{"id":"21e6b6cb-5feb-47fe-9527-595ff574708b","arxiv_id":"1908.01386","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A constructive algorithm recovers the magnetic field curl V of a magnetic Schrödinger operator on a compact submanifold of the cylinder R × T^d, d ≥ 3, from Dirichlet-to-Neumann boundary data, under compact-support and vanishing-moment conditions.","lead":"This mathematics paper shows that the magnetic field inside a compact piece of an infinite cylinder can be recovered, step by step, from measurements made only at the boundary surface. It provides the first reconstruction procedure for the magnetic Schrödinger inverse problem in the cylindrical geometry, together with the global estimate needed to build the special solutions.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 6.5's J-I formula has swapped signs and I^± superscripts; it would give J=0 for m·n>0, contradicting its own m·n=±1 clause, while equations (60)-(61) are correct.","rationale":"The reader's conditional verdict is reasonable: the reconstruction strategy is coherent, the weight of the proof is in the long Carleman and boundary-characterization chain, and the vanishing-moment assumption is a genuine restriction but it is explicitly assumed, not a hidden gap. The parameter-sequence existence in §6.2 also appears routine once one notices that τ can be chosen to make τ^2 non-integral. The most load-bearing issue found in a careful pass is the transposed sign/superscript in Theorem 6.5. This is not fatal to the central theorem because the correct formula is derived in §6.4.1 and used implicitly in §6.3.1, but the theorem as stated is false and must be corrected before the paper can be relied upon. Hence the verdict should remain CONDITIONAL: accept only with the correction of Theorem 6.5 to match equations (60)-(61), and with the reader's existing conditions on the unproved parameter sequence and the quoted base Carleman estimate.","tokens_in":69314,"tokens_out":38078,"duration_ms":392768,"concrete_test":"Take m,n with m·n=1. By the definitions in §6.4.1, I^-_1(m,n)=0, so the printed Theorem 6.5 gives J(m,n)=0, while the same theorem's 'm·n=±1' clause and equation (60) give J(m,n)=I(m,n). Independently recompute the log-series expansion from formula (59) and verify that equation (60), not the Theorem 6.5 statement, is produced. Then run the reconstruction in §6.3.1-§6.3.2 with the corrected formula and confirm that the nonzero Fourier coefficients of curl V are recovered.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The sharpest concrete defect is an internal inconsistency in the bridge between the measured integrals I(m,n) and the needed integrals J(m,n). Theorem 6.5 states that for m·n>0, J(m,n) = sum_j (1/j)(-2π/|m|)^{j-1} I^-_j(m,n), and for m·n<0, J(m,n) = sum_j (1/j)(2π/|m|)^{j-1} I^+_j(m,n). But from the definitions in §6.4.1, when m·n>0 the set T^-_1(m,n) is empty, so I^-_1(m,n)=0, and all higher I^-_j are zero as well; the printed formula would therefore force J=0. This contradicts the same theorem's clause 'if m·n=±1, then J(m,n)=I(m,n)' and is inconsistent with equations (60)-(61), which give the opposite pairing: m·n>0 uses (+2π/|m|)^{j-1}I^+_j, while m·n<0 uses (-2π/|m|)^{j-1}I^-_j. The derivation in §6.3.1 from formula (59) and the log-series expansion yields the appendix version, so the reconstruction can be repaired by replacing the misprinted Theorem 6.5 statement with equations (60)-(61). Still, as printed, the paper contains a false key statement: a reader following Theorem 6.5 literally would conclude that all positive-dot-product contributions vanish, and the reconstruction of curl V would fail at that step.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the Calderón-type inverse problem for the magnetic Schrödinger operator H_{V,W} = (D+V)^2 + W on a compact submanifold M of the infinite cylinder T = R × T^d. The main result, Theorem 1.1, asserts that, for d ≥ 3 and under the hypotheses (†), the magnetic field curl V can be reconstructed in a constructive way from the Dirichlet-to-Neumann map Λ_{V,W}. The proof proceeds by establishing a global Carleman estimate (Theorem 1.2) via conjugation of the magnetic operator to the Laplacian using semiclassical pseudodifferential operators, then constructing complex geometric optics solutions, characterizing their boundary values through a boundary integral equation, and finally extracting the Fourier coefficients of curl V from limits of boundary pairings and Laplace-transform reconstruction. The paper is written in thesis style, with most of the required pseudodifferential calculus and auxiliary results proved in detail.","tokens_in":69485,"tokens_out":7295,"duration_ms":84771,"significance":"If the main result is correct, this is the first constructive reconstruction of the magnetic field and the first global Carleman estimate for the magnetic Schrödinger operator in the cylindrical setting, extending the uniqueness result of Dos Santos Ferreira–Kenig–Salo–Uhlmann and the Euclidean reconstruction of Salo. The paper is careful in stating hypotheses and in flagging its own limitations: the restriction δ < 1, the spectral exclusions τ² ∉ Spec(−Δ_{g0}), the d ≥ 3 requirement in Lemma 6.8, and the explicit remark in Section 4.5 that the uniqueness step is not a direct perturbative consequence. However, the manuscript contains a false key statement in Theorem 6.5 relating the measured integrals I(m,n) to the needed integrals J(m,n); the correct formulas appear in equations (60)–(61), so the central derivation is repairable but not correct as printed.","major_comments":[{"comment":"The displayed formulas in Theorem 6.5 for J(m,n) have interchanged superscripts and reversed signs relative to the derivation. From the definitions of T^±_j in §6.4.1, if m·n > 0 then T^-_1(m,n) is empty, hence I^-_j(m,n) = 0 for all j; the printed formula would therefore force J(m,n) = 0 for every positive dot product, contradicting the same theorem's clause that J(m,n) = I(m,n) when m·n = ±1 and contradicting the correct equations (60)–(61). Since this theorem is the bridge that converts the measured quantities I(m,n) into the integrals J(m,n) used for the Laplace reconstruction, it is load-bearing: a reader following Theorem 6.5 literally would conclude that all positive-dot-product contributions vanish and the reconstruction fails. Please replace the statement of Theorem 6.5 with the formulas in (60)–(61) and re-verify every occurrence of the superscripts and signs.","section":"§6.4.1, Theorem 6.5 and equations (60)–(61)"},{"comment":"The vanishing moment condition ∫_R V(x1,x') dx1 = 0 is a genuine restriction on the admissible magnetic potentials and is not merely a gauge normalization used for convenience. The author's own discussion around equation (16) and Theorem 4.8 states that when the denominator (ξ+i)η + ℏt·m vanishes there is no unique decaying solution and that the vanishing moment condition is the 'simplest way to avoid the problem.' This condition is load-bearing for the conjugation in Theorem 4.2, for the CGO construction, and ultimately for Theorem 1.1. The paper should state prominently, both in the introduction and in the discussion of Theorem 1.1, that the reconstruction is proved only under this mean-zero hypothesis, and should comment on whether the condition can be relaxed or whether it is essential to the method.","section":"§4.1 and conditions (⋆)/(†)"}],"minor_comments":[{"comment":"The notation a_m is used both for Fourier coefficients and for the WKB amplitude introduced in Proposition 6.2. The remark after Proposition 6.2 acknowledges this conflict, but using a different symbol for the amplitude (for example, A_m or α_m) throughout Chapter 6 would substantially improve readability.","section":"§6.2 and §4.3"},{"comment":"Theorem 4.12 states 'There exists ℏ0 ≥ 1' but the proof requires small ℏ and uses bounds of the form 0 < |ℏ| ≤ ℏ0; the intended statement is clearly ℏ0 ∈ (0,1]. Please correct the inequality.","section":"§4.5, Theorem 4.12"},{"comment":"The d ≥ 3 restriction is shown to be sharp for the linear-algebra construction, and this is an important structural limitation of the reconstruction procedure. Since Theorem 1.1 is stated only for d ≥ 3, this restriction is consistent, but it would be helpful to mention it explicitly in the introduction rather than only in the appendix.","section":"§6.4.2, remark after Lemma 6.8"}],"recommendation":"major_revision","confidential_remarks":"The manuscript contains a substantial amount of original and carefully developed material, and the thesis-level detail is a strength. However, the false statement in Theorem 6.5 must be corrected before publication; the correct versions already appear in equations (60)–(61), so I regard this as a repairable but load-bearing error. The paper is also long for a journal submission, and the editor may wish to consider whether the background material in Chapters 2–3 should be condensed in a revised version."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The headline: this is a genuine first — a constructive reconstruction of curl V from the Dirichlet-to-Neumann map for a magnetic Schrödinger operator on the cylinder R×T^d, powered by a global Carleman estimate proved by conjugating the magnetic operator to the Laplacian. It follows Salo's Euclidean blueprint, but the transfer is not routine: the cylinder lacks the rotations Salo uses to reduce the transport equation to a ∂-bar equation, and the author works around this with a Fourier-mode ODE analysis under a vanishing-moment condition. That restriction (∫_R V dx1 = 0) is real and acknowledged; it is the price of the decaying conjugation symbol.\n\nThe strengths: thesis-level detail, all hypotheses explicit, and the author flags his own soft spots — the δ<1 restriction, the τ² ∉ Spec(-Δ_g0) exclusion, the d≥3 linear-algebra obstruction. Chapter 5's boundary characterization via layer potentials is solid, and the ODE lemmas in Chapter 4 are careful. Theorem 4.1 is quoted from Kenig–Salo–Uhlmann with a \"can be carried out\" for the torus; that is a small gap in a thesis claiming proof, but a reasonable quote.\n\nThe one thing you must know before using the paper: Theorem 6.5 is misprinted. The stress-test is right. As printed, the m·n>0 case uses I^-_j, but T^-_1(m,n) is empty in that situation, so I^-_j = 0 and the formula forces J = 0 — contradicting the same theorem's \"m·n=±1 implies J=I\" clause and the appendix derivation. Equations (60)-(61) carry the correct pairing (m·n>0 goes with I^+_j and a plus sign; m·n<0 goes with I^-_j and a minus sign). So the bridge from the measured I's to the needed J's is repairable by swapping superscripts and signs in the statement, but as printed it is a false key statement.\n\nMinor: the τ(m,N,σ) sequence in §6.2 is asserted without proof; it is plausible — the allowed interval is short and contains at most one integer, so dodging the spectrum is easy — but it needs a line.\n\nMy verdict tracks the reader's: sound architecture, honest hypotheses, one genuine typo-level inconsistency in a load-bearing statement. This paper deserves a serious referee; the misprint is a required revision, not a rejection. Anyone working on inverse problems on cylinders or Carleman estimates for magnetic operators gets value from it, and I would cite it. Bring the reconstruction chapter to a reading group; skip the prerequisites unless you need the Carleman machinery.","headline":"A solid, honest thesis that delivers the first reconstruction of curl V on the cylinder, with one genuinely misprinted key theorem that is easily repaired.","tokens_in":70220,"tokens_out":5706,"would_cite":true,"duration_ms":56322,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35R30","35J10","35S05","58J32"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper proves that, under a zero-mean condition on the magnetic potential, the magnetic field $\\operatorname{curl}V$ can be reconstructed constructively from the Dirichlet-to-Neumann map on a compact domain inside…","keywords":["magnetic Schrodinger operator","inverse boundary value problem","Dirichlet-to-Neumann map","Carleman estimate","complex geometric optics","cylindrical manifold","reconstruction of magnetic field","semiclassical pseudodifferential operators"],"falsifier":"Take $d\\ge3$ and a smooth compactly supported magnetic potential $V$ on $M^-$ whose $x_1$-mean is nonzero for some $x'$, while keeping all other hypotheses; then compute whether the conjugated operator $e^{2\\pi\\tau x_1}H_{V,W}e^{-2\\pi\\tau x_1}$ is still invertible from $L^2_\\delta$ to $H^2_{-\\delta}$ with the bound $\\|u\\|_{H^s_{-\\delta}}\\lesssim|\\tau|^{s-1}\\|f\\|_{L^2_\\delta}$. The paper's own ODE analysis shows the transport equation $(\\xi+i)D_{x_1}u_m+\\hbar t\\cdot m\\,u_m=(\\xi+i)F_m+\\hbar t\\cdot G_m$ has no decaying solution when $\\hbar t\\cdot m=0$ and the moment is nonzero, so failure of invertibility in that case would falsify the claimed reduction.","tokens_in":68882,"feed_emoji":"🧲","tokens_out":12374,"duration_ms":102512,"temperature":0.7,"pith_summary":"This paper tries to establish a constructive inverse boundary value theorem: on a compact domain inside the infinite cylinder $\\mathbb{R}\\times\\mathbb{T}^d$ with $d\\ge3$, the magnetic field $\\operatorname{curl}V$ of a magnetic Schrodinger operator $H_{V,W}=(D+V)^2+W$ is determined by the Dirichlet-to-Neumann map $\\Lambda_{V,W}$. The admissible potentials must be smooth, compactly supported in the interior, have zero average along the cylinder axis for each point of the torus, and avoid zero as a Dirichlet eigenvalue. A sympathetic reader would care because this turns a uniqueness statement for magnetic potentials into an actual algorithm: the field, which is gauge-invariant and physically meaningful, can in principle be computed from boundary pairings. The proof's engine is a global Carleman estimate for the conjugated operator, obtained by conjugating the magnetic operator essentially into the Laplacian through semiclassical pseudodifferential operators on the cylinder.","feed_headline":"Boundary map recovers the magnetic field inside a cylinder","feed_subtitle":"A constructive proof extracts the gauge-invariant magnetic field from Dirichlet-to-Neumann measurements on a cylinder.","key_machinery":"The load-bearing mechanism is the conjugation identity $(\\Delta_\\hbar+2\\hbar V_\\hbar)A=B\\Delta_\\hbar+\\hbar^{1+\\varepsilon}R$, in which $\\hbar=\\tau^{-1}$, $A$ and $B$ are invertible semiclassical pseudodifferential operators on $\\mathbb{R}\\times\\mathbb{T}^d$, $V_\\hbar$ is the conjugated magnetic term, and $R$ is a remainder that gains a power of the small parameter. The symbol of $A$ is built from the solution $u$ of a first-order transport equation $(\\xi+i)D_{x_1}u+\\hbar t\\cdot D_{x'}u=(\\xi+i)F+\\hbar t\\cdot G$ in directions where the Laplacian symbol is elliptic; the vanishing-moment condition makes that solution decay. This identity reduces the Carleman estimate for the magnetic operator to the known anisotropic Carleman estimate for the Laplacian, and later the same decay estimates control the error terms in the reconstruction of $\\operatorname{curl}V$.","core_discovery":"The central claim is Theorem 1.1: for $d\\ge3$, if $V\\in C^\\infty_c(M^-)$, $W\\in L^\\infty(M)$, $\\int_{\\mathbb{R}}V(x_1,x')\\,dx_1=0$ for all $x'\\in\\mathbb{T}^d$, and $0$ is not an eigenvalue of $H_{V,W}$ in $M$, then $\\operatorname{curl}V$ can be reconstructed from $\\Lambda_{V,W}$. The supporting Theorem 1.2 states that for $1/2<\\delta<1$ and $\\tau^2\\notin\\operatorname{Spec}(-\\Delta_{g_0})$, the conjugated operator $e^{2\\pi\\tau x_1}H_{V,W}e^{-2\\pi\\tau x_1}$ is invertible from $L^2_\\delta$ onto $H^2_{-\\delta}$ with the one-derivative gain $\\|u\\|_{H^s_{-\\delta}}\\lesssim|\\tau|^{s-1}\\|f\\|_{L^2_\\delta}$ for $s=0,1,2$. Theorem 1.2 is proved by a pseudodifferential conjugation that reduces the magnetic operator to the Laplacian, whose Carleman estimates are already available; the reconstruction then follows the complex-geometric-optics route, producing special solutions whose boundary values are computable from $\\Lambda_{V,W}$ and whose asymptotic integrals determine the Fourier coefficients of $\\operatorname{curl}V$.","pith_inferences":["If the zero-mean condition were dropped, the method's own Fourier-mode analysis indicates there is no decaying solution to the transport equation when $\\hbar t\\cdot m=0$, so the decay estimates (15) and (21) would fail; a different conjugation, or a genuinely different class of amplitudes, would be needed.","The reconstruction is constructive but relies on limits along a sequence $N\\to\\infty$ and on the Laplace transform of an entire function recovered from values on a convergent sequence; a numerical implementation would therefore need boundary data of high accuracy and a stabilization strategy, neither of which the paper addresses.","Because the linear algebra step uses equal-norm lattice points perpendicular to a fixed direction, the method appears tied to $d\\ge3$; extending the reconstruction to $d=2$ would require a new way to generate the curl vectors."],"forward_implications":["Boundary values of the constructed CGO solutions are explicitly computable from $\\Lambda_{V,W}$ via the invertible boundary operator $I+\\operatorname{tr}\\circ S_\\tau(\\Lambda_{V,W}-\\Lambda_{0,0})$.","The measured boundary pairings determine the mixed integrals $I(m,n)=\\int_M e^{2\\pi\\mu_{m,n}x_1}e^{-n(x')}(i|m|,m)\\cdot V\\,\\tilde a_m$, and, via the power-series relation of Theorem 6.5, the linear integrals $J(m,n)$.","The Fourier coefficients of $\\operatorname{curl}V$ are recovered by expressing each curl vector as a linear combination of vectors $(i|m|,m)$ with the same norm, a step the paper proves for $d\\ge3$.","The magnetic potential $V$ itself cannot be recovered, only $\\operatorname{curl}V$: the boundary map is gauge-invariant under $V\\mapsto V+\\nabla\\phi$ with $\\phi|_{\\partial M}=0$."],"supporting_citations":[{"why":"Supplies the global Carleman estimates for the Laplacian on the cylinder that the magnetic Carleman estimate is reduced to by conjugation.","marker":"[8]"},{"why":"Provides the Euclidean magnetic-field reconstruction algorithm and the semiclassical conjugation scheme adapted to the cylinder.","marker":"[18]"},{"why":"Establishes the uniqueness result for the magnetic Schrodinger operator in the cylindrical setting that this thesis extends to reconstruction.","marker":"[2]"},{"why":"Gives the boundary characterization and reconstruction scheme for the electric potential in the cylindrical setting, including the use of the DN map to compute boundary values of CGO solutions.","marker":"[9]"},{"why":"Introduces the boundary-to-interior characterization of special solutions and the scattering-transform idea used in the reconstruction chapter.","marker":"[13]"},{"why":"Records the gauge invariance of the Dirichlet-to-Neumann map, which shows why only the magnetic field can be recovered.","marker":"[22]"},{"why":"Provides the original complex-geometric-optics construction and the global Carleman methodology for the Laplacian.","marker":"[23]"}],"fun_headline_variants":["Magnetic field in cylinder reconstructed from boundary data","Reconstructing curl V from boundary data on a cylinder","Boundary data determine magnetic field in cylinder","Magnetic field recovered from cylinder boundary measurements","Cylinder boundary map yields magnetic field"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is the vanishing-moment condition $\\int_{\\mathbb{R}}V(x_1,x')\\,dx_1=0$ for every $x'$: when the Fourier-mode denominator $(\\xi+i)\\eta+\\hbar t\\cdot m$ vanishes, the paper itself notes that there is no unique decaying solution, and without this condition the conjugation, the decay estimates, and the reconstruction all break.","fun_headline_variants_meta":{"raw":{"variants":["Magnetic field in cylinder reconstructed from boundary data","Reconstructing curl V from boundary data on a cylinder","Boundary data determine magnetic field in cylinder","Magnetic field recovered from cylinder boundary measurements","Cylinder boundary map yields magnetic field"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000306,"raw_usage":{"total_tokens":1844,"prompt_tokens":1126,"completion_tokens":718,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":742,"completion_tokens_details":{"reasoning_tokens":648}},"tokens_in":742,"tokens_out":718,"duration_ms":6660,"temperature":1.0,"reasoning_tokens":648,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T15:16:49.705808+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take $d\\ge3$ and a smooth compactly supported magnetic potential $V$ on $M^-$ whose $x_1$-mean is nonzero for some $x'$, while keeping all other hypotheses; then compute whether the conjugated operator $e^{2\\pi\\tau x_1}H_{V,W}e^{-2\\pi\\tau x_1}$ is still invertible from $L^2_\\delta$ to $H^2_{-\\delta}$ with the bound $\\|u\\|_{H^s_{-\\delta}}\\lesssim|\\tau|^{s-1}\\|f\\|_{L^2_\\delta}$. The paper's own ODE analysis shows the transport equation $(\\xi+i)D_{x_1}u_m+\\hbar t\\cdot m\\,u_m=(\\xi+i)F_m+\\hbar t\\cdot G_m$ has no decaying solution when $\\hbar t\\cdot m=0$ and the moment is nonzero, so failure of invertibility in that case would falsify the claimed reduction.","supporting_citations":[{"cited_title":"Kenig, M","cited_arxiv_id":null,"evidence_quote":"Supplies the global Carleman estimates for the Laplacian on the cylinder that the magnetic Carleman estimate is reduced to by conjugation."},{"cited_title":"Salo , Semiclassical Pseudodifferential Calculus and the Reconst ruction of a Magnetic Field , Comm","cited_arxiv_id":null,"evidence_quote":"Provides the Euclidean magnetic-field reconstruction algorithm and the semiclassical conjugation scheme adapted to the cylinder."},{"cited_title":"Dos Santos Ferreira, C","cited_arxiv_id":null,"evidence_quote":"Establishes the uniqueness result for the magnetic Schrodinger operator in the cylindrical setting that this thesis extends to reconstruction."},{"cited_title":"Kenig, M","cited_arxiv_id":null,"evidence_quote":"Gives the boundary characterization and reconstruction scheme for the electric potential in the cylindrical setting, including the use of the DN map to compute boundary values of CGO solutions."},{"cited_title":"Nachman , Reconstructions from Boundary Measurements , Annals of Mathematics, Vol","cited_arxiv_id":null,"evidence_quote":"Introduces the boundary-to-interior characterization of special solutions and the scattering-transform idea used in the reconstruction chapter."},{"cited_title":"Sun , An inverse boundary value problem for Schr ¨odinger operator with vector potentials , Trans","cited_arxiv_id":null,"evidence_quote":"Records the gauge invariance of the Dirichlet-to-Neumann map, which shows why only the magnetic field can be recovered."},{"cited_title":"Sylvester, G","cited_arxiv_id":null,"evidence_quote":"Provides the original complex-geometric-optics construction and the global Carleman methodology for the Laplacian."}],"review_version":1}