Pith. sign in

REVIEW 3 major objections 5 minor 4 cited by

Solutions of stationary McKean-Vlasov equation on a high-dimensional sphere and other Riemannian manifolds

T0 review · 3 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read For rotationally symmetric interaction kernels on a sphere, the paper proves that a unique negative spherical-harmonic coefficient produces bifurcation branches from the uniform state, and that a relaxed resonance condition forces a…

desk verdict Solid bifurcation theory on the sphere, but the discontinuous phase transition proof has a normalization gap that currently leaves the main discontinuity claim unproven. read the letter →

arxiv 2412.14813 v2 pith:YTB7KXUH submitted 2024-12-19 math.AP math-phmath.MPmath.PR

Anna Shalova , André Schlichting This is my paper · ORCID
classification math.APmath-phmath.MPmath.PR MSC 35B3235Q8358J3582C22
keywords McKean-VlasovequationstationarysolutionsbifurcationphasetransitionsphericalharmonicsconvolutionfreeenergyfunctionalRiemannianmanifolds
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 studies stationary states of the McKean-Vlasov equation, the mean-field PDE for interacting diffusions, placed on a compact Riemannian manifold and in particular on the unit sphere $S^{n-1}$. It establishes that stationary states coincide exactly with fixed points of the Gibbs map and critical points of a free energy functional, and that minimizers exist for every inverse temperature $\gamma$ while being unique for small $\gamma$ under a curvature condition. On the sphere, the paper shows that a rotationally symmetric interaction kernel is diagonalized by spherical harmonics, so each simple negative coefficient $\hat W_k$ produces a bifurcation branch of stationary states emerging from the uniform state at $\gamma_k = -1/\hat W_k$. The main new result is a sufficient spectral condition: if the most-negative modes can be combined into a function whose cube has nonzero spherical average, the system has a discontinuous phase transition at some $\gamma_c$ strictly below the linear-stability threshold $\gamma_\sharp$. The condition is verified for the Onsager model, a spherical opinion-dynamics kernel, and a localized Gaussian heat kernel, while it fails for the noisy transformer kernel, leaving that transition type open.

What carries the argument

The load-bearing object is the spherical harmonics decomposition $\hat W_k$ of the rotationally symmetric kernel, defined through Gegenbauer polynomials, together with the spherical convolution theorem: for any function $u$, the projection of $W * u$ onto the $k$-th harmonic subspace equals $\omega_n \hat W_k$ times the projection of $u$. This diagonalizes the linearized Gibbs map and reduces the bifurcation problem to a simple characteristic value of a compact operator. The phase-transition argument additionally uses the cubic moment $\int_{S^{n-1}} u^3\,d\sigma$ of a competitor built from the minimal modes: the third-order term in the Taylor expansion of the free energy is what makes the uniform state lose global minimality strictly before linear instability. The relaxed resonance condition quantifies how small the bandwidth around the minimal coefficient must be for this cubic term to dominate.

What would settle it

Run a numerical continuation of the stationary McKean-Vlasov equation on $S^2$ with the Onsager kernel $W(x,y)=\sqrt{1-\langle x,y\rangle^2}$: the paper predicts that the uniform state stops being a global minimizer at some $\gamma_c$ strictly below $\gamma_\sharp = -1/\hat W_2$, and if the first non-uniform minimizer instead appears exactly at $\gamma_\sharp$, the strict inequality in Theorem 5.11 would be contradicted. A second check is to compute $\int_{S^{n-1}} Y_{1,0}^3\,d\sigma$ for the noisy transformer kernel; it vanishes because $Y_{1,0}$ is odd, confirming that the relaxed resonance condition is objectively absent there.

Watch

Extended reading notes

Core claim

The central claim is that on $S^{n-1}$, for an interaction kernel of the form $W(x,y)=W(\langle x,y\rangle)$, the stationary McKean-Vlasov equation can be solved spectrally. Using the spherical convolution theorem, the linearization of the Gibbs map at the uniform state has eigenvalues $\lambda_l = -l(n+l-2)(1+\gamma \hat W_l)$, so the uniform state becomes linearly unstable exactly when $\gamma$ exceeds $\gamma_\sharp = -1/\min_l \hat W_l$. The paper proves (Theorem 4.9) that if a single coefficient $\hat W_k<0$ is unique among all coefficients, a nontrivial branch $\rho_\gamma(t) = \bar\rho + f(t)Y_{k,0} + o(f(t))$ bifurcates from the uniform state at $\gamma_k = -1/\hat W_k$. It then proves (Theorem 5.11) that under the relaxed resonance condition, the uniform state is already not a global minimizer at $\gamma_\sharp$, so the phase transition occurs at some $\gamma_c \in (0,\gamma_\sharp)$ and is discontinuous. This transfers the resonance mechanism for discontinuous transitions from the torus to the sphere, where a single self-resonant mode such as $k=2$ or $k=4$ can suffice.

Load-bearing premise

The paper's discontinuous-transition result rests on the relaxed resonance condition: the most-negative spherical-harmonic modes must combine into a bounded function whose cube has a nonzero average over the sphere, with the spread of near-minimal modes kept small enough.

Editorial extensions

If this is right

  • For any compact Riemannian manifold with bounded curvature and a sufficiently regular kernel, stationary McKean-Vlasov states are exactly Gibbs fixed points and critical points of the free energy; minimizers exist for all $\gamma$ and are unique for small $\gamma$ under a geodesic-convexity condition.
  • On $S^{n-1}$, every simple negative spherical-harmonic coefficient $\hat W_k$ yields a bifurcation branch from the uniform state of the form $\bar\rho + f(t)Y_{k,0}+o(f(t))$ at $\gamma_k = -1/\hat W_k$.
  • If the relaxed resonance condition holds, the phase transition is discontinuous and occurs at $\gamma_c < \gamma_\sharp$; in particular the uniform state is globally suboptimal while still linearly stable on $(\gamma_c, \gamma_\sharp)$.
  • For the Onsager kernel and the opinion-dynamics kernel with $p=n+2$, the minimal mode is $k=2$ and is self-resonant, so a discontinuous transition is guaranteed on spheres of arbitrary dimension.
  • For the localized hyperspherical heat kernel with sufficiently small $\varepsilon$, the bandwidth condition of the relaxed resonance condition is satisfied, giving a discontinuous transition for small $\varepsilon$.

Reading between the lines

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

  • Beyond the paper, the strength of the cubic moment $U_3 = \int u^3\,d\sigma$ could be computed numerically for any rotationally symmetric kernel, turning the relaxed resonance condition into a practical spectral test for discontinuous phase transitions on spheres.
  • Because the relaxed resonance condition fails for the noisy transformer kernel, where $k_{\min}=1$ is odd and $\int Y_{1,0}^3\,d\sigma=0$, the transition type for that model likely depends on higher-order terms in the free-energy expansion, which the paper leaves as an open quantitative question.
  • The strict gap $\gamma_c < \gamma_\sharp$ predicts hysteresis in particle simulations: slow cooling from high noise should jump to the non-uniform state at $\gamma_c$, while slow heating from an ordered state could persist until $\gamma_\sharp$.
  • The paper's manifold-level equivalence between stationary solutions, Gibbs fixed points, and critical points suggests that numerical free-energy minimization on the sphere could be used to locate bifurcating branches without solving the PDE directly.
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 / 5 minor

Summary. The paper studies stationary solutions of the McKean-Vlasov equation on compact Riemannian manifolds, with emphasis on the high-dimensional sphere. It claims three main bodies of results: (i) equivalence between weak solutions, zeros of the Gibbs map, and critical points of a free-energy functional, together with existence, uniqueness and small-noise limits on general compact manifolds; (ii) for rotationally symmetric kernels on S^{n-1}, bifurcation branches from the uniform state at values γ_k = -1/\hat W_k determined by spherical-harmonic coefficients, with an explicit description of the bifurcating states in terms of the harmonics Y_{k,0}; and (iii) a sufficient condition, called the relaxed resonance condition, for the existence of a discontinuous phase transition point γ_c strictly below the linear-stability threshold γ_♯. The theoretical results are illustrated on the noisy transformer model, the Onsager model, a spherical opinion-dynamics model, and a localized spherical Gaussian kernel.

Significance. If the results are correct, the paper gives a substantial manifold analogue of the torus analysis of Carrillo–Gvalani–Pavliotis–Schlichting, with explicit spectral conditions and concrete applications to models of current interest. Strengths include the self-contained derivation, the explicit assumptions in Theorems 4.9 and 5.11, and the concrete Gegenbauer/Bessel computations in the examples; no parameters are fitted and the claims are falsifiable. However, the proof of the discontinuous-transition theorem currently rests on an energy comparison whose normalization is inconsistent as written, and the proof of the central equivalence in Proposition 2.5 uses a false regularity assertion. These issues are load-bearing for the paper's main new claims, so the results cannot be accepted in their present form.

major comments (3)
  1. [§2.1, Proposition 2.5, proof of (3)→(2)] The proof asserts that every H^1 function on a compact Riemannian manifold is almost everywhere continuous, which is false in dimension n ≥ 2. The subsequent argument evaluating γ^{-1}log ρ + W*ρ at a 'point of continuity' x_0 is therefore not justified, and the implication that every critical point of F_γ is a zero of the Gibbs map is not proven as written. A Lebesgue-point argument should replace the continuity claim; this is likely repairable, but the repair is not present in the manuscript.
  2. [§5.3, Lemma 5.12] The cancellation of the second-order terms in the energy difference F_{γ_♯}(ρ_♯) - F_{γ_♯}(ρ̄) mixes normalizations. With the scalar product ⟨f,g⟩ = ω_n^{-1} ∫ f g dσ introduced in Section 3.1, an orthonormal spherical harmonic satisfies ∫ Y_{l,0}^2 dσ = ω_n, and Equation (22) yields ∫ u (W*u) dσ = ω_n^2 Σ_l \hat W_l \hat u_l^2, not the expression used in Lemma 3.15 and Lemma 5.12. The entropy expansion is naturally written with respect to the normalized volume m, while the interaction expansion is written with respect to σ; the second-order cancellation at γ_♯ works only under one particular normalization of ∥u∥^2, which is never stated or derived. Consequently the strict inequality F_{γ_♯}(ρ_♯) < F_{γ_♯}(ρ̄) is not established as written. Since Theorem 5.11 derives γ_c < γ_♯ solely from this lemma, the discontinuous-phase-transition conclusion is currently unproven. The calculation appears repairable by working with the density 1 + εξu in L^2(m), but that calculation must be supplied.
  3. [§2.2, Theorem 2.8 and Theorem 1.1] Theorem 2.8 proves the existence of an L^2 minimizer of F_γ, but it does not prove that this minimizer is a critical point of F_γ or a zero of the Gibbs map; Proposition 2.5 requires ρ ∈ H^1(M) ∩ P^+_ac(M). The introduction's Theorem 1.1 states that for every γ there exists a solution ρ_γ ∈ H^1(M) ∩ P_ac(M) of (1), but the proof chain from the minimizer to the weak solution is missing. This is a further load-bearing gap in the existence claim, even though the final statement is plausible and probably standard.
minor comments (5)
  1. [§3.1, Definition 3.6] The normalizing constant is written as c^{n-2/2} in Definition 3.6 and later as c^λ; Definition 1.2 uses α_k without defining it. Please align these notations.
  2. [§5.3, Assumption 5.8] The summation index set K_{α,δ} is not defined; it should presumably be K_{♯,δ} from the preceding display.
  3. [§6.1, Proposition 6.1] The displayed formula for γ_k appears to be missing a fraction or an explicit denominator; please check it against the computed spherical-harmonic coefficient \hat W_{β,k}.
  4. [§6.3] The phrase 'rational interaction kernel' is confusing for the kernel W_p(x,y) = -(1+⟨x,y⟩)^p; it should presumably be 'rotationally symmetric' or 'polynomial-type interaction kernel'.
  5. [§2.4, Theorem 2.19 proof] There is a typo in the displayed estimate: '(\|u\|2 + \|v\|)2' should read '(\|u\|^2 + \|v\|^2)'.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the bifurcation and phase-transition theorems are derived from stated spectral and resonance assumptions via explicit expansions, not from fitted data or self-referential definitions.

full rationale

The paper is a mathematical derivation rather than a data-fitting exercise. Theorem 4.9 is proven by applying the Crandall-Rabinowitz bifurcation theorem to the explicitly linearized Gibbs map; the bifurcation point γ_k = -1/ˆW_k is computed from the spherical-harmonic decomposition (Definition 3.6 and equation (22)), and the uniqueness of the negative coefficient verifies the simple-eigenvalue hypothesis. This is a genuinely sufficient spectral condition, not a restatement of the branch's existence. Theorem 5.11 is a sufficient-condition result: Assumption 5.8 postulates a finite linear combination u with ∫u³ dσ ≠ 0 and small bandwidth δ, and Lemma 5.12 constructs a competitor ρ♯ = ¯ρ(1+εξu) whose energy is lower at γ♯ by a Taylor expansion whose cubic term is controlled by ∫u³. The conclusion γ_c < γ♯ follows from the contrapositive of the external Lemma 5.6, so the resonance assumption is not the conclusion in disguise. The paper's citations of [CGPS20] and [GS20], which include co-author Schlichting, are not load-bearing: they are used as literature parallels and for external torus results, while the actual proofs rely on [Dei13], [Dai13], [Stu05], [CP10] and self-contained computations. The skeptical reviewer's normalization concern about Lemma 5.12 is a correctness or consistency issue rather than circularity, because it concerns whether the stated calculation is carried out coherently, not whether the result is assumed as input. No circular step can be exhibited.

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

The main load-bearing inputs are standard harmonic analysis on spheres, the Sturm geodesic-convexity framework, and the Crandall-Rabinowitz theorem, all externally established. The manifold-level assumptions (bounded symmetric H^1 kernel, curvature bound, lower second-derivative bound) are stated model assumptions. The genuinely new structural input is the resonance/bandwidth condition (Assumption 5.8), which is an ad hoc hypothesis tailored to the Taylor-expansion proof but verified on several natural kernels; its failure for the noisy transformer kernel is acknowledged in Remark 6.2.

assumptions (6)
  • domain assumption Rotational symmetry: W(x,y) = W(⟨x,y⟩) with W: [-1,1]→R, W symmetric, bounded and in H^1 (Assumption 3.13).
    All sphere results, including the convolution diagonalization and the eigenvalue computations, assume this structure; it is not universal for the manifold theorems but is stated at the start of the sphere sections.
  • domain assumption The free-energy characterization: stationary weak solutions of (1) are exactly the zeros of the Gibbs map (5) and the critical points of Fγ (Proposition 2.5).
    This equivalence is the foundation of both Theorems 4.9 and 5.11; the proof as written uses the incorrect assertion that H^1 functions are almost everywhere continuous.
  • standard math Spherical harmonic analysis: the Y_{l,K} form an orthonormal basis of L^2(S^{n-1}), with convolution theorem proj_p(g*f) = ω_n \hat g_p proj_p f (Theorem 3.9, referencing Dai13).
    Used to diagonalize the linearized Gibbs map and to compute spherical-harmonic coefficients in all examples.
  • standard math Crandall-Rabinowitz bifurcation theory for operators of the form F(x,λ)=x-λKx+G(x,λ) with Fredholm and simplicity conditions (Deimling Theorem 28.3).
    This is the external theorem applied in Theorem 4.9; the paper verifies its hypotheses in Lemmas 4.10 and 4.11.
  • ad hoc to paper Relaxed resonance condition: existence of a bounded finite linear combination u of (near-)minimal modes with ∫u^3dσ ≠ 0 and small bandwidth δ relative to (∫u^3)^2 (Assumptions 5.8 and 5.9).
    The condition is introduced specifically to force a negative cubic term in the Taylor expansion of the free energy (Lemma 5.12), yielding a discontinuous transition. It is verified in some examples and explicitly fails for the noisy transformer kernel.
  • standard math Geodesic convexity framework: the entropy is λ-convex for Ric ≥ λ (Sturm 2005), and the interaction energy has the lower second-derivative bound (Theorem 2.14).
    Used to establish uniqueness of minimizers for γ small enough, which is needed to define the transition point and the small-γ regime.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Solutions of stationary McKean-Vlasov equation on a high-dimensional sphere and other Riemannian manifolds." pith.science (2026). https://pith.science/paper/YTB7KXUH

@misc{pith2026241214813,
  author       = {Pith},
  title        = {Pith review of: Solutions of stationary McKean-Vlasov equation on a high-dimensional sphere and other Riemannian manifolds},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/YTB7KXUH}},
  note         = {Machine review of arXiv:2412.14813}
}
read the original abstract

We study stationary solutions of McKean-Vlasov equation on a high-dimensional sphere and other compact Riemannian manifolds. We extend the equivalence of the energetic problem formulation to the manifold setting and characterize critical points of the corresponding free energy functional. On a sphere, we employ the properties of spherical convolution to study the bifurcation branches around the uniform state. We also give a sufficient condition for an existence of a discontinuous transition point in terms of the interaction kernel and compare it to the Euclidean setting. We illustrate our results on a range of system, including the particle system arising from the transformer models and the Onsager model of liquid crystals.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 4 Pith papers

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

  1. Formation of clusters and coarsening in weakly interacting diffusions

    math.AP 2025-10 conditional novelty 7.0 of 10

    Global minimizers of the McKean–Vlasov free energy for short-range attractive potentials on the circle are uniform or single symmetric clusters, and coarsening is governed by exponentially slow mass exchange between clusters.

  2. Cluster formation for weakly interacting kinetic Langevin dynamics

    math.NA 2025-10 conditional novelty 6.0 of 10

    For weakly interacting underdamped Langevin particles with short-range attraction, cluster onset occurs above a friction-independent critical inverse temperature and the onset time scales as (1/(2ψmax)) log N.

  3. Attention's forward pass and Frank-Wolfe

    math.OC 2025-08 conditional novelty 6.0 of 10

    Hardmax self-attention is shown to be a Frank-Wolfe iteration; with positive-definite key-query it converges to Voronoi-cell vertices, and a Markov-chain version of soft attention is metastable there for exponential-i...

  4. Physics- and geometry-aware spatio-spectral graph neural operator for time-independent and time-dependent PDEs

    cs.LG 2025-08 unverdicted novelty 4.0 of 10

    A submission whose abstract describes a new graph neural operator for PDEs but whose full text is a different paper, leaving the claimed method and results unverifiable.

Reference graph

Works this paper leans on

79 extracted references · 74 canonical work pages · cited by 4 Pith papers

  1. [1]

    G. E. Andrews, R. Askey, and R. Roy. Special Functions . Encyclopedia of Mathematics and its Applications. Cambridge University Press, 1999

    work page 1999

  2. [2]

    Ambrosio, N

    L. Ambrosio, N. Gigli, and G. Savar \'e . Gradient flows: in metric spaces and in the space of probability measures . Springer Science & Business Media, 2005

    work page 2005

  3. [3]

    Abramowitz and I

    M. Abramowitz and I. A. Stegun. Handbook of mathematical functions with formulas, graphs, and mathematical tables , volume 55. US Government printing office, 1968

    work page 1968

  4. [4]

    Burger, V

    M. Burger, V. Capasso, and D. Morale. On an aggregation model with long and short range interactions. Nonlinear Anal. Real World Appl. , 8(3):939--958, 2007

    work page 2007

  5. [5]

    A. B. T. Barbaro, J. A. Ca\ nizo, J. A. Carrillo, and P. Degond. Phase transitions in a kinetic flocking model of C ucker- S male type. Multiscale Model. Simul. , 14(3):1063--1088, 2016

    work page 2016

  6. [6]

    Brigati, J

    G. Brigati, J. Dolbeault, and N. Simonov. Logarithmic sobolev and interpolation inequalities on the sphere: constructive stability results. Annales de l'Institut Henri Poincar \'e C , 41(5):1289--1321, 2023

    work page 2023

  7. [7]

    J. A. Bittencourt. Fundamentals of plasma physics . Pergamon Press, Oxford, 1986

    work page 1986

  8. [8]

    Bresch, P.-E

    D. Bresch, P.-E. Jabin, and Z. Wang. Mean field limit and quantitative estimates with singular attractive kernels. Duke Mathematical Journal , 172(13):2591--2641, 2023

    work page 2023

Show all 79 references
  1. [9]

    Bruno, F

    G. Bruno, F. Pasqualotto, and A. Agazzi. Emergence of meta-stable clustering in mean-field transformer models. arXiv preprint arXiv:2410.23228 , 2024

    2024 arXiv

  2. [10]

    A. Braides. A handbook of -convergence. In Handbook of Differential Equations: stationary partial differential equations , volume 3, pages 101--213. Elsevier, 2006

    2006

  3. [11]

    Baernstein, II and B

    A. Baernstein, II and B. A. Taylor. Spherical rearrangements, subharmonic functions, and * -functions in n -space. Duke Math. J. , 43(2):245--268, 1976

    1976

  4. [12]

    Binney and S

    J. Binney and S. Tremaine. Galactic dynamics . Princeton University Press, 2008

    2008

  5. [13]

    C. H. Chan and M. Czubak. The M eyers- S errin theorem on riemannian manifolds: a survey. arXiv preprint arXiv:2405.13322 , 2024

    2024 arXiv

  6. [14]

    J. A. Carrillo, R. C. Fetecau, and H. Park. Existence of ground states for free energies on the hyperbolic space. arXiv preprint arXiv:2409.06022 , 2024

    2024 arXiv

  7. [15]

    J. A. Carrillo and R. S. Gvalani. Phase transitions for nonlinear nonlocal aggregation-diffusion equations. Communications in Mathematical Physics , 382(1):485–545, February 2021

    2021

  8. [16]

    J. A. Carrillo, R. Gvalani, G. Pavliotis, and A. Schlichting. Long-time behaviour and phase transitions for the M c K ean-- V lasov equation on the torus. Archive for Rational Mechanics and Analysis , 235(1):635--690, 2020

    2020

  9. [17]

    T. Chihara. An Introduction to Orthogonal Polynomials . Dover Books on Mathematics. Dover Publications, 2011

    2011

  10. [18]

    Chayes and V

    L. Chayes and V. Panferov. The M c K ean- V lasov equation in finite volume. J. Stat. Phys. , 138(1-3):351--380, 2010

    2010

  11. [19]

    M. G. Crandall and P. H. Rabinowitz. Bifurcation from simple eigenvalues. J. Functional Analysis , 8:321--340, 1971

    1971

  12. [20]

    J. G. Conlon and A. Schlichting. A non-local problem for the Fokker-Planck equation related to the Becker-Döring model . Discret. Contin. Dyn. Syst. , 39(4):1--57, 2019

    2019

  13. [21]

    F. Dai. Approximation theory and harmonic analysis on spheres and balls . Springer, 2013

    2013

  14. [22]

    E. B. Davies. Linear operators and their spectra , volume 106. Cambridge University Press, 2007

    2007

  15. [23]

    Dupuis and R

    P. Dupuis and R. S. Ellis. A weak convergence approach to the theory of large deviations . John Wiley & Sons, 2011

    2011

  16. [24]

    Deimling

    K. Deimling. Nonlinear functional analysis . Springer Science & Business Media, 2013

    2013

  17. [25]

    Degond, A

    P. Degond, A. Frouvelle, and J.-G. Liu. Phase transitions, hysteresis, and hyperbolicity for self-organized alignment dynamics. Arch. Ration. Mech. Anal. , 216(1):63--115, 2015

    2015

  18. [26]

    D. A. Dawson and J. G\" a rtner. Large deviations, free energy functional and quasi-potential for a mean field model of interacting diffusions. Mem. Amer. Math. Soc. , 78(398):iv+94, 1989

    1989

  19. [27]

    M. G. Delgadino, R. S. Gvalani, and G. A. Pavliotis. On the diffusive-mean field limit for weakly interacting diffusions exhibiting phase transitions. Archive for Rational Mechanics and Analysis , 241:91--148, 2021

    2021

  20. [28]

    M. G. Delgadino, R. S. Gvalani, G. A. Pavliotis, and S. A. Smith. Phase transitions, logarithmic sobolev inequalities, and uniform-in-time propagation of chaos for weakly interacting diffusions. Communications in Mathematical Physics , 401(1):275–323, February 2023

    2023

  21. [29]

    Dokmanic and D

    I. Dokmanic and D. Petrinovic. Convolution on the n -sphere with application to pdf modeling. IEEE transactions on signal processing , 58(3):1157--1170, 2009

    2009

  22. [30]

    Duerinckx

    M. Duerinckx. Mean-field limits for some riesz interaction gradient flows. SIAM Journal on Mathematical Analysis , 48(3):2269--2300, 2016

    2016

  23. [31]

    M. Erbar. The heat equation on manifolds as a gradient flow in the W asserstein space. In Annales de l'IHP Probabilit \'e s et statistiques , volume 46, pages 1--23, 2010

    2010

  24. [32]

    R. C. Fetecau, S.-Y. Ha, and H. Park. An intrinsic aggregation model on the special orthogonal group so (3): well-posedness and collective behaviours. Journal of Nonlinear Science , 31:1--61, 2021

    2021

  25. [33]

    R. C. Fetecau and H. Park. Equilibria and energy minimizers for an interaction model on the hyperbolic space. Physica D: Nonlinear Phenomena , 446:133670, 2023

    2023

  26. [34]

    R. C. Fetecau and H. Park. Ground states for aggregation-diffusion models on cartan-hadamard manifolds. arXiv preprint arXiv:2306.04856 , 2023

    2023 arXiv

  27. [35]

    R. C. Fetecau and H. Park. Long-time behaviour of interaction models on riemannian manifolds with bounded curvature. The Journal of Geometric Analysis , 33(7):218, 2023

    2023

  28. [36]

    R. C. Fetecau, H. Park, and F. S. Patacchini. Well-posedness and asymptotic behavior of an aggregation model with intrinsic interactions on sphere and other manifolds. Analysis and Applications , 19(06):965--1017, 2021

    2021

  29. [37]

    Fatkullin and V

    I. Fatkullin and V. Slastikov. Critical points of the onsager functional on a sphere. Nonlinearity , 18(6):2565, 2005

    2005

  30. [38]

    Geshkovski, H

    B. Geshkovski, H. Koubbi, Y. Polyanskiy, and P. Rigollet. Dynamic metastability in the self-attention model. arXiv preprint arXiv:2410.06833 , 2024

    2024 arXiv

  31. [39]

    Geshkovski, C

    B. Geshkovski, C. Letrouit, Y. Polyanskiy, and P. Rigollet. A mathematical perspective on T ransformers, 2024

    2024

  32. [40]

    F. Golse. On the dynamics of large particle systems in the mean field limit. Macroscopic and large scale phenomena: coarse graining, mean field limits and ergodicity , pages 1--144, 2016

    2016

  33. [41]

    D. J. Gates and O. Penrose. The van der W aals limit for classical systems. III . D eviation from the van der W aals- M axwell theory. Comm. Math. Phys. , 17(3):194--209, 1970

    1970

  34. [42]

    I. S. Gradshteyn and I. M. Ryzhik. Table of integrals, series, and products . Academic press, 2014

    2014

  35. [43]

    R. S. Gvalani and A. Schlichting. Barriers of the McKean-Vlasov energy via a mountain pass theorem in the space of probability measures . J. Funct. Anal. , 279(11):108720, 34, 2020

    2020

  36. [44]

    W. K. Hayman. Multivalent functions . Cambridge University Press, 1994

    1994

  37. [45]

    E. Hebey. Sobolev spaces on R iemannian manifolds , volume 1635. Springer Science & Business Media, 1996

    1996

  38. [46]

    Hoeksema, T

    J. Hoeksema, T. Holding, M. Maurelli, and O. Tse. Large deviations for singularly interacting diffusions. In Annales de l'Institut Henri Poincare (B) Probabilites et statistiques , volume 60, pages 492--548. Institut Henri Poincar \'e , 2024

    2024

  39. [47]

    Hegselmann and U

    R. Hegselmann and U. Krause. Opinion dynamics and bounded confidence: models, analysis and simulation. The Journal of artifical societies and social simulation , 5 ( 3 ), 2002

    2002

  40. [48]

    W. J. Suspensions of rod-like molecules: the isotropic-nematic phase transition and flow alignment in 2-d. Master's thesis, University of Bonn, 2006

    2006

  41. [49]

    P.-E. Jabin. A review of the mean field limits for V lasov equations. Kinetic and Related models , 7(4):661--711, 2014

    2014

  42. [50]

    Kielh\"ofer

    H. Kielh\"ofer. Bifurcation theory , volume 156 of Applied Mathematical Sciences . Springer, New York, second edition, 2012. An introduction with applications to partial differential equations

    2012

  43. [51]

    E. F. Keller and L. A. Segel. Model for chemotaxis. Journal of Theoretical Biology , 30(2):225 -- 234, 1971

    1971

  44. [52]

    Kuramoto

    Y. Kuramoto. Rhythms and turbulence in populations of chemical oscillators. Phys. A , 106(1-2):128--143, 1981. Statphys 14 (Proc. Fourteenth Internat. Conf. Thermodynamics and Statist. Mech., Univ. Alberta, Edmonton, Alta., 1980)

    1981

  45. [53]

    D. Lacker. Hierarchies, entropy, and quantitative propagation of chaos for mean field diffusions. Probab. Math. Phys. , 4(2):377--432, 2023

    2023

  46. [54]

    Lucia and J

    M. Lucia and J. Vukadinovic. Exact multiplicity of nematic states for an onsager model. Nonlinearity , 23(12):3157, 2010

    2010

  47. [55]

    Martzel and C

    N. Martzel and C. Aslangul. Mean-field treatment of the many-body F okker- P lanck equation. J. Phys. A , 34(50):11225--11240, 2001

    2001

  48. [56]

    H. P. McKean, Jr. A class of M arkov processes associated with nonlinear parabolic equations. Proc. Nat. Acad. Sci. U.S.A. , 56:1907--1911, 1966

    1907

  49. [57]

    Monmarch \'e

    P. Monmarch \'e . Long-time behaviour and propagation of chaos for mean field kinetic particles. Stochastic Processes and their Applications , 127(6):1721--1737, 2017

    2017

  50. [58]

    Oelschl\"ager

    K. Oelschl\"ager. A martingale approach to the law of large numbers for weakly interacting stochastic processes. Ann. Probab. , 12(2):458--479, 1984

    1984

  51. [59]

    Oelschl \"a ger

    K. Oelschl \"a ger. On the derivation of reaction-diffusion equations as limit dynamics of systems of moderately interacting stochastic processes. Probability Theory and Related Fields , 82(4):565--586, 1989

    1989

  52. [60]

    Otto and C

    F. Otto and C. Villani. Generalization of an inequality by talagrand and links with the logarithmic sobolev inequality. Journal of Functional Analysis , 173(2):361--400, 2000

    2000

  53. [61]

    Otto and M

    F. Otto and M. Westdickenberg. Eulerian calculus for the contraction in the wasserstein distance. SIAM journal on mathematical analysis , 37(4):1227--1255, 2005

    2005

  54. [62]

    H. A. Posch, H. Narnhofer, and W. Thirring. Dynamics of unstable systems. Physical Review A , 42(4):1880–1890, August 1990. Video of simulations: https://www.mediathek.at/atom/018AAB81-08A-0279F-00000484-0189A3E5

    1990

  55. [63]

    F. S. Patacchini and D. Slep c ev. The nonlocal-interaction equation near attracting manifolds. arXiv preprint arXiv:2106.01823 , 2021

    2021 arXiv

  56. [64]

    Rotskoff and E

    G. Rotskoff and E. Vanden-Eijnden. Trainability and accuracy of artificial neural networks: An interacting particle system approach. Communications on Pure and Applied Mathematics , 75(9):1889--1935, 2022

    1935

  57. [65]

    Santambrogio

    F. Santambrogio. Optimal transport for applied mathematicians. Birk \"a user, NY , 55(58-63):94, 2015

    2015

  58. [66]

    S. Serfaty. Mean field limit for Coulomb-type flows . Duke Mathematical Journal , 169(15):2887 -- 2935, 2020. Appendix with Mitia Duerinckx

    2020

  59. [67]

    Sirignano and K

    J. Sirignano and K. Spiliopoulos. Mean field analysis of neural networks: A law of large numbers. SIAM Journal on Applied Mathematics , 80(2):725--752, 2020

    2020

  60. [68]

    A. Stevens. The derivation of chemotaxis equations as limit dynamics of moderately interacting stochastic many-particle systems. SIAM J. Appl. Math. , 61(1):183--212, 2000

    2000

  61. [69]

    K.-T. Sturm. Convex functionals of probability measures and nonlinear diffusions on manifolds. Journal de math \'e matiques pures et appliqu \'e es , 84(2):149--168, 2005

    2005

  62. [70]

    J. Tugaut. Phase transitions of M c K ean- V lasov processes in double-wells landscape. Stochastics , 86(2):257--284, 2014

    2014

  63. [71]

    C. Villani. Topics in optimal transportation , volume 58 of Graduate Studies in Mathematics . American Mathematical Society, Providence, RI, 2003

    2003

  64. [72]

    C. Villani. Optimal transport: old and new , volume 338. Springer Science & Business Media, 2008

    2008

  65. [73]

    M. A. C. Vollmer. Critical points and bifurcations of the three-dimensional O nsager model for liquid crystals. Arch. Ration. Mech. Anal. , 226(2):851--922, 2017

    2017

  66. [74]

    M. Vollmer. A bifurcation analysis of models of liquid crystals . PhD thesis, University of Oxford, 2018

    2018

  67. [75]

    Vaswani, N

    A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, . Kaiser, and I. Polosukhin. Attention is all you need. Advances in neural information processing systems , 30, 2017

    2017

  68. [76]

    Vukadinovic

    J. Vukadinovic. Phase transition for the mckean-vlasov equation of weakly coupled hodgkin-huxley oscillators. Discrete and Continuous Dynamical Systems , 43(11):4113--4138, 2023

    2023

  69. [77]

    Wu and D

    L. Wu and D. Slep c ev. Nonlocal interaction equations in environments with heterogeneities and boundaries. Communications in Partial Differential Equations , 40(7):1241--1281, 2015

    2015

  70. [78]

    Zhao and J

    C. Zhao and J. S. Song. Exact heat kernel on a hypersphere and its applications in kernel svm. Frontiers in Applied Mathematics and Statistics , 4:1, 2018

    2018

  71. [79]

    write newline

    " write newline "" before.all 'output.state := FUNCTION fin.entry add.period write newline FUNCTION new.block output.state before.all = 'skip after.block 'output.state := if FUNCTION new.sentence output.state after.block = 'skip output.state before.all = 'skip after.sentence '...

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.