{"id":"7e18ee0a-f5cd-451c-9f60-d08b298330e4","arxiv_id":"1908.09504","paper_version":4,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Maxwell k-form solution spaces and observable algebras are constructed on globally hyperbolic spacetimes with timelike boundary, under a partially proven assumption on the existence of Green operators.","lead":"This paper sets up Maxwell's equations for differential forms on curved spacetimes with a timelike boundary, including anti-de Sitter-like spaces, and constructs the corresponding quantum algebras of observables. It matters to mathematical physicists working on algebraic quantum field theory in spacetimes with boundaries.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Assumption 16 is not verified for sources supported on the boundary: Appendix A constructs Green operators only on interior-supported forms (Remark 49), so the ultrastatic validation does not cover the domain used by the main Maxwell isomorphisms.","rationale":"The reader's diagnosis is exactly right. The central claim is a theorem whose proof is conditional on Assumption 16, and the only nontrivial evidence for Assumption 16, namely Appendix A, stops at interior-supported sources. I looked for a more damaging internal inconsistency in the main isomorphism proof: the apparent sign typo in Proposition 34 (the string 'G‖δdη = −G‖δdη') and the component-order typo in equation (55) look typographical rather than load-bearing, since the surrounding identities and the boundary-triple degrees are consistent. The use of Remark 22, the vanishing of timelike-compact solutions, is not proved in the paper, but it is a standard energy argument and is not the most exposed point. Thus the most load-bearing concern remains the unverified extension from Γ_c(Λ^kT^*\\mathring M) to Ω^k_c(M). This does not invalidate the paper: the theorems are honest conditional statements, the authors explicitly flag the gap in Remark 49, and the structural results would all follow once the extension is proved (or the main theorems are restricted to interior-supported sources). The verdict should stay CONDITIONAL; closing the boundary-support gap, or explicitly restricting the main theorems to interior-supported test forms with a statement that the boundary-supported extension is open, would make the paper's stated scope accurate.","tokens_in":37420,"tokens_out":9066,"duration_ms":89369,"concrete_test":"Test the missing extension in the simplest ultrastatic model: M = R × R_+ with metric −dt^2 + dx^2, k = 1, ♯ = ‖. Choose α = χ(t)φ(x) dt with φ ∈ C^∞_c([0,∞)) and φ(0) ≠ 0, so supp α ∩ ∂M ≠ ∅. Formally set (G^+_‖ α)(t,x) = θ(t−t') ∫_{R_+} G_D(t−t',x,y) α_t(t',y) dy dt', with G_D the Dirichlet fundamental kernel sin(√(−Δ_D)(t−t'))/√(−Δ_D), i.e. the reflected kernel. Compute t(G^+_‖ α) = G^+_‖ α|_{x=0}, compute tδ(G^+_‖ α), and verify □(G^+_‖ α) = α as distributions, including any boundary-localized terms. If the boundary conditions or the equation fail, Assumption 16 is false as stated on this model; if they hold, repeat for ♯ = ⊥ and for k > 1, then convert the explicit kernels into a continuity argument extending G^±_♯ from interior-supported to general Ω^k_c(M).","verdict_should_be":"UNCHANGED","load_bearing_attack":"Assumption 16 is the load-bearing hinge: it postulates unique advanced and retarded Green operators G^±_♯ : Ω^k_c(M) → Ω^k_{sc,♯}(M) for all compactly supported smooth k-forms on M, with support allowed to meet ∂M. Every structural result in Section 3 — the exact sequence (20) in Proposition 20, the Lorenz-gauge representatives in Propositions 30–31, the classification isomorphisms in Proposition 34, and Corollary 41 — uses this assumption as an input. The only concrete verification, Appendix A, falls short of the stated domain. Theorem 47 is phrased for test sections in Γ_c(Λ^kT^*\\mathring M), the spectral kernel G(τ−τ',x,x') is built on \\mathring M × \\mathring M, and condition (58), which yields the required boundary behaviour, is asserted only for ω ∈ Γ_c(Λ^kT^*\\mathring M). Proposition 48 then identifies the self-adjoint relations Θ_♯ and hence the propagators for the boundary conditions ♯, but still in the interior-supported setting. Remark 49 explicitly says that the extension to supports intersecting ∂M 'can be relaxed' and that a detailed proof 'will be addressed explicitly in a future work.' Thus, for the Maxwell theorems, the stated ultrastatic verification of Assumption 16 does not cover exactly the sources (e.g. α ∈ Ω^k_{tc,δ}(M) with supp α ∩ ∂M ≠ ∅, or the gauge-fixing χ in Proposition 30 whose boundary traces enter the argument) that the proofs manipulate. This is not an internal contradiction, but it makes the central assertions conditional on an unproved functional-analytic extension; the conditions under which boundary-supported sources are permitted are not established.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper develops a systematic treatment of Maxwell equations as a theory for smooth k-forms on globally hyperbolic spacetimes with timelike boundary, in the sense of Aké–Flores–Sánchez. The authors first analyze the D'Alembert–de Rham operator □_k, identify boundary conditions (Dirichlet, □-tangential, □-normal, and Robin-type variants) that make it formally self-adjoint via a Green formula, and, under Assumption 16 (existence of advanced and retarded Green operators), derive an exact sequence characterizing its kernel and cokernel. They then study the Maxwell operator δd, introduce δd-tangential and δd-normal boundary conditions with associated gauge equivalence relations, prove that every gauge equivalence class admits a Lorenz-gauge representative, and establish isomorphisms between the solution spaces Sol_t(M), Sol_nd(M) and quotients of compactly supported coclosed forms by gauge transformations. They also define unital ∗-algebras of observables and prove, under strict inclusions of certain spaces of spacelike-compact exact forms, that these algebras have a nontrivial center. Appendix A uses boundary triples on ultrastatic bounded-geometry spacetimes to verify Assumption 16 for a class of boundary conditions, and Appendices B and C contain the technical decomposition and relative cohomology results used in the proofs.","tokens_in":37758,"tokens_out":10507,"duration_ms":98333,"significance":"If established in full, the paper would provide a rigorous classical and algebraic framework for Maxwell k-forms on Lorentzian manifolds with timelike boundary, including anti-de Sitter type settings, with explicit boundary conditions and a clarification of how the gauge group must be adjusted in the presence of a boundary. The algebraic development is clean: the main isomorphisms are derived in a parameter-free manner from Assumption 16, the dependence on the assumption is explicitly acknowledged, and the boundary-triple verification for ultrastatic spacetimes is a substantial technical contribution. The paper also makes a useful observation, in Corollary 41, that the nontrivial center phenomenon known for Maxwell theory on empty-boundary spacetimes persists under natural boundary conditions. The significance is reduced, however, by the fact that the central classification is conditional on an assumption whose verification in Appendix A is narrower than the assumption as stated.","major_comments":[{"comment":"Assumption 16 is the load-bearing hypothesis for Proposition 20, Propositions 30–31, Proposition 34, and Corollary 41, but the verification in Appendix A covers only test sections with compact support in the interior ˚M: the distributional kernel G in Theorem 47 is built on Γ_c(Λ^kT^*˚M ⊠ Λ^kT^*˚M), condition (58) is asserted for ω ∈ Γ_c(Λ^kT^*˚M), and Remark 49 explicitly defers the extension to supports intersecting ∂M to future work. Since the Maxwell classification in Proposition 34 works with classes in Ω^k_tc,δ(M) and Ω^k_tc,n,δ(M), whose representatives may meet ∂M, the stated ultrastatic verification does not establish Assumption 16 in the form used in the main text. Please either prove the extension, or restate the main theorems with the domain of Assumption 16 narrowed to interior-supported sources and explain how the boundary-supported cases are recovered.","section":"Appendix A, Theorem 47 and Remark 49"},{"comment":"Assumption 16 asserts existence for all f ∈ C∞(∂M), but Proposition 48 verifies the Robin-type relations Θ_{f‖} only for f ≥ 0 and Θ_{f⊥} only for f ≤ 0, with the sign needed for non-negativity of S_{Θ_♯}. Thus the proof of Assumption 16 does not cover arbitrary smooth f; the assumption should be restricted accordingly or the missing sign cases established.","section":"Appendix A, Proposition 48 vs Assumption 16"},{"comment":"In the proof that G‖ descends to the quotient, the displayed chain begins with 'G‖δdη = -G‖δdη', which is inconsistent as written and is not a valid step. The intended identity G‖δdη = δdG‖η (or a corrected version using Corollary 24) is essential for the descent to the quotient, so this proof step must be rewritten.","section":"§3.2, Proposition 34, proof for δd-tangential"},{"comment":"Example 37 considers M = R^m_+ \\ J(p) with the Minkowski metric. The boundary of this set is not smooth: it contains the null cone of p as well as the hyperplane boundary, and the two meet non-transversally. Hence M is not a globally hyperbolic spacetime with smooth timelike boundary in the sense of Theorem 1 and [AFS18], and the example does not demonstrate the strict inclusion dΩ^{k-1}_sc(M) ⊊ Ω^k_sc(M) ∩ dΩ^{k-1}(M) under the paper's standing geometric hypotheses.","section":"§3.2, Example 37"}],"minor_comments":[{"comment":"These propositions are stated for a globally hyperbolic spacetime with timelike boundary without mentioning Assumption 16, although their proofs invoke it through Remark 19 and Lemma 23; the hypotheses should be stated explicitly.","section":"§3.2, Propositions 30, 31, and 34"},{"comment":"In the proof of equation (47), the sentence 'since α ∈ Ω^k_c,δ(M) we can choose α = δβ' is imprecise: the argument tests with δβ for arbitrary β, not that every α is δ-exact. Please rephrase to avoid the implication that every coclosed compactly supported form is exact.","section":"§3.2, Proposition 40"},{"comment":"The symbols tδ and nd used in Appendix A are introduced through a 'slight abuse of notation'; a short explicit definition, for example tδω = t∂Σ δΣ ω and ndω = n∂Σ dΣ ω, would improve readability.","section":"Appendix A, notation after equation (57)"},{"comment":"The reference [DDF19] is listed as 'to appear in Lett. Math. Phys. (2019)'; since the manuscript is dated 2020, the published version should be cited if available.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The main gap, the mismatch between Assumption 16 and the interior-supported verification in Appendix A, is explicitly acknowledged by the authors in Remark 49, so this is not a case of hidden circularity. However, the abstract and the main body state the ultrastatic verification as if it established the assumption in full, and the gap propagates into the central classification results. I would advise the editor that the paper is publishable after the gap is either closed or the claims are made conditional on the restricted domain, and after the proof of Proposition 34 and Example 37 are corrected."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe short version: this is a serious, well-structured paper that extends the DDF19 boundary-triple analysis from scalars to Maxwell k-forms on globally hyperbolic spacetimes with timelike boundary. The algebraic framework—two gauge-equivalence notions, Lorenz-gauge representatives, the solution-space isomorphisms, and the observable algebras with nontrivial center—is new and mostly compelling. But the central results are explicitly conditional on Assumption 16, and the verification in Appendix A covers only test forms with support in the interior. That gap is real and it propagates into Propositions 30, 31, 34 and Corollary 41.\n\nWhat the paper does well: it gives a clean Green-formula analysis of the D'Alembert–de Rham operator with Dirichlet, tangential, normal and Robin-type boundary conditions, proves duality relations and exact sequences conditional on Green operators, then handles Maxwell's operator with two distinct gauge groups, proving every class has a Lorenz-gauge representative. The construction of the observable algebras and the nontrivial-center result is a natural but non-obvious payoff. The authors also flag the weak point themselves in Remark 49, and they thank a referee who found and helped fix a flaw—that is a good sign.\n\nThe main soft spot is exactly the one they flag. Appendix A constructs G± for test sections in Γ_c(Λ^kT*M̊), and Theorem 47's spectral kernel lives on the interior product. Assumption 16 needs G± on all of Ω^k_c(M), including forms whose support touches ∂M. The proofs of the Maxwell theorems manipulate exactly such forms: the α in Proposition 30 can have boundary-supported pieces, and the gauge-fixing χ in Proposition 30 and the β's in Corollary 24 carry boundary traces. So as written, the ultrastatic verification does not establish the domain used by the main theorems. This is not an internal contradiction, and I think the extension is likely true for these particular boundary conditions, but it is not proven here. There are also small typos, e.g. the duplicated sign in the proof of Proposition 34, which make that proof harder to check than it should be.\n\nWho is this for: people working in algebraic QFT on bounded spacetimes, especially AdS and static/ultrastatic models with timelike boundary, and anyone using boundary triples for Green-hyperbolic operators. It is a credible within-program contribution, not a revolution.\n\nMy recommendation: send it to a serious referee. The referee should be asked to verify whether Assumption 16 can be closed, or, failing that, whether the main theorems can be restated for interior-supported test forms without losing the Maxwell applications. I would not desk-reject.","headline":"Solid within-program extension that is honest about its main technical gap: Assumption 16 is not verified for boundary-supported sources, so the central classifications are conditional as written.","tokens_in":38302,"tokens_out":2224,"would_cite":true,"duration_ms":24136,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["81T20","81T05"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper claims that Maxwell's equations on globally hyperbolic spacetimes with timelike boundary can be classified through two boundary-condition-adapted gauge groups, with every gauge class admitting a Lorenz-gauge representative…","keywords":["Maxwell equations","globally hyperbolic spacetimes with timelike boundary","D'Alembert-de Rham operator","advanced and retarded Green operators","Lorenz gauge","boundary triples","algebraic quantum field theory","differential forms"],"falsifier":"Take half-Minkowski spacetime $\\mathbb{R}^{m-1}\\times\\mathbb{R}_+$ with the flat metric, fix a boundary condition among those in Definition 12, and solve $\\Box\\psi=\\omega$ for a smooth $k$-form source $\\omega$ whose compact support meets the boundary $\\partial M$; if for some $k$ and some such $\\omega$ there is no solution $\\psi\\in\\Omega^k_{sc,\\sharp}(M)$ with the required support property, or if two distinct such solutions exist, then Assumption 16 fails and the classification in Propositions 30, 31, and 34 collapses. Conversely, constructing the Green operators for such boundary-touching sources, for instance by proving the reflected-geodesic wavefront set of [GW18], would verify the assumption and complete the paper's program.","tokens_in":37162,"feed_emoji":"⚡","tokens_out":10937,"duration_ms":101546,"temperature":0.7,"pith_summary":"On a spacetime with a timelike boundary, solving Maxwell's equations requires both initial data and boundary conditions, and the gauge group itself must be adapted to those conditions. This paper proves that two distinguished choices — the $\\delta\\mathrm{d}$-tangential condition (the potential restricts to zero on the boundary) and the $\\delta\\mathrm{d}$-normal condition (the normal component of the field strength vanishes) — are compatible with Lorenz gauge fixing: every gauge equivalence class of solutions contains a representative satisfying $\\delta A'=0$ and $\\Box A'=0$. Under the assumption that advanced and retarded Green operators exist for the D'Alembert–de Rham operator with the relevant boundary data, the paper identifies the solution spaces with explicit quotients of compactly supported coclosed forms, and constructs unital $*$-algebras of observables with non-trivial center whenever certain exact spacelike-compact forms are not gauge-trivial. If the assumption holds, this yields a workable quantization scheme for electromagnetism on backgrounds such as anti-de Sitter-type spacetimes, and it shows that a timelike boundary changes the gauge group, not merely the boundary conditions.","feed_headline":"Boundary conditions reduce Maxwell solutions to Lorenz-gauge forms","feed_subtitle":"With the right boundary conditions, Maxwell solutions on timelike-boundary spacetimes reduce to Lorenz-gauge forms.","key_machinery":"The central objects are the advanced and retarded Green operators $G^\\pm_\\sharp$ for the D'Alembert–de Rham operator $\\Box=\\delta d+d\\delta$ on $k$-forms, with $\\sharp$ running over Dirichlet, $\\Box$-tangential, $\\Box$-normal, and Robin-type boundary conditions. Assumption 16 postulates their existence on all compactly supported $k$-forms; from them one forms the causal propagator $G_\\sharp=G^+_\\sharp-G^-_\\sharp$, which fits into the exact sequence $0\\to\\Omega^k_{c,\\sharp}(M)\\to\\Omega^k_c(M)\\to\\Omega^k_{sc,\\sharp}(M)\\to\\Omega^k_{sc}(M)\\to 0$. The mechanism that carries the Maxwell argument is Lemma 23 and Corollary 24: the propagators commute with $d$ and $\\delta$ when the source forms satisfy suitable boundary conditions, up to corrections built from auxiliary forms $\\beta_\\|$ and $\\beta_\\perp$. These identities let the authors push the Lorenz-gauge condition $\\delta A'=0$ through the propagator and read it as coclosedness of the compactly supported initial-data class, which is what produces the quotient isomorphisms. On ultrastatic spacetimes with bounded geometry, the Green operators are constructed using boundary triples, with the boundary conditions encoded as self-adjoint linear relations.","core_discovery":"On a globally hyperbolic spacetime $(M,g)$ with timelike boundary, and for $0<k<m$, the paper establishes that the space $\\mathrm{Sol}_t(M)$ of Maxwell solutions $\\delta dA=0$ with $tA=0$ modulo the gauge group $d\\Omega^{k-1}_t(M)$ is isomorphic, via the causal propagator $G_\\|$, to $\\Omega^k_{tc,\\delta}(M)/\\delta d\\Omega^k_{tc,t}(M)$; analogously, $\\mathrm{Sol}_{nd}(M)$ is isomorphic to $\\Omega^k_{tc,n,\\delta}(M)/\\delta d\\Omega^k_{tc,nd}(M)$, and the same holds for the spacelike-compact versions built from compactly supported forms. In both cases every equivalence class admits a Lorenz-gauge representative $A'$ with $\\Box A'=0$ and $\\delta A'=0$, and the residual gauge freedom is characterized by the spaces $G_t(M)$ and $G_{nd}(M)$. These isomorphisms rely on Assumption 16, the existence of advanced and retarded Green operators for the D'Alembert–de Rham operator with the chosen boundary conditions, which the paper verifies on ultrastatic globally hyperbolic spacetimes with bounded geometry via boundary triples. As an application, the paper constructs unital $*$-algebras of observables $A_t(M)$ and $A_{nd}(M)$ and proves that they have a non-trivial center whenever $d\\Omega^{k-1}_{sc,t}(M)$ is strictly contained in $\\Omega^k_{sc}(M)\\cap d\\Omega^{k-1}_t(M)$, or analogously for the normal case, matching the known failure of general local covariance for gauge theories on boundaryless spacetimes.","pith_inferences":["The paper proves Assumption 16 only for test forms with support in the interior of $M$; until that assumption is proven for supports intersecting the boundary, the classification remains conditional. A plausible route is to show that the Green operators have the reflected-geodesic wavefront set discussed for scalar fields, which would also settle the extension to timelike-compact sources.","A consequence worth drawing is that the variational derivation in the paper singles out the $\\delta\\mathrm{d}$-normal boundary condition as the natural one for ordinary electromagnetism, since it eliminates the boundary term $(t\\alpha, ndA)_{\\partial M}$ in the variation of the action $\\frac{1}{2}(dA,dA)$; the usual gauge group survives only for this condition, and must be restricted for $\\delta\\m","The strict-inclusion condition controlling the center is visible in a concrete example already contained in the paper: in half-Minkowski spacetime with a point removed from the interior, there exists an exact spacelike-compact $1$-form that is not $d$ of a spacelike-compact $0$-form, so the non-trivial center is not a purely abstract possibility.","If a generalization of Lemma 23 could be proven for Robin-type Maxwell boundary conditions, the same quotient-and-algebra construction would likely extend to that larger family; the paper explicitly leaves this as an open obstruction."],"forward_implications":["With Assumption 16 in force, $\\mathrm{Sol}_t(M)$ and $\\mathrm{Sol}_{nd}(M)$ (and their spacelike-compact versions) are completely described by the quotients $\\Omega^k_{tc,\\delta}(M)/\\delta d\\Omega^k_{tc,t}(M)$ and $\\Omega^k_{tc,n,\\delta}(M)/\\delta d\\Omega^k_{tc,nd}(M)$, so computing Maxwell solutions reduces to computing compactly supported coclosed forms and exact forms with boundary data.","Every gauge equivalence class of Maxwell solutions admits a Lorenz-gauge representative, and the residual gauge freedom is exactly the kernel of the Maxwell operator in the restricted gauge spaces $G_t(M)$ and $G_{nd}(M)$.","On ultrastatic globally hyperbolic spacetimes with bounded geometry, the needed Green operators exist for Dirichlet, $\\Box$-tangential, $\\Box$-normal, and Robin-type boundary conditions with $f$ of definite sign, so the classification is unconditional there.","The algebras $A_t(M)$ and $A_{nd}(M)$ have a non-trivial center whenever the inclusion of spacelike-compact exact forms is strict, so the presence of a timelike boundary does not cure the known obstruction to general local covariance in Abelian gauge theories.","The presymplectic forms $\\sigma_t$ and $\\sigma_{nd}$ on the spacelike-compact solution spaces are isomorphic to the forms $\\tilde{G}_\\|$ and $\\tilde{G}_\\perp$ on the quotient spaces, giving a phase-space description that can serve as input for quantization."],"supporting_citations":[{"why":"Supplies the definition and structural theorem, including Cauchy surfaces and the splitting $M\\simeq\\mathbb{R}\\times\\Sigma$, for globally hyperbolic spacetimes with timelike boundary that grounds the whole paper.","marker":"[AFS18]"},{"why":"The scalar-wave analogue whose boundary-triple construction is adapted in Appendix A to prove existence of Green operators on ultrastatic spacetimes.","marker":"[DDF19]"},{"why":"Provides the Green-hyperbolic operator framework, including support properties and extension of fundamental solutions, that Assumption 16 and Proposition 20 rely on.","marker":"[Bär15]"},{"why":"The standard source for advanced and retarded fundamental solutions and the exact-sequence description of solution spaces on boundaryless globally hyperbolic spacetimes that the paper generalizes.","marker":"[BGP07]"},{"why":"Supplies the optimal-observable and Lorenz-gauge machinery for Maxwell $k$-forms that is adapted to the boundary setting in Propositions 30, 31, and 40.","marker":"[Ben16]"},{"why":"Provides the presymplectic-form and symplectic-quotient analysis of linear gauge theories used in Propositions 33 and 35 and in Remark 36.","marker":"[HS13]"},{"why":"Gives the D'Alembert–de Rham operator on $k$-forms and its decomposition into time and spatial elliptic parts used in Appendix A and in the setup of the Maxwell operator.","marker":"[Pfe09]"},{"why":"The wavefront-set analysis of fundamental solutions with boundary reflection that Remark 49 points to as the route for extending Green operators to supports touching the boundary.","marker":"[GW18]"}],"fun_headline_variants":["Lorenz gauge reps for Maxwell on timelike boundary spacetimes","Timelike-boundary Maxwell: gauge reductions and non-trivial center","Maxwell observables on timelike boundaries gain non-trivial center","Boundary conditions ensure Lorenz gauge for Maxwell solutions"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is Assumption 16, that unique advanced and retarded Green operators exist for the D'Alembert–de Rham operator on all compactly supported $k$-forms with the chosen boundary conditions, including forms whose support touches the boundary — the paper only proves existence for interior-supported forms on ultrastatic spacetimes, so if that gap cannot be closed, the classification is conditional.","fun_headline_variants_meta":{"raw":{"variants":["Lorenz gauge reps for Maxwell on timelike boundary spacetimes","Timelike-boundary Maxwell: gauge reductions and non-trivial center","Maxwell observables on timelike boundaries gain non-trivial center","Boundary conditions ensure Lorenz gauge for Maxwell solutions"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000513,"raw_usage":{"total_tokens":2628,"prompt_tokens":1217,"completion_tokens":1411,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":833,"completion_tokens_details":{"reasoning_tokens":1339}},"tokens_in":833,"tokens_out":1411,"duration_ms":11379,"temperature":1.0,"reasoning_tokens":1339,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:09:15.679325+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take half-Minkowski spacetime $\\mathbb{R}^{m-1}\\times\\mathbb{R}_+$ with the flat metric, fix a boundary condition among those in Definition 12, and solve $\\Box\\psi=\\omega$ for a smooth $k$-form source $\\omega$ whose compact support meets the boundary $\\partial M$; if for some $k$ and some such $\\omega$ there is no solution $\\psi\\in\\Omega^k_{sc,\\sharp}(M)$ with the required support property, or if two distinct such solutions exist, then Assumption 16 fails and the classification in Propositions 30, 31, and 34 collapses. Conversely, constructing the Green operators for such boundary-touching sources, for instance by proving the reflected-geodesic wavefront set of [GW18], would verify the assumption and complete the paper's program.","supporting_citations":[],"review_version":1}