Pith. sign in

REVIEW 3 major objections 4 minor 88 references

KPZ Superdiffusion of Local Correlators in Diffusive Random Quantum Circuits

T0 review · 3 major / 4 minor · reviewed 2026-08-15 · deepseek-v4-flash

Pith's one-line read For fixed spacetime disorder, the single-particle Green's function in a one-dimensional random unitary circuit is governed by a directed wave in a random medium, with KPZ wandering $t^{2/3}$ and free-energy fluctuations $t^{1/3}$.

desk verdict The strong-noise directed-wave mapping is solid and the weak-noise crossover prediction is genuinely new, but the headline claim for quenched Green's functions rests on an unproven ansatz—still worth refereeing. read the letter →

arxiv 2608.06459 v1 pith:FL6DD3ZZ submitted 2026-08-06 quant-ph cond-mat.dis-nncond-mat.stat-mechcond-mat.str-el

classification quant-phcond-mat.dis-nncond-mat.stat-mechcond-mat.str-el
keywords Kardar-Parisi-Zhanguniversalitydirectedpolymerinarandommediumwaveunitarycircuitssingle-particleGreen'sfunctionoperatorhydrodynamicsmacroscopicfluctuationtheoryquantumthermalization
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 argues that a basic phase-sensitive observable—the single-particle Green's function of a particle-number-conserving random quantum circuit coupled to a bath—does not simply diffuse or decay exponentially in a fixed realization of the disorder. Instead, for fixed spacetime disorder, the Green's function is governed by a strong-disorder fixed point: a directed wave (or, after phase averaging, a directed polymer) moving in a random medium. The paper predicts that the center of the normalized distribution $p(x,t)\propto |G(x,t)|^2$ wanders on a length scale $O(t^{2/3})$, and that sample-to-sample fluctuations of $-\log \sum_x |G(x,t)|^2$ grow as $t^{1/3}$. At weak noise there is a crossover to this fixed point at a parametrically long time $O(\gamma^{-3/2})$. If right, this means that intrinsically quantum, phase-sensitive correlators in an otherwise diffusive system are controlled by a different universality class than classical conserved densities.

What carries the argument

The central objects are the projected single-coherence subspace and, at weak noise, the source-manifold (projected operator-dynamics) ansatz. In strong noise the dynamics closes on states $|x\rangle=\sigma_x^+$; each half-layer gives a random $2\times 2$ matrix, leading to a discrete directed wave in a random medium and the continuum equation above. In weak noise, the ansatz represents $G$ as a path sum over a coherence trajectory $X$ and an exclusion-process fluid history $n$: $G\simeq\sum_X\sum_n P_U[n|X]A_U[X|n]$. Phase averaging imposes local occupation agreement and reduces the norm to $\mathcal{N}\simeq\sum_X P_{\rm coh}[X]\,Z_{\rm 2-copy}[X]$. The workhorse is then the point-absorber / macroscopic fluctuation theory (MFT) action for the $r$-symbol density, whose saddle gives the void size, and the Feynman\textendash polaron treatment of the internal coherence motion, which yields the coarse-grained directed-polymer action. These ingredients convert microscopic gate randomness into an $O(1)$ coarse-grained random potential with the KPZ/DPRM scaling.

What would settle it

A decisive calculation is to evolve one fixed weak-noise circuit for times $t\gg\gamma^{-3/2}$, then measure across samples the wandering variance $\mathrm{Var}\langle x\rangle$ and the free-energy variance $\mathrm{Var}[-\log\sum_x|G|^2]$; if these do not grow as $t^{4/3}$ and $t^{2/3}$ respectively, the central claim is falsified.

Watch

Extended reading notes

Core claim

The central claim, stated on the paper's own terms, is that for a fixed circuit realization and fixed noise rate $\gamma$, the infinite-temperature Green's function $G_\gamma(x,t)=\langle \sigma^-_x(0)\sigma^+_0(t)\rangle$ is described at long times by a directed wave in a random medium (DWRM). In the strong-noise limit this is derived exactly: large $\gamma$ projects the operator dynamics onto a single moving coherence $\sigma^+_x$, whose transfer matrix is a contractive non-Hermitian Schr\"odinger evolution with random complex hoppings and a random complex potential, coarse-graining to $\partial_t G = D\partial_x^2 G - \Gamma G + \eta(x,t)G$. The KPZ exponents follow from the DWRM fixed point: the center wanders as $t^{2/3}$ and the free energy $F=-\log\sum_x |G|^2$ has $t^{1/3}$ fluctuations. At weak noise, the paper extends the projected operator-dynamics ('source-manifold') ansatz of Ref. [39] to a coherence dressed by a diagonal exclusion-process fluid with quenched random hopping rates; phase averaging turns the norm $\mathcal{N}=\mathbb{E}_\phi |G|^2$ into a sub-stochastic sum over coherence paths with a two-copy exclusion weight, whose stationary saddle is a slow 'void' of size $\xi\sim\sqrt{D_0/\gamma}$. Restoring hopping disorder makes the void coordinate a directed polymer in a random medium with $O(1)$ random potential on the rescaled spacetime scale $\xi^2/D_{\rm void}\sim \gamma^{-3/2}$, giving the same KPZ exponents. Tensor-network simulations of individual circuits at $\gamma=2/3$, and of the full two-copy transfer matrix at weak noise, show the predicted $t^{2/3}$ and $t^{1/3}$ scalings.

Load-bearing premise

The load-bearing premise is the paper's 'projected operator-dynamics' ansatz in the weak-noise regime: after projecting onto operators with one local coherence, the surrounding diagonal operators evolve as a symmetric exclusion process with the same random hopping rates as the gates. If that representation fails, the directed-polymer mapping and the KPZ exponents do not follow.

Editorial extensions

If this is right

  • If the claim holds, the single-particle Green's function in a fixed circuit is not self-averaging at the diffusive scale: it shows KPZ wandering, so phase-sensitive local probes see strong-disorder fluctuations even where densities diffuse.
  • The weak-noise crossover time $\gamma^{-3/2}$ is parametrically larger than the void-formation time $\gamma^{-1}$, so experiments or simulations at small $\gamma$ must wait beyond that scale to see the KPZ regime.
  • The mapping predicts $\mathrm{Var}[F]\sim t^{2/3}$ and wandering $\langle x\rangle^2\sim t^{4/3}$ simultaneously, giving a sharp numerical signature in tensor-network or analog-simulation data.
  • The result implies that annealing over circuit phases, as in earlier ensemble studies, misses the typical single-sample behavior; phase-quenched disorder is relevant for this observable.

Reading between the lines

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

  • A consequence the author leaves implicit is that the same strong-disorder fixed point should govern other charge-carrying phase-sensitive correlators, such as fermionic or bosonic single-particle Green's functions in diffusive random circuits, wherever a similar single-coherence subspace can be projected out.
  • A direct testable extension is to measure the center wandering of a local spin coherence in a cold-atom quantum simulator; the predicted $t^{2/3}$ wandering of the spin-correlation cloud should be visible once the bath coupling is fixed and the disorder realization is held.
  • The weak-noise analysis explicitly treats the phase-annealed norm; whether genuine single-sample Green's functions at weak noise show the same DWRM behavior is still a prediction rather than a demonstrated fact, so a reader should treat that part as the open edge of the paper's claim.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 4 minor

Summary. The paper studies the single-particle Green's function G(x,t)=<sigma^-_x(0)sigma^+_0(t)> in one-dimensional, particle-number-conserving random unitary circuits with local noise. For a fixed circuit realization, it argues that G is governed, in both the strong- and weak-noise regimes, by directed waves in a random medium, leading to KPZ scaling: the center of the normalized distribution |G|^2 wanders as t^{2/3}, and sample-to-sample fluctuations of -log sum_x |G|^2 grow as t^{1/3}, with a weak-noise crossover time of order gamma^{-3/2}. In the strong-noise limit, the paper derives an exact mapping to a discrete non-Hermitian Schr\"odinger equation with random complex potential and confirms the predicted scaling by tensor-network simulations at moderate noise. In the weak-noise regime, it introduces a projected 'source-manifold' ansatz for the q=1 operator sector, derives a DPRM description for the phase-annealed norm N=E_phi|G|^2 via a macroscopic fluctuation theory, and tests the resulting scaling by numerically exact two-copy simulations. A replica treatment in the Supplementary Material extends the weak-noise mapping to balanced moments of the genuine Green's function, yielding a DWRM description. The paper explicitly states that direct verification of the weak-noise DWRM scaling for genuine single-sample Green's functions remains an outstanding task.

Significance. If the weak-noise claim for genuine single-sample Green's functions holds, the paper identifies a new strong-disorder universality class for a phase-sensitive observable in a diffusive random quantum circuit, contrasting with the standard hydrodynamics of conserved densities. The strong-noise mapping is exact in the gamma-to-infinity limit and is a clean, self-contained contribution. The weak-noise MFT derivation is transparent and yields a specific, parameter-free crossover scale gamma^{-3/2}; the numerics for the phase-annealed norm N provide a direct benchmark of the projected-dynamics ansatz for that observable. However, the central advertised claim for genuine G at weak noise rests on an ansatz that is explicitly labeled as the paper's principal assumption and is carried over from the author's own Ref. [39]; it is not derived from the microscopic circuit, and the numerical evidence at weak noise tests only N, not the interference in G.

major comments (3)
  1. [Main text, 'Effective dynamics at weak noise', Eq. (5)] The weak-noise treatment of the genuine single-particle Green's function is built on the source-manifold ansatz, which the manuscript explicitly labels 'our principal assumption' and which is carried over from Ref. [39]. This ansatz projects the q=1 sector onto a single local coherence plus a diagonal SEP fluid, and it is not derived from the microscopic circuit. Because Eq. (5) is the starting point for all subsequent weak-noise results, including the DPRM description of N and the DWRM description of G, the paper's abstract claim that G is governed by directed waves 'in both the strong- and weak-noise limits' is not yet established for genuine G. The paper's own Discussion confirms this: verifying the weak-noise DWRM scaling for genuine single-sample Green's functions is listed as an outstanding task. The revision should either provide numerical evidence for genuine G at weak noise, derive the ansatz from the circuit, or substantially temper the abstract and introduction to present the weak-noise claim for G as a conjecture.
  2. [Supplementary Material II, Eq. (S14) and Eq. (S65)] The replica extension from the phase-annealed norm N to balanced moments of the genuine Green's function inherits the source-manifold ansatz and adds further uncontrolled approximations: the point-absorber reduction at all replica numbers and the Feynman-polaron replacement on collision-free segments. The first sentence of SM II states that the derivation assumes the projected operator dynamics ansatz. The point-absorber reduction is justified only by the estimate N_QSS^u ~ gamma log(1/gamma) in the End Matter and SM II, which is itself computed within the ansatz. If multi-coherence sectors contribute beyond this estimate, the DWRM mapping for G does not follow. The claim in Eq. (S65) that G_F matches all moments of G upon coarse-graining is therefore conditional on these approximations and should be stated as such in the main text.
  3. [Main text, 'Effective dynamics at weak noise' and Fig. 2] The numerical confirmation at weak noise applies only to the phase-annealed norm N, not to genuine G. Fig. 2 simulates the full two-copy transfer-matrix dynamics for N, which is exact for N but is insensitive to the quenched interference that distinguishes directed waves from directed polymers. The claim that the numerics 'provide a direct benchmark of both the projected-dynamics ansatz and the weak-noise scaling' is therefore only valid for the phase-annealed observable. Since the abstract's headline prediction concerns G, the distinction between N and G should be made prominent throughout, and any statement that the KPZ predictions for G are 'confirmed' should be limited to the strong-noise regime and to the phase-annealed proxy at weak noise.
minor comments (4)
  1. [Introduction, first paragraph] The phrase 'amcentral problem' appears to be a typo for 'a central problem'.
  2. [Main text, 'MFT formulation'] The sentence 'We call the resulting low-entropy region avoid' appears to contain a typo; it should likely read 'we call the resulting low-entropy region a void.'
  3. [Fig. 1 and Fig. 2] The figures would benefit from explicit error bars and a description of the fitting windows used to extract the t^{2/3} and tau^{2/3} guides; the collapse in Fig. 2 uses a free reference time t0, and the sensitivity of the collapse to t0 should be discussed.
  4. [Supplementary Material II, Eq. (S38)] The notation in Eq. (S38) is somewhat compressed; defining P^{(0)}_{I_i}[X_a] and the Feynman-polaron action S_{F,I} explicitly before use would improve readability.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the derivation is self-contained apart from an explicit, separately benchmarked source-manifold ansatz that is not disguised as a first-principles result.

full rationale

The paper's strong-noise result is derived microscopically from the projected single-coherence transfer matrix and then mapped to Eq. (3), a directed wave in a random medium; the KPZ exponents are inherited from known DWRM universality and the numerical checks use the predicted exponents as guides rather than as fitted inputs. The weak-noise result is the only place where a self-referential element appears: Eq. (5) is the source-manifold ansatz, explicitly called 'our principal assumption,' and carried over from the author's own Ref. [39]. This is a genuine assumption and a load-bearing one for the claim about genuine single-sample Green's functions, but it is not circular by construction: the paper does not present the ansatz as derived from the microscopic circuit, it benchmarks the phase-annealed norm N against full two-copy transfer-matrix simulations, and it states in the Discussion that 'One outstanding task is to verify the weak-noise DWRM scaling directly for genuine single-sample Green's functions, rather than for the phase-annealed norm N.' Thus the weak-noise DWRM prediction for G is conditional and incomplete, which is a limitation or correctness risk, but not a case of a fitted parameter being renamed as a prediction, nor of an output being equivalent to an input by definition. The crossover time gamma^{-3/2} is derived from the void scales xi ~ (D0/gamma)^{1/2} and D_void ~ (D0 gamma)^{1/2}, and the data collapse at that scale supports the derivation. No circular step can be exhibited in the paper's equations.

Assumptions & free parameters 1 free parameters · 6 assumptions · 0 invented entities

The theoretical input consists of the microscopic circuit model, the projected-operator ansatz at weak noise, the quasi-stationary void and point-absorber approximation, the Feynman-polaron variational treatment, and the imported KPZ/DPRM universality. No genuinely new entities are postulated; the void is a collective density depression. The only numerical fitting parameter is the reference time t0 used in the weak-noise collapse.

free parameters (1)
  • t0 (time reference in weak-noise collapse) = chosen per gamma, >= tau_void
    In Fig. 2, times are measured relative to a reference t0 chosen to suppress short-time transients; the data collapse and apparent crossover depend on this choice. The scaling exponents and the crossover exponent are not fitted, so this does not affect the theoretical predictions.
assumptions (6)
  • domain assumption U(1)-symmetric two-qubit Haar-random gates and a weak-symmetric depolarizing bath at rate gamma define the microscopic model.
    Standard model for diffusive random quantum circuits; it defines the system under study.
  • ad hoc to paper At weak noise, the q=1 sector is captured by the projected operator-dynamics ansatz: one local coherence plus a diagonal SEP fluid.
    Stated as 'Our principal assumption' in the weak-noise section; carried over from Ref. [39] and not derived from the microscopic circuit.
  • domain assumption The quasi-stationary void and the point-absorber rule of Eq. (9) capture the dominant histories, with u-symbols negligible.
    Justified by the scaling estimate N_QSS^u ~ gamma log(1/gamma) << 1; it is an approximation, not an exact statement.
  • domain assumption The Feynman-polaron variational ansatz for the void internal motion is valid.
    Used to construct the slow-coordinate effective action; controlled by the small parameter ell/xi << 1.
  • standard math KPZ/DPRM universality: a directed polymer with O(1) diffusivity and short-range correlated O(1) random potential in 1+1D has wandering exponent 2/3 and free-energy fluctuations t^{1/3}.
    Known results from Refs. [46-53]; imported as an input, not derived here.
  • domain assumption At strong noise, the random-diffusivity term in the continuum limit is irrelevant and only multiplicative noise remains.
    RG power-counting statement in SM section I.4; standard for DWRM.

how reviews work

0 comments
Cite this review

Pith. "Pith review of KPZ Superdiffusion of Local Correlators in Diffusive Random Quantum Circuits." pith.science (2026). https://pith.science/paper/FL6DD3ZZ

@misc{pith2026260806459,
  author       = {Pith},
  title        = {Pith review of: KPZ Superdiffusion of Local Correlators in Diffusive Random Quantum Circuits},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FL6DD3ZZ}},
  note         = {Machine review of arXiv:2608.06459}
}
abstract

We study the single-particle Green's function $G(x,t)=\langle \sigma^-_x(0)\sigma^+_0(t)\rangle$ in one-dimensional particle-number-conserving random unitary circuits coupled to an external bath. For fixed spacetime disorder, we argue that $G(x,t)$ is governed, in both the strong- and weak-noise limits, by directed waves in a random medium. We find Kardar-Parisi-Zhang (KPZ) scaling in the wandering statistics of the normalized spatial distribution $p(x,t)\propto |G(x,t)|^2$ and in the associated free energy. In particular, its center $\langle x(t)\rangle\equiv\sum_x x\, p(x,t)$ wanders on a length-scale $\mathcal{O}(t^{2/3})$, while sample-to-sample fluctuations of $-\log\sum_x |G(x,t)|^2$ scale as $t^{1/3}$. At weak noise $\gamma \ll 1$, the crossover to the strong-disorder fixed point occurs at a parametrically long time $\mathcal{O}(\gamma^{-3/2})$. These predictions are confirmed numerically using tensor-network simulations of the noisy operator dynamics in individual circuits at moderate noise, and of a phase-annealed proxy retaining hopping disorder at weak noise.

Figures

Figures reproduced from arXiv: 2608.06459 by the authors.

Figure 1
Figure 1. FIG. 1 [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2 [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

88 extracted references · 40 canonical work pages

  1. [60]

    Nahum, S

    A. Nahum, S. Roy, S. Vijay, and T. Zhou, Phys. Rev. B 106, 224310 (2022)

  2. [39]

    McCulloch, J

    E. McCulloch, J. A. Jacoby, and S. Gopalakr- ishnan, arXiv e-prints , arXiv:2604.27074 (2026), arXiv:2604.27074 [quant-ph]

  3. [1]

    J. Lux, J. M¨ uller, A. Mitra, and A. Rosch, Phys. Rev. A89, 053608 (2014)

  4. [2]

    J. F. Wienand, S. Karch, A. Impertro, C. Schweizer, E. McCulloch, R. Vasseur, S. Gopalakrishnan, M. Aidels- burger, and I. Bloch, Nature Physics20, 1732 (2024)

  5. [3]

    M. K. Joshi, F. Kranzl, A. Schuckert, I. Lovas, C. Maier, R. Blatt, M. Knap, and C. F. Roos, Science376, 720 (2022)

  6. [4]

    Rosenberg, T

    E. Rosenberg, T. Andersen, R. Samajdar, A. Petukhov, J. Hoke, D. Abanin, A. Bengtsson, I. Drozdov, C. Erickson, P. Klimov,et al., Science384, 48 (2024), https://www.science.org/doi/pdf/10.1126/science.adi7877

  7. [5]

    D. Wei, A. Rubio-Abadal, B. Ye, F. Machado, J. Kemp, K. Srakaew, S. Hollerith, J. Rui, S. Gopalakrishnan, N. Y. Yao,et al., Science376, 716 (2022)

  8. [6]

    Y. Le, Y. Zhang, S. Gopalakrishnan, M. Rigol, and D. S. Weiss, Nature618, 494 (2023)

Show all 88 references
  1. [7]

    Gross and I

    C. Gross and I. Bloch, Science357, 995 (2017)

  2. [8]

    Scholl, H

    P. Scholl, H. J. Williams, G. Bornet, F. Wallner, D. Barredo, L. Henriet, A. Signoles, C. Hainaut, T. Franz, S. Geier, A. Tebben, A. Salzinger, G. Z¨ urn, T. Lahaye, M. Weidem¨ uller, and A. Browaeys, PRX Quantum3, 020303 (2022)

  3. [9]

    Ollitrault, European Journal of Physics29, 275 (2008), arXiv:0708.2433 [nucl-th]

    J.-Y. Ollitrault, European Journal of Physics29, 275 (2008), arXiv:0708.2433 [nucl-th]

  4. [10]

    Jaiswal and V

    A. Jaiswal and V. Roy, Advances in High Energy Physics 2016, 9623034 (2016)

  5. [11]

    M¨ uller and S

    M. M¨ uller and S. Sachdev, Phys. Rev. B78, 115419 (2008), arXiv:0801.2970 [cond-mat.str-el]

  6. [12]

    Lucas, J

    A. Lucas, J. Crossno, K. C. Fong, P. Kim, and S. Sachdev, Phys. Rev. B93, 075426 (2016), arXiv:1510.01738 [cond-mat.str-el]

  7. [13]

    Lucas, Phys

    A. Lucas, Phys. Rev. B93, 245153 (2016), arXiv:1604.03955 [cond-mat.str-el]

  8. [14]

    Crossno, J

    J. Crossno, J. K. Shi, K. Wang, X. Liu, A. Harzheim, A. Lucas, S. Sachdev, P. Kim, T. Taniguchi, K. Watan- abe, T. A. Ohki, and K. C. Fong, Science351, 1058 (2016), arXiv:1509.04713 [cond-mat.mes-hall]

  9. [15]

    B. N. Narozhny, I. V. Gornyi, A. D. Mirlin, and J. Schmalian, Annalen der Physik529, 1700043 (2017), arXiv:1704.03494 [cond-mat.mes-hall]

  10. [16]

    Lucas and K

    A. Lucas and K. Chung Fong, Journal of Physics Condensed Matter30, 053001 (2018), arXiv:1710.08425 [cond-mat.str-el]

  11. [17]

    G. Bal, A. Lucas, and M. Luskin, Journal of Math- ematical Physics62, 011503 (2021), arXiv:2004.10790 [math.AP]

  12. [18]

    McCulloch, J

    E. McCulloch, J. De Nardis, S. Gopalakrishnan, and R. Vasseur, Phys. Rev. Lett.131, 210402 (2023)

  13. [19]

    Singh, E

    H. Singh, E. McCulloch, S. Gopalakrishnan, and R. Vasseur, arXiv e-prints , arXiv:2405.13892 (2024), arXiv:2405.13892 [cond-mat.stat-mech]

  14. [20]

    Gopalakrishnan, E

    S. Gopalakrishnan, E. McCulloch, and R. Vasseur, Proceedings of the National Academy of Sciences121, e2403327121 (2024)

  15. [21]

    McCulloch, R

    E. McCulloch, R. Vasseur, and S. Gopalakrishnan, arXiv e-prints , arXiv:2601.02475 (2026), arXiv:2601.02475 [cond-mat.stat-mech]

  16. [22]

    Bauer, D

    M. Bauer, D. Bernard, and T. Jin, SciPost Physics3, 033 (2017), arXiv:1706.03984 [cond-mat.stat-mech]

  17. [23]

    Bernard and T

    D. Bernard and T. Jin, Phys. Rev. Lett.123, 080601 (2019), arXiv:1904.01406 [cond-mat.stat-mech]

  18. [24]

    Bernard and T

    D. Bernard and T. Jin, Communications in Mathematical Physics384, 1141 (2021), arXiv:2006.12222 [math-ph]

  19. [25]

    Albert, D

    M. Albert, D. Bernard, T. Jin, S. Scopa, and S. Wei, arXiv e-prints , arXiv:2601.16883 (2026), 6 arXiv:2601.16883 [cond-mat.stat-mech]

  20. [26]

    J. A. Sobota, Y. He, and Z.-X. Shen, Rev. Mod. Phys. 93, 025006 (2021)

  21. [27]

    S. Loth, M. Etzkorn, C. P. Lutz, D. M. Eigler, and A. J. Heinrich, Science329, 1628 (2010)

  22. [28]

    Liang, L

    K. Liang, L. Bi, Q. Zhu, H. Zhou, and S. Li, ACS Applied Optical Materials1, 924 (2023)

  23. [29]

    E. G. Sorte, B. V. Fine, and B. Saam, Phys. Rev. B83, 064302 (2011)

  24. [30]

    S. W. Morgan, V. Oganesyan, and G. S. Boutis, Phys. Rev. B86, 214410 (2012)

  25. [31]

    Maldacena, S

    J. Maldacena, S. H. Shenker, and D. Stanford, Journal of High Energy Physics2016, 106 (2016)

  26. [32]

    D. A. Roberts and B. Swingle, Phys. Rev. Lett.117, 091602 (2016)

  27. [33]

    Nahum, S

    A. Nahum, S. Vijay, and J. Haah, Phys. Rev. X8, 021014 (2018)

  28. [35]

    Rakovszky, F

    T. Rakovszky, F. Pollmann, and C. W. von Keyserlingk, Phys. Rev. X8, 031058 (2018)

  29. [36]

    Khemani, A

    V. Khemani, A. Vishwanath, and D. A. Huse, Phys. Rev. X8, 031057 (2018)

  30. [37]

    Basko, I

    D. Basko, I. Aleiner, and B. Altshuler, Annals of Physics 321, 1126 (2006)

  31. [38]

    Subexponential decay of local cor- relations from diffusion-limited dephasing,

    E. McCulloch, J. A. Jacoby, C. von Keyserlingk, and S. Gopalakrishnan, “Subexponential decay of local cor- relations from diffusion-limited dephasing,” (2025), arXiv:2504.05380 [quant-ph]

  32. [40]

    Medina, M

    E. Medina, M. Kardar, Y. Shapir, and X. R. Wang, Phys. Rev. Lett.62, 941 (1989)

  33. [41]

    Medina, M

    E. Medina, M. Kardar, and H. Spohn, Phys. Rev. Lett. 66, 2176 (1991)

  34. [42]

    L. Saul, M. Kardar, and N. Read, Phys. Rev. A45, 8859 (1992)

  35. [43]

    Blum and Y

    T. Blum and Y. Y. Goldschmidt, Nuclear Physics B380, 588 (1992)

  36. [44]

    Kardar, arXiv e-prints , cond-mat/9411022 (1994), arXiv:cond-mat/9411022 [cond-mat]

    M. Kardar, arXiv e-prints , cond-mat/9411022 (1994), arXiv:cond-mat/9411022 [cond-mat]

  37. [45]

    Kim and D

    H. Kim and D. A. Huse, Phys. Rev. B83, 052405 (2011)

  38. [46]

    Kardar, G

    M. Kardar, G. Parisi, and Y.-C. Zhang, Phys. Rev. Lett. 56, 889 (1986)

  39. [47]

    Halpin-Healy and Y.-C

    T. Halpin-Healy and Y.-C. Zhang, Physics Reports254, 215 (1995)

  40. [48]

    Kardar, Nuclear Physics B290, 582 (1987)

    M. Kardar, Nuclear Physics B290, 582 (1987)

  41. [49]

    D. S. Fisher and D. A. Huse, Phys. Rev. B43, 10728 (1991)

  42. [50]

    Kardar and Y.-C

    M. Kardar and Y.-C. Zhang, Phys. Rev. Lett.58, 2087 (1987)

  43. [51]

    Kardar, Phys

    M. Kardar, Phys. Rev. Lett.55, 2923 (1985)

  44. [52]

    D. A. Huse, C. L. Henley, and D. S. Fisher, Physical review letters55, 2924 (1985)

  45. [53]

    Bertini and N

    L. Bertini and N. Cancrini, Journal of Statistical Physics 78, 1377 (1995)

  46. [54]

    See Supplemental Material

  47. [55]

    Buˇ ca and T

    B. Buˇ ca and T. Prosen, New Journal of Physics14, 073007 (2012)

  48. [56]

    V. V. Albert and L. Jiang, Phys. Rev. A89, 022118 (2014)

  49. [57]

    P. Sala, S. Gopalakrishnan, M. Oshikawa, and Y. You, Phys. Rev. B110, 155150 (2024)

  50. [58]

    Strong-to-weak symmetry breaking phases in steady states of quantum operations,

    N. Ziereis, S. Moudgalya, and M. Knap, “Strong-to-weak symmetry breaking phases in steady states of quantum operations,” (2025), arXiv:2509.09669 [cond-mat.stat- mech]

  51. [59]

    The discrete-time dynamics hasO(1) hop amplitudes even asγ→∞

    There is no Zeno effect here. The discrete-time dynamics hasO(1) hop amplitudes even asγ→∞

  52. [61]

    von Keyserlingk, F

    C. von Keyserlingk, F. Pollmann, and T. Rakovszky, Physical Review B105(2022), 10.1103/phys- revb.105.245101

  53. [62]

    Ahlbrecht, C

    A. Ahlbrecht, C. Cedzich, R. Matjeschk, V. B. Scholz, A. H. Werner, and R. F. Werner, Quantum Information Processing11, 1219 (2012)

  54. [63]

    Joye, Communications in Mathematical Physics307, 65 (2011)

    A. Joye, Communications in Mathematical Physics307, 65 (2011)

  55. [64]

    Atν= 1/2, particle and hole voids are degenerate. Once a void has formed, however, its polarization switches only on a nonperturbatively long timescaleO(exp(c/γ)) [39], much longer than the timescales algebraic in 1/γconsid- ered here. We may therefore regard the void polariza...

  56. [65]

    Bertini, A

    L. Bertini, A. De Sole, D. Gabrielli, G. Jona-Lasinio, and C. Landim, Reviews of Modern Physics87, 593 (2015), arXiv:1404.6466 [cond-mat.stat-mech]

  57. [66]

    Derrida, SciPost Physics Lecture Notes (2025), 10.21468/scipostphyslectnotes.106

    B. Derrida, SciPost Physics Lecture Notes (2025), 10.21468/scipostphyslectnotes.106

  58. [67]

    Mallick, H

    K. Mallick, H. Moriya, and T. Sasamoto, Phys. Rev. Lett.129, 040601 (2022)

  59. [68]

    S. Saha, S. Jangid, T. Arnoulx de Pirey, J. U. Klamser, and T. Sadhu, arXiv e-prints , arXiv:2601.02319 (2026), arXiv:2601.02319 [cond-mat.stat-mech]

  60. [69]

    Saha and T

    S. Saha and T. Sadhu, SciPost Phys.17, 033 (2024)

  61. [70]

    R. P. Feynman, R. W. Hellwarth, C. K. Iddings, and P. M. Platzman, Phys. Rev.127, 1004 (1962)

  62. [71]

    Vidal, Phys

    G. Vidal, Phys. Rev. Lett.93, 040502 (2004)

  63. [72]

    Verstraete, J

    F. Verstraete, J. J. Garc´ ıa-Ripoll, and J. I. Cirac, Phys. Rev. Lett.93, 207204 (2004)

  64. [73]

    S. R. White and A. E. Feiguin, Phys. Rev. Lett.93, 076401 (2004)

  65. [74]

    Hauschild, J

    J. Hauschild, J. Unfried, S. Anand, B. Andrews, M. Bintz, U. Borla, S. Divic, M. Drescher, J. Geiger, M. Hefel, K. H´ emery, W. Kadow, J. Kemp, N. Kirchner, V. S. Liu, G. M¨ oller, D. Parker, M. Rader, A. Romen, S. Scalet, L. Schoonderwoerd, M. Schulz, T. Soejima, P. Thoma, Y....

  66. [75]

    Both external noise and phase annealing substantially re- duce entanglement growth in the two-copy transfer ma- trix dynamics

  67. [76]

    Bernard, F

    D. Bernard, F. H¨ ubner, and S. Scopa, arXiv e-prints , arXiv:2607.28255 (2026), arXiv:2607.28255 [cond- mat.stat-mech]

  68. [77]

    Tilma and E

    T. Tilma and E. C. G. Sudarshan, Journal of Physics A: Mathematical and General35, 10467 (2002)

  69. [78]

    7 End Matter Point-absorber approximation—We now check the validity of the point-absorber approximation Eq

    Since unbalanced moments vanish both forG F U andG U. 7 End Matter Point-absorber approximation—We now check the validity of the point-absorber approximation Eq. (9), by estimating the number ofusymbols in the quasi- stationary state. In the static void, the reset-symbol den- ...

  70. [79]

    Mapping to discrete directed waves Consider a gate acting on a bondj= (x,x+ 1). In the sectors with zero and two particles it acts by phasese iϕ0,j ande iϕ2,j, while its one-particle block may be parameterized as [77] u(1) j =e iχj eiαj p1−p j eiβj√pj −e−iβj√pj e−iαj p1−p j .(...

  71. [80]

    Hence Vj,L =Vj,R = 4 3π−1

    Noise statistics: means For the Haar distribution, we have the following averages, h= √p|sinδ|= 2 3 2 π = 4 3π, h2 =p sin2δ= 1 4, r2 = 1−p cos2δ= 1 4.(S6) Becauseζis uniform, we have re±iζ = 0. Hence Vj,L =Vj,R = 4 3π−1. Defineδh j =h j−hand the centered complex noise ηj,L≡V j...

  72. [81]

    The two-point correlation functions of the complex noiseη(Eq

    Noise statistics: covariances Lets,s ′∈{L,R}. The two-point correlation functions of the complex noiseη(Eq. (S7)) are ηm j,sηm′ j′,s′ =δ mm′δjj′ C20 ss′, ηm j,sηm′∗ j′,s′ =δ mm′δjj′ C11 ss′,(S9) with covariance matrices C20 = B A A B ,C 11 = A B B A .(S10) The random hopping a...

  73. [82]

    (S5) coarse-grains to ∂tψ=D∂ 2 xψ+ [−Γ 0 +η(x,t)]ψ+∂ x[δD(x,t)∂xψ] +....(S12) HereD, Γ 0, and the continuum noise covariances are nonuniversal but finite andO(1)

    Continuum limit On scales large compared with one bond and one half-layer, Eq. (S5) coarse-grains to ∂tψ=D∂ 2 xψ+ [−Γ 0 +η(x,t)]ψ+∂ x[δD(x,t)∂xψ] +....(S12) HereD, Γ 0, and the continuum noise covariances are nonuniversal but finite andO(1). The random-diffusivity term contain...

  74. [83]

    The first follows from a sitewise rephasing freedom of the gate ensemble

    Local phase constraints and replica pairing As in theq= 1 calculation in the main text, the phase average imposes two local constraints. The first follows from a sitewise rephasing freedom of the gate ensemble. By Haar invariance, aU(1)-symmetric two-qubit gate may be dressed ...

  75. [84]

    We first illustrate this for a collision of two coherence paths

    Point-absorber reduction We next show that, at weak noise, the occupation constraint imposed by the phase average reduces to the same point-absorber rule in every replica, including at replica collisions. We first illustrate this for a collision of two coherence paths. Conside...

  76. [85]

    We now reintroduce the neglected outputs, i.e., those producingu’s, and estimate the number ofusymbols in the stationary state

    Validity of the point-absorber reduction It remains to verify that the omitted outputs containingusymbols carry vanishing weight asγ→0. We now reintroduce the neglected outputs, i.e., those producingu’s, and estimate the number ofusymbols in the stationary state. Within the po...

  77. [86]

    qY a=1 ZU[Xa] qY b=1 ZU[Xb] # ×Eϕ

    Feynman-polaron reduction of the balanced moments at weak noise The discussion above is independent of the circuit geometry and applies equally to brickwork and Poissonized circuits. For convenience, we now specialize to a Poissonized circuit, in which two-site gates act at in...

  78. [87]

    (S32)), we have A(0) ϕ [X] = p P0[X]eiΦϕ[X], p P0[X]≃exp " − Z dτ ˙X2 8D0 # ,Φ ϕ[X] = Z dτϕ(X(τ),τ).(S45) The internal coordinate relaxes on the timescaleτ ℓ∼ℓ 2/D0

    Single-replica path sum Equation (S42) is the balanced moment of a single-replica amplitude, Mq≃E ϕ h GF ϕ 2qi , G F ϕ = X X Z Dye−SF[X,y]A(0) ϕ [X].(S43) WritingX=y+zand taking the continuum limit gives GF ϕ = Z DyDzexp ( − Z dτ ˙y2 8Dvoid + D0 8ℓ4z2 +λ 0 ) A(0) ϕ [X].(S44) F...

  79. [88]

    Phase-induced damping fluctuations We now determine the effects of the microscopic phase disorder on the slow polaron coordinateyupon integrating out the confined internal motion. Consider a spacetime blockBof spatial widthO(ℓ) and durationTsatisfying τℓ≪T≪ ℓ2 Dvoid .(S48) Wit...

  80. [89]

    We now restore the quenched hopping disorder, whose fluctuations we have already seen (in the main text) produce anO(1) real random energy on the (ξ,eτ) scale

    Restoring hopping disorder We have found that phase disorder produces an order-one random block phase but only a parametrically small fluctuation of the real damping rate. We now restore the quenched hopping disorder, whose fluctuations we have already seen (in the main text) ...

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.