Boundary preparation of strong Mpemba states in holography

Changqing Liu; Chikun Ding; Liping Li; Sheng Long; Xinwu Liu; Yuebing Zhou; Yu Shi

REVIEW

Boundary preparation of strong Mpemba states in holography

Yu Shi, Xinwu Liu, Liping Li, Chikun Ding, Yuebing Zhou, Changqing Liu, Sheng Long

High Energy Physics - Theory · Submitted 19 Oct 2026

Pith's verdict: A clean, internally consistent holographic preparation result—the density-history cubic term is genuinely new—whose main external risk is the unreleased spectral-count numerics, not the acknowledged probe limit.

T0 · AI review

How this review was made: An AI referee (deepseek/deepseek-v4-flash) read the paper on October 20, 2026 and answered a fixed list of questions about it. T0 is the starting level every paper gets. The levels describe how deeply Pith has checked the paper, never the authors.

T0 AI review An AI read the paper and answered a fixed list of questions.
T1 Audit Automatic checks that the references and files the paper relies on exist and match.
T2 Compute The paper's code or calculations re-run with the same results.
T3 Formal statement The claims written out in a formal math language a computer can read.
T4 Proof Those claims proved, and the proof checked by computer.

How Pith checks papers →

· 1 major objection · 1 minor · Ask about this paper · The compiled record · Read the full review →

Challenge this review Comment (0) Read on arXiv ↗

Cite ▾

arXiv ID2610.26218 (version 3)

Pith IDLED7KO2H, Pith's permanent record of this paper and its review

CitationChangqing Liu, Chikun Ding, Liping Li, Sheng Long, Xinwu Liu, Yuebing Zhou, Yu Shi. "Boundary preparation of strong Mpemba states in holography." arXiv:2610.26218 (2026). https://pith.science/paper/LED7KO2H

Cite this reviewBibTeX for Pith's review of this paper

Claim paper

Why Pith says this

Highlight any text to comment on it.

Editor's note

read the letter

The real claim is that finite charged boundary drives leave a calculable cubic slow-mode amplitude, and that a compensating pulse computed from the low-order equations suppresses it. That is demonstrated, not just proposed. The density-history term G_prep^ρ and its 85–89% dominance is the genuinely new ingredient. The two-pulse prediction β3 = −G/H1, the full-nonlinear reduction from O(ε^3) to O(ε^5), and the roughly 600-fold suppression at ε = 0.05 are the kind of checkable numerical results that make the paper worth engaging.

What the paper does well: the Table I decomposition checks out internally, including the interference term 2Re[(G_prep^ρ)*G_a] ≈ 3.303 × 10⁻³ and the phase of β3. The cubic coefficient is assembled from the linear driven field and the Ward history, not fitted to the cancellation target; only the fifth-order coefficient g5 is fitted, and it is labeled as such. The distance-positivity proof and the high-frequency spectral bound are serious analytical work. The four-protocol full-nonlinear verification and the three-pulse two-mode suppression are substantial extensions beyond the earlier relaxation-only setup.

Soft spots, in proportion: the load-bearing spectral count in App. A.4—exactly two simple QNMs with Re λ > −2.9—rests on in-house winding computations with no released code and no certified error control. The fast-window claims, envelope scalings, and two-mode interpretation all inherit that count. I would want an independent winding computation or a released repository with a commit hash before trusting the cleanest numbers. The code-and-data-on-request policy is genuinely weak given the 10⁻¹³-level bounds. Also, the extension claim relative to Ref. [31] is conditional because that baseline is an unpublished preprint. The probe limit is not my main concern: the authors state it plainly and backreaction is an extension, not an internal contradiction. Control-error sensitivity and the display cutoff are handled honestly; the quoted slow amplitudes in the common-input section are explicitly estimates, not upper bounds.

This paper is for people working on holographic thermalization and Mpemba effects. It deserves a serious referee. My recommendation is to send it to peer review, requesting code/data release and an independent spectral check; if those survive, accept.

Referee report

Significance

If the results hold, the paper makes a substantive contribution: it converts strong-Mpemba preparation from a statement about initial states into a computable control problem with a falsifiable pulse prescription, and it identifies a genuinely new mechanism—the density-history contribution to the cubic slow amplitude—that dominates the compensation at ρ_f=3. Strengths include the in-text checkable perturbative derivation of G and β3 from the linear driven trajectory and exact Ward history; the independent confirmation of β3 by full nonlinear tuning in both quadratures; the validated spectral gap analysis; the explicit positivity and static-uniqueness proofs for the distance functional; and the careful numerical resolution reporting, including grid-reﬁnement checks and conservative bounds. The paper is also unusually honest about its scopes and limitations (probe limit, uncontrolled norms, display cutoffs, spectral-count numerics). The central claim—that the density-history and Maxwell-feedback terms are calculable from low-order data and predict the compensating pulse—is well supported internally.

Major objections (1)

[Sec. V] The total-time advantage in Table II rests on the specific common-input preparation (single pulse of amplitude 0.031997738..., width 0.5) and on the chosen pulse family (47) with ν=3. The claimed advantage is therefore conditional on a particular input and on the analytic minimization within that two-pulse family; the paper states this scope but does not quantify how the advantage degrades for other common inputs or other pulse families. Since Sec. V is presented as a general result ('Charge-neutral controls retain a total-time advantage'), I recommend either softening the claim to the tested family or adding a short argument or example showing that the advantage is not an artifact of the particular input amplitude and phase reset chosen.

How Pith checks papers → · Full audit of this paper →

The paper

The paper claims

Boundary drives that end at the same charge density can leave different relaxation channels excited, and this paper shows that a cubic correction computed from low-order equations suppresses the slowest channel by roughly 600-fold in a…

The reading

What a finite boundary drive does to a system is not fixed by where it ends: two drives that end at the same charge density can excite different decay channels, and the paper demonstrates this in a holographic superconductor on a fixed AdS black brane. The central result is that preparation history leaves a third-order slow tail even after the linear slow-mode response is cancelled by a two-pulse drive: the asymptotic slow amplitude is $C_1 = \epsilon^3(H_1\beta + G) + O(\epsilon^5)$, where the complex coefficient $G = G_{\mathrm{prep}}^a + G_{\mathrm{prep}}^\rho + G_{\mathrm{post}}^a$ collects density history during the drive, radial Maxwell feedback during the drive, and nonlinear feedback after release. For the tested protocols at final density $\rho_f = 3$, the density-history term dominates, carrying 85–89% of the modulus of $G$ and interfering constructively with the net radial feedback. Because $G$ follows from the low-order equations alone, the compensating second-pulse correction $\beta_3 = -G/H_1$ is a prediction rather than a fit, and full nonlinear evolution confirms the residual slow amplitude drops from order $\epsilon^3$ to order $\epsilon^5$, about 600-fold at $\epsilon = 0.05$. The payoff is a longer window of fast-channel relaxation that grows logarithmically in $1/\epsilon$, achieved without changing the quasinormal-mode spectrum; the same construction, extended to three pulses, suppresses two slow modes, and charge-neutral protocols from a common input retain a total-time advantage.

What carries the argument

The load-bearing object is the asymptotic slow-mode amplitude $C_1(w_0) = \ell_1 w_0 + \int_0^\infty e^{-\lambda_1 s}\ell_1 N_3(w(s))\,ds$, a variation-of-constants formula that projects the full nonlinear evolution onto the slowest quasinormal mode $\lambda_1 \simeq -0.177 - 0.373i$; it differs from the instantaneous projection precisely because the cubic Maxwell nonlinearity $N_3(w) = ia_\psi w + \frac{i}{2}a'_\psi Iw$ keeps feeding the slow channel after release. The cancellation condition $C_1 = 0$ defines an invariant graph, a strong stable manifold near the normal state, whose nonstationary points the authors call strong Mpemba states, and a state on this graph relaxes at a faster channel's rate even when its instantaneous slow overlap is nonzero. The graph is built from two exact ingredients: the constrained Maxwell equation, which is quadratic in the scalar and nonlocal in radius, and the Ward identity $\dot\rho = 2\,\mathrm{Im}[J^*\langle O\rangle]$, which turns the boundary source into a second-order density excursion during the drive. Around a two-pulse protocol that already cancels the linear response, the density excursion and the radial Maxwell feedback combine into the cubic coefficient $G = G_{\mathrm{prep}}^a + G_{\mathrm{prep}}^\rho + G_{\mathrm{post}}^a$, and the identity $\beta_3 = -G/H_1$ converts this low-order calculation into a pulse correction that full nonlinear evolution then confirms. The spectral closures, a converged winding count identifying exactly two simple quasinormal modes in $\mathrm{Re}\,\lambda > -2.9$ and a high-frequency bound, guarantee that cancelling this pair is what exposes the faster channels.

Extended reading notes

Core claim

The paper's central claim is that the right control target for strong-Mpemba preparation is not the instantaneous slow-mode overlap at release but the complete asymptotic slow-mode amplitude $C_1$, which includes nonlinear Maxwell feedback that continues after the source is removed. Working in the probe Abelian–Higgs model on a fixed AdS$_4$ black brane at final density $\rho_f = 3$, the authors show that after linear slow-mode cancellation the asymptotic amplitude is $C_1 = \epsilon^3(H_1\beta + G) + O(\epsilon^5)$, with $G = G_{\mathrm{prep}}^a + G_{\mathrm{prep}}^\rho + G_{\mathrm{post}}^a$ the sum of three cubic contributions. The density-history term $G_{\mathrm{prep}}^\rho$, generated by the charge excursion during the drive through the Ward identity, has modulus 85–89% of $|G|$ for the tested protocols and interferes constructively with the net radial feedback, so replacing the density by its final value throughout preparation would omit the dominant effect. Setting the bracket to zero fixes the compensating pulse $\beta_3 = -G/H_1$ from the low-order equations alone, and full nonlinear evolution verifies both quadratures of that prediction, the $\epsilon^5$ scaling of the residual, and the roughly 600-fold suppression of the slow amplitude at $\epsilon = 0.05$. The same construction, with two complex controls, suppresses two slow modes with a three-pulse preparation, and a charge-neutral version from a common input retains a total-time advantage over passive relaxation.

Load-bearing premise

The probe limit carries the entire quantitative structure: the spectrum, the coefficient $G$, and the compensating pulse are all computed on a fixed AdS$_4$ black brane at $\rho_* = 3$, treating the scalar and Maxwell fields as matter that moves on the geometry without bending it, and the authors state that the backreacted extension is left open.

Editorial extensions

If this is right

The fast-relaxation window doubles under compensation: the coefficient of the $\log(1/\epsilon)$ growth of the envelope-crossover time goes from $2/(\gamma_2-\gamma_1)$ under linear cancellation to $4/(\gamma_2-\gamma_1)$ under cubic compensation, so the advantage grows as the drive weakens.
Suppression is stackable: a three-pulse preparation cancels both resolved slow quasinormal modes within numerical resolution, exposing two faster modes whose coherent interference can make the instantaneous logarithmic decay rate oscillate even while the relaxation distance falls monotonically.
Control precision sets the achievable window: finite pulse errors restore a slow amplitude linearly, with $|C_1| < 10^{-8}$ requiring $|\delta\alpha_2| < 2.78 \times 10^{-8}$ at fixed final charge.
Charge conservation does not kill the effect: for a fixed common input with zero net charge transfer, a regular circle of pulse controls suppresses the leading slow response and reaches $D = 10^{-10}$ at $t \simeq 4.0$–$4.4$ including control time, against 34.5 for passive relaxation.
The quasinormal-mode spectrum is unchanged by compensation, so the faster relaxation is a redistribution of excitation across existing decay channels rather than a change in the decay rates themselves.

Reading between the lines

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

A natural testable extension is to rerun the two-pulse protocol with the metric dynamical: under full backreaction the spectrum, the coefficient $G$, and hence $\beta_3$ should all shift, and measuring whether the density-history term still dominates $G$ would show how much of the compensation mechanism survives geometry response.
The doubling pattern suggests a general ladder: each additional two orders of slow-amplitude suppression should lengthen the logarithmic fast window by $2/(\gamma_2-\gamma_1)\,\log(1/\epsilon)$, so higher-order pulse families would extend the effect beyond the cubic correction demonstrated here.
The boundary diagnostic derived in the appendix, a three-sample filter that isolates the slow-tail amplitude from the boundary order parameter alone, would be the natural observable to port to cold-atom or trapped-ion Mpemba experiments where bulk access is unavailable.
Because $G$ is protocol-dependent in magnitude and phase, the construction doubles as a characterization tool: measuring the residual slow amplitude across pulse widths and separations reconstructs $G$, and any deviation from the predicted $\epsilon^5$ scaling would expose a slow mode missed by the spectral count.

Original abstract

We study the boundary preparation of strong Mpemba states in a probe Abelian Higgs model on a fixed AdS black brane. Different drives ending at the same charge density can prepare states with different slow mode amplitudes. Two pulses cancel the slowest mode at linear order, but charge exchange during preparation and radial Maxwell feedback regenerate a slow tail at cubic order. We derive the complete cubic response from the lower order field equations, including feedback after source removal, and use it to predict a compensating pulse. Full nonlinear evolution confirms that the compensation reduces the residual slow amplitude from third to fifth order in the source strength. For a representative protocol, the slow amplitude is reduced by a factor of about 600 relative to linear cancellation. This extends the fast relaxation window without changing the final quasinormal mode spectrum. Further nonlinear tuning suppresses the estimated residual amplitude to the level of numerical resolution. The prepared states can overtake a weakly driven reference despite having a larger excess free energy at source removal. Three pulses also suppress a second slow channel over the resolved time window and reveal interference between faster modes. Starting from a common initial state, controls with zero net charge transfer reach a given relaxation threshold sooner than passive relaxation, even when the full control duration is included.

The compiled record

The paper's own sentences, arranged, kept beside the review and separate from it. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. Numbers in the tables are the authors', checked by code. No sentence in this record was written by a model.

Compiled by classifier-1.13.0 via TypeSafe API, question bank abstract-p1-20260923, 2026-10-20: 1 request, 8,126 input tokens.

What the abstract itself states, as probabilities: claims first 0.04 · states limitation 0.52 · human participants 0.01 · controlled comparison 0.90 · randomized assignment 0.01 · states quantitative result 0.98 · compares with prior results 0.66 · mentions code or data availability 0.02.

The abstract, sentence by sentence

arXiv listing abstract v1. Results and the main claim are underlined; show every role.

Finite boundary drives ending at the same charge density can select different relaxation channels. We demonstrate this in a probe Abelian--Higgs model on a fixed AdS black brane with a stable normal endpoint. Charge exchange during preparation modifies scalar propagation and source coupling. Density history and radial Maxwell feedback during and after the drive leave a cubic slow amplitude even after linear slow-mode cancellation. For the tested protocols at $\rho_f=3$, the density-history term interferes constructively with the net radial feedback; its modulus is 85--89\% of that of the total cubic coefficient. Calculated from the low-order equations alone, this total coefficient predicts the compensating pulse. Full nonlinear evolution confirms the reduction from $O(\epsilon^3)$ to $O(\epsilon^5)$ in the source amplitude $\epsilon$, with approximately 600-fold suppression for the representative protocol at $\epsilon=0.05$. Compensation extends the window of fast-channel relaxation without changing the final quasinormal-mode spectrum; early distance crossing already arises from linear interference. Fully tuned two- and three-pulse preparations suppress one and two slow modes, respectively, within numerical resolution. Local analysis establishes finite-pulse reachability of strong Mpemba states with vanishing selected asymptotic slow-mode coefficients. Charge-neutral controls retain a total-time advantage from a common input when the full subsequent control duration is included.

background contribution method result interpretation main claim

What the classifier reports for each sentence. The number is the probability it gave the label.
#	sentence	role	the authors' own work	states the main conclusion	strength
s0	Finite boundary drives ending at the same charge density can select different relaxation channels.	result 0.77	0.37	0.43	hedged possibility 0.58
s1	We demonstrate this in a probe Abelian--Higgs model on a fixed AdS black brane with a stable normal endpoint.	method 0.83	0.91	0.07	asserted 1.00
s2	Charge exchange during preparation modifies scalar propagation and source coupling.	result 0.69	0.58	0.06	asserted 1.00
s3	Density history and radial Maxwell feedback during and after the drive leave a cubic slow amplitude even after linear slow-mode cancellation.	result 0.99	0.86	0.16	asserted 1.00
s4	For the tested protocols at $\rho_f=3$, the density-history term interferes constructively with the net radial feedback; its modulus is 85--89\% of that of the total cubic coefficient.	result 1.00	0.94	0.11	asserted 1.00
s5	Calculated from the low-order equations alone, this total coefficient predicts the compensating pulse.	result 0.64	0.88	0.09	asserted 0.99
s6	Full nonlinear evolution confirms the reduction from $O(\epsilon^3)$ to $O(\epsilon^5)$ in the source amplitude $\epsilon$, with approximately 600-fold suppression for the representative protocol at $\epsilon=0.05$.	result 1.00	0.95	0.18	asserted 1.00
s7	Compensation extends the window of fast-channel relaxation without changing the final quasinormal-mode spectrum; early distance crossing already arises from linear interference.	interpretation 0.58	0.89	0.26	asserted 1.00
s8	Fully tuned two- and three-pulse preparations suppress one and two slow modes, respectively, within numerical resolution.	result 0.96	0.94	0.20	asserted 0.99
s9	Local analysis establishes finite-pulse reachability of strong Mpemba states with vanishing selected asymptotic slow-mode coefficients.	result 0.72	0.89	0.38	asserted 1.00
s10	Charge-neutral controls retain a total-time advantage from a common input when the full subsequent control duration is included.	result 0.94	0.78	0.10	asserted 1.00

What kind of paper the classifier reports this is: theoretical result 0.81 · evidence from simulation or computation 0.94.

What was examined, and what was not

sources read

arXiv listing abstract v1

not examined

full text
figures and tables
references and citing sentences
relations to other papers

levels run

P abstract

not run

S sections
T body sentences
C citations
R cross-paper relations

Every quoted sentence on this record is the authors'. Every label is from Pith's closed vocabulary, with the probability shown beside it. Every number in the measurements table is the authors', checked by code. Record as JSON

Discussion

Disagree with the review, or know something it missed? Leave a comment, highlight any passage above to comment on it, or challenge this review formally.