Pith. sign in

REVIEW 5 major objections 7 minor 1 cited by

Non-equilibrium steady state for a three-mode energy cascade model

T0 review · 5 major / 7 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read By solving a Feynman-Kac equation for a Lyapunov pre-factor, the paper proves that a three-mode stochastic model of resonant NLS dynamics has a unique non-equilibrium steady state and polynomial convergence to it.

desk verdict A genuinely new Feynman-Kac-Lyapunov construction; the main theorem's arbitrary-γ statement is not supported (only small γ is proven), but this is fixable and the paper deserves refereeing. read the letter →

arxiv 2505.16018 v1 pith:Z3GPTULE submitted 2025-05-21 math.PR math-phmath.DSmath.MP

Zaher Hani , Yao Li , Andrea Nahmod , Gigliola Staffilani This is my paper
classification math.PRmath-phmath.DSmath.MP MSC 60H1037A3060J60
keywords non-equilibriumsteadystateinvariantmeasureenergycascadeFeynman-KacLyapunovfunctionstochasticdifferentialequationergodicitywaveturbulence
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

The paper proves that a three-mode stochastic model of resonant NLS dynamics, with forcing and dissipation on the two end modes, has a unique non-equilibrium steady state (an invariant probability measure), and that convergence to it is at least polynomial in time. The model is a drastic reduction of the infinite-dimensional wave-turbulence cascade, but it keeps the central difficulty: the middle mode can only exchange energy when phase angles are aligned, so the system may overheat at high internal energy or freeze as $I_2 \to 0$. To control both dangers the authors construct Lyapunov functions with a pre-factor obtained by solving an elliptic Feynman-Kac equation, which assigns large values precisely to the bad phase configurations. The result is presented as a first rigorous statistical-steady-state viewpoint on energy cascades, complementary to the kinetic-equation derivation of Kolmogorov-Zakharov spectra.

What carries the argument

The Feynman-Kac-Lyapunov method: instead of guessing a Lyapunov function for the full five-dimensional diffusion, the authors look for a Lyapunov function of the form $x^\alpha f(y)$, where $y$ is a fast, lower-dimensional subsystem, and require the pre-factor $f$ to satisfy $L_y f + \alpha c f = -1$ with $c$ the coupling coefficient. By the Feynman-Kac representation, a positive solution exists when the fast subsystem is geometrically ergodic and the average of $c$ under its invariant measure is negative; eigenvalue perturbation theory then gives the principal eigenvalue. In the high-energy regime the fast subsystem is the single phase angle $\theta$ following $d\theta = I_2(1+2\cos\theta)\,dt + \sqrt{\gamma I_2}\,dB$, and in the low-energy regime it is the broken system $(I_1,I_3,\theta_1,\theta_3)$ at $I_2=0$. The pre-factor makes the Lyapunov function large exactly on the bad-phase set where $\sin\theta<\gamma$, so that the drift becomes negative uniformly.

What would settle it

Compute, for a fixed moderate $\gamma$ (say $\gamma=0.5$), the invariant density $u^*(\theta)$ from Lemma 4.4 for the reduced equation $d\theta=(1+2\cos\theta)\,ds+\sqrt{\gamma}\,dB$ and evaluate $\mathbb{E}_\pi[\sin\Theta^*]$. If this mean is $\le \gamma$, then $\pi(-2(\sin\theta-\gamma))\ge 0$, so no positive $f$ can solve $L_1 f=-1$, and the high-energy Lyapunov construction collapses for that $\gamma$; Theorem 1 as stated, with arbitrary $\gamma>0$, would then be false.

Watch

Extended reading notes

Core claim

The central claim is Theorem 1: for any $\beta_0>1$, $\gamma>0$, finite $T_1$, and $T_3$ sufficiently large, the stochastic system (1.8) has a unique invariant probability measure $\pi$ on $\Omega=\mathbb{R}_+^3\times\mathbb{T}^2$, and its transition kernels satisfy two explicit total-variation bounds with polynomial decay at rate $(t+1)^{-\beta_0}$. The proof combines two Lyapunov functions: $V=(I_1+I_2+I_3)^{\beta_0}+I_1^{-\beta_1}f(\theta_1)+I_3^{-\beta_1}f(\theta_3)$ controlling high internal energy, and $W=I_2^{-\alpha}U$ controlling the near-death of the middle mode, with $U$ built from a positive solution of a Feynman-Kac equation on a bounded domain of the reduced $(I_1,I_3,\theta_1,\theta_3)$ system. The mechanism is a time-scale separation: a fast subsystem (the phase angle, or the four-variable reduced system at $I_2=0$) mixes quickly and has negative average of the coefficient that feeds $I_2$, which is exactly the condition that makes the Feynman-Kac pre-factor exist and forces the Lyapunov drift negative.

Load-bearing premise

The high-energy Lyapunov bound assumes a positive pre-factor $f$ on the circle that makes $L^I f\le -\epsilon I$ for all large $I$; the paper proves this only for the noise weight $g=\sqrt{I}$ and for sufficiently small $\gamma$, while the theorem quantifies $\gamma>0$ arbitrarily.

Editorial extensions

If this is right

  • A unique non-equilibrium steady state exists for the three-mode chain whenever $T_3\gg T_1$ and $\gamma$ is small, with at least polynomial convergence rate $\beta_0>1$.
  • The invariant measure has finite moments with respect to the constructed Lyapunov function, so expectations of energy-type observables under the NESS are controlled.
  • The method gives a template for high-dimensional SDEs with degenerate noise on slow modes, provided one can prove geometric ergodicity and a negative mean coupling for a fast subsystem.
  • The paper establishes stochastic stability as a first step toward connecting NESS statistics with Kolmogorov-Zakharov cascade spectra, a connection the authors explicitly leave for future work.

Reading between the lines

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

  • Editorial inference: if the Feynman-Kac-Lyapunov construction survives for longer chains, the same two-regime argument (fast phase stabilization at high $I_2$; four-variable reduced dynamics at $I_2=0$) may yield NESS and polynomial ergodicity for the $n$-mode toy model, provided the reduced systems remain geometrically ergodic with negative mean coupling.
  • Editorial inference: a natural testable extension is numerical computation of the energy flux through the middle mode under the NESS; comparison of the flux scaling with the KZ exponents $|k|^{-d}$ and $|k|^{-d+2/3}$ would make the connection to wave turbulence quantitative rather than motivational.
  • Editorial inference: the method implies a concrete sufficient condition for stabilization of degenerate SDEs: a fast ergodic subsystem whose invariant measure makes the linearized coupling negative. This suggests applying the same pre-factor construction to other partially damped systems, such as shell models, where the same overheating/freezing dichotomy appears.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

5 major / 7 minor

Summary. The paper constructs a stochastic three-mode system derived from the resonant NLS toy model of Colliander--Keel--Staffilani--Takaoka--Tao, with stochastic heat baths attached to the first and third modes and a damping-dissipation structure designed to allow energy transfer through the middle mode. The main result, Theorem 1, asserts that for β0 > 1, γ > 0, and 0 < T1 < ∞, provided T3 is sufficiently large, there exist a noise weight g and a Lyapunov function V such that the SDE admits an invariant probability measure π with finite V-moment, and the transition semigroup converges polynomially in total variation. The proof combines a high-energy Lyapunov function based on a pre-factor f(θ) solving a Feynman--Kac equation (Assumption (H)) with a low-energy Lyapunov function based on a pre-factor U solving another Feynman--Kac equation (Assumption (L)). The paper also analyzes the reduced systems that arise in each regime, proving geometric ergodicity, explicit invariant densities for the angle processes, and sign/variance estimates for the energy-transfer coefficient h.

Significance. If the main theorem were established in the stated generality, this would be a notable contribution: it would give the first rigorous construction and polynomial ergodicity of a non-equilibrium steady state for a reduced model of the NLS energy cascade, using a genuinely novel Feynman--Kac-Lyapunov method. The paper contains substantial, detailed analysis of the reduced systems, including explicit invariant densities, concentration estimates, and multiplicative-ergodicity arguments, and it does not rely on fitted parameters or circular numerical inputs. However, as explained in the major comments, the theorem as stated is not fully supported by the proofs: the parameter range for γ exceeds what Assumption (H) is proved for, the verification of Assumption (L) invokes an elliptic-regularity theorem outside its hypotheses, and the treatment of the absorbing boundary I2 = 0 and the gluing of the low-energy pre-factor leave gaps in the uniqueness and Lyapunov-drift claims. These issues are substantive but appear repairable.

major comments (5)
  1. [§1.2, Theorem 1; §4, Assumption (H); §4.2, Theorem 4.9] Theorem 1 quantifies γ > 0 arbitrarily, but the high-energy Lyapunov construction relies entirely on Assumption (H)(d), which is verified only for the specific weight g = √I and for γ sufficiently small (Corollary 4.6 and Lemma 4.7). The small-γ restriction is not cosmetic: if γ ≥ 1, then sinθ − γ ≤ 0 everywhere, and at any interior minimum of a positive C^2 function f one has f' = 0 and f'' ≥ 0, giving L_I f ≥ 2β1 I(γ−1)f ≥ 0, which contradicts the required inequality L_I f ≤ −εI < 0. The theorem and the abstract's 'unique NESS' claim therefore exceed the proven parameter range, and Theorem 1 should be restated with 0 < γ ≪ 1 (as the model description already assumes).
  2. [§4.2, Theorem 4.9] The displayed formula for L_I in the proof of Theorem 4.9 reads L_I f = −4β1 I(sinθ−γ)f + 2I(1+2cosθ)f' + (γ/2) I f'', whereas Assumption (H)(d) and the calculation in Lemma 4.1 use L_I f = −2β1 I(sinθ−γ)f + I(1+2cosθ)f' + (γ/2) g² f''. With g = √I the divided operator should be L_1 f = −2β1(sinθ−γ)f + (1+2cosθ)f' + (γ/2)f'', which is exactly the form L_θ + β1 c(θ) for the generator L_θ of (4.9) at I2 = 1 and c(θ) = −2(sinθ−γ). The factor-of-two discrepancy makes the stated operator inconsistent with both Assumption (H) and the proof's use of Theorem 2.7; this step must be corrected before the verification of (H) is valid.
  3. [§5.1–§5.5, Theorem 5.12 and Assumption (L)] The proof of Assumption (L) applies Theorem 2.7 to the reduced system (5.1) on the domain Ω_M = {I1 + I3 ≤ M}. However, the diffusion coefficients of (5.1) vanish at I1 = 0 and I3 = 0, and the boundary of Ω_M has corners at (0,M) and (M,0); Theorem 2.7 and its input Theorem 2.5 require uniform ellipticity and a regular C^4 boundary. The manuscript does not regularize the domain, does not establish the Feynman–Kac equation on a nested family of smooth subdomains, and does not justify that the characteristic boundary can be ignored. Since Assumption (L) is the sole input to the low-energy Lyapunov function W in Theorem 5.1, this gap directly affects the proof of the main theorem.
  4. [§1.2 and §6.2, state space and uniqueness] The set {I2 = 0} is absorbing for (1.8), and on that set the system reduces to (5.1), which by Theorem 5.6 has its own invariant probability measure. Consequently the full Markov process on Ω = R_+^3 × T^2 has at least two distinct invariant measures: the constructed π supported on {I2 > 0} with ∫V dπ < ∞, and an invariant measure supported on {I2 = 0}. The uniqueness assertion in Section 7 and the convergence estimates in Theorem 1 must therefore be restricted to the recurrent class {I2 > 0} (equivalently, to initial conditions with finite V); as written, the statement 'the 3-mode chain admits a unique NESS' is false on the full state space.
  5. [§5.1, Theorem 5.1] The low-energy pre-factor U is defined as Q on {I1 + I3 ≤ M} and as I1 + I3 on {I1 + I3 > M}, with only the boundary value q = M matched at the interface. The normal derivative of Q at I1 + I3 = M is not controlled, so U is generally only C^0, not C^1, across that surface. Applying the generator to W = ψ(I2)U then produces a local-time term supported on the interface that is not estimated in the proof; the inequalities L W ≤ −δW and the final Lyapunov inequality are therefore not established in the usual sense on Ω. The proof should either use a C^2 smoothing that preserves the drift estimates or explicitly control the jump in the normal derivative.
minor comments (7)
  1. [§1.2, Eq. (1.8)] The definition of the angle variables contains an obvious typo: 'θ3 = 2(φ3 − φ3)' should read 'θ3 = 2(φ3 − φ2)'.
  2. [§4.2, definition of g(I,θ)] The definition of the interpolation function g contains repeated 'g' where 'I' is meant: '1 if g ≤ 1' should be '1 if I ≤ 1', and similarly for the other branches.
  3. [Lemma 4.4] In the formula for the constant C0, the denominator contains 'exp(−α(r+2πr))'; this should be 'exp(−α(r+2 sin r))'.
  4. [Lemma 5.8] The displayed normalization constant of the generalized Gamma density appears incorrect; the density as written does not integrate to 1 with that constant. The moment formulas used later (e.g., ⟨I3⟩) are consistent with the correct constant, so this appears to be a typographical error that should still be fixed.
  5. [Lemma 5.9] With α defined as in the proof, the correct relation is α = 2I3/γ, not α = 4I3/γ; the displayed stationary density should be normalized accordingly.
  6. [§5.1, proof of Theorem 5.1] The computation writes 'L_b(I1+I3) = 2I1I2(sinθ1−γ) + 2I3I2(sinθ3−γ) + γ(T1+T3−I1^3−I3^3)', but L_b is the generator of the reduced system (5.1) in which I2 = 0; the I2-dependent terms belong to the full generator, not to L_b.
  7. [Theorem 5.12] The statement of Theorem 5.12 omits the small coefficient α that appears in Assumption (L) and in the proof via Theorem 2.7; since the multiplicative-ergodicity argument only yields a solution for sufficiently small α, the theorem should either include α explicitly or state that the coefficient may be taken small.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the Lyapunov pre-factors are solved for via Feynman-Kac equations, and the noise coefficient g is a constructed model parameter, not a fitted input.

full rationale

The derivation chain is self-contained with respect to circularity. The Lyapunov pre-factors f (high-energy) and Q/U (low-energy) are constructed as positive solutions of Feynman-Kac Cauchy-Dirichlet problems (Section 3.1 and Assumptions (H), (L)); their existence is proved from negativity of averaged coefficients (Corollary 4.6, Theorem 5.7, Theorem 5.12) via Theorems 2.7 and 2.3, rather than assumed from the desired conclusion. The noise weight g = sqrt(I) is chosen and then verified in Theorem 4.9 to satisfy Assumption (H); since Theorem 1 is an existence statement over g, this is a legitimate construction of a model parameter, not a fitted input renamed as a prediction. The starting model is taken from the published independent work [31] and standard ergodicity results [24, 38, 40]; no load-bearing claim rests solely on a self-citation. The one notable limitation is that Theorem 1 states gamma > 0 arbitrarily, while the proof of Assumption (H) in Theorem 4.9 explicitly requires gamma sufficiently small (with Corollary 4.6 and Lemma 4.7 being small-gamma statements). That is an overclaim or correctness gap in the parameter range, not a circular reduction: the needed positivity of the pre-factor is proven under stated hypotheses, not defined so that the desired invariant measure exists by construction.

Assumptions & free parameters 6 free parameters · 5 assumptions · 2 invented entities

The paper introduces no free parameters fitted to data, but it chooses Lyapunov exponents β0, β1, α, thresholds ζ and M, and a special noise weight g to close the estimates. Several standard stochastic-analysis theorems are assumed as black boxes; the continuous-time multiplicative ergodic theorem (Theorem 2.3) is the most delicate and is only partially supported by the cited literature. The model itself rests on a domain assumption from [31] and an acknowledged ad hoc heat bath design.

free parameters (6)
  • β0
    Exponent in the high-energy Lyapunov term (I1+I2+I3)^β0, chosen with β0>1 and β0<β1+2 to balance estimates in Lemma 4.1.
  • β1
    Exponent in I1^{−β1} and I3^{−β1}, chosen with β1+1<β0<β1+2 so that the low-energy penalty dominates in Case II of Lemma 4.1.
  • α
    Small exponent in the low-energy Lyapunov function W=I2^{−α}U; must be small enough for the Feynman-Kac eigenvalue perturbation argument in Theorem 2.7.
  • ζ
    Small threshold for the low-energy regime where ψ(I2)=I2^{−α}; chosen so that I2 A(Q) < ε/2 in Theorem 5.1.
  • M
    Boundary size for the Feynman-Kac domain I1+I3≤M; chosen large enough that L(I1+I3) is negative at the boundary.
  • γ = sufficiently small (required by proof)
    Damping strength. The theorem states γ>0 arbitrary, but the proof of Assumption (H) requires γ small (Theorem 4.9, Corollary 4.6). This mismatch is a soundness issue.
assumptions (5)
  • standard math Continuous-time multiplicative ergodic theorem with maximal isolated eigenvalue and eigenfunction for L+αF
    Theorem 2.3, used in Theorem 2.7 and Appendix A, is stated for continuous-time Markov processes and attributed to [33], but the authors note that the analytic eigenvalue expansion for continuous time is not available; the proof relies on Λ(α)=o(α) via [32] without a complete derivation.
  • standard math Hörmander's theorem: the Lie bracket condition implies a smooth transition density
    Used in Lemma 6.2 to establish joint continuity of the transition kernel; standard result from [44].
  • standard math Freidlin-Wentzell large deviation theory identifies the limiting invariant measure for small noise
    Used in Lemma 4.5 and Lemma 4.7 to show concentration near 2π/3 and variance O(γ); standard result from [20].
  • domain assumption The three-mode ODE (1.4) captures the resonant dynamics of the NLS system
    The model is inherited from Colliander-Keel-Staffilani-Takaoka-Tao [31]; no rigorous approximation theorem connecting (1.4) to full NLS is proved here.
  • ad hoc to paper The designed heat bath couplings and noise weight g=√I are legitimate stand-ins for physical reservoirs
    Authors acknowledge in Section 1.3 that the Gibbs measure is not invariant even at equal temperatures and that natural heat baths are left to future work. The choice of g is made to close the proof.
invented entities (2)
  • Noise weight function g(I2,θ)
    purpose: Amplifies noise near the unstable phase equilibrium 4π/3 so that a single pre-factor f works uniformly for large I2
    No physical derivation. Chosen as g=√I for I≥2 in Section 4 to make Assumption (H) provable.
  • Ad hoc heat bath coupling at I1 and I3
    purpose: Injects energy at the low mode and extracts it from the high mode to create a nonequilibrium steady state
    The coupling is a technical choice: the Gibbs measure is not invariant even when T1=T3, and the authors say more natural baths are future work (Section 1.3).

how reviews work

0 comments
Cite this review

Pith. "Pith review of Non-equilibrium steady state for a three-mode energy cascade model." pith.science (2026). https://pith.science/paper/Z3GPTULE

@misc{pith2026250516018,
  author       = {Pith},
  title        = {Pith review of: Non-equilibrium steady state for a three-mode energy cascade model},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/Z3GPTULE}},
  note         = {Machine review of arXiv:2505.16018}
}
read the original abstract

Motivated by the central phenomenon of energy cascades in wave turbulence theory, we construct non-equilibrium statistical steady states (NESS), or invariant measures, for a simplified model derived from the nonlinear Schr\"odinger (NLS) equation with external forcing and dissipation. This new perspective to studying energy cascades, distinct from traditional analyses based on kinetic equations and their cascade spectra, focuses on the underlying statistical steady state that is expected to hold when the cascade spectra of wave turbulence manifest. In the full generality of the (infinite dimensional) nonlinear Schr\"odinger equation, constructing such invariant measures is more involved than the rigorous justification of the Kolmogorov-Zakharov (KZ) spectra, which itself remains an outstanding open question despite the recent progress on mathematical wave turbulence. Since such complexity remains far beyond the current knowledge (even for much simpler chain models), we confine our analysis to a three-mode reduced system that captures the resonant dynamics of the NLS equation, offering a tractable framework for constructing the NESS. For this, we introduce a novel approach based on solving an elliptic Feynman-Kac equation to construct the needed Lyapunov function.

Figures

Figures reproduced from arXiv: 2505.16018 by the authors.

Figure 1
Figure 1. Left: Illustration of Case I and II in the proof of Lemma 4.1 and some contour plots of V (x). Middle: Illustration of different subcases of Case I in the proof of Lemma 4.1. Right: Illustration of different cases in the proof of Lemma 4.2. LV = β0(I1 + I2 + I3) β0−1 [PITH_FULL_IMAGE:figures/full_fig_p016_1.png] view at source ↗
Figure 2
Figure 2. Illustration of the dynamics of equation (5.5) and sets A1, A2 used in the proof. Therefore, when ρ = (4π/3 − 0.2)2 = 15.9104, we have µ([2π/3 + 0.08, 2π/3 + 0.1] ∪ [2π/3 − 0.2, 2π/3 − 0.16]) := µ(A1) ≤e −0.063I3 R 15.9104 0.0256 (16γt)−1dt < e−0.025I3/γ . When ρ = 4 × (4π/3 − 7π/6)2 = π 2/9 ≈ 1.0966, we have µ([2π/3 + 0.08, 7π/6] ∪ [5π/3, 2π] ∪ [0, 2π/3 − 0.16]) := µ(A2) ≤e −0.063I3 R 1.0966 0.0259 (16γt)−1dt < e−0… view at source ↗

Discussion (0). Sign in to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Non-uniqueness of stationary measures for stochastic systems with almost surely invariant manifolds

    math.DS 2025-05 conditional novelty 8.0 of 10

    A positive transverse Lyapunov exponent forces new stationary measures off an almost surely invariant manifold, yielding exactly two ergodic stationary measures for weakly damped Lorenz 96.

Reference graph

Works this paper leans on

45 extracted references · 42 canonical work pages · cited by 1 Pith paper

  1. [1]

    Avanti Athreya, Tiffany Kolba, and Jonathan C Mattingly, Propagating lyapunov functions to prove noise–induced stabilization, Electron. J. Probab 17 (2012), no. 96, 1–38

    work page 2012

  2. [2]

    Jacob Bedrossian, A note on cascade flux laws for the stochastically-driven nonlinear schr¨ odinger equation, Nonlinearity 3 (2024), no. 6

    work page 2024

  3. [3]

    Jacob Bedrossian, Alex Blumenthal, Keagan Callis, and Kyle Liss, Existence of stationary measures for partially damped sdes with generic, euler-type nonlinearities , arXiv preprint arXiv:2407.16592 (2024)

    work page Pith review arXiv 2024

  4. [4]

    1, 241–303

    Jacob Bedrossian, Alex Blumenthal, and Samuel Punshon-Smith, Almost-sure exponential mixing of passive scalars by the stochastic navier–stokes equations , The Annals of Probability 50 (2022), no. 1, 241–303

    work page 2022

  5. [5]

    1, 101–178

    Jacob Bedrossian and Kyle Liss, Stationary measures for stochastic differential equations with degen- erate damping, Probability Theory and Related Fields 189 (2024), no. 1, 101–178

    work page 2024

  6. [6]

    1, 47–92

    Henri Berestycki, Louis Nirenberg, and SR Srinivasa Varadhan, The principal eigenvalue and maxi- mum principle for second-order elliptic operators in general domains , Communications on Pure and Applied Mathematics 47 (1994), no. 1, 47–92

    work page 1994

  7. [7]

    Buckmaster, P

    T. Buckmaster, P. Germain, Z. Hani, and J. Shatah, Onset of the wave turbulence description of the longtime behavior of the nonlinear schr¨ odinger equation, Invent. Math. 225 (2021), no. 3, 787–855

    work page 2021

  8. [8]

    Nicolas Camps and Gigliola Staffilani, Modified scattering for the cubic schr¨ odinger equation on dio- phantine waveguides, Preprint, arXiv:2404.16817 (2024)

    work page Pith review arXiv 2024

Show all 45 references
  1. [9]

    1, 012210

    Eric A Carlen, David A Huse, and Joel L Lebowitz, Stationary states of boundary-driven quantum systems: Some exact results , Physical Review A 111 (2025), no. 1, 012210

    2025

  2. [10]

    Charles Collot, Helge Dietert, and Pierre Germain, Stability and cascades for the kolmogorov-zakharov spectrum of wave turbulence , Arch. Ration. Mech. Anal. (2024), no. 248. 50 HANI, LI, NAHMOD, STAFFILANI

    2024

  3. [11]

    Pi 9 (2021)

    Yu Deng and Zaher Hani, On the derivation of the wave kinetic equation for nls , Forum Math. Pi 9 (2021)

    2021

  4. [12]

    , Derivation of the wave kinetic equation: Full range of scaling laws , Memoirs of the AMS (to appear) (2023)

    2023

  5. [13]

    , Full derivation of the wave kinetic equation , Invent. Math. 233 (2023), no. 2, 543–724

    2023

  6. [14]

    , Long time justification of wave turbulence theory , Preprint, arXiv:2311.10082 (2023)

    2023 arXiv

  7. [15]

    3, 439–461

    RE Lee DeVille, Paul A Milewski, Ricardo J Pignol, Esteban G Tabak, and Eric Vanden-Eijnden, Nonequilibrium statistics of a reduced model for energy transfer in waves , Communications on Pure and Applied Mathematics: A Journal Issued by the Courant Institute of Mathematical Sc...

    2007

  8. [16]

    J-P Eckmann, C-A Pillet, and Luc Rey-Bellet, Non-equilibrium statistical mechanics of anharmonic chains coupled to two heat baths at different temperatures , Communications in Mathematical Physics 201 (1999), 657–697

    1999

  9. [17]

    1, 237–267

    J-P Eckmann and L-S Young, Nonequilibrium energy profiles for a class of 1-d models , Communica- tions in mathematical physics 262 (2006), no. 1, 237–267

    2006

  10. [18]

    Jean-Pierre Eckmann, Claude-Alain Pillet, and Luc Rey-Bellet, Entropy production in nonlinear, thermally driven hamiltonian systems , Journal of statistical physics 95 (1999), 305–331

    1999

  11. [19]

    19, American Mathematical Society, 2022

    Lawrence C Evans, Partial differential equations , vol. 19, American Mathematical Society, 2022

    2022

  12. [20]

    260, Springer Science & Business Media, 2012

    Mark I Freidlin and Alexander D Wentzell, Random perturbations of dynamical systems , vol. 260, Springer Science & Business Media, 2012

    2012

  13. [21]

    109, Princeton university press, 1985

    Mark Iosifovich Freidlin, Functional integration and partial differential equations , no. 109, Princeton university press, 1985

    1985

  14. [22]

    Pierre Germain, Joonhyun La, and Angeliki Menegaki, Stability of rayleigh-jeans equilibria in the kinetic fpu equation , arXiv:2409.01507 (2024)

    2024 arXiv

  15. [23]

    220 (2022)

    Filippo Giuliani and Marcel Guardia, Sobolev norms explosion for the cubic nls on irrational tori , Nonlinear Anal. 220 (2022)

    2022

  16. [24]

    Martin Hairer, Convergence of markov processes, Lecture notes 18 (2010), 26

    2010

  17. [25]

    Martin Hairer and Jonathan C Mattingly, Ergodicity of the 2d navier-stokes equations with degenerate stochastic forcing, Annals of Mathematics (2006), 993–1032

    2006

  18. [26]

    Pi 3 (2015)

    Zaher Hani, Benoit Pausader, Nikolay Tzvetkov, and Nicola Visciglia, Modified scattering for the cubic schr¨ odinger equation on product spaces and applications, Forum Math. Pi 3 (2015)

    2015

  19. [27]

    3, 1893–1942

    ALEXANDRU HENING and DANG H NGUYEN, Coexistence and extinction for stochastic kol- mogorov systems, The Annals of Applied Probability 28 (2018), no. 3, 1893–1942

    2018

  20. [28]

    David P Herzog and Jonathan C Mattingly, Noise-induced stabilization of planar flows i , Electron. J. Probab 20 (2015), no. 111, 1–43

    2015

  21. [29]

    Alexander Hrabski, Yulin Pan, Gigliola Staffilani, and Bobby Wilson, Energy transfer for solutions to the nonlinear schr¨ odinger equation on irrational tori, To appear in the Abel Symposium Proceedings (2023)

    2023

  22. [30]

    4, 1712–1730

    WEN HUANG, MIN JI, ZHENXIN LIU, and YINGFEI YI, Integral identity and measure estimates for stationary fokker–planck equations , The Annals of Probability 43 (2015), no. 4, 1712–1730

    2015

  23. [31]

    Staffilani H

    G. Staffilani H. Takaoka T. Tao J. Colliander, M. Keel, Transfer of energy to high frequencies in the cubic defocusing nonlinear schr¨ odinger equation., Invent. Math. 181 (2010), 39–113

    2010

  24. [32]

    132, Springer Science & Business Media, 2013

    Tosio Kato, Perturbation theory for linear operators , vol. 132, Springer Science & Business Media, 2013

    2013

  25. [33]

    1, 304–362

    Ioannis Kontoyiannis and Sean P Meyn, Spectral theory and limit theorems for geometrically ergodic markov processes, The Annals of Applied Probability 13 (2003), no. 1, 304–362

    2003

  26. [34]

    , Electronic Communications in Probability [electronic only] 10 (2005), 61–123

    Ioannis Kontoyiannis and SP Meyn, Large deviations asymptotics and the spectral theory of multi- plicatively regular markov processes. , Electronic Communications in Probability [electronic only] 10 (2005), 61–123

    2005

  27. [35]

    Stefano Lepri, Roberto Livi, and Antonio Politi, Thermal conduction in classical low-dimensional lattices, Physics reports 377 (2003), no. 1, 1–80

    2003

  28. [36]

    9, 1777–1811

    Yao Li and Yingfei Yi, Systematic measures of biological networks i: Invariant measures and entropy , Communications on Pure and Applied Mathematics 69 (2016), no. 9, 1777–1811

    2016

  29. [37]

    3, 639–660

    Genqian Liu, Strongly continuous semigroups and stochastic representation , Journal of the London Mathematical Society 65 (2002), no. 3, 639–660. NESS 51

    2002

  30. [38]

    2, 185–232

    Jonathan C Mattingly, Andrew M Stuart, and Desmond J Higham, Ergodicity for sdes and approxi- mations: locally lipschitz vector fields and degenerate noise , Stochastic processes and their applications 101 (2002), no. 2, 185–232

    2002

  31. [39]

    1, 189–220

    Jonathan C Mattingly, Toufic Suidan, and Eric Vanden-Eijnden, Simple systems with anomalous dissipation and energy cascade, Communications in mathematical physics 276 (2007), no. 1, 189–220

    2007

  32. [40]

    Sean P Meyn and Richard L Tweedie, Markov chains and stochastic stability , Springer Science & Business Media, 2012

    2012

  33. [41]

    Sergey Nazarenko, Wave turbulence,, Lecture notes in Physics 825 (2011)

    2011

  34. [42]

    Luc Rey-Bellet and Lawrence E Thomas, Exponential convergence to non-equilibrium stationary states in classical statistical mechanics , Communications in mathematical physics 225 (2002), 305–329

    2002

  35. [43]

    Barbara Stachurska, On a nonlinear integral inequality , Zeszyty Nauk. Uniw. Jagiello´ n. Prace Mat. (1971), no. 15, 151–157

    1971

  36. [44]

    112, Cambridge University Press, 2008

    Daniel W Stroock, Partial differential equations for probabilists, no. 112, Cambridge University Press, 2008

    2008

  37. [45]

    V. S. L’vov V. E. Zakharov and G. Falkovich, Kolmogorov spectra of turbulence i: Wave turbulence , Springer Science & Business Media,, 2012

    2012

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.