REVIEW 2 major objections 1 minor
Stable Determination of Coefficients in Nonlinear Dynamical Schr\"odinger Equations by Carleman Estimates
T0 review · 2 major / 1 minor · reviewed 2026-08-05 · deepseek-v4-flash
Pith's one-line read For nonlinear Schrödinger equations with analytic nonlinearities, stationary coefficients are stably and uniquely recoverable from boundary flux measurements, provided the coefficients are known near the boundary.
desk verdict Abstract-only, but the high-order linearization plus Carleman approach for nonlinear Schrödinger is a sensible, likely-correct extension; deserves a real referee. 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
High-order linearization: repeatedly differentiating the solution-to-boundary-data map with respect to small-amplitude excitations, producing a hierarchy of linear Schrödinger equations whose sources are built from lower-order derivatives; analyticity of the nonlinearity guarantees all derivatives exist and that the $k$-th Taylor coefficient of the boundary response is isolated by the $k$-th linearized problem. Carleman estimates: exponentially weighted $L^2$ estimates for the linear Schrödinger equation that provide quantitative unique continuation and stability, transferring smallness of the boundary data difference to smallness of the coefficient difference.
What would settle it
Fix the cubic nonlinear Schrödinger equation $i u_t + \Delta u + q(x) u + |u|^2 u = 0$ on a bounded domain with zero Dirichlet boundary data, and let two potentials $q_1,q_2$ agree near the boundary but differ inside. For a small boundary input of amplitude $\epsilon$, compute the third-order term in $\epsilon$ of the Neumann data. If these third-order Neumann data coincide for every such input while $q_1\ne q_2$ in the interior, the uniqueness claim is false; if they differ, the difference as a function of the potential difference tests the stated stability bound. A one-dimensional numerical
Extended reading notes
Core claim
The paper treats the inverse problem of recovering time-independent coefficients in a dynamical nonlinear Schrödinger equation $i\partial_t u + \Delta u + q(x)u + f(x,u,\bar u)=0$ on a bounded domain in $\mathbb{R}^n$, with zero Dirichlet boundary data and small initial data; here $q$ is a stationary potential and $f$ is a locally analytic nonlinearity that may itself depend on stationary coefficients. The main discovery is that these stationary coefficients are uniquely and stably determined by the Neumann data (the normal derivative $\partial_\nu u$ on the boundary) measured on a subset of the boundary, assuming the coefficients are already known close to the boundary. This is proved in tw
Load-bearing premise
The nonlinearity must be locally analytic (representable by a convergent power series around zero), because the entire method relies on taking infinitely many derivatives of the solution map at zero to separate the effects of the coefficients.
Editorial extensions
If this is right
- The same boundary data set used for a linear Schrödinger inverse problem is sufficient for the nonlinear problem, as long as the nonlinearity is analytic and the coefficients are known near the boundary.
- High-order linearization is constructive: sending multiple small-amplitude inputs and taking differences of the recorded Neumann data isolates each Taylor order, so the nonlinearity's coefficients can be recovered one order at a time.
- The geometric-condition case and the arbitrary-subset case quantify the trade-off between measurement access and required coefficient regularity: larger measurement sets need milder assumptions, while small patches need stronger bounds on the size and smoothness of the coefficients.
- The small-data well-posedness result provides the rigorous foundation for the solution map used in the recovery, ensuring the high-order derivatives taken in the linearization are well defined.
Reading between the lines
- If the nonlinearity is only $C^k$ rather than analytic, the hierarchy stops after $k$ steps, so coefficients beyond the $k$-th Taylor order likely become invisible; the stability modulus should degrade accordingly. This is an inference, not a claim of the paper.
- The same strategy of high-order linearization plus Carleman estimates may apply to other nonlinear evolution equations (e.g., nonlinear wave or heat equations) with analytic nonlinearities, so the method may be general beyond Schrödinger equations.
- Because coefficients must be known near the measured boundary, the result implicitly describes a two-stage experimental protocol: first calibrate a boundary layer, then recover the interior coefficients from subsequent boundary flux measurements; the interior recovery's stability should weaken with distance from the boundary.
- For a cubic nonlinearity, the third-order linearized problem corresponds to a four-wave mixing term; this gives a concrete numerical testbed: simulate two potentials that agree near the boundary and check whether their third-order Neumann data are distinguishable.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper (arXiv:2508.07231, abstract only) claims stable and unique determination of stationary coefficients in a class of nonlinear dynamical Schrödinger equations with locally analytic nonlinear terms, given knowledge of the coefficients near the boundary and measured Neumann data on a boundary subset. Two measurement regimes are discussed: a subset satisfying a geometric condition, and an arbitrary subset under stronger regularity and size assumptions on the coefficients. The proposed method combines high-order linearization with Carleman estimates for the linear Schrödinger equation, together with small-data well-posedness. No proofs, theorem statements, or derivations are available in the reviewed material; the full text was not provided.
Significance. If the claimed result is correct, it would be a meaningful advance in inverse problems for nonlinear Schrödinger equations, going beyond linear uniqueness/stability by using analyticity of the nonlinearity to generate infinitely many independent linearized problems. The arbitrary-subset result, under stronger assumptions, is particularly notable. The approach is methodologically credible: high-order linearization is a well-established technique, and Carleman estimates are the standard tool for stable coefficient recovery. The explicit mention of locally analytic nonlinear terms and the geometric condition indicates awareness of the technical requirements. However, because the manuscript contains only the abstract, the significance is necessarily conditional on the details of the Carleman estimates, the well-posedness theory, and the stability inequalities, none of which can be inspected here.
major comments (2)
- [Abstract (entire available text)] The central claim—stable and unique determination of coefficients—is stated but not supported by any proof or precise statement in the available material. The Carleman estimate, the geometric condition on the measurement subset, the function spaces, the smallness assumptions, and the exact stability inequality are all absent. These are load-bearing components of the result. Without them, the correctness of the claim cannot be assessed. This is not an allegation of error, but a fundamental limitation of the reviewable material.
- [Abstract, 'locally analytic nonlinear terms'] The local analyticity of the nonlinearity is the structural assumption that makes high-order linearization possible. The abstract does not specify the precise class of nonlinearities (e.g., polynomial, entire, or with prescribed growth bounds), nor does it state how the high-order linearized problems are derived and why they are well-posed. Since the recovery procedure depends on extracting arbitrarily many Taylor coefficients of the boundary map, the exact analyticity assumption is central. The manuscript needs to state this assumption precisely and justify the linearization procedure at each order.
minor comments (1)
- [Abstract, terminology] The phrase 'stable and unique determination' is used without defining the measurement map or the notion of stability (e.g., Lipschitz, log-Lipschitz, conditional Hölder). The roles of 'small initial data' and 'trivial boundary data' should be clarified: are they necessary for well-posedness or for the linearization procedure?
Circularity Check
No circularity identified from the abstract; the recovery argument is not visibly self-referential.
full rationale
The abstract states an inverse problem in which stationary coefficients are recovered from boundary Neumann data via high-order linearization and Carleman estimates for the linear Schrödinger equation. The key structural assumptions are locality and analyticity of the nonlinear term, knowledge of the coefficients near the boundary, and either a geometric condition or stronger regularity/size assumptions. None of these inputs appears to be defined in terms of the output coefficient itself, and no fitted parameter is renamed as a prediction. The use of Carleman estimates is an external analytic tool, not a circular premise. The local analyticity assumption is explicitly stated and is a hypothesis of the theorem rather than a hidden redefinition of the target coefficients. Because the full text is not available, the correctness of the Carleman estimate and the stability proof cannot be verified, but unverifiability is a correctness risk, not evidence of circularity. No self-citation, definitional circularity, or construction-induced reduction is visible from the abstract alone. Therefore the circularity score is 0.
Assumptions & free parameters
assumptions (5)
- domain assumption The nonlinear term is locally analytic in the solution variable.
- domain assumption The initial data are sufficiently small and the boundary data are trivial.
- domain assumption The coefficients are known in a neighborhood of the boundary.
- domain assumption The Neumann data are measured on a subset of the boundary satisfying a geometric condition (or on an arbitrary subset under stronger regularity and size assumptions).
- standard math Carleman estimates for the linear Schrödinger equation hold on the considered domain.
Cite this review
Pith. "Pith review of Stable Determination of Coefficients in Nonlinear Dynamical Schr\"odinger Equations by Carleman Estimates." pith.science (2026). https://pith.science/paper/FQNRDJUF
@misc{pith2026250807231,
author = {Pith},
title = {Pith review of: Stable Determination of Coefficients in Nonlinear Dynamical Schr\"odinger Equations by Carleman Estimates},
year = {2026},
howpublished = {\url{https://pith.science/paper/FQNRDJUF}},
note = {Machine review of arXiv:2508.07231}
}
read the original abstract
We consider the inverse problem of recovering stationary coefficients in a class of dynamical Schr\"odinger equations with locally analytic nonlinear terms. Upon treating the well-posedness for small initial data and trivial boundary data, we proceed to establish stable and unique determination provided knowledge of the coefficients near the boundary and the measured Neumann data of the solution. We discuss both the case of measurement on a subset of the boundary large enough to satisfy a certain geometrical condition and, under stronger assumptions on the regularity and size of the coefficients, the case of measurement on arbitrary subsets of the boundary. Our argument relies on high-order linearization and Carleman estimates for the linear Schr\"odinger equation.
Reviewed August 5, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.