{"id":"86a4c6f3-a2dc-4c4d-b340-6c4ec1a38991","arxiv_id":"2607.21815","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Unit directions of Killing fields in Einstein-Lambda spacetimes are characterized from Cauchy-surface data by a new nonlinear PDE system (uKID), derived two ways and shown to admit at most n(n+1)/2 - 1 independent solutions.","lead":"This paper derives the unit Killing initial data (uKID) equations: a nonlinear PDE system on a spatial slice whose solutions are exactly the unit-normalized directions of Killing vector fields in the Einstein-Lambda vacuum development. It provides two independent derivations and shows the system is finite-type, giving a new way to check for spacetime symmetries from initial data without evolving the Einstein equations.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The λ-cancellation defining uKID(2)_AB in (21) is asserted without proof; an expansion using (17)–(20) appears to leave residual D lnλ terms, so Theorem 3's central equivalence needs a computer-algebra check.","rationale":"The reader's CONDITIONAL verdict is sensible, but the single most load-bearing concern is not the inherited KID bijection. For the stated Einstein Λ-vacuum setting, the standard KID-to-Killing theorem is well established; the extended-data caveat is real but is already framed by the authors as conditional on known results. The truly load-bearing internal step is the λ-cancellation in (21). If that cancellation fails, the uKID system is not even well-defined, and no external theorem can rescue the central claim. The paper's own text flags several unverified computations, but the uKID(2) cancellation is more fundamental than the omitted terms in Theorem 4(b) or the sketch of Proposition 10, since it underpins Theorem 3 itself. The proposed CAS check is decisive and inexpensive. Because the reader already required conditional acceptance pending completion of long computations, our concern does not move the verdict; it identifies a specific computation that must be completed. Hence UNCHANGED, with partial agreement on the reader's weakest assumption.","tokens_in":953,"tokens_out":1984,"duration_ms":307551,"concrete_test":"Run a computer algebra system (e.g., xAct, as cited in [17]) on the full expression for uKID(2)_AB in (21), substituting λ^{-1}KID(2)_AB[λu] from (17) and L_A[λu]=L_A[u]+D_A lnλ from (20), for a generic Riemannian metric g_AB, symmetric π_AB, symmetric ∇0π_AB, and arbitrary nonvanishing u^A, u0 satisfying u0^2=1+u_Au^A. Ask the CAS to simplify and collect all terms containing D_A lnλ and D_A D_B lnλ. If any such terms remain, the definition (21) is inconsistent and Theorem 3 must be revised.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Theorem 3 rests on the claim that the right-hand side of (21) is independent of the arbitrary scalar λ, so that uKID(2)_AB is a well-defined operator on u alone. The paper states this cancellation without a displayed computation. Directly expanding (21) with L_A[λu] = L_A[u] + D_A lnλ (eq. 20) and the KID scaling (17), the terms u0 D_A D_B lnλ, u0(D_A lnλ)(D_B lnλ), 2(D_(A u0)(D_B) lnλ), and 2u^Cπ_C(A D_B) lnλ appear to cancel against corresponding terms from D_(A L_B)[λu], L_A[λu]L_B[λu], -2L_(A[u]L_B)[λu], -2L_(A[λu]D_B)u0, and 2u^Cπ_C(A L_B)[λu]. However, a residual term of the form -u0(D_A lnλ L_B[u] - L_A[u] D_B lnλ) survives; it is not symmetric under A↔B and does not vanish upon symmetrization. If this expansion is correct, the definition (21) depends on the arbitrary λ and uKID(2) is not well-defined, breaking Theorem 3 before any KID-extension theorem is invoked. The authors provide no explicit verification of the cancellation, and the reader's own re-derivation covered uKID(0) and uKID(1) but did not check uKID(2). This is the most load-bearing unverified algebraic step in the central construction.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces a new system of PDEs, the unit Killing initial data (uKID) equations, intended to characterize, from initial data, unit vector fields that are proportional to a Killing vector in an Einstein Λ-vacuum development. The system is derived by rescaling the classical KID equations to remove the scaling degree of freedom, yielding a non-linear system (22) whose solutions are claimed to be in local bijection with unit timelike or spacelike Killing directions. The authors also provide prolonged connection forms (Theorem 4), claiming finite type with solution space dimension n(n+1)/2−1, and an independent derivation of uKID via propagation identities and ADN strictly hyperbolic theory (Section 3, Appendix A). The direct derivation in Section 2 is explicit and checkable; the central algebraic cancellation underlying the λ-independence of uKID(2) in (21) is valid.","tokens_in":22424,"tokens_out":21570,"duration_ms":183029,"significance":"If the advertised properties hold, the uKID system is a genuinely useful new tool for intrinsic characterizations of spacetimes in which only the normalized Killing direction is geometrically available, complementing the classical KID equations. The direct derivation is self-contained: I verified the scaling identity (20) and the complete cancellation of λ-dependent terms in the definition of uKID(2), so the central bijection in Theorem 3 is sound. The finite-type prolongation and the propagation-identity derivation are ambitious and potentially valuable, but as written they are not fully demonstrated. The paper would be strengthened by completing those proofs or clearly marking them as conditional.","major_comments":[{"comment":"The proof of the prolonged uKID system (25) is not complete. The text says that certain left-hand sides equal uKID expressions 'up to' omitted terms that are 'long and otherwise unenlightening', and concludes that the system can be solved algebraically. This does not establish the claimed equivalence between (22) and (25). Since Remark 3's dimension bound and the advertised finite-type property depend on this theorem, please supply the full computation, a computer-algebra verifiable supplement, or a more detailed proof.","section":"Theorem 4(b), Section 2.2"},{"comment":"The well-posedness of quasi-linear ADN strictly hyperbolic systems on manifolds is explicitly not proved; Proposition 10 is only sketched. Lemma 11 and Theorem 6 rely on this proposition, so the independent derivation of the uKID equations in Section 3 is conditional on an unproved generalization. Please either prove Proposition 10 or restate Theorem 6 as conditional on it.","section":"Appendix A, Proposition 10; Section 3, Lemma 11/Theorem 6"},{"comment":"The abstract claims a 'bijection' between solutions of the uKID equations and unit vector fields proportional to a Killing vector, but Theorem 3 only guarantees a locally existing scalar λ on Σ. On a non-simply-connected Σ, uKID(0)=0 makes L_A closed but not necessarily exact, so a solution need not correspond to a global Killing vector. Please qualify the abstract and the discussion accordingly ('locally proportional', 'local bijection').","section":"Abstract and Theorem 3"}],"minor_comments":[{"comment":"The sentence 'B_ab[u]=0 implies that locally u_a is the gradient of a scalar' is imprecise: it is the derivative along u, namely u̇_a, that is closed and hence locally a gradient. The scalar is then used as ln λ. Please correct the wording.","section":"Proof of Proposition 1"},{"comment":"Since the λ-independence of uKID(2) is the crux of the construction, please display the simplified λ=1 form of (21) after cancellation, and perhaps introduce notation such as ℓ_A = L_A[u] and φ_A = D_A lnλ to make the cancellation transparent.","section":"Equation (21)"},{"comment":"The proof is dismissed as 'completely analogous' to Proposition 1. It would be helpful to spell out how uKID(0)=0 gives the local scalar λ (via closedness of L_A[u]) and how uKID(1), uKID(2) then reduce to KID(1), KID(2) for v=λu.","section":"Theorem 3 proof"},{"comment":"The statement 'We have checked that this procedure reproduces the uKID equations (22)' is not supported by a displayed calculation. Please include the calculation or provide a supplementary file, so that the claimed independent derivation can be verified.","section":"Section 3.2, end"},{"comment":"The theorem assumes that solutions of the extended KID system (16), with ∇0π kept independent, are in bijection with Killing vectors of the development. This is a non-trivial imported assumption, acknowledged in Remark 2, but it is not stated in the theorem. Please make it an explicit hypothesis.","section":"Remark 2 / Theorem 3 statement"},{"comment":"There are several typos ('prolongued', 'unitu a', etc.) and the notation L_A[u] versus L_A[λu] in (21) is easy to misread. A short remark distinguishing the two would improve readability.","section":"General presentation"}],"recommendation":"major_revision","confidential_remarks":"The central mathematical claim survives the main stress test: the apparent residual λ-dependence in uKID(2), eq. (21), is not real. Expanding with (17)–(20) shows exact cancellation of all D lnλ terms, including the cross terms involving L_A[u]. The paper's main construction is therefore sound. However, the paper advertises a finite-type prolongation and an independent propagation-identity derivation whose proofs are incomplete as written; those need to be made checkable before publication. The local-versus-global bijection wording in the abstract should also be corrected."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear Colleague,\n\nThe main result survives. I re-did the λ-cancellation in the definition of uKID(2)_AB (Eq. 21) by hand, which is the step the stress-test attacks, and the attack does not land. The cross terms in the combination -L_A[λu]L_B[λu] + 2L_(A[u]L_B)[λu] cancel each other exactly, and the remaining u0 D_AD_B lnλ and u0(D_A lnλ)(D_B lnλ) pieces are killed by the same terms in λ^{-1}KID(2)_AB[λu]. The supposed residual -u0(D_A lnλ L_B[u] - L_A[u]D_B lnλ) would, if present, be antisymmetric, and the operator is symmetric. So the definition is λ-free as claimed, and Theorem 3 sits on solid algebra. The L_A trick (19)-(20) is clean.\n\nWhat is actually new: the uKID system (22), the prolongation (25), and the finite-type bound n(n+1)/2 - 1. The motivation is concrete — static perfect-fluid stars give you only the unit direction of the Killing vector from curvature, and uKID is the missing piece for intrinsic initial-data characterizations of that sort. The paper is also honest: it credits Ehlers for the spacetime-level unit Killing condition and claims only the initial-data system as its own. The direct derivation in Section 2 is the real content; the propagation-identity machinery in Section 3 is corroboration, not the foundation.\n\nThe soft spots are where the reader put them. Theorem 4(b) asserts the equivalence but leaves the core of the computation as 'long and otherwise unenlightening.' Since that result carries the dimension bound, a referee should ask for the computation to be completed or machine-checked. Theorem 6 inherits Proposition 10, a manifold-level ADN well-posedness statement the authors present as only a sketch. The claim that the propagation-identity procedure reproduces (22) is asserted, not shown. And the prolonged system divides by u0, so the u0 = 0 case of Theorem 3(b) is outside the canonical form. The abstract's unqualified 'bijection' also glosses over the Poincaré lemma step, which makes the correspondence local on non-simply-connected surfaces.\n\nNone of this breaks the central claim. The paper deserves a serious referee. Send it out, with the request that the omitted verifications be supplied or backed by a CAS worksheet and that the local/global wording be tightened.","headline":"The uKID equations are a genuine new system and the direct derivation in Section 2 checks out; the stress-test's claimed failure of the λ-cancellation in uKID(2) is itself a calculation error, but the paper leans on omitted computations and a sketched well-posedness result that referees should push to finish.","tokens_in":23109,"tokens_out":20907,"would_cite":true,"duration_ms":167973,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["83C05","53C20"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper introduces a new initial-data system, the unit Killing initial data (uKID) equations, whose solutions are in bijection with unit vector fields proportional to Killing vectors in the Einstein-Λ-vacuum development.","keywords":["unit Killing initial data","uKID equations","Killing initial data","Einstein-Λ vacuum","finite type PDE systems","prolongation to connection form","propagation identity","spacetime symmetries"],"falsifier":"Find any extended initial-data set satisfying uKID(0)=uKID(1)=uKID(2)=0 for which no local scalar λ makes (λu0,λuA) a solution of the KID equations; equivalently, on a non-simply-connected initial surface, exhibit a uKID solution whose associated closed 1-form has nontrivial period, so no global λ exists and no global Killing vector arises.","tokens_in":21874,"feed_emoji":"📐","tokens_out":7207,"duration_ms":65588,"temperature":0.7,"pith_summary":"The paper's aim is to find the initial-data version of a simple but useful idea: if a Killing vector is rescaled to unit length, only its direction carries geometric information. The authors construct a system of differential equations on a spacelike initial surface whose solutions are exactly those unit directions that come from Killing vectors of the vacuum spacetime that will develop from the surface. This matters because in several physically relevant settings, such as static perfect-fluid stars, only the unit-normalized symmetry direction can be recovered from curvature invariants, while the full Killing vector cannot. The paper also shows the system is finite type by giving explicit prolongations to connection form, and derives the same equations independently through a propagation-identity argument. If the construction stands, initial-data characterizations of stationary spacetimes can be formulated without first solving the Einstein equations.","feed_headline":"Unit symmetry directions read off initial data alone","feed_subtitle":"New uKID equations match unit parts of Killing vectors before the spacetime is evolved.","key_machinery":"The load-bearing construction is the elimination of the scaling degree of freedom v=λu. From the KID operator the authors define a spatial operator L_A[v] that transforms as L_A[λu] = L_A[u] + D_A ln λ, so its antisymmetrized derivative and a carefully chosen combination with the rescaled KID equations become invariant under λ. That yields the explicit uKID system (21)-(22). The finite-type character is established by giving explicit first-order prolongations to parallel-transport-like connection form (Theorems 2 and 4), whose auxiliary bundle ranks give the dimension bound n(n+1)/2−1; the independent derivation uses a propagation identity whose mixed-order principal symbol is shown to be st","core_discovery":"The central claim is that the unit-normalized part of a Killing vector satisfies its own closed, nonlinear system of equations, both in spacetime and on an initial-data surface. In spacetime, a unit timelike vector u solves the projected, trace-free part of the Killing equation together with the closedness of the 1-form ˙u_a minus the gradient of the scaling factor if and only if u is locally proportional to a Killing vector (Proposition 1). On an extended initial-data surface (Σ, g, π, ∇0π), the corresponding statement is Theorem 3: the three tensor equations uKID(0)=0, uKID(1)=0, uKID(2)=0 hold for a unit vector (u0,uA) if and only if there exists a local scalar λ such that (λu0,λuA) satis","pith_inferences":["The same λ-cancellation trick should apply to other scaling-covariant overdetermined systems, e.g. conformal Killing initial data, producing unit-normalized conformal variants.","Because the rescaling factor λ is produced locally via exactness of a closed form, the uKID-to-Killing bijection is only local on non-simply-connected initial surfaces; global obstructions would require holonomy or cohomology conditions not addressed in the paper.","The paper keeps ∇0π as an independent tensor to stay matter-model agnostic; if the KID-to-Killing isomorphism fails for some matter model in this extended-data sense, the uKID correspondence would inherit that failure.","For static stellar models, uKID equations may supply the missing initial-data ingredient that turns a curvature-based characterization of the unit symmetry direction into a full intrinsic characterization of the spacetime."],"forward_implications":["Initial-data sets can be tested for hidden spacetime symmetries by solving uKID equations instead of first evolving the Einstein equations.","The solution space of unit symmetry directions has dimension at most n(n+1)/2−1, one less than the full Killing-vector space.","The uKID conditions work for timelike and spacelike unit symmetry directions, with the same equations up to a substitution u → i u.","The prolonged connection form gives a practical way to count and extract unit Killing directions from initial data.","The propagation-identity derivation extends the known framework and supplies a well-posedness result for the coupled evolution of the projected Killing fields."],"fun_headline_variants":["Unit Killing fields captured by new initial-data equations","uKID: isolate unit symmetry from initial data alone","Stripping scale: unit Killing vectors emerge from initial data","Forget scaling: unit part of Killing vectors solved on surface"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The central claim rests on the imported theorem that solutions of the KID equations on the initial surface extend to actual Killing vectors of the Einstein-Λ-vacuum development, a result the paper uses rather than proves for its extended-data setting.","fun_headline_variants_meta":{"raw":{"variants":["Unit Killing fields captured by new initial-data equations","uKID: isolate unit symmetry from initial data alone","Stripping scale: unit Killing vectors emerge from initial data","Forget scaling: unit part of Killing vectors solved on surface"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000184,"raw_usage":{"total_tokens":1135,"prompt_tokens":707,"completion_tokens":428,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":451,"completion_tokens_details":{"reasoning_tokens":373}},"tokens_in":451,"tokens_out":428,"duration_ms":9139,"temperature":1.0,"reasoning_tokens":373,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T06:37:24.140524+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Find any extended initial-data set satisfying uKID(0)=uKID(1)=uKID(2)=0 for which no local scalar λ makes (λu0,λuA) a solution of the KID equations; equivalently, on a non-simply-connected initial surface, exhibit a uKID solution whose associated closed 1-form has nontrivial period, so no global λ exists and no global Killing vector arises.","supporting_citations":[],"review_version":1}