Pith. sign in

REVIEW 4 major objections 4 minor 65 references

A small fraction of antiferromagnetic bonds is enough to drive the spin stiffness of a two-dimensional Heisenberg ferromagnet to zero at long wavelengths, while the sample remains magnetized.

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

2026-08-01 10:59 UTC pith:WQZ5VDDW

load-bearing objection A clever, honestly written mechanism paper whose central numerical input is too fragile to carry the conclusion. the 4 major comments →

arxiv 2607.20036 v1 pith:WQZ5VDDW submitted 2026-07-22 cond-mat.dis-nn

Vanishing spin stiffness in weakly disordered two-dimensional Heisenberg ferromagnets

Jacopo Niedda , Aldo Coraggio , Giacomo Bracci-Testasecca , Antonello Scardicchio This is my paper
classification cond-mat.dis-nn MSC 82D3082B2882B44
keywords spin stiffnessdisordered Heisenberg ferromagnetantiferromagnetic bondsreplica field theoryrenormalization groupdynamical exponentmagnon density of statesrandom stiffness
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.

The paper establishes that weak bond disorder—a small density p of antiferromagnetic bonds—makes the spin stiffness of a 2D Heisenberg ferromagnet vanish in the thermodynamic limit, even though the system remains magnetized. The mechanism is the emergence of logarithmically correlated spatial fluctuations of the local spin stiffness from short-ranged, uncorrelated microscopic disorder; this log-correlated field is exactly marginal in two dimensions. A one-loop replica renormalization-group calculation shows that the stiffness decreases and the disorder strength grows without bound, producing a scale-dependent dynamical exponent z > 2 and anomalously soft magnons with a singular density of states. Numerical diagonalization of the semiclassical spin-wave Hamiltonian supports the anomalous low-energy scaling. A sympathetic reader would care because it overturns the standard expectation that weak bond disorder merely renormalizes the Goldstone-mode stiffness without changing the hydrodynamic form of the theory.

Core claim

The central claim is that weak bond disorder—a fraction p of antiferromagnetic bonds—promotes the local spin stiffness ρ_s(x) to a quenched random field with logarithmic spatial correlations C(r) ≈ -(Δ/2π) ln(μr), even though the microscopic disorder is short-ranged. Because such log-correlated disorder is exactly marginal in two dimensions, a one-loop replica RG calculation gives dρ_s/dℓ = -Δ/(4πρ_s) and dΔ/dℓ = Δ²/(2πρ_s²): the stiffness decreases and the disorder grows indefinitely, driving the theory to an infinite-disorder, zero-stiffness regime. The magnon dispersion becomes ω(k) ~ k^{2+λ(ℓ)} with λ = Δ/(4πρ_s²), a scale-dependent dynamical exponent z = 2 + λ > 2, and a singular densit

What carries the argument

The key machinery is the local spin stiffness ρ_s(x) as a random field with covariance C(r) ≈ -(Δ/2π) ln(μr), measured via a local-twist protocol on classical ground states. In momentum space the disorder propagator is 1/(q²+μ²), so the disorder-magnon interaction is marginal in d=2. The one-loop RG in the conserved combination ρ_s²Δ = const reduces the flow to a single running coupling λ = Δ/(4πρ_s²) satisfying dλ/dℓ = 4λ²; the resulting Landau pole sets the length scale L* and gives the running exponent z(ℓ) = 2 + λ(ℓ). The argument is carried by the interplay between the dimensionless stiffness and the emergent logarithmic correlations—the double marginality.

Load-bearing premise

The load-bearing premise is that the numerically fitted stiffness covariance is exactly logarithmic with a Gaussian distribution over all relevant scales; if screening or higher-order correlations cut off or alter this log form, the marginality argument and the flow to zero stiffness do not hold.

What would settle it

Compute the local-stiffness correlation C(r) on larger lattices and for r/L beyond the fitted window (log(r/L) > -1.5): if C(r) flattens or decays faster than logarithmically, the marginality is lost. Alternatively, compute the fourth cumulant of ρ_s(x); if it is not small compared to the second, the Gaussian replica calculation is invalid. A practical experimental check: measure the magnon density of states in a weakly frustrated 2D magnet; a singular low-energy enhancement ϱ(ω) ~ ω^{-λ/(2+λ)} would support the claim, while a flat or gapped DOS would contradict it.

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

Share X Bluesky LinkedIn Reddit HN

If this is right

  • For any arbitrarily small p, the thermodynamic-limit ferromagnet has zero spin stiffness but a finite order parameter—a state with no rigidity, analogous to the loss of superfluidity in disordered films.
  • Magnons become anomalously soft, with dispersion ω ~ k^{2+λ} and a singular density of states ϱ(ω) ~ ω^{d/z-1}, which should be visible in low-energy spectroscopy of 2D magnetic materials and quantum simulators.
  • The flow reaches an infinite-disorder fixed point; the one-loop length scale L* ~ exp(1/4λ_0) is astronomically large for small p, so for realistic samples the effect shows up as a small, measurable increase in the dynamical exponent.
  • The mechanism is generic: any two-dimensional system with a continuous order parameter and dilute local frustration should show the same vanishing rigidity.
  • The droplet-stiffness connection yields a negative droplet exponent θ_st = 2 - z < 0, linking the result to spin-glass phenomenology.

Where Pith is reading between the lines

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

  • A testable extension: measure the local-twist stiffness correlations in the 2D XY model with dilute frustration; if the log-correlation is generic, the same zero-stiffness flow should appear there.
  • The Gaussian assumption for δρ is fitted to the two-point function; checking the fourth cumulant of the stiffness field would tell whether the replica field theory needs non-Gaussian corrections that could change the beta functions.
  • The End Matter's dipole screening caveat suggests magnetization might be protected beyond L_p ~ e^{c/p²}; if so, the vanishing-stiffness phenomenon could persist on intermediate scales even in a strictly magnetized thermodynamic state—a scenario that could be tested by including dipolar interactions in ground-state calculations.
  • If the zero-stiffness regime is robust, it implies that the Heisenberg ferromagnet at p→0 is not a conventional ferromagnet in the thermodynamic limit, reshaping the phase diagram of disordered 2D magnets; a finite-temperature extension could reveal a new type of Griffiths regime.

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 / 4 minor

Summary. The paper studies the classical and semiclassical 2D Heisenberg ferromagnet with a small concentration p of antiferromagnetic bonds. It claims that, although the ground state remains magnetized, the local spin stiffness becomes a quenched random field with logarithmically correlated spatial fluctuations. Using this as input, a replica field theory with one-loop RG (Eqs. 12–15) predicts that the stiffness flows to zero while the effective disorder strength grows, giving a scale-dependent dynamical exponent z = 2 + λ > 2 and a singular magnon density of states. The authors support the prediction with numerical diagonalization of the spin-wave Hamiltonian and with a local-twist protocol that measures stiffness correlations. The central quantitative comparison is explicitly acknowledged to be premature, since for p = 0.03 the RG flow does not grow over accessible sizes. The paper also discusses connections to droplet theory and to Chakravarty's work.

Significance. If the central claim were established, it would overturn the common expectation that weak bond disorder merely renormalizes the spin stiffness in a 2D ferromagnet, and it would provide a concrete mechanism for the numerically observed z > 2 anomalous magnon scaling. The paper contains a clean one-loop RG calculation with explicit diagrams and integrals in the End Matter, and it makes falsifiable predictions (stiffness vanishing, z > 2, singular density of states) that could be tested experimentally or in quantum simulators. However, the result rests on an empirical input, Eq. (4), that is fitted over a narrow range without error bars or a derivation from the microscopic model; the paper itself admits that the one-loop RG does not quantitatively reproduce the motivating numerics. The conceptual framework is interesting and the RG computation is a useful contribution, but the load-bearing premise is currently not sufficiently supported.

major comments (4)
  1. [Spin-stiffness spatial fluctuations, Eq. (4) and Fig. 2] The entire RG flow to zero stiffness relies on the logarithmic form of C(r). The fit in Fig. 2 covers only log(r/L) between about −3.5 and −1.5, less than 1.5 decades, with no error bars or goodness-of-fit. Over such a short interval, a power law C(r) ~ r^{-α} with α ≃ 0.2–0.4 is statistically difficult to exclude, and the paper itself states in ref. [38] that any α > 0 makes the disorder irrelevant, so the flow to zero stiffness would not occur. Moreover, the measurement uses heavily overlapping 10×10 patches that share spins and boundary conditions, which can artificially induce long-range correlations. The authors should provide error bars, test alternative functional forms, and correct or justify the overlapping-patch protocol.
  2. [One-loop RG flow and Discussion, Eq. (13)–(15)] The claimed numerical confirmation is not supported by the RG as presented. For p = 0.03 the paper states λ₀ = z(0) − 2 ≃ 0.0028 and L* ∼ 10³⁸, so λ remains essentially unchanged for L ≤ 90, whereas Fig. 1 shows a much larger and p-dependent z − 2, with a plateau around 0.3. The text admits that a quantitative comparison requires going beyond one loop or identifying other sources of correlated disorder. This is a load-bearing gap: the abstract and introduction say numerical diagonalization confirms the anomalous scaling, but the one-loop theory does not reproduce the measured exponent or its size dependence. The authors should either soften the confirmation claim or provide a quantitative mechanism for the observed z − 2.
  3. [End Matter, Dipolar deformations, Eq. (20)] The End Matter computation derives a logarithmic divergence of |m|², i.e. of the squared magnetization fluctuations, not of the spin-stiffness correlator C(r) in Eq. (4). The connection between long-ranged magnetization deformations and logarithmic stiffness correlations is asserted, not derived. Since Eq. (4) is the central input to the replica calculation, this is not a minor omission. The authors should either derive Eq. (4) from the microscopic dipole problem or state clearly that it is an independent phenomenological assumption supported only by the limited numerical fit.
  4. [Effective field theory, Eqs. (5)–(6)] The replica treatment assumes the quenched stiffness disorder is exactly Gaussian. The paper provides no test of Gaussianity of δρ(x), e.g. no measurement of higher cumulants such as the four-point correlation. If the stiffness field is non-Gaussian, the replica action and the one-loop flow could receive additional contributions from higher cumulants that are marginal or relevant. At minimum, the authors should present numerical evidence for Gaussian statistics or bound the size of higher cumulants on the scales used for the fit.
minor comments (4)
  1. [Fig. 1 inset] The inset axes and caption are difficult to read; the meaning of the symbols (n = 1...7) and the inferred exponent should be stated more clearly, and error bars on the fitted z values should be shown.
  2. [Spin-stiffness spatial fluctuations] The text says '(L/2)² overlapping 10×10 patches' but does not specify the total number of disorder realizations or how the average in C(r) is taken over patches and samples. This information is needed to assess the statistical significance of the fit.
  3. [One-loop RG, Eq. (27)–(29)] The regularization scheme is a sharp UV cutoff Λ, but the text also uses μ ∼ L^{-1} as the IR regulator. The relation between the two cutoffs and the identification ℓ = ln(L/L₀) should be made explicit.
  4. [Discussion] The phrase 'for historical reasons' in the last paragraph is unclear; the sentence should be rephrased. Also, the reference to Ref. [11] as 'Science 392, 624 (2026)' should be checked for completeness.

Circularity Check

1 steps flagged

One definitional identification (λ0 = z(0)−2) makes the advertised 'explanation' of z>2 partly built-in; the replica RG itself is otherwise a self-contained calculation from the fitted log-correlated input.

specific steps
  1. self definitional [Eqs. (13)-(15) and following paragraph, p.5]
    "The identification of the running coupling λ with the measured z−2 would seem to allow more than a qualitative comparison with the one-loop RG. However, this is premature. For p=0.03, the microscopic (10×10 plaquette) value of λ0=z(0)−2≃0.0028 starts very small and does not significantly grow for L≤90 since L≪L∗∼10^38. So, while the RG explains z>2 at one loop, a quantitative comparison requires probably going beyond one loop"

    In Eq.13 the running coupling λ is defined as −d lnρs/dℓ ≡ Δ/(4πρs²), and Eq.15 defines z(ℓ)=2+λ(ℓ). The paragraph then sets the microscopic value λ0 equal to z(0)−2. Since z(0) is not an independently measured input at the 10×10 plaquette scale (the SW fits in Fig.1 are for L=40,60,90), the initial z>2 is the fitted combination Δ/(4πρs²) renamed as an exponent. The one-loop flow dλ/dℓ=4λ² is so slow (L*∼10^38) that, as the paper concedes, it does not generate the measured z−2≃0.3 plateau at L≤90. Thus the advertised 'explanation' of the anomalous exponent partially reduces to the definitional identification λ0=z(0)−2, with the quantitative content deferred to higher loops.

full rationale

The derivation chain is largely self-contained once Eq.4 is accepted: the replicated field theory (Eqs.5-11), one-loop diagrams and RG equations (Eqs.12-13) are presented explicitly, and the flow ρs→0 follows from those equations. The log-correlated stiffness correlation is a numerical input (fitted over log(r/L)∈[-3.5,-1.5], with no error bars or power-law alternative), not derived from the dipole argument, which yields only |m|² log divergence (Eq.20). That is a robustness weakness, but an input being fitted is not by itself circularity. The same applies to self-citations [20,21]: they supply the motivating z>2 observation and ground-state methods, but the paper reproduces the anomaly in Fig.1, so they are not the load-bearing proof. The one concrete by-construction step is the identification λ0=z(0)−2 after defining z=2+λ; the paper also concedes quantitative comparison is premature and the one-loop growth is negligible at accessible sizes. The End Matter admissions (neglect of dipole screening, no two-loop calculation) further weaken the quantitative status but are stated limitations, not circular reductions. Overall: minor but real definitional circularity in the comparison, with independent RG content.

Axiom & Free-Parameter Ledger

3 free parameters · 6 axioms · 0 invented entities

The central RG flow (12) is driven by two fitted numbers (Δ, ρ_s) and an assumed Gaussian log-correlated disorder. The microscopic origin of the log correlations is supported only by a heuristic dipole argument. The classical FM background is itself debated. This is a moderate load of assumptions for a Letter-length claim.

free parameters (3)
  • Δ (amplitude of logarithmic stiffness correlations) = 0.2 (p=0.03), ~0.3 (p=0.05)
    Fitted from the slope of C(r) vs ln(r/L) in Fig. 2 right; sets bare disorder strength g²=Δ in the RG (Eqs. 11-12). Central because λ0=Δ/(4πρ_s²) determines the Landau pole scale.
  • ρ_s (mean local spin stiffness) = 2.4 (p=0.03), 2.1 (p=0.05)
    Measured via the twist protocol; enters the RG equations (12) and the identification z-2 = Δ/(4πρ_s²).
  • μ (infrared regulator) = ~1/L
    Chosen as 1/L to match the numerical log fit (4); cutoff scale in the RG; not independently determined.
axioms (6)
  • ad hoc to paper The quenched stiffness disorder is Gaussian with covariance C(r) = -(Δ/2π) ln(μr) (Eq. 4-6).
    Assumed without derivation from the microscopic model and without checking higher cumulants; the log form is fitted over a limited scale range (Fig. 2).
  • domain assumption The classical ground state at p≪1/2 is globally magnetized and supports a small-canting spin-wave expansion.
    The paper acknowledges a Villain-type argument that FM order may be unstable at any p>0; it relies on domain sizes L_p ~ e^{c/p²} being effectively infinite (End Matter, Eq. 21).
  • domain assumption One-loop perturbative RG with UV cutoff captures the IR flow even into strong disorder.
    The flow to infinite disorder is obtained from one-loop beta functions (12); two-loop terms (acknowledged incomplete in Acknowledgments) and non-perturbative effects are not controlled.
  • standard math Replica trick n→0 is valid for this quenched average.
    Standard replicas; the replicated action is given in Eq. (10).
  • ad hoc to paper Stiffness fluctuations inherit the log correlations of the dipolar magnetization deformations; no explicit derivation of Eq. (4) from Eq. (20).
    End Matter derives |m|² ~ 2πµ²p² ln(ℓ/a) but does not derive C(r); the text 'attributes' the log stiffness correlations to these deformations (p.2, p.6).
  • domain assumption The quadratic magnon (large-s) Hamiltonian is sufficient; magnon interactions (1/s corrections) are neglected.
    The authors state results are confined to the quadratic Hamiltonian and are valid for quantum and classical models except for s=1/2 interaction corrections (Discussion).

pith-pipeline@v1.3.0-alltime-deepseek · 11284 in / 18767 out tokens · 173124 ms · 2026-08-01T10:59:52.516513+00:00 · methodology

0 comments
Cite this review

Pith. "Pith review of Vanishing spin stiffness in weakly disordered two-dimensional Heisenberg ferromagnets." pith.science (2026). https://pith.science/paper/WQZ5VDDW

@misc{pith2026260720036,
  author       = {Pith},
  title        = {Pith review of: Vanishing spin stiffness in weakly disordered two-dimensional Heisenberg ferromagnets},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WQZ5VDDW}},
  note         = {Machine review of arXiv:2607.20036}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

We show that a small fraction of antiferromagnetic bonds qualitatively alters the long-wavelength dynamics of two-dimensional Heisenberg ferromagnets. Although the classical ground state remains magnetized, weak bond frustration generates logarithmically correlated spatial fluctuations of the local spin stiffness, despite the microscopic disorder being short ranged. A replica field theory calculation shows that the effective disorder strength grows under coarse graining, while the spin stiffness decreases, yielding anomalously soft magnons with a scale-dependent dynamical exponent $z > 2$. Numerical diagonalization of the semiclassical spin-wave Hamiltonian confirms the anomalous low-energy scaling. The flow is toward an infinite-disorder, zero stiffness regime.

Figures

Figures reproduced from arXiv: 2607.20036 by Aldo Coraggio, Antonello Scardicchio, Giacomo Bracci-Testasecca, Jacopo Niedda.

Figure 1
Figure 1. Figure 1: FIG. 1: Increase of the dynamical exponent [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2 [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3: The two relevant diagrams for the one-loop [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4: One-loop RG flux lines. One can recognize the [PITH_FULL_IMAGE:figures/full_fig_p004_4.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 · 1 canonical work pages

  1. [1]

    S. Yan, D. A. Huse, and S. R. White, Science332, 1173–1176 (2011)

    2011

  2. [2]

    Kimchi, A

    I. Kimchi, A. Nahum, and T. Senthil, Physical Review X 8, 10.1103/physrevx.8.031028 (2018)

    doi:10.1103/physrevx.8.031028 2018

  3. [3]

    L. Liu, H. Shao, Y.-C. Lin, W. Guo, and A. W. Sandvik, Phys. Rev. X8, 041040 (2018)

    2018

  4. [4]

    S. Dey, E. C. Andrade, and M. Vojta, Phys. Rev. B101, 020411(R) (2020)

    2020

  5. [5]

    Gonget al., Nature546, 265 (2017)

    C. Gonget al., Nature546, 265 (2017)

    2017

  6. [6]

    Huanget al., Nature546, 270 (2017)

    B. Huanget al., Nature546, 270 (2017)

    2017

  7. [7]

    Bedoya-Pintoet al., Science374, 616 (2021)

    A. Bedoya-Pintoet al., Science374, 616 (2021)

    2021

  8. [8]

    Cenkeret al., Nature Physics17, 20 (2021)

    J. Cenkeret al., Nature Physics17, 20 (2021)

    2021

  9. [9]

    Jenkins, L

    S. Jenkins, L. R´ ozsa, U. Atxitia, R. F. L. Evans, K. S. Novoselov, and E. J. G. Santos, Nature Communications 13, 6917 (2022)

    2022

  10. [10]

    Y. J. Baeet al., Nature609, 282 (2022)

    2022

  11. [11]

    Palet al., Science392, 624 (2026)

    B. Palet al., Science392, 624 (2026)

    2026

  12. [12]

    Ebadi, T

    S. Ebadi, T. T. Wang, H. Levine, A. Keesling, G. Se- meghini, A. Omran, D. Bluvstein, R. Samajdar, H. Pich- ler, W. W. Ho,et al., Nature595, 227 (2021)

    2021

  13. [13]

    Scholl, M

    P. Scholl, M. Schuler, H. J. Williams, A. A. Eberharter, D. Barredo, K.-N. Schymik, V. Lienhard, L.-P. Henry, T. C. Lang, T. Lahaye,et al., Nature595, 233 (2021)

    2021

  14. [14]

    C. Chen, G. Bornet, M. Bintz, G. Emperauger, L. Leclerc, V. S. Liu, P. Scholl, D. Barredo, J. Hauschild, S. Chatterjee,et al., Nature616, 691 (2023)

    2023

  15. [15]

    In the ground state, the magnetization field components must satisfy the Euler-Lagrange equations: ∇2mα = 0, α={x, y}, in any region without defects

    for the local magnetization field⃗ m(⃗ r): H=J X ij (⃗ mi −⃗ mj)2 ∼ Z d2r X α={x,y,z} (∇mα)2 .(16) We choose the reference frame such that in absence of defectsm z i = 1, and so we assume that a few de- fects producesmallfluctuations in the orthogonal plane: |mx,y i | ≪1. In the ground state, the magnetization field components must satisfy the Euler-Lagra...

  16. [16]

    Auerbach,Interacting Electrons and Quantum Mag- netism(Springer, New York, NY, 1994)

    A. Auerbach,Interacting Electrons and Quantum Mag- netism(Springer, New York, NY, 1994)

    1994

  17. [17]

    M. Fava, J. L. Jacobsen, and A. Nahum, Proceedings of the National Academy of Sciences121, e2401292121 (2024), https://www.pnas.org/doi/pdf/10.1073/pnas.2401292121

    work page doi:10.1073/pnas.2401292121 2024

  18. [18]

    S. Li, H. Shao, and A. W. Sandvik, Phys. Rev. Lett.134, 086501 (2025)

    2025

  19. [19]

    B. I. Halperin and P. C. Hohenberg, Phys. Rev.188, 898 (1969)

    1969

  20. [20]

    B. I. Halperin and W. M. Saslow, Phys. Rev. B16, 2154 (1977)

    1977

  21. [21]

    L. L. Viteritti, R. Rende, G. Bracci-Testasecca, J. Niedda, R. Moessner, G. Carleo, and A. Scardicchio, arXiv:2507.05073 (2024)

    Pith/arXiv arXiv 2024

  22. [22]

    Bracci-Testasecca, J

    G. Bracci-Testasecca, J. Niedda, A. Coraggio, R. Moess- ner, and A. Scardicchio, Semiclassical picture of the heisenberg spin glass in two dimensions: from weak localization to hydrodynamics (2026), arXiv:2603.22077 [cond-mat.dis-nn]

    Pith/arXiv arXiv 2026

  23. [23]

    Villain, Journal of Physics C: Solid State Physics10, 4793 (1977)

    J. Villain, Journal of Physics C: Solid State Physics10, 4793 (1977)

    1977

  24. [24]

    G. N. Parker and W. M. Saslow, Physical review. B, Con- densed matter38 16, 11718 (1988)

    1988

  25. [25]

    Vannimenus, S

    J. Vannimenus, S. Kirkpatrick, F. D. M. Haldane, and C. Jayaprakash, Phys. Rev. B39, 4634 (1989)

    1989

  26. [26]

    Giamarchi and P

    T. Giamarchi and P. Le Doussal, Phys. Rev. B52, 1242 (1995)

    1995

  27. [27]

    Balluffi,Introduction to Elasticity Theory for Crystal Defects(2012)

    R. Balluffi,Introduction to Elasticity Theory for Crystal Defects(2012)

    2012

  28. [28]

    Sutton,Physics of Elasticity and Crystal Defects (2020)

    A. Sutton,Physics of Elasticity and Crystal Defects (2020)

    2020

  29. [29]

    Oitmaa and O

    J. Oitmaa and O. P. Sushkov, Physical review letters87 16, 167206 (2001)

    2001

  30. [30]

    Arrachea and M

    L. Arrachea and M. J. Rozenberg, Phys. Rev. Lett.86, 5172 (2001)

    2001

  31. [31]

    Baity-Jesi and G

    M. Baity-Jesi and G. Parisi, Phys. Rev. B91, 134203 (2015)

    2015

  32. [32]

    Baity-Jesi, V

    M. Baity-Jesi, V. Mart ´ ın-Mayor, G. Parisi, and S. Perez- Gaviro, Phys. Rev. Lett.115, 267205 (2015)

    2015

  33. [33]

    N. N. Bogoljubov, Journal of Physics (USSR)11, 23 (1947)

    1947

  34. [34]

    Colpa, Physica A: Statistical Mechanics and its Appli- cations93, 327 (1978)

    J. Colpa, Physica A: Statistical Mechanics and its Appli- cations93, 327 (1978)

    1978

  35. [35]

    Blaizot and G

    J.-P. Blaizot and G. Ripka,Quantum Theory of Finite Systems(MIT Press, Cambridge, Massachusetts, 1986)

    1986

  36. [36]

    For a small twist angleθ, we write F(θ)−F(0) =ρ sθ2 +O(θ 4)

    Theglobalspin stiffness, also referred to as the helic- ity modulus [63, 64], is usually measured from the free- energy cost associated with an imposed twist of bound- ary conditions. For a small twist angleθ, we write F(θ)−F(0) =ρ sθ2 +O(θ 4). AtT= 0, the free energy coincides with the energyF(θ) =E(θ)

  37. [37]

    M´ ezard, G

    M. M´ ezard, G. Parisi, and M. A. Virasoro,Spin glass the- ory and beyond: An Introduction to the Replica Method and Its Applications, Vol. 9 (World Scientific Publishing Company, 1987)

    1987

  38. [38]

    Efetov,Supersymmetry in Disorder and Chaos(Cam- bridge University Press, 1996)

    K. Efetov,Supersymmetry in Disorder and Chaos(Cam- bridge University Press, 1996)

    1996

  39. [39]

    More generally, power-law correlationsC(r)∼∆r −α would require [∆] =−αand are again irrelevant for α >0

  40. [40]

    V. J. Emery and S. A. Kivelson, Nature374, 434 (1995)

    1995

  41. [41]

    Ghosal, M

    A. Ghosal, M. Randeria, and N. Trivedi, Phys. Rev. B 65, 014501 (2001)

    2001

  42. [42]

    B. I. Halperin and D. R. Nelson, Phys. Rev. Lett.41, 121 (1978)

    1978

  43. [43]

    D. R. Nelson and B. I. Halperin, Phys. Rev. B19, 2457 (1979)

    1979

  44. [44]

    A. P. Young, Phys. Rev. B19, 1855 (1979)

    1979

  45. [45]

    J. L. Cardy and S. Ostlund, Phys. Rev. B25, 6899 (1982)

    1982

  46. [46]

    D. S. Fisher and D. A. Huse, Phys. Rev. Lett.56, 1601 (1986)

    1986

  47. [47]

    D. S. Fisher and D. A. Huse, Journal of Physics A: Math- ematical and General20, L1005 (1987)

    1987

  48. [48]

    D. S. Fisher and D. A. Huse, Phys. Rev. B38, 386 (1988)

    1988

  49. [49]

    A. K. Hartmann and A. P. Young, Phys. Rev. B64, 180404 (2001)

    2001

  50. [50]

    A. K. Hartmann, Phys. Rev. B77, 144418 (2008)

    2008

  51. [51]

    Kawamura and H

    H. Kawamura and H. Yonehara, Journal of Physics A: Mathematical and General36, 10867–10880 (2003)

    2003

  52. [52]

    Weigel and M

    M. Weigel and M. J. Gingras, Physical review letters96, 097206 (2006)

    2006

  53. [53]

    Chakravarty, Phys

    S. Chakravarty, Phys. Rev. Lett.66, 481 (1991)

    1991

  54. [54]

    Altshuler, V

    B. Altshuler, V. Kravtsov, and I. Lerner, Zh. Eksp. Teor. Fiz91, 2276 (1986)

    1986

  55. [55]

    Vojta and J

    T. Vojta and J. Schmalian, Physical Review B—Condensed Matter and Materials Physics72, 045438 (2005)

    2005

  56. [56]

    Mohan, R

    P. Mohan, R. Narayanan, and T. Vojta, Physical Review B—Condensed Matter and Materials Physics81, 144407 (2010)

    2010

  57. [57]

    Vojta, Journal of Physics A: Mathematical and Gen- eral39, R143 (2006)

    T. Vojta, Journal of Physics A: Mathematical and Gen- eral39, R143 (2006)

    2006

  58. [58]

    Laumann, A

    C. Laumann, A. Scardicchio, and S. L. Sondhi, Phys. Rev. E77, 061139 (2008)

    2008

  59. [59]

    D. M. Basko, I. L. Aleiner, and B. L. Altshuler, Annals of physics321, 1126 (2006)

    2006

  60. [60]

    Sierant, M

    P. Sierant, M. Lewenstein, A. Scardicchio, L. Vidmar, and J. Zakrzewski, Reports on Progress in Physics88, 026502 (2025)

    2025

  61. [61]

    Cieplak and M

    M. Cieplak and M. Cieplak, Journal of Physics C: Solid 9 State Physics18, 1481 (1985)

    1985

  62. [62]

    J. V. Jos´ e, Phys. Rev. B20, 2167 (1979)

    1979

  63. [63]

    J. V. Jos´ e, Phys. Rev. Lett.46, 1591 (1981)

    1981

  64. [64]

    B. S. Shastry and B. Sutherland, Phys. Rev. Lett.65, 243 (1990)

    1990

  65. [65]

    Bonˇ ca, J

    J. Bonˇ ca, J. P. Rodriguez, J. Ferrer, and K. S. Bedell, Phys. Rev. B50, 3415 (1994)

    1994