Pith. sign in

REVIEW 4 major objections 7 minor 65 references

Shakti artificial spin ice responds to magnetic-field driving in a history-dependent way, while square artificial spin ice does not; even the minimal supercell shows both deterministic and stochastic sequence memory controlled by the coupli

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · deepseek-v4-flash

2026-08-03 13:59 UTC pith:TDKB5C4K

load-bearing objection Square-vs-Shakti memory comparison is new and worth thinking about, but the stochastic-memory claims are not reproducible until the update protocol is specified. the 4 major comments →

arxiv 2512.21767 v1 pith:TDKB5C4K submitted 2025-12-25 cond-mat.stat-mech cond-mat.mtrl-sci

Memory of topologically constrained disorder in Shakti artificial spin ice

Priyanka Priyanka , Cristiano Nisoli , Yair Shokef This is my paper
classification cond-mat.stat-mech cond-mat.mtrl-sci
keywords artificial spin iceShakti latticesequence-dependent memorytopological ordervertex frustrationreturn-point memorymagnetic hysteresisathermal dynamics
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

This paper aims to establish that sequence-dependent memory — a response to external driving that depends on the order in which the drive is applied — can reside in topologically constrained disorder, not only in ordered phases. It contrasts the Landau-ordered square artificial spin ice, which responds predictably and without history dependence, with the Shakti lattice, which for interaction ratio α<2 has a degenerate, topologically ordered ground state and for α>2 an ordered antiferromagnetic one. Using a nearest-neighbour vertex model and athermal quasistatic dynamics, the authors show that even the minimal Shakti supercell gives deterministic sequence memory (selection among degenerate states) and stochastic memory (bifurcations to multiple final states), depending on α and on the field protocol. They further find that returning along the same path used to reach the maximal field preserves memory better than a loop path. A sympathetic reader would care because this suggests a concrete nanoscale system whose final magnetic configuration encodes the history of applied fields, with the coupling ratio α as a control knob.

Core claim

The central claim is that Shakti artificial spin ice shows a sequence-dependent response to in-plane magnetic fields, whereas square artificial spin ice does not: for square ice every protocol ends in the same final magnetization, while in Shakti the outcome depends on the path and history of the field. For α<2 the ground state is a degenerate manifold of 24 supercell configurations with topological order; for α>2 it is the antiferromagnetic state inherited from square ice. In both regimes Shakti exhibits path-dependent final states under unidirectional pulse and bidirectional field protocols, with spin-flip thresholds at H=1, α, 2α−1, 3α−2, and the dynamics can bifurcate into multiple stoch

What carries the argument

The load-bearing object is the Shakti vertex model with nearest-neighbour couplings J∥ and J⊥, whose ratio α=J⊥/J∥ selects the ground state: for α<2, a degenerate, topologically ordered manifold of 24 minimal-supercell configurations (mappable to the F-model or a dimer model); for α>2, the ordered antiferromagnetic state. The argument proceeds by classifying the 24 configurations by symmetry and magnetization, and analytically enumerating single-spin-flip trajectories with thresholds H=1, α, 2α−1, 3α−2 under six field protocols (1D1W, 1D2W, 1DPulse, 2D1W, 2D2W, 2DLoop). The degenerate manifold carries the memory: different spin-flip sequences select different members of the manifold (determi

Load-bearing premise

The entire analysis rests on a nearest-neighbour, athermal, quasistatic vertex model with one interaction ratio α=J⊥/J∥; if long-range dipolar interactions or thermal fluctuations in real Shakti samples reorder the vertex energies, the predicted sequence memory could vanish or change.

What would settle it

Drive a real Shakti array with two unequal unidirectional field pulses (amplitudes H1≠H2) in both orders and image the final states with magnetic force microscopy: the model predicts that for α<2 all 24 supercell ground states show order-dependent final states (stochastic above threshold), while for α>2 at most 8 of 24 do. A second decisive check is to measure the return-memory field threshold versus α and look for the predicted discontinuity at α=2.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • Square artificial spin ice cannot encode drive history, whereas Shakti can, making the latter a candidate platform for nanoscale sequence memory.
  • Below α=2 all 24 supercell ground states show sequence-dependent responses to asymmetric pulses once a threshold is crossed; above α=2 only a subset (up to 8 of 24) do.
  • The return-to-ground-state threshold as a function of α is discontinuous at α=2, marking the transition between the degenerate topological and ordered antiferromagnetic ground states.
  • For larger systems, overlapping supercells partially cancel single-cell differences, but Monte Carlo results show the sequence dependence and defect fractions persist and saturate with system size.
  • Retracing the same field path (2D2W) preserves memory of the initial degenerate state up to larger fields than the loop protocol (2DLoop), which destroys memory by intertwining x- and y-spin flips.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • If real dipolar arrays behave qualitatively like the nearest-neighbour model, the stochastic bifurcations imply that large Shakti samples driven identically may reach different final states run-to-run — usable for stochastic computing or physical unclonable functions, but an obstacle for deterministic memory.
  • The paper's proposed mechanism — that degenerate ground-state manifolds, not ordering, carry sequence memory — suggests a design rule: other vertex-frustrated lattices with dimer-model ground states (e.g., Santa Fe or kagome spin ice) should show analogous path-dependent responses, with defect kinetics setting whether the memory is deterministic or stochastic.
  • The singular behavior at α=2 is a testable experimental knob: tuning the perpendicular-to-parallel coupling ratio in a fabricated sample should sharply alter memory thresholds and fluctuation statistics, providing a clean control parameter for memory strength.
  • A natural extension would be to probe memory of disorder under reversed protocols (negative fields), where the degenerate manifold's Z2 symmetry may produce additional sequence-dependent selection rules that the current unidirectional analysis does not explore.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 7 minor

Summary. The paper investigates sequence-dependent memory in square and Shakti artificial spin ices driven by quasistatic in-plane magnetic field protocols. Within a nearest-neighbor vertex model with coupling ratio α = J⊥/J∥, the authors find that square ice exhibits no path dependence, whereas Shakti shows pronounced sequence-dependent response. For the Shakti lattice, exact enumeration of the L=4 supercell reveals deterministic and stochastic forms of memory depending on whether α is above or below 2, and zero-temperature Monte Carlo simulations for larger systems are used to argue that retracing the same input path (2D2W) enhances memory retention relative to a different return path (2DLoop). The paper is primarily a phenomenological numerical survey, with a central claim that Shakti's degenerate, topologically constrained ground-state manifold generates memory.

Significance. If the central claim holds, the paper identifies a new type of history-dependent response in a frustrated artificial spin ice, with potential implications for programmable metamaterials and memory devices. The exact enumeration of all L=4 supercell configurations is a useful benchmark, and the comparison between square and Shakti lattices is conceptually clean. However, the computational evidence is not fully reproducible: the Monte Carlo algorithm is underspecified, and several quantitative claims lack error bars. The paper also does not provide code or data. These issues are fixable and do not invalidate the core observation, but they currently limit the confidence in the reported stochastic bifurcations and the quantitative system-size trends.

major comments (4)
  1. [Sec. 2.2, Figs. 13 and 14] The zero-temperature single-flip Metropolis dynamics is not fully specified. The text says 'quasistatically' and 'single-flip Monte Carlo using the Metropolis algorithm', but it does not state the update order (random sequential, ordered sweep), the treatment of ΔE=0 moves, the number of sweeps per field step, or the field step size. These details are decisive for the central phenomenology: Fig. 13 shows four different configurations reached from the same state and field, and the paper repeatedly invokes 'stochastic' outcomes and 'bifurcations'. Without a precise stochastic process, branch weights are undefined. Please provide a complete algorithmic specification, and if the stochasticity is intended to represent thermal noise or a protocol convention, state that explicitly.
  2. [Figs. 8, 14, 18] The Monte Carlo results for larger systems are shown without error bars or the number of independent runs. For example, Fig. 14a plots Δ versus system size with no uncertainty, Fig. 14b plots δ with no spread, and Fig. 18 shows fractions of Type II2 defects versus L with no statistical errors. The claims of a singularity at α=2 (Fig. 14 inset) and of monotonic increase for α>2 (Fig. 18b) rest on these unquantified points. Please include error bars or a statistical analysis (e.g., standard error over independent initial conditions and random update orders).
  3. [Sec. 5.2, 5.3, and abstract] The claim that 'retracing the same input path leads to enhanced memory retention' is not defined as a quantitative measure. The text compares 2D2W and 2DLoop, but the figures (Figs. 16, 17) show different outcome categories (initial state, degenerate, coexistence, fluctuation, unique excited) without a single index for 'memory retention'. Is the metric the threshold field to return to the initial state, or to any degenerate state? Fig. 8 displays cutoff fields but without error bars and with ambiguous line styles. Please define the memory measure and present a direct, error-barred comparison between the two protocols.
  4. [Sec. 4.1, Fig. 4] The enumeration and classification of the degenerate ground states are hard to follow. The text says the 24 configurations reduce to 12 modulo the Z2 flip, but then groups (i)–(iv) are described with sub-configurations while complementary configurations are later denoted (i)–(iv) and treated as distinct. For example, 'group (iii) contains four configurations that in an extended lattice are equivalent by translational symmetry' is not reconciled with the reduction by Z2. Please clarify the equivalence relations, the labels, and how the 24 states map onto the 12 reduced states and the four groups.
minor comments (7)
  1. [Eq. (1)] The Hamiltonian uses a cross product \(\vec S_i \times \vec S_j\) for the perpendicular interaction, which is ambiguous for Ising spins in a plane. Please define the operator: presumably it is the dot product of one spin with the perpendicular component of the other, or an equivalent scalar expression.
  2. [Eqs. (3) and (4)] The integrals defining Δ and δ lack explicit limits. Since the thresholds scale with α, the normalization factor 1/(α−1) is introduced, but the integration intervals are not stated. Please specify the field range (e.g., from 0 to saturation) and clarify whether the integrals are over the full protocol.
  3. [Fig. 8 caption] The caption mentions red, blue, green, solid thick, solid thin, and broken lines, but the text refers to 'green line' and 'broken line' as if they are different. Please make the legend unambiguous and consistent with the text.
  4. [Sec. 5.3] Typo: '2D2Loop' in the sentence about Fig. 18 should be '2DLoop'.
  5. [Fig. 19 caption] Typo: 'Sequnece dependent' should be 'Sequence dependent'.
  6. [Sec. 5.1, Fig. 14] The text says 'The dashed lines represent exact values obtained for L=4' but the figure caption and the main text are not clear about which line is dashed and which represents the L=32 average. Please clarify.
  7. [Sec. 6] The phrase 'topologically protected disorder' is used, but the paper does not compute or discuss any topological invariant. Consider rephrasing to 'topologically constrained' or provide the specific sense in which the degenerate manifold is topologically protected.

Circularity Check

0 steps flagged

No significant circularity: central memory claims are computed from the stated Hamiltonian and not fitted; self-citations are background.

full rationale

The paper's central claims—square ice has no sequence-dependent response, Shakti has sequence-dependent response, the α=2 singularity, and the 2D2W vs 2DLoop memory-retention difference—are all generated from the explicit vertex model of Eq. (1) under quasistatic field protocols. No parameter is fitted to the target outcome. The ratio α is scanned as a control parameter, and thresholds such as Hc = 3α−2 are derived from the vertex-energy table (Table 1), not inferred from the claimed response. The L=4 results are exact enumerations of all states, and L=32 results are Monte Carlo simulations of the same model; neither is a fit to the reported memory quantities. Self-citations to Refs. [36,39] (Chern-Morrison-Nisoli; Lao et al., with Nisoli as coauthor) are used to characterize Shakti's ground state as topologically ordered or dimer-like, but this characterization is background framing, not the load-bearing derivation of the memory results. The discussion's mention of Ref. [39] as a 'perhaps' mechanism for stochasticity is explicitly speculative and does not define the observed outcome. A separate correctness concern—the zero-temperature update protocol is not fully specified (random spin-selection order, ΔE=0 acceptance), which affects the reported 'stochastic' bifurcations—is a reproducibility issue, not a circularity: the results are not equivalent to the inputs by construction. Because the core computational findings are self-contained and the self-citations are not load-bearing, the circularity score is low.

Axiom & Free-Parameter Ledger

0 free parameters · 4 axioms · 0 invented entities

The central results are computational and rest on the chosen model and dynamics. No new free parameters are fitted to data; α is a variable material ratio. The paper introduces no new physical entities. The key assumptions are the two-coupling nearest-neighbor model, athermal quasistatic dynamics, and the validity of the generated ground-state configurations.

axioms (4)
  • domain assumption The nearest-neighbor vertex model with couplings J∥ and J⊥ and α = J⊥/J∥ captures the physics of real artificial spin ice (Section 2.1, 'tried and true' approximation).
    Long-range dipolar interactions are omitted; the paper treats this as a model input, not a derived limit.
  • domain assumption Dynamics are athermal and quasistatic: the system always remains in a local energy minimum under single-spin-flip Metropolis updates (Section 2.2).
    Finite temperature and correlated multi-spin avalanches are ignored; this is stated as the chosen protocol, not justified from experiments.
  • domain assumption The ground states for Shakti are known and can be generated by the described defect-ordering algorithm (Section 2.1 and 4.1).
    The paper relies on the earlier mappings to F-model/dimer model (Refs. [36,39]) to define the degenerate ground state, but does not rederive them.
  • standard math The field thresholds and vertex energies are computed from the vertex table (Table 1), assuming the Hamiltonian of Eq. 1.
    This is a direct algebraic consequence of the model, but the model itself is a simplification.

pith-pipeline@v1.3.0-alltime-deepseek · 22430 in / 7480 out tokens · 77543 ms · 2026-08-03T13:59:00.792838+00:00 · methodology

0 comments
read the original abstract

Complex behaviors often sit at a critical threshold between order and disorder. But not all disorder is created equal. Disorder can be trivial or constrained, and correlated disorder can even be topological. Crucially, constrained disorder can harbor memory, leading to non-trivial, sequence-dependent responses to external manipulations. And yet the fascinating subject of "memory of disorder" remains poorly explored, as memory is often associated to the retention of metastable order. In recent years artificial frustrated materials -- in particular arrays of frustrated nanomagnets known as artificial spin ices -- have been employed to study complex disorders and its wealth of exotic behaviors, yet their memory properties have received much less attention. Here, we investigate both analytically and numerically the sequence-dependent responses of two somehow opposite yet related artificial spin ices: the Landau-ordered square spin ice and the disordered but topologically-ordered Shakti spin ice. We find that Shakti exhibits a pronounced sequence-dependent response, whereas in the square lattice, such path dependence is absent. Within Shakti, even the minimal periodic supercell demonstrates both deterministic and stochastic forms of sequence memory, depending on the interaction strength. Extending our study to cyclic driving, we find that retracing the same input path leads to enhanced memory retention. These results open new perspectives on how topological constraints and correlated disorder generate robust memory effects in frustrated artificial materials, hitherto examined mainly in terms of their ground-state kinetics and thermodynamics.

Figures

Figures reproduced from arXiv: 2512.21767 by Cristiano Nisoli, Priyanka Priyanka, Yair Shokef.

Figure 1
Figure 1. Figure 1: a) Dilution of the square lattice (top) yields the Shakti lattice (bottom). [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: The protocols used to drive the artificial spin ice by an in-plane external [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Response of square artificial spin ice to one-way protocols. a) Magnetization at [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Degenerate ground state configurations of the Shakti supercell. Excited [PITH_FULL_IMAGE:figures/full_fig_p011_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Response starting from the degenerate state. Minimum energy states obtained [PITH_FULL_IMAGE:figures/full_fig_p012_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Response starting from the antiferromagnetic state. Minimum energy states [PITH_FULL_IMAGE:figures/full_fig_p014_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Memory retention at the end of the unidirectional two-way protocol 1D2W [PITH_FULL_IMAGE:figures/full_fig_p015_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Memory retention threshold field as a function of the interaction ratio [PITH_FULL_IMAGE:figures/full_fig_p016_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Number of initial configurations with sequence dependence after the [PITH_FULL_IMAGE:figures/full_fig_p017_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Response of the degenerate state to the bidirectional one way protocol. [PITH_FULL_IMAGE:figures/full_fig_p018_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Magnetization as function of the final external field [PITH_FULL_IMAGE:figures/full_fig_p019_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: Magnetization as function of the final external field [PITH_FULL_IMAGE:figures/full_fig_p020_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: The configurational changes during the bidirectional one way protocol along [PITH_FULL_IMAGE:figures/full_fig_p021_13.png] view at source ↗
Figure 14
Figure 14. Figure 14: a) The difference in magnetization between the XY and YX bidirectional one [PITH_FULL_IMAGE:figures/full_fig_p022_14.png] view at source ↗
Figure 15
Figure 15. Figure 15: Configurational changes at the end of the different bidirectional protocols [PITH_FULL_IMAGE:figures/full_fig_p023_15.png] view at source ↗
Figure 16
Figure 16. Figure 16: Memory retention at the end of the bidirectional two-way protocol 2D2W [PITH_FULL_IMAGE:figures/full_fig_p024_16.png] view at source ↗
Figure 17
Figure 17. Figure 17: Memory retention at the end of the bidirectional loop protocol 2DLoop [PITH_FULL_IMAGE:figures/full_fig_p026_17.png] view at source ↗
Figure 18
Figure 18. Figure 18: System size dependence of the fraction of vertices of Type [PITH_FULL_IMAGE:figures/full_fig_p027_18.png] view at source ↗
Figure 19
Figure 19. Figure 19: Summary of results for degenerate and antiferromagnet states. Green boxes [PITH_FULL_IMAGE:figures/full_fig_p029_19.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

65 extracted references · 37 canonical work pages · 1 internal anchor

  1. [1]

    Lindeman C M 2025Soft Matter214890–4897 URLhttps://doi.org/10.1039/D5SM00470E

    work page doi:10.1039/d5sm00470e

  2. [2]

    Mukherji S, Kandula N, Sood A K and Ganapathy R 2019Phys. Rev. Lett.122158001 URL https://doi.org/10.1103/PhysRevLett.122.158001

    work page doi:10.1103/physrevlett.122.158001

  3. [3]

    Shohat D, Hexner D and Lahini Y 2022Proc. Natl. Acad. Sci.119e2200028119 URLhttps: //doi.org/10.1073/pnas.2200028119

    work page doi:10.1073/pnas.2200028119

  4. [4]

    Fiocco D, Foffi G and Sastry S 2014Phys. Rev. Lett.112025702 URLhttps://doi.org/10. 1103/PhysRevLett.112.025702

  5. [5]

    Commun.68805 URL https://doi.org/10.1038/ncomms9805

    Regev I, Weber J, Reichhardt C, Dahmen K A and Lookman T 2015Nat. Commun.68805 URL https://doi.org/10.1038/ncomms9805

    work page doi:10.1038/ncomms9805

  6. [6]

    Mungan M, Sastry S, Dahmen K and Regev I 2019Phys. Rev. Lett.123178002 URLhttps: //doi.org/10.1103/PhysRevLett.123.178002

    work page doi:10.1103/physrevlett.123.178002

  7. [7]

    Keim N C, Hass J, Kroger B and Wieker D 2020Phys. Rev. Research2012004 URLhttps: //doi.org/10.1103/PhysRevResearch.2.012004

    work page doi:10.1103/physrevresearch.2.012004

  8. [8]

    Adv.7eabg7685 URLhttps://doi.org/10.1126/sciadv

    Keim N C and Paulsen J D 2021Sci. Adv.7eabg7685 URLhttps://doi.org/10.1126/sciadv. abg7685 30

    doi:10.1126/sciadv

  9. [9]

    Mungan M, Kumar D, Patinet S and Vandembroucq D 2025Phys. Rev. Lett.134178203 URL https://doi.org/10.1103/PhysRevLett.134.178203

    doi:10.1103/physrevlett.134.178203

  10. [10]

    Jules T, Reid A, Daniels K E, Mungan M and Lechenault F 2022Phys. Rev. Research4013128 URLhttps://doi.org/10.1103/PhysRevResearch.4.013128

    work page doi:10.1103/physrevresearch.4.013128

  11. [11]

    Ding J and van Hecke M 2022J. Chem. Phys.156204902 URLhttps://doi.org/10.1063/5. 0087863

    doi:10.1063/5

  12. [12]

    Commun.15 4008 URLhttps://doi.org/10.1038/s41467-024-47780-w

    Sirote-Katz C, Shohat D, Merrigan C, Lahini Y, Nisoli C and Shokef Y 2024Nat. Commun.15 4008 URLhttps://doi.org/10.1038/s41467-024-47780-w

    work page doi:10.1038/s41467-024-47780-w

  13. [13]

    Liu J, Teunisse M, Korovin G, R V I, Jin L, Bense H and Hecke M 2024Proc. Natl. Acad. Sci. 121e2308414121 URLhttps://doi.org/10.1073/pnas.2308414121

    work page doi:10.1073/pnas.2308414121

  14. [14]

    Meulblok C M, Singh A, Labousse M and Hecke M V 2025arXiv:2503.07764URLhttps: //doi.org/10.48550/arXiv.2503.07764

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.2503.07764

  15. [15]

    Phys.52219–342 URLhttps://doi.org/10.1080/ 0001873031000093582

    Ritort F and Sollich P 2003Adv. Phys.52219–342 URLhttps://doi.org/10.1080/ 0001873031000093582

  16. [16]

    Keim N C, Paulsen J D, Zeravcic Z, Sastry S and Nagel S R 2019Rev. Mod. Phys.91035002 URLhttps://doi.org/10.1103/RevModPhys.91.035002

    work page doi:10.1103/revmodphys.91.035002

  17. [17]

    Mungan M, Clement E, Vandembroucq D and Sastry S 2025arXiv:2510.25367URLhttps: //doi.org/10.48550/arXiv.2510.25367

    work page doi:10.48550/arxiv.2510.25367

  18. [18]

    Yasuda H, Buskohl P R, Gillman A, Murphey T D, Stepney S, Vaia R A and Raney J R 2021 Nature59839–48 URLhttps://doi.org/10.1038/s41586-021-03623-y

    doi:10.1038/s41586-021-03623-y 2021

  19. [19]

    Paulsen J D and Keim N C 2025Annu. Rev. Condens. Matter Phys.1661–81 URLhttps: //doi.org/10.1146/annurev-conmatphys-032822-035544

    work page doi:10.1146/annurev-conmatphys-032822-035544

  20. [21]

    Henley C L 2010Annu. Rev. Condens. Matter Phys.1179–210 URLhttps://doi.org/10.1146/ annurev-conmatphys-070909-104138

  21. [22]

    Matter23164212 URLhttps://doi.org/10.1088/ 0953-8984/23/16/164212

    Henley C L 2011J Phys: Condens. Matter23164212 URLhttps://doi.org/10.1088/ 0953-8984/23/16/164212

  22. [23]

    Lieb E H 1967Phys. Rev. Lett.181046 – 1048 URLhttps://doi.org/10.1103/PhysRevLett. 18.1046

    doi:10.1103/physrevlett

  23. [24]

    Lett.13247005 URLhttps://doi.org/10.1209/0295-5075/132/47005

    Nisoli C 2020Europhys. Lett.13247005 URLhttps://doi.org/10.1209/0295-5075/132/47005

    work page doi:10.1209/0295-5075/132/47005

  24. [25]

    Zhang X, Fitez G, Subzwari S, Bingham N S, Chioar I A, Saglam H, Ramberger J, Leighton C, Nisoli C and Schiffer P 2023Science380526–531 URLhttps://doi.org/10.1126/science. add6575

    doi:10.1126/science

  25. [26]

    1038/nature04447

    Wang R F, Nisoli C, Freitas R S, Li J, McConville W, Cooley B J, Lund M S, Samarth N, Leighton C, Crespi V H and Schiffer P 2006Nature439303–306 URLhttps://doi.org/10. 1038/nature04447

  26. [27]

    Nisoli C, Moessner R and Schiffer P 2013Rev. Mod. Phys.851473–1490 URLhttps://doi.org/ 10.1103/RevModPhys.85.1473

    doi:10.1103/revmodphys.85.1473

  27. [28]

    Phys.13200–203 URLhttps://doi.org/10.1038/ nphys4059

    Nisoli C, Kapaklis V and Schiffer P 2017Nat. Phys.13200–203 URLhttps://doi.org/10.1038/ nphys4059

  28. [29]

    Skjærvø S H, Marrows C H, Stamps R L and Heyderman L J 2020Nat. Rev. Phys.213–28 URL https://doi.org/10.1038/s42254-019-0118-3

    work page doi:10.1038/s42254-019-0118-3

  29. [30]

    Phys.6359–363 URL https://doi.org/10.1038/nphys1628

    Ladak S, Read D E, Perkins G K, Cohen L F and Branford W R 2010Nat. Phys.6359–363 URL https://doi.org/10.1038/nphys1628

    work page doi:10.1038/nphys1628

  30. [31]

    Phys.768–74 URLhttps://doi.org/10.1038/nphys1794

    Mengotti E, Heyderman L J, Rodr ´ ıguez A F, Nolting F, H¨ ugli R V and Braun H B 2011Nat. Phys.768–74 URLhttps://doi.org/10.1038/nphys1794

    doi:10.1038/nphys1794

  31. [32]

    Phys.15045029 URLhttps: //doi.org/10.1088/1367-2630/15/4/045029

    Wysin G M, Moura-Melo W, M´ ol L and Pereira A 2013New J. Phys.15045029 URLhttps: //doi.org/10.1088/1367-2630/15/4/045029

    work page doi:10.1088/1367-2630/15/4/045029

  32. [33]

    Lib´ al A, Reichhardt C and Olson Reichhardt C 2012Phys. Rev. E86021406 URLhttps: //doi.org/10.1103/PhysRevE.86.021406 31

    work page doi:10.1103/physreve.86.021406

  33. [34]

    Gilbert I, Chern G W, Fore B, Lao Y, Zhang S, Nisoli C and Schiffer P 2015Phys. Rev. B92 104417 URLhttps://doi.org/10.1103/PhysRevB.92.104417

    work page doi:10.1103/physrevb.92.104417

  34. [35]

    Nisoli C, Wang R, Li J, McConville W F, Lammert P E, Schiffer P and Crespi V H 2007Phys. Rev. Lett.98217203 URLhttps://doi.org/10.1103/PhysRevLett.98.217203

    work page doi:10.1103/physrevlett.98.217203

  35. [36]

    Chern G W, Morrison M J and Nisoli C 2013Phys. Rev. Lett.111177201 URLhttps: //doi.org/10.1103/PhysRevLett.111.177201

    work page doi:10.1103/physrevlett.111.177201

  36. [37]

    Phys.15045009 URLhttps://doi.org/ 10.1088/1367-2630/15/4/045009

    Morrison M J, Nelson T R and Nisoli C 2013New J. Phys.15045009 URLhttps://doi.org/ 10.1088/1367-2630/15/4/045009

    work page doi:10.1088/1367-2630/15/4/045009

  37. [38]

    Phys.10 670–675 URLhttps://doi.org/10.1038/nphys3037

    Gilbert I, Chern G W, Zhang S, O’Brien L, Fore B, Nisoli C and Schiffer P 2014Nat. Phys.10 670–675 URLhttps://doi.org/10.1038/nphys3037

    work page doi:10.1038/nphys3037

  38. [39]

    Phys.14723–727 URLhttps://doi.org/10

    Lao Y, Caravelli F, Sheikh M, Sklenar J, Gardeazabal D, Watts J D, Albrecht A M, Scholl A, Dahmen K, Nisoli C and Schiffer P 2018Nat. Phys.14723–727 URLhttps://doi.org/10. 1038/s41567-018-0077-0

  39. [40]

    Stopfel H, ¨Ostman E, Chioar I A, Greving D, Arnalds U B, Hase T P A, Stein A, Hj¨ orvarsson B and Kapaklis V 2018Phys. Rev. B98014435 URLhttps://doi.org/10.1103/PhysRevB.98. 014435

    doi:10.1103/physrevb.98

  40. [41]

    Phys.6113 URL https://doi.org/10.1038/s42005-023-01236-7

    Rodr ´ ıguez-Gallo C, Ortiz-Ambriz A, Nisoli C and Tierno P 2023Comm. Phys.6113 URL https://doi.org/10.1038/s42005-023-01236-7

    doi:10.1038/s42005-023-01236-7

  41. [42]

    Phys.10623–624 URLhttps://doi.org/10.1038/nphys3072

    Stamps R L 2014Nat. Phys.10623–624 URLhttps://doi.org/10.1038/nphys3072

    work page doi:10.1038/nphys3072

  42. [43]

    Phys.12162–165 URLhttps://doi.org/10.1038/nphys3520

    Gilbert I, Lao Y, Carrasquillo I, O’Brien L, Watts J D, Manno M, Leighton C, Scholl A, Nisoli C and Schiffer P 2016Nat. Phys.12162–165 URLhttps://doi.org/10.1038/nphys3520

    work page doi:10.1038/nphys3520

  43. [44]

    Today6954–59 URLhttps://doi.org/10.1063/ PT.3.3237

    Gilbert I, Nisoli C and Schiffer P 2016Phys. Today6954–59 URLhttps://doi.org/10.1063/ PT.3.3237

  44. [45]

    Sultana R, Mondal A K, Bhat V S, Stenning K, Li Y, Arroo D M, Vasdev A, McCarter M R, De Long L E, Hastings J T, Gartside J C and Jungfleisch M B 2025J. Appl. Phys.138061101 URLhttps://doi.org/10.1063/5.0274799

    work page doi:10.1063/5.0274799

  45. [46]

    Nisoli C, Li J, Ke X, Garand D, Schiffer P and Crespi V H 2010Phys. Rev. Lett.105047205 URL https://doi.org/10.1103/PhysRevLett.105.047205

    work page doi:10.1103/physrevlett.105.047205

  46. [47]

    Phys.775–79 URLhttps: //doi.org/10.1038/nphys1853

    Morgan J P, Stein A, Langridge S and Marrows C H 2011Nat. Phys.775–79 URLhttps: //doi.org/10.1038/nphys1853

    work page doi:10.1038/nphys1853

  47. [48]

    Zhang S, Gilbert I, Nisoli C, Chern G W, Erickson M J, O’brien L, Leighton C, Lammert P E, Crespi V H and Schiffer P 2013Nature500553–557 URLhttps://doi.org/10.1038/ nature12399

  48. [49]

    Phys.15055012 URL https://doi.org/10.1088/1367-2630/15/5/055012

    Porro J, Bedoya-Pinto A, Berger A and Vavassori P 2013New J. Phys.15055012 URL https://doi.org/10.1088/1367-2630/15/5/055012

    work page doi:10.1088/1367-2630/15/5/055012

  49. [50]

    Sendetskyi O, Scagnoli V, Leo N, Anghinolfi L, Alberca A, L¨ uning J, Staub U, Derlet P M and Heyderman L J 2019Phys. Rev. B99214430 URLhttps://doi.org/10.1103/PhysRevB.99. 214430

    doi:10.1103/physrevb.99

  50. [51]

    Coulais C, Teomy E, de Reus K, Shokef Y and van Hecke M 2016Nature535529 URL https://doi.org/10.1038/nature18960

    doi:10.1038/nature18960

  51. [52]

    Phys.22023030 URL https://doi.org/10.1088/1367-2630/ab69b5

    Meeussen A S, O˘ guz E C, van Hecke M and Shokef Y 2020New J. Phys.22023030 URL https://doi.org/10.1088/1367-2630/ab69b5

    work page doi:10.1088/1367-2630/ab69b5

  52. [53]

    Pisanty B, O˘ guz E C, Nisoli C and Shokef Y 2021SciPost Phys.10136 URLhttps://doi.org/ 10.21468/SciPostPhys.10.6.136

    work page doi:10.21468/scipostphys.10.6.136

  53. [54]

    Merrigan C, Nisoli C and Shokef Y 2021Phys. Rev. Res.3023174 URLhttps://doi.org/10. 1103/PhysRevResearch.3.023174

  54. [55]

    Branford W, Ladak S, Read D E, Zeissler K and Cohen L 2012Science3351597–1600 URL https://doi.org/10.1126/science.1211379

    work page doi:10.1126/science.1211379

  55. [56]

    Chern G W 2017Phys. Rev. Appl.8064006 URLhttps://doi.org/10.1103/PhysRevApplied. 8.064006

    doi:10.1103/physrevapplied

  56. [57]

    Le B, Park J, Sklenar J, Chern G W, Nisoli C, Watts J, Manno M, Rench D W, Samarth N, 32 Leighton C and Schiffer P 2017Phys. Rev. B95060405 URLhttps://doi.org/10.1103/ PhysRevB.95.060405

  57. [58]

    Phys.24023020 URLhttps://doi.org/10

    Caravelli F, Chern G W and Nisoli C 2022New J. Phys.24023020 URLhttps://doi.org/10. 1088/1367-2630/ac4c0a

  58. [59]

    Rev.162162 URLhttps://doi.org/10.1103/PhysRev.162.162

    Lieb E H 1967Phys. Rev.162162 URLhttps://doi.org/10.1103/PhysRev.162.162

    doi:10.1103/physrev.162.162

  59. [60]

    M¨ oller G and Moessner R 2009Phys. Rev. B80140409 URLhttps://doi.org/10.1103/ PhysRevB.80.140409

  60. [61]

    1038/nature20155

    Perrin Y, Canals B and Rougemaille N 2016Nature540410–413 URLhttps://doi.org/10. 1038/nature20155

  61. [62]

    Sch´ anilec V, Brunn O, Hor´ aˇ cek M, Kr´ atk` y S, Meluz ´ ın P,ˇSikola T, Canals B and Rougemaille N 2022Phys. Rev. Lett.129027202 URLhttps://doi.org/10.1103/PhysRevLett.129.027202

    work page doi:10.1103/physrevlett.129.027202

  62. [63]

    Phys.14035009 URL https://doi.org/10.1088/1367-2630/14/3/035009

    Kapaklis V, Arnalds U B, Harman-Clarke A, Papaioannou E T, Karimipour M, Korelis P, Taroni A, Holdsworth P C W, Bramwell S T and Hj¨ orvarsson B 2012New J. Phys.14035009 URL https://doi.org/10.1088/1367-2630/14/3/035009

    work page doi:10.1088/1367-2630/14/3/035009

  63. [64]

    Commun.126514 URLhttps://doi.org/10.1038/s41467-021-26734-6

    Zhang X, Duzgun A, Lao Y, Subzwari S, Bingham N S, Sklenar J, Saglam H, Ramberger J, Batley J T, Watts J D, Bromley D, Chopdekar R V, O’Brien L, Leighton C, Nisoli C and Schiffer P 2021Nat. Commun.126514 URLhttps://doi.org/10.1038/s41467-021-26734-6

    work page doi:10.1038/s41467-021-26734-6

  64. [65]

    Chern G W, Mellado P and Tchernyshyov O 2011Phys. Rev. Lett.106207202 URLhttps: //doi.org/10.1103/PhysRevLett.106.207202

    work page doi:10.1103/physrevlett.106.207202

  65. [66]

    O˘ guz E C, Ortiz-Ambriz A, Shem-Tov H, Babi` a-Soler E, Tierno P and Shokef Y 2020Phys. Rev. Lett.124238003 URLhttps://doi.org/10.1103/PhysRevLett.124.238003 33

    work page doi:10.1103/physrevlett.124.238003