Pith. sign in

REVIEW 2 major objections 5 minor 1 cited by

Implementing transferable annealing protocols for combinatorial optimisation on neutral atom quantum processors: a case study on smart-charging of electric vehicles

T0 review · 2 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

Pith's one-line read A single annealing schedule, trained on 50 small triangular graphs, transfers to larger unseen graphs of the same geometry and solves a real smart-charging problem.

desk verdict The small-graph transfer result is real and worth citing, but the N=100 hardware claim needs a classical post-processing baseline before it can be taken at face value. read the letter →

arxiv 2411.16656 v3 pith:D7L2567W submitted 2024-11-25 quant-ph

classification quant-ph
keywords transferableannealingprotocolsvariationalquantumBayesianoptimizationmaximumindependentsetRydbergatomarraysunitdiskgraphssmartcharging
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

This paper tries to establish that one quantum annealing schedule can be trained once on small problems and then reused on larger, unseen problems that share the same geometry, removing the per-instance parameter tuning that makes variational quantum optimization expensive. The authors optimize the schedule with Bayesian methods on 50 small maximum-independent-set instances drawn from a triangular atomic layout, then apply the same schedule to smart-charging instances and to larger triangular graphs. In noiseless emulation the transferred schedule prepares maximum independent sets with an average probability of 98.3(4)% on unseen 9- to 23-node smart-charging instances, and on neutral-atom hardware it retrieves maximum independent sets up to 100 atoms once measured bitstrings are repaired by a classical post-processing step. If these results hold, variational annealing becomes a one-time training cost per graph geometry rather than a per-instance optimization cost.

What carries the argument

The load-bearing object is a VQAA schedule: a time-dependent annealing drive parameterized by the Rabi frequency and detuning at a small number of fixed time points plus the total duration, with monotonic cubic spline interpolation between points. The drive is trained by Bayesian optimization on a cost that averages the MIS-preparation error over a family of graphs and penalizes spread, which pushes the optimizer toward schedules that work uniformly across the family. The geometric encoding is supplied by the Rydberg blockade effect, in which atoms closer than a blockade radius cannot both be excited, so a unit-disk graph can be embedded as an atom arrangement on a triangular lattice. The mechanism that makes transfer possible is parameter concentration: for VQAA-like drives the individual cost landscapes of graphs in a family overlap enough that one path through the phase diagram, ending in the maximum-independent-set phase and avoiding small-gap regions, prepares MIS configurations for all graphs in the family.

What would settle it

Run the trained protocol, with no re-optimization, on a fresh set of triangular-embeddable graphs of sizes between 10 and 100 and compare the noiseless-emulated average MIS probability to the paper's fitted exponential decay (roughly 37% at size 100). If the success probability at intermediate sizes drops far faster than that decay, or if on hardware after detection-error correction no maximum independent set is found at size 100, the size-transfer claim fails.

Watch

Extended reading notes

Core claim

The central claim is that optimal variational annealing parameters concentrate for graph families with a shared lattice geometry, so a single control protocol generalizes within the family and from small to large graphs. Specifically, a variational quantum annealing protocol with two control fields (Rabi frequency and detuning) sampled at fixed times and interpolated by monotonic splines is trained by Bayesian optimization on 50 triangular-lattice graphs of 5 to 9 nodes, with the cost defined as the average MIS-preparation error over the training set plus its standard deviation. The optimized schedule starts in the independent-set phase and ends inside the reduced maximum-independent-set phase of the phase diagram for the two control parameters, steering clear of regions with vanishing spectral gaps. On the training set it reaches a MIS probability of about 99.7%; on unseen smart-charging graphs of 9 to 23 nodes it reaches 98.3(4)% in noiseless emulation; and on hardware, after classical repair of constraint-violating bitstrings, it finds a maximum independent set at least once for every graph tested up to 100 atoms. The paper also reports that at large sizes the probability of sampling a maximum independent set decays roughly exponentially with graph size, while near-optimal solutions remain substantially more likely.

Load-bearing premise

The method rests on every target problem being drawable as a graph on the same triangular grid used in training, with edges only between nearby grid points, and on the physical atom array reproducing those connections faithfully.

Editorial extensions

If this is right

  • Once a schedule is trained for a lattice geometry, new instances of that geometry can be solved with a single fixed pulse sequence, skipping the closed-loop optimization that would otherwise cost hundreds or thousands of shots per instance.
  • For smart-charging instances that fit a triangular layout, the end-to-end pipeline turns a maximum-independent-set problem into a small number of experimental shots: the first MIS is typically found within a few shots, and with classical post-processing all MISs are found at least once for graphs up to 100 atoms.
  • At sizes beyond the training range, the protocol still acts as a useful low-energy sampler: even where the MIS probability has decayed to about 37% on average at size 100 in noiseless emulation, MIS or MIS-1 configurations are sampled frequently enough that roughly 14 shots suffice for a near-optimal solution at size 500 under the fitted decay.
  • Hardware noise, especially detection errors, degrades the raw distributions, but the classical repair step (removing conflicting nodes and greedily adding nodes up to depth 2) restores maximum independent sets, making the hybrid quantum-classical pipeline practical at current coherence times.

Reading between the lines

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

  • In our reading, the same training recipe should transfer to other regular layouts: training on square-lattice or Shastry-Sutherland graphs should produce a family-specific schedule, with the optimal ending detuning shifting as the geometry changes, as the paper's landscape comparison already hints.
  • The use of the standard deviation of the cost over the training family as a regularizer is a simple and possibly general idea; a testable extension would be to apply the same cost construction to QAOA parameter sets and see whether it tightens the weaker concentration the paper observes there.
  • If the graph-embedding gadgets mentioned in the outlook are placed on a regular lattice, the transferable-schedule approach could be applied to denser, non-unit-disk industrial constraints, making the method a candidate warm-start provider for classical solvers rather than only a standalone sampler.
  • The fitted exponential decay is itself a quantitative prediction: running the same protocol at intermediate sizes such as 40, 60, or 90 nodes and comparing the observed cumulative MIS probabilities to the fitted curves would test whether the decay law holds and how much of it is hardware-limited.
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

2 major / 5 minor

Summary. The paper proposes a transferable variational quantum annealing protocol for maximum independent set (MIS) problems on unit-disk graphs. Using Bayesian optimization, a single VQAA schedule is trained on 50 small triangular-lattice graphs and then applied, without further optimization, to unseen smart-charging instances (9–23 nodes) and to larger triangular graphs up to N=100. Noiseless emulation yields 98.3(4)% average MIS probability on the industrial test set and about 37% at N=100; hardware experiments yield 29% on the small instances and 0.05% raw MIS probability at N=100, with MIS configurations recovered after classical post-processing. The paper also fits exponential decay laws for MIS-k probabilities and uses them to extrapolate sampling requirements at larger sizes.

Significance. If substantiated, this is a useful demonstration that a single annealing schedule can transfer across instances sharing a common geometry, reducing the variational optimization overhead that currently limits practical quantum optimization. The out-of-sample design — training on small graphs and testing on unseen, larger graphs — avoids the most common circularity failure, and the paper deserves credit for benchmarking emulated noiseless, noisy-emulated, and experimental results, for explicitly modeling detection errors and decoherence, and for building on the open-source Pulser framework. The main caveats are that the headline N=100 hardware claim currently rests on classical post-processing without control baselines, and that the scaling extrapolation is fitted to the same data it is used to predict.

major comments (2)
  1. [Sec. IV C and App. H] The N=100 experimental claim is not isolated from classical post-processing. Raw experimental P(MIS) is reported as 0.05% at N=100, i.e. about one expected MIS in the ~1000-shot sample, and the statement that the authors 'are able to experimentally find the MIS at least once for each graph' refers to distributions after the App. H repair procedure, which removes an average of 4.75 constraint-violating nodes and greedily adds up to 2 nodes. No control baseline is provided for this post-processor: it is not applied to random bitstrings with the measured marginal excitation density, to shots from a trivial or untrained schedule, or to a purely classical independent-set sampler. Without such a baseline, the N=100 data cannot distinguish a genuinely transferable quantum schedule from a classical repair algorithm that would retrieve a MIS from almost any input distribution. The abstract's statement that experimental results 'validate the effectiveness of our approach, scaling to problems with up to 100 qubits' is therefore stronger than the evidence presented and should either be qualified to the post-processed hybrid pipeline or supported by the missing controls.
  2. [Sec. IV B, Eq. (3), Table I] The exponential decay model in Eq. (3) is fitted to the same emulated and experimental data that it is then used to extrapolate, for example in the N=500 sampling estimates in Sec. IV B. The piecewise form has free thresholds b_k and decay constants N_k, no goodness-of-fit or uncertainty is reported, and the experimental fits use only five sizes (N=30, 50, 70, 80, 100). The text should present Eq. (3) as an empirical interpolation over the measured range rather than as a predictive scaling law, and any extrapolation should be accompanied by error bars and a clear statement of model risk.
minor comments (5)
  1. [Sec. IV A and IV C] The time-to-solution comparison is stated inconsistently: Sec. IV A says the VQAA time to solution is 'around three orders of magnitude higher than with CPLEX', while Sec. IV C says it is 'sill three orders of magnitude below the state-of-the-art CPLEX method'; these cannot both be correct, and the intended comparison (including the typo 'sill') should be fixed.
  2. [Eq. (3) and Table I] The threshold parameters b_k are listed only for the experimental and post-processed fits, not for the emulated fits; please clarify whether the emulated fits use fixed thresholds, b_k=0, or separately fitted values.
  3. [Sec. II B] The text states that QAOA-like local minima 'do not approach 0 in value' and that VQAA landscapes reach 0.04–0.06; a sentence explaining why the averaged normalized approximation ratio saturates above zero would help the reader interpret the concentration plots.
  4. [Sec. III B and Conclusion] The paper acknowledges that cliques larger than 3 cannot be encoded on the triangular layout and that the industrial test set is restricted to 2D unit-disk instances; this limitation should be reflected more explicitly in the abstract's claim that the method applies to 'real-world scenarios'.
  5. [Fig. 5] The experimental points in Figs. 5(b) and 5(c) are shown without error bars; since each size appears to be represented by a single graph with ~1000 shots, the sampling uncertainty should be displayed or at least stated in the caption.

Circularity Check

2 steps flagged · score 6.0 of 10

Main transfer result is genuinely out-of-sample, but the N=500 sampling estimates in Secs. IV B/C are fitted exponentials evaluated at N=500, so those forward-looking scaling predictions reduce to their own fits.

  1. fitted input called prediction [Section IV B, 'Expected scaling of MIS-k probabilities', Eq. (3) and Table I]
    "To understand the decay of the summed probabilities over MIS −k configurations with graph size, we can nonetheless perform a naive fit using a piecewise exponential decay model, fk(N ) = ... . The decay constants extracted from the emulated data are summarised in Table I. Following these fits, we can extrapolate that measuring a MIS with F = 99% probability at size N = 500 would require around nshots ≥ log(1 − F )/ log(1 − exp(−N/N emu_0 )) ≈ 600 shots."

    The 'prediction' is obtained by inserting N=500 into the piecewise exponential f_0(N) of Eq. (3), using the decay constant N_0^{emu} fitted to the same cumulative MIS probabilities at N=10-100 shown in Fig. 5(a). No independent model, dataset, or benchmark enters: the 600-shot estimate is the fitted curve evaluated outside its fitting range, so the prediction is the fit by construction. The paper then draws the scalability conclusion ('the transferable protocol proves useful') from this self-evaluation of the fit.

  2. fitted input called prediction [Section IV C, 'Experimental scaling and classical post-processing', after Fig. 5(b)]
    "The P(MIS) extrapolated for N = 500 is now close to 0 and the quality of solutions that one can hope to retrieve after few hours of sampling has downgraded to MIS-6."

    This 'extrapolated' P(MIS) at N=500 is the piecewise-exponential model of Eq. (3) evaluated with the experimental decay constants N_k^{expt} from Table I, which were fitted to the raw experimental cumulative probabilities in Fig. 5(b). The downgrade to MIS-6 is therefore a direct evaluation of the fitted curve, not an independent measurement or physical derivation, so it cannot provide additional support for the scaling narrative beyond the data already used to determine the constants.

full rationale

The central transferability demonstration is out-of-sample and therefore not circular: a single VQAA schedule is trained by Bayesian optimization on 50 small triangular graphs (Sec. II C) and then applied without re-optimization to unseen smart-charging graphs of sizes 9-23 (Sec. III B) and to a separate triangular test set up to N=100 (Sec. IV A). The high noiseless P(MIS) on those unseen instances and the raw experimental data are evidence, not artifacts of the training cost. The phase-diagram and minimum-gap statements are sanity checks on the optimizer's trajectory, not derivations of it, and self-citations to Pulser, tensor-network emulation, and prior smart-charging work are tool/context citations rather than load-bearing uniqueness arguments. The circularity is confined to the forward-looking scaling estimates: Eq. (3) fits piecewise exponentials to the same cumulative-probability data displayed in Fig. 5, and the N=500 shot-count estimates (about 600 shots for MIS; P(MIS) close to 0 experimentally) are simply those fitted curves evaluated at N=500. Since no independent data or model is invoked, these numbers are the fit itself renamed as an extrapolated prediction and cannot validate the scaling claim. The experimental N=100 result is heavily mediated by App. H classical post-processing (raw P(MIS)=0.05% and no control baseline for the repair step); that is a validation weakness rather than definitional circularity, but it does underscore that the headline 'up to 100 qubits' claim is not isolated from classical assistance. Overall, the central derivation is self-contained while the secondary extrapolations reduce to fits, giving partial circularity.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

The central empirical result rests on the Rydberg encoding, on the existence of a quasi-adiabatic path, on distributional similarity between training and test graphs, and on the fitted exponential decay model used for the scaling extrapolation. No new physical entities are introduced. The main free choices are training cost hyperparameters, schedule hyperparameters, and the decay constants in Eq. (3).

free parameters (3)
  • Exponential decay constants N_k and thresholds b_k in Eq. (3) = N_0^emu=1.0(1)x10^2, N_0^expt=1.3(0)x10, b_0^expt=1, N_0^proc=2.1(1)x10, b_0^proc=3 (full set in Table I)
    The exponential decay model is a naive fit to emulated and experimental data, then used to extrapolate sampling requirements at N=500.
  • Weight of the standard-deviation term in training cost = 1
    In Sec. II C the training cost adds the standard deviation with weight set to 1; this affects which schedule is selected and is not derived or swept.
  • VQAA schedule hyperparameters m and Tmax = m=3, Tmax=4 us
    The number of control points and maximum duration constrain the optimized protocol; sensitivity to these choices is not reported.
assumptions (5)
  • domain assumption The Rydberg blockade Hamiltonian faithfully encodes the UD-MIS cost function for the considered geometries (App. A).
    The mapping assumes two-level atoms, a single C6 interaction scale, and negligible next-nearest-neighbor terms; the paper notes square geometry retains 1/8 spurious NNN interactions.
  • domain assumption A single quasi-adiabatic schedule of about 4 us can prepare MIS ground states for all triangular-lattice graphs in the family (Sec. II C and App. D).
    The optimized path avoids small gaps, but for triangular lattices first-order transitions with exponentially small gaps are expected, so this is an empirical assumption tested on the training set rather than a proven property.
  • ad hoc to paper Parameters optimized on graphs of size 5 to 9 transfer to graphs of size up to 100 when the geometry is shared (Sec. II B and IV A).
    This is the core transferability premise, supported by observed concentration in two-parameter landscapes and by out-of-sample tests, not by a derivation.
  • domain assumption The classical MIS size S_G computed with CPLEX is exact and defines both the cost and the reported success probabilities (Sec. III B and IV A).
    If CPLEX is wrong, normalization by S_G and all MIS probability values would be off; the paper trusts CPLEX for graphs up to 100 nodes.
  • ad hoc to paper The piecewise exponential model in Eq. (3) describes how MIS-k probabilities decay with graph size (Sec. IV B).
    The authors call this a naive fit; the thresholds and decay constants are fitted to the same data later used for extrapolation, so it is a modeling assumption rather than a theorem.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Implementing transferable annealing protocols for combinatorial optimisation on neutral atom quantum processors: a case study on smart-charging of electric vehicles." pith.science (2026). https://pith.science/paper/D7L2567W

@misc{pith2026241116656,
  author       = {Pith},
  title        = {Pith review of: Implementing transferable annealing protocols for combinatorial optimisation on neutral atom quantum processors: a case study on smart-charging of electric vehicles},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/D7L2567W}},
  note         = {Machine review of arXiv:2411.16656}
}
abstract

In the quantum optimization paradigm, variational quantum algorithms face challenges with hardware-specific and instance-dependent parameter tuning, which can lead to computational inefficiencies. The promising potential of parameter transferability across problem instances with similar local structures has been demonstrated in the context of the quantum approximate optimization algorithm. In this paper we build on these advancements by extending the concept to annealing-based protocols, employing Bayesian optimization to design robust quasi adiabatic schedules. Our study reveals that, for maximum independent set problems on graph families with shared geometries, optimal parameters naturally concentrate, enabling efficient transferability between similar instances and from smaller to larger ones. Experimental results on the Orion Alpha platform validate the effectiveness of our approach, scaling to problems with up to $100$ qubits. We apply this method to address a smart-charging optimization problem on a real dataset. These findings highlight a scalable, resource-efficient path for hybrid optimization strategies applicable in real-world scenarios.

Figures

Figures reproduced from arXiv: 2411.16656 by the authors.

Figure 1
Figure 1. FIG. 1 [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2 [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3 [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: FIG. 4 [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]
Figure 5
Figure 5. Figure 5: (a) shows the average cumulative probabilities for various types of configurations: non-independent sets (non-IS), independent sets (IS, with the largest labelled as MIS−k, indicating they are missing k ele￾ments to qualify as MIS), and MIS. While for unseen graphs of …

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Identifying hard native instances for the maximum independent set problem on neutral atoms quantum processors

    quant-ph 2025-02 conditional novelty 5.0 of 10

    Density and treewidth make natively embeddable unit-disk graph MIS instances harder for CPLEX, and current neutral-atom quantum devices are still about three orders of magnitude slower than classical solvers.

Reference graph

Works this paper leans on

70 extracted references · 37 canonical work pages · cited by 1 Pith paper

  1. [1]

    Algorithms for quantum computation: discrete logarithms and factoring

    Peter W Shor. Algorithms for quantum computation: discrete logarithms and factoring. In Proceedings 35th annual symposium on foundations of computer sci- ence, pages 124–134. Ieee, 1994

  2. [2]

    Quantum mechanics helps in searching for a needle in a haystack

    Lov K Grover. Quantum mechanics helps in searching for a needle in a haystack. Physical review letters , 79 (2):325, 1997

  3. [3]

    Colloquium: Quantum annealing and analog quantum computa- tion

    Arnab Das and Bikas K Chakrabarti. Colloquium: Quantum annealing and analog quantum computa- tion. Reviews of Modern Physics , 80(3):1061–1081, 2008

  4. [4]

    A quantum approximate optimization algo- rithm

    Edward Farhi, Jeffrey Goldstone, and Sam Gut- mann. A quantum approximate optimization algo- rithm. ArXiv:1411.4028, art. arXiv:1411.4028, 11

  5. [5]

    A review on quantum approximate op- timization algorithm and its variants

    Kostas Blekos, Dean Brand, Andrea Ceschini, Chiao- Hui Chou, Rui-Hao Li, Komal Pandya, and Alessan- dro Summer. A review on quantum approximate op- timization algorithm and its variants. Physics Re- ports, 1068:1–66, 6 2024. ISSN 0370-1573. doi: 10.1016/j.physrep.2024.03.002

  6. [6]

    Leo Zhou, Sheng-Tao Wang, Soonwon Choi, Hannes Pichler, and Mikhail D. Lukin. Quantum ap- proximate optimization algorithm: Performance, mechanism, and implementation on near-term de- vices. Phys. Rev. X , 10:021067, 6 2020. doi: 13 10.1103/PhysRevX.10.021067

  7. [7]

    Sched- ule path optimization for adiabatic quantum comput- ing and optimization

    Lishan Zeng, Jun Zhang, and Mohan Sarovar. Sched- ule path optimization for adiabatic quantum comput- ing and optimization. Journal of Physics A: Mathe- matical and Theoretical, 49(16):165305, 2016

  8. [8]

    Quantum adiabatic evolution algorithms with differ- ent paths

    Edward Farhi, Jeffrey Goldstone, and Sam Gutmann. Quantum adiabatic evolution algorithms with differ- ent paths. arXiv preprint quant-ph/0208135 , 2002

Show all 70 references
  1. [9]

    A study of heuristic guesses for adiabatic quantum computation

    Alejandro Perdomo-Ortiz, Salvador E Venegas- Andraca, and Al´ an Aspuru-Guzik. A study of heuristic guesses for adiabatic quantum computation. Quantum Information Processing, 10:33–52, 2011

  2. [10]

    Different strategies for optimization using the quantum adiabatic algo- rithm

    Elizabeth Crosson, Edward Farhi, Cedric Yen-Yu Lin, Han-Hsuan Lin, and Peter Shor. Different strategies for optimization using the quantum adiabatic algo- rithm. arXiv preprint arXiv:1401.7320 , 2014

  3. [11]

    Quantum computing with neutral atoms

    Lo ¨ ıc Henriet, Lucas Beguin, Adrien Signoles, Thierry Lahaye, Antoine Browaeys, Georges-Olivier Rey- mond, and Christophe Jurczak. Quantum computing with neutral atoms. Quantum, 4:327, 2020

  4. [12]

    Graph algo- rithms with neutral atom quantum processors

    Constantin Dalyac, Lucas Leclerc, Louis Vignoli, Mehdi Djellabi, Wesley da Silva Coelho, Bruno Ximenez, Alexandre Dareau, Davide Dreon, Vin- cent E Elfving, Adrien Signoles, et al. Graph algo- rithms with neutral atom quantum processors. The European Physical Journal A , 60(9)...

  5. [13]

    A survey on trans- fer learning

    Sinno Jialin Pan and Qiang Yang. A survey on trans- fer learning. IEEE Trans. Knowl. Data Eng. , 22(10): 1345–1359, October 2010

  6. [14]

    Transferability in deep learning: A sur- vey, 2022

    Junguang Jiang, Yang Shu, Jianmin Wang, and Ming- sheng Long. Transferability in deep learning: A sur- vey, 2022. URL https://arxiv.org/abs/2201.05867

  7. [15]

    J. A. Montanez-Barrera, Dennis Willsch, and Kris- tel Michielsen. Transfer learning of optimal qaoa pa- rameters in combinatorial optimization, 2024. URL https://arxiv.org/abs/2402.05549

  8. [16]

    Symmetry- informed transferability of optimal parameters in the quantum approximate optimization algorithm

    Isak Lyngfelt and Laura Garc ´ ıa-´Alvarez. Symmetry- informed transferability of optimal parameters in the quantum approximate optimization algorithm. Phys. Rev. A (Coll. Park.) , 111(2), February 2025

  9. [17]

    Transferability of optimal qaoa parameters between random graphs, 2021

    Alexey Galda, Xiaoyuan Liu, Danylo Lykov, Yuri Alexeev, and Ilya Safro. Transferability of optimal qaoa parameters between random graphs, 2021. URL https://arxiv.org/abs/2106.07531

  10. [18]

    Quantum computing with Rydberg atoms : control and modelling for quantum sim- ulation and practical algorithms

    Lucas Leclerc. Quantum computing with Rydberg atoms : control and modelling for quantum sim- ulation and practical algorithms . Theses, Univer- sit´ e Paris-Saclay, September 2024. URL https:// pastel.hal.science/tel-04745992

  11. [20]

    Syn- thetic three-dimensional atomic structures assembled atom by atom

    Daniel Barredo, Vincent Lienhard, Sylvain De Lese- leuc, Thierry Lahaye, and Antoine Browaeys. Syn- thetic three-dimensional atomic structures assembled atom by atom. Nature, 561(7721):79–82, 2018

  12. [21]

    Nogrette, H

    F. Nogrette, H. Labuhn, S. Ravets, D. Barredo, L. B´ eguin, A. Vernier, T. Lahaye, and A. Browaeys. Single-atom trapping in holographic 2d arrays of mi- crotraps with arbitrary geometries. Phys. Rev. X , 4: 021034, 5 2014. doi:10.1103/PhysRevX.4.021034

  13. [22]

    Many-body physics with individually controlled rydberg atoms

    Antoine Browaeys and Thierry Lahaye. Many-body physics with individually controlled rydberg atoms. Nature Physics, 16(2):132–142, 2020

  14. [23]

    Morgado and S

    M. Morgado and S. Whitlock. Quantum simula- tion and computing with Rydberg-interacting qubits. A VS Quantum Science , 3(2):023501, 6 2021. doi: 10.1116/5.0036562

  15. [24]

    B´ e guin, A

    L. B´ e guin, A. Vernier, R. Chicireanu, T. La- haye, and A. Browaeys. Direct measurement of the van der waals interaction between two rydberg atoms. Phys. Rev. Lett. , 110(26), 6 2013. doi: 10.1103/physrevlett.110.263201

  16. [25]

    ˇSibali´ c, J.D

    N. ˇSibali´ c, J.D. Pritchard, C.S. Adams, and K.J. Weatherill. ARC: An open-source library for calcu- lating properties of alkali rydberg atoms. Computer Physics Communications, 220:319–331, 11 2017. doi: 10.1016/j.cpc.2017.06.015

  17. [26]

    Hannes Pichler, Sheng-Tao Wang, Leo Zhou, Soonwon Choi, and Mikhail D. Lukin. Quantum optimization for maximum in- dependent set using rydberg atom arrays. ArXiv:1808.10816, art. arXiv:1808.10816, 8 2018. doi:10.48550/ARXIV.1808.10816

  18. [27]

    McClean, Sergio Boixo, Vadim N

    Jarrod R. McClean, Sergio Boixo, Vadim N. Smelyan- skiy, Ryan Babbush, and Hartmut Neven. Barren plateaus in quantum neural network training land- scapes. Nat. Commun. , 9(1), 11 2018. ISSN 2041-

  19. [28]

    Cerezo, and Patrick J

    Zo¨ e Holmes, Kunal Sharma, M. Cerezo, and Patrick J. Coles. Connecting ansatz expressibility to gradi- ent magnitudes and barren plateaus. PRX Quan- tum, 3(1), 1 2022. ISSN 2691-3399. doi: 10.1103/prxquantum.3.010313

  20. [29]

    Coles, and M

    Martin Larocca, Piotr Czarnik, Kunal Sharma, Gopikrishnan Muraleedharan, Patrick J. Coles, and M. Cerezo. Diagnosing barren plateaus with tools from quantum optimal control. Quantum, 6:824, 9

  21. [30]

    Evaluating an- alytic gradients on quantum hardware

    Maria Schuld, Ville Bergholm, Christian Gogolin, Josh Izaac, and Nathan Killoran. Evaluating an- alytic gradients on quantum hardware. Phys. Rev. A , 99(3), 3 2019. ISSN 2469-9934. doi: 10.1103/physreva.99.032331

  22. [31]

    Izmaylov, Robert A

    Artur F. Izmaylov, Robert A. Lang, and Tzu-Ching Yen. Analytic gradients in variational quantum al- gorithms: Algebraic extensions of the parameter- shift rule to general unitary transformations. Phys. Rev. A , 104(6), 12 2021. ISSN 2469-9934. doi: 10.1103/physreva.104.062443

  23. [32]

    Cora, and Nando de Freitas

    Eric Brochu, Vlad M. Cora, and Nando de Freitas. A tutorial on bayesian optimization of expensive cost functions, with application to active user modeling and hierarchical reinforcement learning, 2010. URL https://arxiv.org/abs/1012.2599

  24. [33]

    Op- timal quantum control with poor statistics

    Fr´ ed´ eric Sauvage and Florian Mintert. Op- timal quantum control with poor statistics. PRX Quantum , 1:020322, 12 2020. doi: 10.1103/PRXQuantum.1.020322

  25. [34]

    Jernej Rudi Finˇ zgar, Martin J. A. Schuetz, J. Kyle Brubaker, Hidetoshi Nishimori, and Helmut G. Katz- graber. Designing quantum annealing schedules us- ing bayesian optimization. Physical Review Re- search, 6(2), 4 2024. ISSN 2643-1564. doi: 14 10.1103/physrevresearch.6.023063

  26. [35]

    A quantum adiabatic evolution algorithm applied to random instances of an NP-complete prob- lem

    Edward Farhi, Jeffrey Goldstone, Sam Gutmann, Joshua Lapan, Andrew Lundgren, and Daniel Preda. A quantum adiabatic evolution algorithm applied to random instances of an NP-complete prob- lem. Science, 292(5516):472–475, 4 2001. doi: 10.1126/science.1057726

  27. [36]

    Hannes Pichler, Sheng-Tao Wang, Leo Zhou, Soon- won Choi, and Mikhail D. Lukin. Quantum opti- mization for maximum independent set using ryd- berg atom arrays, 2018. URL https://arxiv.org/ abs/1808.10816

  28. [37]

    Pulser: An open-source package for the de- sign of pulse sequences in programmable neutral-atom arrays

    Henrique Silv´ erio, Sebasti´ an Grijalva, Constantin Dalyac, Lucas Leclerc, Peter J Karalekas, Nathan Shammah, Mourad Beji, Louis-Paul Henry, and Lo ¨ ıc Henriet. Pulser: An open-source package for the de- sign of pulse sequences in programmable neutral-atom arrays. Quantum, ...

  29. [38]

    Antiferromagnetic metal and mott transition on shastry-sutherland lattice

    Hai-Di Liu, Yao-Hua Chen, Heng-Fu Lin, Hong-Shuai Tao, and Wu-Ming Liu. Antiferromagnetic metal and mott transition on shastry-sutherland lattice. Sci- entific Reports, 4(1), 4 2014. ISSN 2045-2322. doi: 10.1038/srep04829

  30. [39]

    L¨ auchli

    Vincent Lienhard, Sylvain de L´ es´ eleuc, Daniel Barredo, Thierry Lahaye, Antoine Browaeys, Michael Schuler, Louis-Paul Henry, and An- dreas M. L¨ auchli. Observing the space- and time-dependent growth of correlations in dynamically tuned synthetic ising models with antiferro...

  31. [40]

    Quantum phase transitions

    Subir Sachdev. Quantum phase transitions. Physics world, 12(4):33, 1999

  32. [41]

    Universal signatures of quantum critical points from finite-size torus spectra: A window into the operator content of higher-dimensional conformal field theories

    Michael Schuler, Seth Whitsitt, Louis-Paul Henry, Subir Sachdev, and Andreas M L¨ auchli. Universal signatures of quantum critical points from finite-size torus spectra: A window into the operator content of higher-dimensional conformal field theories. Physical Review Letters,...

  33. [42]

    Quantum annealing: The fastest route to quantum computa- tion? The European Physical Journal Special Topics , 224(1):75–88, 2015

    Christopher R Laumann, Roderich Moessner, An- tonello Scardicchio, and Shivaji Lal Sondhi. Quantum annealing: The fastest route to quantum computa- tion? The European Physical Journal Special Topics , 224(1):75–88, 2015

  34. [43]

    Phase diagram of the quantum ising model on a triangular lattice un- der external field

    Yuan Da Liao, Han Li, Zheng Yan, Hao-Tian Wei, Wei Li, Yang Qi, and Zi Yang Meng. Phase diagram of the quantum ising model on a triangular lattice un- der external field. Physical Review B, 103(10):104416, 2021

  35. [44]

    Qualifying quantum approaches for hard industrial optimization problems

    Constantin Dalyac, Loic Henriet, Emmanuel Jeandel, Wolfgang Lechner, Simon Perdrix, Marc Porcheron, and Margarita Veshchezerova. Qualifying quantum approaches for hard industrial optimization problems. a case study in the field of smart-charging of elec- tric vehicles. EPJ Qua...

  36. [45]

    A comprehensive review of dc fast-charging stations with energy storage: Architectures, power convert- ers, and analysis

    Md Ahsanul Hoque Rafi and Jennifer Bauman. A comprehensive review of dc fast-charging stations with energy storage: Architectures, power convert- ers, and analysis. IEEE Transactions on Trans- portation Electrification , 7(2):345–368, 2020. doi: 10.1109/TTE.2020.3015743

  37. [46]

    Vehicle-to-x (v2x) implementation: An overview of predominate trial configurations and technical, social and regulatory challenges

    Christine Gschwendtner, Simon R Sinsel, and An- negret Stephan. Vehicle-to-x (v2x) implementation: An overview of predominate trial configurations and technical, social and regulatory challenges. Renewable and Sustainable Energy Reviews , 145:110977, 2021. doi:10.1016/j.rser.2...

  38. [47]

    Combining dantzig-wolfe and benders decompositions to solve a large-scale nuclear outage planning prob- lem

    Rodolphe Griset, Pascale Bendotti, Boris Detienne, Marc Porcheron, Halil S ¸en, and Fran¸ cois Vanderbeck. Combining dantzig-wolfe and benders decompositions to solve a large-scale nuclear outage planning prob- lem. European Journal of Operational Research , 298 (3):1067–1083, 2022

  39. [48]

    Louis Hakimi

    Kazuo Nakajima and S. Louis Hakimi. Complex- ity results for scheduling tasks with discrete starting times”. journal of algorithms. Journal of Algorithms , 3:344–361, 1982

  40. [49]

    Mark Keil

    J. Mark Keil. On the complexity of scheduling tasks with discrete starting times. Operations Research Let- ters, 12:293–295, 1992

  41. [50]

    www.data.gouv.fr/fr/ datasets/belib-reseau-parisien-de-bornes- de-recharges-accelerees-22-kw-ac-dc-pour- vehicules-electriques/, 2017

    Belib r´ eseau parisien de bornes de recharges pour v´ ehicules ´ electriques. www.data.gouv.fr/fr/ datasets/belib-reseau-parisien-de-bornes- de-recharges-accelerees-22-kw-ac-dc-pour- vehicules-electriques/, 2017

  42. [51]

    Cloud on-demand emulation of quantum dynamics with tensor networks, 2023

    Kemal Bidzhiev, Aleksander Wennersteen, Mourad Beji, Mario Dagrada, Mauro D’Arcangelo, Sebastian Grijalva, Anne-Claire Le Henaff, Anton Quelle, and Alvin Sashala Naik. Cloud on-demand emulation of quantum dynamics with tensor networks, 2023

  43. [52]

    Halld´ orsson

    Ravi Boppana and Magn´ us M. Halld´ orsson. Approxi- mating maximum independent sets by excluding sub- graphs. BIT, 32(2):180–196, 6 1992. ISSN 1572-9125. doi:10.1007/bf01994876

  44. [53]

    Probing many-body dy- namics on a 51-atom quantum simulator

    Hannes Bernien, Sylvain Schwartz, Alexander Keesling, Harry Levine, Ahmed Omran, Hannes Pich- ler, Soonwon Choi, Alexander S Zibrov, Manuel En- dres, Markus Greiner, et al. Probing many-body dy- namics on a 51-atom quantum simulator. Nature, 551 (7682):579–584, 2017

  45. [54]

    Yassine Naghmouchi and Wesley da Silva Coelho

    M. Yassine Naghmouchi and Wesley da Silva Coelho. Mixed-integer linear programming solver using ben- ders decomposition assisted by a neutral-atom quan- tum processor. Phys. Rev. A , 110:012434, Jul 2024. doi:10.1103/PhysRevA.110.012434. URL https:// link.aps.org/doi/10.1103/P...

  46. [55]

    Lukin, Sheng-Tao Wang, and Hannes Pichler

    Minh-Thi Nguyen, Jin-Guo Liu, Jonathan Wurtz, Mikhail D. Lukin, Sheng-Tao Wang, and Hannes Pichler. Quantum optimization with arbitrary connectivity using rydberg atom arrays. PRX Quantum, 4:arXiv:2209.03965, 2 2023. doi: 10.1103/PRXQuantum.4.010316

  47. [56]

    Towards graph encoding for quantum computing with neutral atoms, 2024

    Dimitri Delpech de Saint Guilhem, Pascale Bendotti, and Joseph Mikael. Towards graph encoding for quantum computing with neutral atoms, 2024. URL https://fondation-hadamard.fr/media/ filer public/a4/f7/a4f795b0-1a5a-4efb-a327- c44887361100/pgmo-2024-booklet1 .pdf

  48. [57]

    Scaling advantage in approximate optimization with quantum annealing

    Humberto Munoz Bauza and Daniel A Lidar. Scaling advantage in approximate optimization with quantum annealing. arXiv preprint arXiv:2401.07184 , 2024

  49. [58]

    Thomas M. J. Fruchterman and Edward M. Rein- gold. Graph drawing by force-directed placement. Software: Practice and Experience, 21(11):1129–1164,

  50. [59]

    Consistency of the adiabatic theorem

    Mohammad HS Amin. Consistency of the adiabatic theorem. Phys. Rev. Lett., 102(22):220401, 2009. doi: 10.1103/PhysRevLett.102.220401

  51. [60]

    Adiabatic quan- tum computation

    Tameem Albash and Daniel A Lidar. Adiabatic quan- tum computation. Reviews of Modern Physics , 90(1): 015002, 2018. doi:10.1103/RevModPhys.90.015002

  52. [61]

    The quan- tum adiabatic algorithm applied to random optimiza- tion problems: The quantum spin glass perspective

    Victor Bapst, Laura Foini, Florent Krzakala, Guil- hem Semerjian, and Francesco Zamponi. The quan- tum adiabatic algorithm applied to random optimiza- tion problems: The quantum spin glass perspective. Physics Reports, 523(3):127–205, 2013

  53. [62]

    Quantum computation by adiabatic evolution

    Edward Farhi, Jeffrey Goldstone, Sam Gutmann, and Michael Sipser. Quantum computation by adiabatic evolution. arXiv:quant-ph/0001106., 2000

  54. [63]

    Quantum optimization of maximum inde- pendent set using rydberg atom arrays

    Sepehr Ebadi, Alexander Keesling, Madelyn Cain, Tout T Wang, Harry Levine, Dolev Bluvstein, Giu- lia Semeghini, Ahmed Omran, J-G Liu, Rhine Sama- jdar, et al. Quantum optimization of maximum inde- pendent set using rydberg atom arrays. Science, 376 (6598):1209–1215, 2022. 16 A...

  55. [69]

    Embedding UD graphs into a plane can be achieved in a relative straightforward way using force-directed graph drawingmethods such as the Fruchterman-Reingold algorithm [58]

    Therefore, embedding a graph corresponding to a square without diagonal connections is especially tricky as the diagonal terms, supposedly not representing edges, will still contribute with strength 1 /8 regardless of the chosen blockade radius. Embedding UD graphs into a plan...

  56. [70]

    F ormalism The Quantum Approximate Optimisation Algorithm (QAOA) [5] is a prominent method within the quantum optimisation paradigm [4, 6], inspired by the Trotterised version of adiabatic evolution. In QAOA, the quantum unitary evolution operator is parameterised by two sets ...

  57. [71]

    We perform a closed loop with p = 1 on the N = 6 graph of Fig

    Closed loop and miscalibration We apply the QAOA-like approach as a way to benchmark both the ability of the QPU to perform quantum dynamics under such protocols and the ability for the Bayesian optimiser to find optimal driving parameters even in the presence of noise. We per...

  58. [1723]

    doi:10.1038/s41467-018-07090-4

  59. [1991]

    URL https://onlinelibrary.wiley.com/doi/abs/ 15 10.1002/spe.4380211102

    doi:https://doi.org/10.1002/spe.4380211102. URL https://onlinelibrary.wiley.com/doi/abs/ 15 10.1002/spe.4380211102

  60. [2014]

    doi:10.48550/ARXIV.1411.4028

  61. [2021]

    doi:10.1140/epjqt/s40507-021- 00100-3

    ISSN 2196-0763. doi:10.1140/epjqt/s40507-021- 00100-3

  62. [2022]

    doi:10.22331/q-2022-09-29- 824

    ISSN 2521-327X. doi:10.22331/q-2022-09-29- 824

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.