{"id":"2c3323b5-1674-4fbc-b0bb-7ed43157c1c9","arxiv_id":"2412.02616","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":8,"one_line_summary":"Anisotropic, nonthermal fluctuations can transiently stabilize order parameter configurations that are not minima of the equilibrium free energy, through mutual feedback between order and fluctuations.","lead":"This paper shows that after a sudden temperature jump, slow anisotropic fluctuations of two coupled order parameters can temporarily create a new stable state that does not exist in the equilibrium energy landscape. The authors propose this 'nonthermal order by disorder' mechanism as a general explanation for light-induced hidden phases in superconductors, kagome metals, and orbitally ordered materials.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The central claim depends on the Gaussian closure for fluctuation renormalization; without a non-Gaussian benchmark, the transient minima at φ≈π/2 could be artifacts of that closure.","rationale":"The paper's central claim is that transient stabilization of nonthermal order arises because nonthermal fluctuations renormalize the effective free-energy landscape and create minima absent from the bare potential. The derivation of those minima is performed entirely within a Gaussian ansatz for P[ϕ](t), and the fluctuation force constants in Eq. (7) are linear in the two-point functions. Thus the existence of the transient minimum at φ ≈ π/2 is a direct consequence of the closure. This is the most load-bearing assumption because every case—cooperative, competitive, and C4-symmetric—uses the same mechanism, and the Methods explicitly state that the large-N expansion cannot be used for N = 2. The authors acknowledge the approximation but provide no independent check, and the code is not public. A direct stochastic TDGL simulation on a lattice, without the Gaussian ansatz, would settle this: if the plateau persists and fourth-order cumulants remain small, the claim is robust; if the plateau vanishes, the central conclusion does not extend beyond the closure. The rest of the argument is internally consistent: the effective potential is a valid descriptor of a gradient-like dynamics with anisotropic damping, and the transient minima shown in the figures are indeed absent from the bare equilibrium potential. I therefore see no more serious flaw than the one the reader identified, and the conditional verdict remains appropriate.","tokens_in":23938,"tokens_out":5746,"duration_ms":66809,"concrete_test":"Simulate the same stochastic time-dependent Ginzburg-Landau dynamics—Eq. (2) with V from Eq. (1), K from Eq. (12), noise from Eq. (4), and T(t) from Eq. (5)—on a 3D lattice without imposing the Gaussian closure, using an Euler-Maruyama scheme with the parameters of Fig. 4 (Γ1 = 0.9, Γ2 = 0.5, K1 = 0.1, K2 = 5.0, Ti = 0.025, Tq = 4, τ = 0.3, Tc = 0.5, r0 = 15, u1 = 0.9, u2 = 1.0, and a lattice cutoff corresponding to Λc = 2π). Average over many noise realizations and check whether ⟨R(t)⟩ and ⟨φ(t)⟩ exhibit the plateau at R ≈ 3.85, φ ≈ π/2 for t ≈ 22–30; also measure the fourth-order cumulant of the field distribution at the plateau. If the plateau persists and the non-Gaussian cumulants are small relative to the Gaussian values, the closure is adequate; if not, the transient minima are an artifact of the Gaussian approximation.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The proof of principle is built around the effective potential F̄(ϕ̄1, ϕ̄2) of Eq. (9), whose fluctuation-induced quadratic coefficients r_fl^αβ in Eq. (7) are linear in the two-point functions n_αβ(t) of Eq. (8). The Gaussian ansatz for P[ϕ](t) is invoked in the Methods around Eq. (15) to apply Wick's theorem and discard all higher-order correlations. Every reported case—cooperative, competitive, and C4-symmetric—relies on the same mechanism: after the quench, anisotropic fluctuations n11 > n22 renormalize the effective coefficients so that a transient minimum appears at φ ≈ π/2 (Figs. 4–7 and 8–9). Because this minimum is generated by precisely the fluctuation variables that are truncated at Gaussian order, the result is not independently supported: the authors note in the Methods that the large-N expansion is not available for N = 2 (Ref. [86]), and no comparison with a non-Gaussian calculation, microscopic simulation, or public code is provided. If the actual field distribution develops sizeable fourth-order cumulants during the strongly nonlinear stage when R ∼ 0, the effective quadratic coefficients in Eq. (7) are no longer sufficient, and the transient minimum could disappear or shift substantially.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript examines the relaxational dynamics of a two-component Ginzburg-Landau model with Z2×Z2 or C4-symmetric potential, coupled to its own order-parameter fluctuations through a Gaussian closure. The equations of motion for the homogeneous averages φ̄α and the correlation functions Dαβ(k,t) are solved numerically after an incoherent temperature quench (and, in one case, a coherent drive). The authors define an effective potential F̄(φ̄1,φ̄2) whose gradient reproduces the average-field dynamics at fixed fluctuation content, and show that anisotropic fluctuations n11>n22, caused by different stiffnesses or relaxation rates of the two order parameters, renormalize F̄ such that a transient minimum appears at values such as R≈3.85, φ≈π/2, which are not minima of the bare equilibrium potential. The state persists as a plateau in the dynamics, leading the authors to propose a nonequilibrium order-by-disorder mechanism (NOBD). The final sections discuss relevance for photoinduced superconductivity, density-wave order, and orbital systems.","tokens_in":24221,"tokens_out":22251,"duration_ms":203545,"significance":"If the central claim survives a test of the Gaussian closure, the paper provides a conceptually new and rather general mechanism for transient nonthermal order that does not rely on pre-existing metastable states. The connection to order-by-disorder is well motivated, and the explicit numerical demonstrations for three model situations are useful. The manuscript is clearly written and the equations are internally consistent; the authors are transparent about the Gaussian approximation and its lack of a small parameter. The main weakness is that the transient minima are generated by precisely the fluctuation variables that are truncated at Gaussian order, and no non-Gaussian benchmark or error estimate is provided. Thus, while the ideas are attractive, the proof of principle remains conditional.","major_comments":[{"comment":"The central result rests entirely on the Gaussian closure for the probability distribution P[ϕ](t). In the Methods, the authors state that Wick's theorem is used to obtain Eq. (15) and that the large-N expansion is not performed because N=2 and the potentials lack continuous symmetry. Since the transient minima in Figs. 4-9 arise solely through the fluctuation-induced coefficients rfl_αβ in Eq. (7), which are linear in the two-point functions nαβ(t), the closure is load-bearing. The quartic couplings in the model are not small (e.g., u2R^4/4 ~ 64 for R~4), so there is no perturbative control. I ask the authors to provide a quantitative test of the closure, for example by numerically integrating the full Langevin dynamics in a reduced geometry or a few-mode truncation, or by comparing with a cumulant truncation that retains fourth-order cumulants. In the absence of such a test, the possibility remains that the plateau at φ≈π/2 is an artifact of the Gaussian approximation. A discussion of the expected magnitude of non-Gaussian correlations in the regime of Figs. 4-9 would be the minimum response.","section":"Methods, Eq. (15); Results, Figs. 4-9"},{"comment":"The effective potential F̄ is defined so that its gradient reproduces the equations of motion, with the fluctuation coefficients rfl_αβ(t) evaluated at the current time. Thus F̄ is a time-dependent pseudo-potential, not a true free energy functional; any point where the dynamics is slow is, by construction, a stationary point of F̄ at that instant. The manuscript should state this explicitly and qualify statements such as 'transient free energy minima' and 'stabilization of ordered states at transient free energy minima' in the abstract and introduction as referring to the instantaneous pseudo-potential. The physical effect is the long-lived dynamical plateau in the coupled φ̄-n dynamics, not the existence of a minimum of an independently defined free energy. This clarification is important because the significance claim hinges on the contrast with equilibrium free-energy minima.","section":"Effective potential, Eq. (9)"}],"minor_comments":[{"comment":"The text states 'the critical value r2,0/r1,0 = u2/(2u1 − u2)'; with the given parameters r2,0=12, r1,0=15, u1=0.9, u2=1.0, the left side is 0.8 while u2/(2u1−u2)=1.25. The correct critical condition for the dominant order is r1,0/r2,0 = u2/(2u1−u2), or equivalently r2,0/r1,0 = (2u1−u2)/u2. Please correct the formula in the main text and in the caption of Fig. 6.","section":"Almost competitive order section and caption of Fig. 6"},{"comment":"The numerical integration of Eqs. (6) and (15) is not described. Please specify the time-stepping scheme, the discretization of the k-integration, the treatment of the ultraviolet cutoff Λc, and the convergence checks (e.g., dependence on time step and k-grid).","section":"Methods"},{"comment":"The caption's description of the line styles ('solid (dotted) lines are characterized by Γ=0.5 (Γ=0.9)') is ambiguous because the text also refers to dotted gray and green lines. Please label the curves directly or use a more explicit legend.","section":"Isotropic fluctuations section, caption of Fig. 3"},{"comment":"The notation ¯Ki,α = rα,0 K δ_{i,α} is confusing because the indices i and α label different spaces. Consider writing the stiffness explicitly as K(ϕ1,ϕ2) = (K/2)[r1,0 (∂1ϕ1)^2 + r2,0 (∂2ϕ2)^2].","section":"C4 model, Eq. (13)"},{"comment":"The statement that data 'can be made available after request' is weak for a computational proof of principle. Making the code and data publicly available (e.g., in a repository) would strengthen reproducibility.","section":"Data and code availability"}],"recommendation":"major_revision","confidential_remarks":"The manuscript addresses a question of current interest and the proposed mechanism is elegant. The main technical concern is the uncontrolled Gaussian closure; if the authors can provide a non-Gaussian check, the paper would be a strong contribution. The critical-ratio typo and the numerical details are easily fixed. The paper is within the scope of the journal. I also note that the 'Note added' is slightly unusual but acceptable."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The thing to know: this is a conceptually new mechanism, not just a tweak. The paper shows that after a quench, anisotropic fluctuations (n11 > n22) can renormalize the effective free energy so that a transient minimum appears at φ ~ π/2, even when the equilibrium free energy has no metastable states. That is a real departure from the earlier fluctuation-softening literature (Ref. [29]) and from relaxation-time separation arguments (Ref. [56]).\n\nWhat is good: the derivation is clean and internally consistent. The effective potential is defined to reproduce the equations of motion, and the plateau in the dynamics (R ~ 3.85, φ ~ π/2 etc.) is actually there in the numerical solution. No parameter was fitted to the plateau; the fluctuation variables n_αβ evolve dynamically rather than being imposed. The authors also explicitly note the overlap with Ref. [87] and are candid about the limitations of their approach.\n\nThe soft spot: the whole result rests on a Gaussian closure for the probability distribution. All correlations beyond two-point functions are dropped. The authors acknowledge that the large-N expansion is not available for N=2 (Ref. [86]), and they provide no comparison with a non-Gaussian calculation or a microscopic simulation. The stress-test worry is legitimate: if the actual distribution develops substantial fourth-order cumulants during the strongly nonlinear stage near R ~ 0, the transient minimum could shift or disappear. That said, this is a proof-of-principle in toy Ginzburg-Landau models, and the mechanism is plausible; the closure is not obviously wrong, just uncontrolled.\n\nA minor point: calling the state a minimum of F_bar is partly definitional, since F_bar is constructed so its gradient matches the equations of motion. But that does not cheapen the result; the plateau in the physical dynamics is what matters.\n\nAlso, no public code or data. For a theory paper of this type, that is a reproducibility concern, though the equations are given in full.\n\nOverall: a serious, honest paper. It deserves peer review. I would ask the authors to either provide code/data or add a non-Gaussian benchmark (e.g., a higher-order closure or a small microscopic simulation) to test the robustness of the mechanism. I would cite it if I worked on nonequilibrium order.","headline":"A clean proof-of-principle that anisotropic nonthermal fluctuations can transiently stabilize states that are not equilibrium minima; the main caveat is the uncontrolled Gaussian closure, but the paper is honest about it and deserves serious refereeing.","tokens_in":24775,"tokens_out":2314,"would_cite":true,"duration_ms":23576,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["64.60.-i","05.70.Ln"],"model":"deepseek-v4-flash","headline":"Nonthermal fluctuations can trap an order parameter in a state that has no equilibrium counterpart.","keywords":["nonthermal order","order by disorder","quench dynamics","Ginzburg-Landau theory","fluctuation renormalization","hidden states","intertwined orders","model A dynamics"],"falsifier":"Simulate the same two-order-parameter Langevin dynamics on a finite lattice without the Gaussian closure, for example by direct stochastic sampling of the noise, and check whether the plateau at $R\\approx 3.85$, $\\varphi\\approx\\pi/2$ (or $R\\approx 3$, $\\varphi\\approx\\pi/2$) still appears after the quench; if the plateau disappears, the trap was created by the Gaussian closure. Alternatively, measure the momentum-resolved fluctuation anisotropy with time-resolved resonant elastic x-ray scattering during the plateau: the mechanism requires $D_{11}$ and $D_{22}$ to be distinctly elongated along the softer direction, so the absence of that anisotropy would rule out nonthermal order by disorder.","tokens_in":23724,"feed_emoji":"⚡","tokens_out":7785,"duration_ms":72880,"temperature":0.7,"pith_summary":"After an ultrafast quench, the slow order-parameter fields of a material can be trapped for a while in an ordered configuration that is not even a local minimum of the equilibrium free energy. The paper proves this is possible in a minimal two-order-parameter Ginzburg-Landau theory, provided the two fields have different stiffnesses or relaxation rates, so that the quench leaves an anisotropic bath of fluctuations. Those nonthermal fluctuations renormalize the effective free-energy landscape and open a new transient minimum; the trapped order, in turn, changes the fluctuation spectrum, so order and fluctuations keep each other alive until the fluctuations decay. The authors call this nonthermal order by disorder, a nonequilibrium analogue of equilibrium order-by-disorder selection, and argue it may underlie light-induced hidden states in cuprates, kagome metals, and orbitally ordered compounds.","feed_headline":"Quench fluctuations can stabilize orders absent from equilibrium","feed_subtitle":"Nonthermal fluctuations can trap two coupled order parameters in a state that equilibrium forbids.","key_machinery":"The central object is the fluctuation-renormalized effective potential $\\bar{F} = \\bar{F}_0 + \\bar{F}_{\\mathrm{fl}}$ of Eq. (9), whose derivative by construction reproduces the equation of motion for the average order parameter. The fluctuation correction $\\bar{F}_{\\mathrm{fl}}$ is built from the fluctuation force constants $r^{\\mathrm{fl}}_{\\alpha\\beta}$ of Eq. (7), which depend on the integrated fluctuation populations $n_{\\alpha\\beta}(t)=\\int D_{\\alpha\\beta}(k,t)\\,dk/(2\\pi)^d$. The dynamics closes under the Gaussian approximation: Wick's theorem turns the average-field equation into Eq. (6) and the correlation functions into Eq. (15), with an effective quadratic coefficient $r^{\\mathrm{eff}}_{\\alpha\\beta}=\\bar{r}_{\\alpha\\beta}+r^{\\mathrm{fl}}_{\\alpha\\beta}$. The load-bearing asymmetry is the combination of unequal stiffnesses $K_\\alpha$ and unequal relaxation rates $\\Gamma_\\alpha$, which makes $n_{11}\\neq n_{22}$ after the quench; the quartic anharmonic terms then make the mass term for the fluctuations depend self-consistently on $\\bar{\\phi}$, so order and fluctuations lock into a transiently stable configuration.","core_discovery":"The paper's central claim is that a genuine nonequilibrium drive can transiently stabilize an ordered state even when the equilibrium free energy has no metastable minima at all. Working with two coupled order parameters $\\bar{\\phi}_1,\\bar{\\phi}_2$ governed by a $\\mathbb{Z}_2\\times\\mathbb{Z}_2$-symmetric Landau potential plus a gradient stiffness, the authors show that after a temperature quench the fluctuation-renormalized effective potential $\\bar{F}(\\bar{\\phi}_1,\\bar{\\phi}_2)$ develops minima that are absent from the bare potential $V$: in the cooperative case the order parameter is trapped at $R\\approx 3.85$, $\\varphi\\approx\\pi/2$, and in the almost-competitive case at $R\\approx 3$, $\\varphi\\approx\\pi/2$. The trap appears only when the quench produces an anisotropic fluctuation population, here $n_{11}>n_{22}$, realized by making the stiffnesses and relaxation rates of the two fields unequal ($K_1<K_2$, $\\Gamma_1>\\Gamma_2$). Because the quartic coupling makes the fluctuation mass term depend on $\\bar{\\phi}$ itself, the trapped order and the anisotropic fluctuations feed back on each other and mutually stabilize during a finite time window before the equilibrium state is recovered. The same mechanism is demonstrated in a $C_4$-symmetric compass-model continuum theory, where a coherent pulse first breaks the $C_4$ symmetry and a subsequent quench stabilizes the nonthermal order.","pith_inferences":["If the Gaussian closure is relaxed, the plateau lifetime and even its existence could change; a stochastic-lattice simulation would tell whether the trap survives beyond Gaussian statistics.","The same effective-potential mechanism should generalize to more than two coupled order parameters or to vector order parameters with anisotropic stiffness, where the transient landscape could contain several competing nonthermal minima.","The NOBD picture reframes hidden-state searches: instead of mapping only the minima of the equilibrium free energy, experiments should look for transient anisotropic fluctuation distributions, since those are what define and stabilize the nonthermal state."],"forward_implications":["A pure temperature quench can create a transient ordered state in a two-order-parameter system even though the equilibrium landscape has no local minima, so hidden states need not be pre-existing metastable states.","The transient state is self-stabilizing: during the plateau time window, the nonthermal fluctuations and the nonthermal order renormalize each other's mass terms and delay the return to equilibrium.","The mechanism works for cooperative, almost-competitive, and symmetry-lowered $C_4$ setups, and the required ingredient is only an anisotropy in stiffness and/or relaxation, making it a generic route to nonthermal order.","The momentum-space signature is an anisotropic, elongated distribution of the correlation functions $D_{11}(k,t)$ and $D_{22}(k,t)$, observable with time-resolved resonant elastic x-ray scattering or ultrafast transmission electron microscopy.","For materials with intertwined orders, the model predicts which order parameter is transiently enhanced based on the sign of the stiffness and relaxation-rate asymmetry, connecting to light-induced superconductivity and density-wave experiments."],"supporting_citations":[{"why":"Supplies the known fluctuation-induced softening and slowdown of the order-parameter dynamics in a rotationally invariant $\\phi^4$ theory, the effect the authors generalize to the anisotropic two-field case.","marker":"[29]"},{"why":"Provides the experimental pump-probe context in tritelluride materials where the Gaussian approximation for order-parameter fluctuations was applied, supporting the closure used here.","marker":"[26]"},{"why":"Introduces the $\\mathbb{Z}_2\\times\\mathbb{Z}_2$ Ginzburg-Landau description of VO$_2$ that motivates one of the physical interpretations of the two order parameters.","marker":"[61]"},{"why":"Gives the 90-degree compass model whose continuum limit is the $C_4$-symmetric model used for the coherent-control demonstration.","marker":"[34]"},{"why":"Defines model A relaxational time-dependent Ginzburg-Landau dynamics, the equation of motion used throughout the paper.","marker":"[57]"},{"why":"Reports the light-induced persistent superconducting phase in LBCO that the cooperative-order scenario is designed to explain.","marker":"[72]"},{"why":"Reports the transient enhancement of charge-density-wave order at the expense of superconductivity in YBCO, the competitive-order example.","marker":"[73]"}],"fun_headline_variants":["Transient order from nonthermal fluctuations","Quench stabilizes order absent from equilibrium","Nonequilibrium fluctuations trap forbidden order","Anisotropic quench yields transient order","Quench-driven transient order beyond equilibrium"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument assumes the probability distribution of the fluctuating order-parameter fields stays Gaussian around its average, so all fluctuation effects are captured by two-point correlations; if non-Gaussian correlations grow after the quench, the transient minima could be an artifact of the closure.","fun_headline_variants_meta":{"raw":{"variants":["Transient order from nonthermal fluctuations","Quench stabilizes order absent from equilibrium","Nonequilibrium fluctuations trap forbidden order","Anisotropic quench yields transient order","Quench-driven transient order beyond equilibrium"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000325,"raw_usage":{"total_tokens":1847,"prompt_tokens":999,"completion_tokens":848,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":615,"completion_tokens_details":{"reasoning_tokens":785}},"tokens_in":615,"tokens_out":848,"duration_ms":8645,"temperature":1.0,"reasoning_tokens":785,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T23:15:32.608590+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Simulate the same two-order-parameter Langevin dynamics on a finite lattice without the Gaussian closure, for example by direct stochastic sampling of the noise, and check whether the plateau at $R\\approx 3.85$, $\\varphi\\approx\\pi/2$ (or $R\\approx 3$, $\\varphi\\approx\\pi/2$) still appears after the quench; if the plateau disappears, the trap was created by the Gaussian closure. Alternatively, measure the momentum-resolved fluctuation anisotropy with time-resolved resonant elastic x-ray scattering during the plateau: the mechanism requires $D_{11}$ and $D_{22}$ to be distinctly elongated along the softer direction, so the absence of that anisotropy would rule out nonthermal order by disorder.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the known fluctuation-induced softening and slowdown of the order-parameter dynamics in a rotationally invariant $\\phi^4$ theory, the effect the authors generalize to the anisotropic two-field case."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the experimental pump-probe context in tritelluride materials where the Gaussian approximation for order-parameter fluctuations was applied, supporting the closure used here."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduces the $\\mathbb{Z}_2\\times\\mathbb{Z}_2$ Ginzburg-Landau description of VO$_2$ that motivates one of the physical interpretations of the two order parameters."},{"cited_title":"Moessner, Canadian Journal of Physics 79, 1283 (2001), URL https://doi.org/10.1139/p01-123","cited_arxiv_id":null,"evidence_quote":"Gives the 90-degree compass model whose continuum limit is the $C_4$-symmetric model used for the coherent-control demonstration."},{"cited_title":"Transient selection of competing order in frustrated spin Peierls systems after a quench","cited_arxiv_id":"2409.19955","evidence_quote":"Defines model A relaxational time-dependent Ginzburg-Landau dynamics, the equation of motion used throughout the paper."},{"cited_title":"Domr¨ ose, T","cited_arxiv_id":null,"evidence_quote":"Reports the light-induced persistent superconducting phase in LBCO that the cooperative-order scenario is designed to explain."}],"review_version":1}