REVIEW 4 major objections 7 minor 22 references
A Li-Yau gradient estimate for the Finslerian logarithmic Schr\"{o}dinger equation
T0 review · 4 major / 7 minor · reviewed 2026-08-15 · deepseek-v4-flash
Pith's one-line read A Li-Yau gradient estimate holds for the Finslerian logarithmic Schrödinger equation under mixed weighted Ricci curvature bounded below, yielding Harnack and boundedness results.
desk verdict A serious but conditional Finslerian extension of Wang's Li–Yau estimate; the main theorem leans on an unstated comparison theorem from an unreviewed preprint, and the applications violate the strict A > a_+ hypothesis. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The central mechanism is the maximum-principle cut-off argument applied to $H=\phi L$ with $L=t[F^2(\nabla f)+(A+a)f+2(E+b)-2f_t]$ and a cut-off $\phi$ equal to 1 on $B_p(R)$ and 0 outside $B_p(2R)$. At the maximum point, the proof controls $\phi\Delta_{\nabla u}L$ from below by $BL$; the constant $B$ comes from a new Laplacian comparison theorem for Finsler distance functions, together with the reverse triangle inequality for the misalignment $\alpha$ and the uniform tensor bound $F(U)+F^*(T)+F(\mathrm{div}_C(V))\le K_0$. The quadratic-in-$L$ term in the Bochner inequality, involving $h=F^2(\nabla f)/L$, is what forces the $4N[\cdots]$ structure of the final estimate. Large constants $A,E$ are chosen so that lower-order coefficient terms become nonnegative, allowing the two-case split that completes the bound.
What would settle it
Take a forward complete Finsler manifold with $m\mathrm{Ric}_N\ge -K$, for instance a Finsler metric on $\mathbb{R}^n$ with a prescribed S-curvature, and an explicit positive bounded solution of (1.1). Compute both sides of (1.2) on a grid in $B_p(R)\times(0,\infty)$; if the left side ever exceeds the right side while the tensor bound $F(U)+F^*(T)+F(\mathrm{div}_C(V))\le K_0$ holds, the theorem is false. A more targeted check is to verify the cut-off Laplacian bound $\Delta_{\nabla u}\phi \ge -B$ used in the proof on that metric, since that is the step where the comparison theorem enters.
Extended reading notes
Core claim
The paper proves Theorem 1.1: if $u$ is a positive bounded solution of $(\Delta-\partial_t)u + a(x,t)u\log u + b(x,t)u = 0$ on a forward complete non-compact Finsler manifold with finite misalignment, $m\mathrm{Ric}_N \ge -K$ on $B_p(2R)$, and $F(U)+F^*(T)+F(\mathrm{div}_C(V))\le K_0$, then with $f=\log(u/D)\le 0$ the quantity $F^2(\nabla u)/u^2 + (A+a)f - 2u_t/u$ is bounded on $B_p(R)\times(0,\infty)$ by the explicit right-hand side of (1.2): a combination of $4N[1/t + [a]_+ + B + N C_1^2/R^2 + 2[M-a\log D]_+/N]$ and a term built from $-[A-2K-2-|\log D|]_+$ and $[\Delta_{\nabla u}a + a_t]_+/(4(A-[a]_+))$. The proof runs the maximum principle on $\phi L$, where $L=t[F^2(\nabla f)+(A+a)f+2(E+b)-2f_t]$, using the Bochner formula and a new Laplacian comparison theorem to control $\Delta_{\nabla u}\phi$. As applications the paper derives a local Harnack inequality (Theorem 4.1) and, for the stationary equation $\Delta u+2u\log u+V(x)u=0$ with $m\mathrm{Ric}_N\ge 0$, an explicit boundedness estimate for positive solutions (Theorem 4.2).
Load-bearing premise
The estimate depends on a comparison theorem for the Laplacian of distance on Finsler manifolds being available, and on the non-Riemannian correction tensors being uniformly bounded; if either condition fails, the lower bound on the cut-off Laplacian and therefore the final inequality do not follow.
Editorial extensions
If this is right
- For the heat-type equation $(\Delta-\partial_t)u + b(x,t)u=0$, the gradient estimate directly implies the local Harnack inequality $u(x_1,t_1)\le u(x_2,t_2)(t_2/t_1)^{2N}\exp\{(t_2-t_1)T+S(x_1,x_2,t_2-t_1)\}$ for $x_1,x_2\in B_p(R)$ and $0\le t_1\le t_2$.
- For the stationary equation $\Delta u+2u\log u+V(x)u=0$ with $m\mathrm{Ric}_N\ge 0$ and bounded $V$, $\Delta_{\nabla u}V$, and $F_{\nabla u}(\nabla_{\nabla u}V)$, every positive solution is bounded above by an explicit exponential of the bounds on $V$ and its Hessian and gradient.
- The estimate is local and quantitative: it holds on $B_p(R)\times(0,\infty)$ and degenerates in the expected way as $t\to 0$ through $1/t$ and as $R\to\infty$ through the constant $B$.
- Because the coefficients $a$ and $b$ are allowed to be time-dependent, the theorem covers parabolic equations not accessible to earlier constant-coefficient treatments.
Reading between the lines
- Not pursued in the paper: the same cut-off argument should adapt to other nonlinear Finslerian parabolic equations whose nonlinearity is a Lipschitz perturbation of the Laplacian, as long as the same comparison and tensor bounds hold.
- The two-case split in the proof suggests the constant $4N$ and the term involving $-(A-2K-2-|\log D|)_+$ are not optimized; a reader could try to sharpen them by choosing $A$ and the cut-off differently.
- A direct consequence the author leaves implicit is that letting $R\to\infty$ under global curvature and tensor bounds converts the local estimate into a global one, in the spirit of the stationary boundedness application; this is a testable route to Liouville-type results under mild growth assumptions on solutions.
- Restricting the new Laplacian comparison theorem to a Riemannian metric should make the non-Riemannian tensor terms vanish and recover the Riemannian predecessor estimate, which would isolate exactly what the Finsler generalization adds.
Signed reviews
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The manuscript claims a local Li-Yau type gradient estimate (Theorem 1.1) for positive bounded solutions of the Finslerian logarithmic Schrödinger equation (1.1) on a forward complete non-compact Finsler metric measure space, assuming mixed weighted Ricci curvature bounded below and a uniform bound on certain non-Riemannian tensors. The proof combines Bochner-type formulas with a maximum principle and a cut-off function, and it relies on a Laplacian comparison theorem due to B. Shen (reference [16]). Two applications are given: a Harnack inequality (Theorem 4.1) and an a priori boundedness result for a stationary equation (Theorem 4.2).
Significance. If the proof is correct, this would be a meaningful generalization of Wang's Riemannian gradient estimates to Finsler geometry with general time-dependent coefficients, and the explicit constants in the estimate would provide a useful tool. The manuscript makes a serious attempt at handling the nonlinear Finsler Laplacian, and the algebraic core of Theorem 3.1 appears plausible. However, the central estimate is conditional on an unstated external comparison theorem from an unreviewed preprint, and the applications as written violate the hypotheses of the main theorem. The paper also leaves several technical gaps in the maximum-principle step and in the definition of constants. For these reasons I cannot recommend acceptance in the present form.
major comments (4)
- [Section 3, cut-off argument before (3.23)] The lower bound for Δ∇uφ is obtained by quoting 'the Laplacian comparison theorem ... established by B. Shen. See Theorem 1.1 in [16]', but the theorem is not stated in the manuscript, and [16] is an arXiv preprint rather than a peer-reviewed reference. The bound is used for the solution-dependent direction ∇u, and the constant C0 is asserted to depend on K0, the bound on F(U)+F*(T)+F(div_C(V)). It is not apparent from the text whether Shen's theorem applies to this nonlinear setting, to this class of reference directions, and to cut-offs on forward geodesic balls with the stated constants. Because the B·L term in (3.23) is what makes the maximum-principle argument close, the validity of (1.2) is directly conditional on this unverified external result. The author should state Shen's theorem precisely and either prove the needed version or show explicitly that the hypotheses of [16, Theorem 1.1] are satisfied in the present situation.
- [Section 3, 'By Calabi's trick' before (3.23)] The proof says: 'By Calabi’s trick [5], we can assume further that φ is smooth on the forward metric ball Bp(2R).' On a Finsler manifold the distance function r is only Lipschitz, and no approximation argument is supplied to justify the pointwise maximum-principle identities ∇∇uH(x0,t0)=0, Δ∇uH(x0,t0)≤0 at a maximum point that may lie at a non-smooth point of r. Since these identities are used to derive (3.23) and hence the final estimate, this is a load-bearing regularity gap. The discussion in Remark 3.4 asserts that one can approximate by smooth functions, but it does not explain how the maximum-principle step with the cut-off is recovered in the limit, particularly for the solution-dependent reference direction ∇u.
- [Section 4, Theorems 4.1 and 4.2] Theorem 1.1 requires A > a_+ on Bp(2R)×(0,∞) and A to be a positive constant. In the proof of Theorem 4.1 the author applies Theorem 1.1 'in case 2 with A = a = 0', and in Theorem 4.2 sets A = a = 2. Both choices violate the hypothesis A > a_+ (since a_+=0 in the former and a_+=2 in the latter), and the expression (1.2) contains 4(A-[a]_+) in a denominator. Even if the case-2 bound (3.29) could be used formally without that hypothesis, the theorems as stated do not apply, and the transition from (1.2) to the Harnack and boundedness claims is not justified by the stated hypotheses.
- [Section 4, Theorem 4.1 statement] The Harnack constant T in Theorem 4.1 is displayed as T = 4N[ B + NC1^2/R^2 + 2E/N + 1/2 (-[-2K-2-|logD|]_+) ], but the constant E is not defined in the theorem statement; it is introduced only inside the proof of Theorem 1.1 as a constant chosen large enough to satisfy (3.20) and (3.21). The theorem therefore does not give an explicit Harnack constant in terms of the data, and it is unclear how E depends on a, b, K0, α, D, and the other bounds. This needs to be clarified or the statement revised so that T is expressed entirely through the hypotheses.
minor comments (7)
- [Theorem 1.1 statement] After the inequality (1.2), the line 'ut/u = F(∇f)^2 + (A+a)f − 2ft' is incorrect as a definition of ut/u; it appears to be a garbled version of L/t = F^2(∇f)+(A+a)f+2(E+b)−2ft. Please correct the formula and make the roles of L, E, and M in the theorem statement coherent.
- [Notation, beginning of Section 1 and use in (1.2)] The notation [h]_+ is defined as a global supremum over Bp(2R)×(0,∞), but in (1.2) it is used pointwise, e.g., [a(x,t)]_+. These two uses are inconsistent; please introduce separate notation for the pointwise positive part and the global sup.
- [Definition 2.2] In the definition of global misalignment, the expression 'αM(x,M)' should presumably be 'αM(x,U)'; as written it is not meaningful.
- [Section 3, after (3.22)] The sentence 'Choose the same cut-off function φ as below' is confusing because no cut-off function was introduced before; it should say 'as follows' or the passage should be reworded.
- [Theorem 4.2 statement] The theorem writes '(L,F,µ)' for the manifold, whereas the rest of the paper uses M; please use consistent notation.
- [Section 3, cut-off Laplacian bound] The constants C(N,α) and C0(K0,α) appear in the lower bound for Δ∇uφ but are never defined or characterized. Even aside from the external theorem, give at least the defining inequalities or the exact expressions used.
- [Lemma 2.10(3) proof] The proof of Lemma 2.10(3) is very abbreviated: the displayed integration by parts in s is not carried out, and the reader is told to use the weak formulation of Definition 2.2 without seeing the details. This is acceptable as a sketch but should be expanded for a journal submission.
Circularity Check
No circularity: the gradient estimate is derived from a maximum-principle computation with externally cited geometric comparison results, not from its own conclusion.
full rationale
The paper's central estimate (1.2) is obtained by differentiating the Li–Yau-type functional L = t[F^2(∇f)+(A+a)f+2(E+b)-2f_t], applying the Finslerian Bochner formula, the mixed weighted Ricci curvature lower bound mRic_N ≥ -K, and algebraic inequalities. The target quantity appears on the left-hand side, but it is not assumed: inequality (3.1) is a lower bound for Δ^{∇u}L - L_t obtained from the equation and curvature assumptions, and the final estimate follows by a standard maximum-principle/cut-off argument. The constants A, E, and M are chosen from the assumed bounds on the coefficient functions and curvature (e.g., E large enough to satisfy (3.20)-(3.21)); they are not fitted to force the final inequality. The cut-off argument invokes the Laplacian comparison theorem of B. Shen, reference [16], which is an external result by a different author; even if that theorem is stated only by citation and the preprint is not independently verified, such dependence is a correctness risk, not circularity. The paper does not define its estimate in terms of the cited theorem's conclusion, does not rename a known result, and contains no load-bearing self-citation. Thus no circular step is present.
Assumptions & free parameters
free parameters (4)
- A =
chosen as a positive constant with A > a_+; applications use A = 0 and A = 2
- E =
large enough to satisfy (3.20) and (3.21)
- M =
unspecified; depends on bounds of a, b, a_t, Δ∇u a, Δ∇u b, F∇u(∇∇u a), F∇u(∇∇u b)
- C1, C2 =
positive constants from the cut-off function, with -C1 ≤ φ'/√φ ≤ 0 and φ'' ≥ -C2
assumptions (5)
- standard math Standard Finsler geometry background, Chern connection, curvature, Laplacian and Hessian identities
- domain assumption Laplacian comparison theorem on non-compact Finsler manifolds from Shen [16], Theorem 1.1
- domain assumption Mixed weighted Ricci curvature lower bound mRic_N ≥ -K and the non-Riemannian tensor bound F(U) + F*(T) + F(div_C(V)) ≤ K0
- domain assumption Regularity of global and local solutions, Theorems 2.8 and 2.9, and mollification Lemma 2.10 from [16]
- ad hoc to paper Without loss of generality assume L > 0 in the proof of Theorem 1.1
Cite this review
Pith. "Pith review of A Li-Yau gradient estimate for the Finslerian logarithmic Schr\"{o}dinger equation." pith.science (2026). https://pith.science/paper/TJAPXE5L
@misc{pith2026250706237,
author = {Pith},
title = {Pith review of: A Li-Yau gradient estimate for the Finslerian logarithmic Schr\"odinger equation},
year = {2026},
howpublished = {\url{https://pith.science/paper/TJAPXE5L}},
note = {Machine review of arXiv:2507.06237}
}
read the original abstract
By leveraging a new Laplacian comparison theorem, we derive a Li-Yau type gradient estimate for a particular nonlinear parabolic equation, namely, the Finslerian logarithmic Schrodinger equation on a non-compact, complete Finsler manifold with mixed weighted Ricci curvature bounded from below. In our framework, all coefficients are time-dependent functions defined on the manifold. As applications, we establish both a Harnack inequality and an a priori estimate for the positive solutions of this specific equation
Reference graph
Works this paper leans on
-
[16]
Shen, B., Operators on nonlinear metric measure spaces I: A new Laplacian comparison theo- rem on Finsler manifolds and a generalized Li–Yau approach to gradient estimates of Finslerian Schr¨ odinger equation, arXiv:2312.06617 [math.DG], accessed 22 Apr. 2024
arXiv 2024
-
[5]
Hopf’s maximum principle with an application to Riemannian geometry, Duke Math
Calabi, E., An extension of E. Hopf’s maximum principle with an application to Riemannian geometry, Duke Math. J. 25 (1958), no. 1. 24 ZISU ZHAO
work page 1958
-
[1]
Abate, M. and Patrizio, G., Finsler Metrics: A Global Approach with Applications to Geo- metric Function Theory, Springer, Berlin, 1994
work page 1994
-
[2]
and Shen, Z., An Introduction to Riemann–Finsler Geometry , Springer, New York, 2012
Bao, D.-W., Chern, S.-S. and Shen, Z., An Introduction to Riemann–Finsler Geometry , Springer, New York, 2012
work page 2012
-
[3]
and Lieb, E.H., Sharp uniform convexity and smoothness inequalities for trace norms, Invent
Ball, K., Carlen, E.A. and Lieb, E.H., Sharp uniform convexity and smoothness inequalities for trace norms, Invent. Math. 115 (1994), no. 1, 463–482
work page 1994
-
[4]
Brasher, J.D., Nonlinear wave mechanics, information theory, and thermodynamics, Int. J. Theor. Phys. 30 (1991), no. 7, 979–984
work page 1991
-
[6]
Cao, X. and Zhang, Z., Differential Harnack estimates for parabolic equations, in: Complex and Differential Geometry, Springer, Berlin, 2011, pp. 87–98
work page 2011
-
[7]
and Yan, D., Ground states for logarithmic Schr¨ odinger equations on locally finite graphs, J
Chang, X., Wang, R. and Yan, D., Ground states for logarithmic Schr¨ odinger equations on locally finite graphs, J. Geom. Anal. 33 (2023), no. 7. doi:10.1007/s12220-023-01267-0
Show all 22 references
-
[8]
and Khanh, N.N., Gradient estimates of Hamilton–Souplet–Zhang type for a general heat equation on Riemannian manifolds, Arch
Dung, N.T. and Khanh, N.N., Gradient estimates of Hamilton–Souplet–Zhang type for a general heat equation on Riemannian manifolds, Arch. Math. 105 (2015), no. 5, 479–490. doi:10.1007/s00013-015-0828-4
2015 doi
-
[9]
and Ma, B., Gradient estimates for a nonlinear parabolic equation on Riemannian manifolds, Arch
Huang, G. and Ma, B., Gradient estimates for a nonlinear parabolic equation on Riemannian manifolds, Arch. Math. 94 (2010), no. 3, 265–275. doi:10.1007/s00013-009-0091-7
2010 doi
-
[10]
Ma, L., Gradient estimates for a simple elliptic equation on complete non-compact Riemannian manifolds, J. Funct. Anal. 241 (2006), no. 1, 374–382. doi:10.1016/j.jfa.2006.06.006
2006 doi
-
[11]
et al., Logarithmic Schr¨ odinger-like equation as a model for magma transport, Europhys
Martino, S.D. et al., Logarithmic Schr¨ odinger-like equation as a model for magma transport, Europhys. Lett. 63 (2003), no. 3, 472–475. doi:10.1209/epl/i2003-00547-6
2003 doi
-
[12]
Ohta, S., Nonlinear geometric analysis on Finsler manifolds, Eur. J. Math. 3 (2017), no. 4, 916–952. doi:10.1007/s40879-017-0143-7
2017 doi
-
[13]
Ohta, S., Comparison Finsler Geometry, Springer, Cham, 2021
2021
-
[14]
and Sturm, K.-T., Bochner–Weitzenb¨ ock formula and Li–Yau estimates on Finsler manifolds, Adv
Ohta, S. and Sturm, K.-T., Bochner–Weitzenb¨ ock formula and Li–Yau estimates on Finsler manifolds, Adv. Math. 252 (2014), 429–448. doi:10.1016/j.aim.2013.10.018
2014 doi
-
[15]
Shen, Z., Lectures on Finsler Geometry, World Scientific, Singapore, 2001
2001
-
[17]
and Zhu, Y., Li–Yau estimates for a nonlinear parabolic equation on Finsler metric measure spaces, Manuscripta Math
Shen, B. and Zhu, Y., Li–Yau estimates for a nonlinear parabolic equation on Finsler metric measure spaces, Manuscripta Math. 176 (2025), no. 22. doi:10.1007/s00229-025-01615-0
2025 doi
-
[18]
and Zhang, Q.S., Sharp gradient estimate and Yau’s Liouville theorem for the heat equation on noncompact manifolds, Bull
Souplet, P. and Zhang, Q.S., Sharp gradient estimate and Yau’s Liouville theorem for the heat equation on noncompact manifolds, Bull. London Math. Soc. 38 (2006), no. 6, 1045–
2006
-
[19]
Wang, J., Gradient estimates for a class of elliptic and parabolic equations on Riemannian manifolds, Front. Math. 18 (2023), no. 5, 999–1024. doi:10.1007/s11464-021-0420-0
2023 doi
-
[20]
Xia, Q., Li–Yau’s estimates on Finsler manifolds, J. Geom. Anal. 33 (2022), no. 2. doi:10.1007/s12220-022-01103-x
2022 doi
-
[21]
Yasue, K., Quantum mechanics of nonconservative systems, Ann. Phys. 114 (1978), nos. 1–2, 479–496. doi:10.1016/0003-4916(78)90279-8 School of Mathematical Science, Xiamen University, Xiamen, Fujian, China Email address : steven1998zhao@outlook.com
1978 doi
-
[1053]
doi:10.1112/S0024609306018947
Reviewed August 15, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.