

    There- fore, the feasible and infeasible subspacesH F and HQ are invariant under the total Hamiltonian

    Forϵ= 0, we haveH ϵ init =H 0 init =H feas init . There- fore, the feasible and infeasible subspacesH F and HQ are invariant under the total Hamiltonian. Due to the invariance, eigenvalues whose eigenvectors are supported on different invariant subspaces cross exactly; see Figure 8 (a)

  5. [5]

    This is the desired feasible trajectory shown in blue in Figure 8 (a)

    Forϵ= 0, we assume that the restricted feasible Hamiltonian HF F(t) =P F H(t)P F has a continuous eigenvalue trajectory that starts from the chosen initial feasible eigenstate and terminates at the optimal feasible state|x ⋆⟩at s(T) = 1. This is the desired feasible trajectory shown in blue in Figure 8 (a). This assumption holds for mixers that connect al...

  6. [6]

    the algo- rithm of the century

    Forϵ >0, the termsϵH init F QandϵH init QF weakly cou- pleH F andH Q. Hence, provided the correspond- ing coupling matrix elements are nonzero, the exact crossings associated with theϵ= 0 generically open intoϵ-avoided level crossings; see Figure 8 (b). We assume that the two levels forming eachϵ-avoided level crossing remain isolated from all other level...

    1975

  7. [7]

    , n}and edge setE

    Maximum Independent Set Problem LetG= (V, E) be an undirected graph with vertices V={1, . . . , n}and edge setE. An independent set is a subset of vertices that contains no adjacent pair. In other words, if two verticesiandjare connected by an edge (i, j)∈E, then they cannot both be selected. We encode the choice of vertices by binary variables xi ∈ {0,1}...

  8. [8]

    In this for- mulation, the constraints are implemented through the mixer by restricting evolution within a feasible subspace of the full binary search space

    0–1 Knapsack Problem Figure 8 uses an instance of the 0–1 Knapsack Problem to illustrate the penalty-free formulation. In this for- mulation, the constraints are implemented through the mixer by restricting evolution within a feasible subspace of the full binary search space. The Knapsack Problem provides a simple setting for this construction: its feasi-...

Show all 67 references
  1. [9]

    The structure of X3C provides prior knowledge about the Hamming weight of the optimal solution

    Exact Cover by 3-Sets Figure 9 uses an instance of the Exact Cover by 3- Sets Problem (X3C) to illustrate that the penalty-based formulation is a spectral dual of the penalty-free case. The structure of X3C provides prior knowledge about the Hamming weight of the optimal solut...

  2. [10]

    R. E. Bixby, A brief history of linear and mixed- integer programming computation, inOptimization Sto- ries(European Mathematical Society-EMS-Publishing House GmbH, 2012) pp. 107–121

    2012

  3. [11]

    J. K. Strayer,Linear programming and its applications (Springer Science & Business Media, 2012)

    2012

  4. [12]

    R. J. Vanderbei,Linear Programming: Foundations and Extensions, 5th ed. (Springer, 2020)

    2020

  5. [13]

    Conforti, G

    M. Conforti, G. Cornu´ ejols, and G. Zambelli, Integer programming models, inInteger Programming(Springer,

  6. [14]

    Lucas, Ising formulations of many np problems, Fron- tiers in physics2, 5 (2014)

    A. Lucas, Ising formulations of many np problems, Fron- tiers in physics2, 5 (2014)

    2014

  7. [15]

    Gabbassov, G

    E. Gabbassov, G. Rosenberg, and A. Scherer, Lagrangian duality in quantum optimization: Overcoming qubo lim- itations for constrained problems, Physical Review Re- search7, 023305 (2025)

    2025

  8. [16]

    Glover, G

    F. Glover, G. Kochenberger, and Y. Du, A tutorial on formulating and using qubo models, arXiv preprint arXiv:1811.11538 (2018)

    2018 arXiv

  9. [17]

    De Santis, S

    D. De Santis, S. Tirone, S. Marmi, and V. Giovannetti, Optimized qubo formulation methods for quantum com- puting, Quantum Science and Technology (2024)

    2024

  10. [18]

    Hadfield, Z

    S. Hadfield, Z. Wang, B. O’gorman, E. G. Rieffel, D. Ven- turelli, and R. Biswas, From the quantum approximate optimization algorithm to a quantum alternating opera- tor ansatz, Algorithms12, 34 (2019)

    2019

  11. [19]

    Z. Wang, N. C. Rubin, J. M. Dominy, and E. G. Rief- fel, Xy mixers: Analytical and numerical results for the quantum alternating operator ansatz, Physical Review A 101, 012320 (2020)

    2020

  12. [20]

    J. Cook, S. Eidenbenz, and A. B¨ artschi, The quantum al- ternating operator ansatz on maximum k-vertex cover, in 2020 IEEE International Conference on Quantum Com- puting and Engineering (QCE)(IEEE, 2020) pp. 83–92

    2020

  13. [21]

    F. G. Fuchs, K. O. Lye, H. Møll Nilsen, A. J. Stasik, and G. Sartor, Constraint preserving mixers for the quantum approximate optimization algorithm, Algorithms15, 202 (2022)

    2022

  14. [22]

    Y. Hao, Q. Ding, X. Yuan, and X. Wang, Constraint- aware quantum optimization via hamming weight opera- tors, Science China Physics, Mechanics & Astronomy69, 250314 (2026)

    2026

  15. [23]

    B¨ artschi and S

    A. B¨ artschi and S. Eidenbenz, Grover mixers for qaoa: Shifting complexity from mixer design to state prepara- tion, in2020 IEEE International Conference on Quan- tum Computing and Engineering (QCE)(IEEE, 2020) pp. 72–82

    2020

  16. [24]

    Bucher, J

    D. Bucher, J. Stein, S. Feld, and C. Linnhoff-Popien, Penalty-free approach to accelerating constrained quan- tum optimization, Physical Review A112, 062605 (2025)

    2025

  17. [25]

    Bucher, D

    D. Bucher, D. Porawski, M. Janetschek, J. Stein, C. O’Meara, G. Cortiana, and C. Linnhoff-Popien, Effi- cient qaoa architecture for solving multi-constrained opti- mization problems, in2025 IEEE International Confer- ence on Quantum Computing and Engineering (QCE), Vol. 1 (IEE...

    2025

  18. [26]

    Herman, R

    D. Herman, R. Shaydulin, Y. Sun, S. Chakrabarti, S. Hu, P. Minssen, A. Rattew, R. Yalovetzky, and M. Pistoia, Constrained optimization via quantum zeno dynamics, Communications Physics6, 219 (2023)

    2023

  19. [27]

    K. A. Pawlak, J. M. Epstein, D. Crow, S. Gand- hari, M. Li, T. C. Bohdanowicz, and J. King, Quan- tum subspace correction for constraints, arXiv preprint arXiv:2310.20191 (2023)

    2023 arXiv

  20. [28]

    F. G. Fuchs and R. P. Bassa, Lx-mixers for qaoa: Op- timal mixers restricted to subspaces and the stabilizer formalism, Quantum8, 1535 (2024)

    2024

  21. [30]

    Farhi, J

    E. Farhi, J. Goldstone, S. Gutmann, J. Lapan, A. Lund- gren, and D. Preda, A quantum adiabatic evolution al- gorithm applied to random instances of an np-complete problem, Science292, 472 (2001)

    2001

  22. [31]

    Feinstein, I

    N. Feinstein, I. Shalashilin, S. Bose, and P. Warburton, Robustness of diabatic enhancement in quantum anneal- ing, Quantum Science and Technology10, 025011 (2025)

    2025

  23. [32]

    Muthukrishnan, T

    S. Muthukrishnan, T. Albash, and D. A. Lidar, Tunneling and speedup in quantum optimization for permutation- symmetric problems, Physical Review X6, 031010 (2016)

    2016

  24. [33]

    Fry-Bouriaux, D

    L. Fry-Bouriaux, D. T. O’Connor, N. Feinstein, and P. A. Warburton, Locally suppressed transverse-field protocol for diabatic quantum annealing, Physical Review A104, 052616 (2021)

    2021

  25. [34]

    F. A. Potra and S. J. Wright, Interior-point methods, Journal of computational and applied mathematics124, 281 (2000)

    2000

  26. [35]

    D. P. Bertsekas,Constrained optimization and Lagrange multiplier methods(Academic press, 2014)

    2014

  27. [36]

    Farhi, J

    E. Farhi, J. Goldstone, and S. Gutmann, A quan- tum approximate optimization algorithm, arXiv preprint arXiv:1411.4028 (2014)

    2014 arXiv

  28. [37]

    N. Moll, P. Barkoutsos, L. S. Bishop, J. M. Chow, A. Cross, D. J. Egger, S. Filipp, A. Fuhrer, J. M. Gam- betta, M. Ganzhorn,et al., Quantum optimization us- ing variational algorithms on near-term quantum devices, Quantum Science and Technology3, 030503 (2018)

    2018

  29. [38]

    L. K. Grover, From schr¨ odinger’s equation to the quan- tum search algorithm, American Journal of Physics69, 769 (2001)

    2001

  30. [39]

    Roland and N

    J. Roland and N. J. Cerf, Quantum search by local adi- abatic evolution, Physical Review A65, 042308 (2002)

    2002

  31. [40]

    Durr and P

    C. Durr and P. Hoyer, A quantum algorithm for finding the minimum, arXiv preprint quant-ph/9607014 (1996)

    1996 arXiv

  32. [41]

    Rajak, S

    A. Rajak, S. Suzuki, A. Dutta, and B. K. Chakrabarti, Quantum annealing: An overview, Philosophical Trans- actions of the Royal Society A: Mathematical, Physical and Engineering Sciences381(2023)

    2023

  33. [42]

    Farhi, J

    E. Farhi, J. Goldstone, S. Gutmann, and M. Sipser, Quantum computation by adiabatic evolution, arXiv:quant-ph/0001106 (2000)

    2000 arXiv

  34. [43]

    Albash and D

    T. Albash and D. A. Lidar, Adiabatic quantum compu- tation, Reviews of Modern Physics90, 015002 (2018)

    2018

  35. [44]

    Aharonov, W

    D. Aharonov, W. Van Dam, J. Kempe, Z. Landau, S. Lloyd, and O. Regev, Adiabatic quantum computation is equivalent to standard quantum computation, SIAM review50, 755 (2008)

    2008

  36. [45]

    McDowall, K

    K. McDowall, K. Georgopoulos, and P. Wallden, Spectral gap informed ramp qaoa, arXiv preprint 25 arXiv:2604.24580 (2026)

    2026 arXiv

  37. [46]

    Nzongani, D

    U. Nzongani, D. L. Mermoud, and A. Braida, Scal- ing qaoa: transferring optimal adiabatic schedules from small-scale to large-scale variational circuits, arXiv preprint arXiv:2602.14986 (2026)

    2026

  38. [47]

    Deshpande, A

    A. Deshpande, A. V. Gorshkov, and B. Fefferman, Im- portance of the spectral gap in estimating ground-state energies, PRX Quantum3, 040327 (2022)

    2022

  39. [48]

    Dooley, G

    S. Dooley, G. Kells, H. Katsura, and T. C. Dorlas, Sim- ulating quantum circuits by adiabatic computation: Im- proved spectral gap bounds, Physical Review A101, 042302 (2020)

    2020

  40. [49]

    Lin and Y

    L. Lin and Y. Tong, Near-optimal ground state prepara- tion, Quantum4, 372 (2020)

    2020

  41. [50]

    Roca-Jerat, T

    S. Roca-Jerat, T. Sancho-Lorente, J. Rom´ an-Roche, and D. Zueco, Circuit complexity through phase transitions: Consequences in quantum state preparation, SciPost Physics15, 186 (2023)

    2023

  42. [51]

    ˇZunkoviˇ c, M

    B. ˇZunkoviˇ c, M. Ballarin, L. Wright, and M. Lubasch, Scalable, self-verifying variational quantum eigen- solver using adiabatic warm starts, arXiv preprint arXiv:2602.17612 (2026)

    2026

  43. [52]

    X. Wang, Y. Chai, X. Feng, Y. Guo, K. Jansen, and C. T¨ uys¨ uz, Imaginary hamiltonian variational ansatz for combinatorial optimization problems, Physical Review A 111, 032612 (2025)

    2025

  44. [53]

    Hopkins and V

    A. Hopkins and V. Kendon, Multi-stage quantum walks for finding ising ground states, arXiv preprint arXiv:2511.01312 (2025)

    2025 arXiv

  45. [54]

    Gabbassov and A

    E. Gabbassov and A. Kempf, Adiabatic dynamics of entanglement, Quantum Science and Technology10, 045054 (2025)

    2025

  46. [55]

    D. H. Wolpert and W. G. Macready, No free lunch theo- rems for optimization, IEEE transactions on evolutionary computation1, 67 (2002)

    2002

  47. [56]

    H. P. Williams,Model building in mathematical program- ming(John Wiley & Sons, 2013)

    2013

  48. [57]

    Rieffel, J

    E. Rieffel, J. M. Dominy, N. Rubin, and Z. Wang, Xy- mixers: analytical and numerical results for qaoa, Phys. Rev. A101, 012320 (2020)

    2020

  49. [58]

    B¨ artschi and S

    A. B¨ artschi and S. Eidenbenz, Deterministic prepara- tion of dicke states, inInternational Symposium on Fun- damentals of Computation Theory(Springer, 2019) pp. 126–139

    2019

  50. [59]

    B¨ artschi and S

    A. B¨ artschi and S. Eidenbenz, Short-depth circuits for dicke state preparation, arXiv preprint arXiv:2207.09998 (2022)

    2022 arXiv

  51. [60]

    Suzuki, Fractal decomposition of exponential opera- tors with applications to many-body theories and monte carlo simulations, Physics Letters A146, 319 (1990)

    M. Suzuki, Fractal decomposition of exponential opera- tors with applications to many-body theories and monte carlo simulations, Physics Letters A146, 319 (1990)

    1990

  52. [61]

    Ostmeyer, Optimised trotter decompositions for classi- cal and quantum computing, Journal of Physics A: Math- ematical and Theoretical56, 285303 (2023)

    J. Ostmeyer, Optimised trotter decompositions for classi- cal and quantum computing, Journal of Physics A: Math- ematical and Theoretical56, 285303 (2023)

    2023

  53. [62]

    Lemelin, C

    A. Lemelin, C. Pere, O. Landon-Cardinal, and C. Coti, Mid-circuit measurement as an algorithmic primitive, arXiv preprint arXiv:2506.00118 (2025)

    2025 arXiv

  54. [63]

    A. B. Magann, K. M. Rudinger, M. D. Grace, and M. Sarovar, Feedback-based quantum optimization, Physical Review Letters129, 250502 (2022)

    2022

  55. [64]

    Gabbassov, Stochastic schr¨ odinger equations for quan- tum reverse diffusion, Physical Review Research8, 023329 (2026)

    E. Gabbassov, Stochastic schr¨ odinger equations for quan- tum reverse diffusion, Physical Review Research8, 023329 (2026)

    2026

  56. [65]

    L. V. Kantorovich, Mathematical methods of organiz- ing and planning production, Management science6, 366 (1960)

    1960

  57. [66]

    D. Avis, A. Hertz, and O. Marcotte, eds.,Graph Theory and Combinatorial Optimization(Springer, New York, NY, 2005)

    2005

  58. [67]

    Korte and J

    B. Korte and J. Vygen,Combinatorial Optimization: Theory and Algorithms, 5th ed., Algorithms and Com- binatorics, Vol. 21 (Springer, Berlin, Heidelberg, 2012)

    2012

  59. [68]

    D. E. Knuth,The Art of Computer Programming, Vol- ume 4, Fascicle 5: Mathematical Preliminaries Redux; Introduction to Backtracking; Dancing Links(Addison- Wesley Professional, Boston, MA, 2020)

    2020

Pith tools

Reviewed August 2, 2026 · model on record in the stance chip above.