{"id":"2a7272c5-328d-4461-bc01-fb92666081c4","arxiv_id":"2608.07653","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Celephais constructs spectrally accurate binary-neutron-star and black-hole-neutron-star initial data with arbitrary spin orientations, using a sparse Jacobian, adaptive hp-refinement, and PN-informed eccentricity reduction.","lead":"This paper introduces Celephais, a spectral code that builds initial data for merging neutron-star and black-hole-neutron-star binaries with tilted spins, using sparse-matrix tricks and adaptive resolution to cut cost. A generalist might read it because efficient initial data is a bottleneck for large gravitational-wave simulation surveys.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Approximate Bernoulli relation for tilted spins is the load-bearing approximation; its neglected spin-curl term is never quantified.","rationale":"The reader identified the same weakest assumption, and I agree it is the most load-bearing point. The paper's novelty is the construction of initial data with arbitrary spin orientations; the hydrostatic-equilibrium model is where the new physical regime enters. The numerical machinery (sparse Jacobian assembly, MUMPS-preconditioned Newton-Krylov, parity-sector preconditioning, adaptive hp-refinement) is described in detail and benchmarked on realistic configurations, so the main risk is not the linear algebra but the physical modeling of the spinning fluid. Equation (21) is honestly labeled approximate, but honesty about an unquantified approximation does not establish its validity. The paper's validations, including the FUKA compatibility test, PN binding-energy comparisons, eccentricity reduction, and the single precessing-BNS evolution, do not specifically constrain the error from the dropped spin-curl term: a slightly out-of-equilibrium star can still produce a stable inspiral and a PN-like precession axis. The proposed concrete check, evaluating the full Euler residual on the constructed solution, directly measures whether the approximation is justified for tilted spins. Since the verdict was already CONDITIONAL and this concern does not by itself require a stronger verdict before the check is performed, I recommend leaving the verdict unchanged.","tokens_in":22485,"tokens_out":9876,"duration_ms":100092,"concrete_test":"On the converged precessing-BNS solution used in Table I/Fig. 10, compute inside each stellar domain the L2 norm of the neglected spin-curl term V^j(D_j \\hat u_i - D_i \\hat u_j) and compare it with the L2 norm of the retained Bernoulli gradient D_i(h\\alpha/W + \\hat u_j V^j) from Eq. (17). Report this ratio for a tilt sweep (e.g., 0, 30, 60, and 81 degrees). If the ratio exceeds a few percent anywhere in the star, the approximate equilibrium is biased and the central claim needs qualification; if it is below 1 percent, the reader's concern is resolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that Celephais constructs spectrally resolved initial data for arbitrary spin orientations. For spinning neutron stars, the fluid velocity is built from an irrotational potential plus a rigid-rotation term (Eq. 20), and hydrostatic equilibrium is closed with the approximate Bernoulli relation (Eq. 21), explicitly dropping the spin-curl term V^j(D_j \\hat u_i - D_i \\hat u_j) in Eq. (17). For a spin tilted 81 degrees from the orbital axis (Table I), this term need not be small: the spin vector and the orbital velocity field are not aligned, so the antisymmetrized derivative of \\hat u_i can contribute to the enthalpy balance. The paper never estimates the size of the neglected term, nor does it report a stellar-oscillation or enthalpy-residual diagnostic that would reveal an equilibrium bias. The single evolution and PN precession-axis comparison do not isolate this error, since a mildly non-equilibrium star can still inspiral and precess approximately as expected. If the neglected term is non-negligible, the 'arbitrary spin orientation' claim is not established for the very configuration meant to demonstrate it.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents Celephais, a Kadath-based spectral initial-data code for binary neutron star (BNS) and black hole-neutron star (BHNS) systems with generic spin orientations. The key numerical innovations are a dependency-filtered, batched sparse Jacobian assembler; a MUMPS factorization reused as a refreshed right preconditioner in a Jacobian-free Newton-Krylov solve; and an anisotropic hp-refinement scheme driven by spectral tails. The validation suite includes compatibility with FUKA data, binding-energy comparisons with post-Newtonian sequences, a precessing BNS eccentricity-reduction test, and a full evolution of the same system from which the precession axis is reconstructed and compared with a 3.5PN simple-precession model. The main claims are that the code solves a representative precessing BNS on a laptop in minutes, that adaptive refinement recovers uniform-grid constraint accuracy with about three times fewer unknowns for a mass-ratio-20 BHNS, and that the code delivers spectrally resolved BNS/BHNS initial data with arbitrary spin orientations.","tokens_in":1967,"tokens_out":1984,"duration_ms":84846,"significance":"If the claims hold, Celephais would be a valuable public tool: it extends the Kadath/FUKA ecosystem toward generic spin orientations, addresses the memory bottleneck of dense spectral Jacobians, and provides a concrete AMR strategy for high-mass-ratio binaries. The paper is commendably concrete about its numerical machinery and about several of its own limitations, and it makes good use of external checks rather than relying on internal self-consistency alone. The efficiency benchmarks and the sparse-preconditioning strategy are plausible and well described. However, the headline claim of arbitrary spin orientations rests on an explicitly approximate treatment of spinning neutron-star hydrostatics whose error is never quantified, and the tilted-spin validation is limited to a single BNS configuration. These issues are load-bearing for the paper's central claim and require additional quantitative evidence.","major_comments":[{"comment":"The approximate Bernoulli relation (21) drops the spin-curl term V^j(D_j \\hat u_i - D_i \\hat u_j) from Eq. (17). For the tilted-spin BNS of Table I (81 degrees spin-orbit tilt, chi=0.3), this term does not vanish by construction, yet the manuscript never estimates its magnitude or reports a stellar enthalpy-residual diagnostic. Because hydrostatic equilibrium of the spinning star is the physical basis for the claim of arbitrary spin orientations, please provide a quantitative bound on the neglected term for the configurations presented, e.g. by evaluating the full enthalpy balance (17) on the solved data or by comparing against a solve that retains the spin-curl term. The dynamical checks in Secs. V B-V D are not sensitive enough to exclude a small equilibrium bias, since a mildly non-equilibrium star can still inspiral and precess approximately as expected.","section":"Sec. II A, Eqs. (17)-(21)"},{"comment":"The binding-energy comparison with post-Newtonian sequences is presented as a validation of quasiequilibrium quality, but the vertical bars in Fig. 7 are the virial discrepancy (36), not statistical or systematic uncertainties in E_b. The numerical sequences are said to follow the PN trends, yet no resolution study of E_b or an error budget is given. To make this check quantitative, please report E_b as a function of resolution or provide an independent uncertainty estimate; with the present presentation, the agreement cannot be distinguished from a curve-fit effect.","section":"Sec. V A, Fig. 7"},{"comment":"The adaptive-refinement gain stated as 'about three times fewer unknowns' is measured exclusively by the Hamiltonian-constraint norm ||H||_2. The XCTS system is a coupled set of elliptic equations, and the momentum constraint and the stellar-surface conditions are equally part of the solved problem. Reporting only the Hamiltonian norm may overstate the efficiency gain. Please report at least the momentum-constraint norm or the full nonlinear residual for the same refinement sequences.","section":"Sec. IV A, Figs. 5 and 6"},{"comment":"No tilted-spin BHNS test is presented. The BHNS examples in Secs. IV and V A are either nonspinning or have aligned black-hole spin, while the only precessing configuration is the BNS of Table I and Secs. V B-V D. Since the abstract and conclusion claim arbitrary spin orientations for both BNS and BHNS initial data, the BHNS part of that claim is currently unvalidated. Please add at least one misaligned-spin BHNS case, or explicitly restrict the claim of arbitrary spin orientations to BNS systems.","section":"Sec. V"},{"comment":"The full evolution is performed at a single resolution, and the precession-axis comparison is conditional on the measured (2,2) frequency input to the PN model (Eqs. (59)-(65)). The manuscript explicitly acknowledges that waveform-phase convergence is not established, and that the simple-precession model is only valid in the slow-motion regime. This is acceptable as a usability check, but it means the comparison cannot isolate a possible bias in the initial stellar model. A second resolution and an early-time diagnostic of the fluid state (e.g. the enthalpy residual or the spin-curl term of Eq. (17)) would materially strengthen the claim that the generic-spin initial data are true quasiequilibrium data.","section":"Secs. V C and V D"}],"minor_comments":[{"comment":"The first paragraph contains typos: 'estalishing' should be 'establishing' and 'machienry' should be 'machinery'.","section":"Sec. III A"},{"comment":"The sentence 'Therefore, we intented to use one fourth of np' contains a typo: 'intented' should be 'intended'.","section":"Sec. III C"},{"comment":"The sentence 'the black-hole is treated with an excision horizon with' is grammatically incomplete; it appears to be missing a word or clause.","section":"Sec. VI"},{"comment":"The caption contains a garbled phrase 'coe/cientenergy'; it should likely read 'coefficient energy'.","section":"Fig. 5 caption"},{"comment":"The legend text 'orbital-axisdirection(O'ShaughnessyhLaLbi)' is garbled and should refer cleanly to the O'Shaughnessy et al. radiation-axis construction.","section":"Fig. 10 legend"},{"comment":"The text uses 'Macbook Pro' in one place and 'MacBook Pro' in Table I; please standardize the spelling.","section":"Sec. II D / Table I"},{"comment":"The precessing eccentricity estimate uses an aligned-spin approximation for \\dot{a}, which the text acknowledges as a 'principal limitation'. Consider reporting how much this approximation affects the initial eccentricity for a moderately misaligned case, or at least note that the measured eccentricity below 10^-2 already includes this approximation.","section":"Sec. V B, Eq. (51)"}],"recommendation":"major_revision","confidential_remarks":"This is a solid code paper that fits the scope of the journal. The main risk is not the numerical machinery but the unquantified physical approximation in Eqs. (20)-(21) for tilted neutron-star spins, combined with the absence of any tilted-spin BHNS test. I would be comfortable with acceptance after the authors provide a quantitative estimate of the neglected spin-curl term (or an equivalent diagnostic) and add the missing BHNS validation. The code availability statement is important; the paper should make clear whether the repository will be public before or only after acceptance."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThe result is real: this is a Kadath-based XCTS initial-data solver that adds a structural dependency filter, batched forward-mode automatic differentiation for the sparse Jacobian, a refreshed MUMPS factorization as a right preconditioner for Newton-Krylov, anisotropic hp-refinement driven by spectral-tail indicators, and a PN eccentricity estimator generalized to arbitrary spin orientations. The q=20 BHNS benchmark reaches the uniform-grid constraint accuracy with about three times fewer unknowns, and the laptop-scale timings (under 17 GB on an M4 Max) are a genuine practical step forward for survey campaigns.\n\nThe validation is honest and proportionate. FUKA compatibility is a good sanity check, the binding-energy curves track PN, the eccentricity-reduction example works, and the single precessing BNS evolution to merger shows the data are usable. The paper explicitly lists its own limitations: no phase-convergence study, a single eccentricity-reduction demonstration, and an aligned-spin projection for the PN radial-infall estimate.\n\nThe main soft spot is the approximate Bernoulli relation for tilted NS spins. Eq. (21) drops the spin-curl term from Eq. (17), and there is no estimate of how large that term is for the 81-degree configuration that showcases the headline capability. A mildly non-equilibrium star could still inspiral and precess roughly as expected, so the evolution test does not isolate this. This is a known modeling approximation from previous generic-spin BNS work, not a numerical error, but because the paper's central claim is arbitrary spin orientations, the size of the neglected term should be quantified, ideally with a dedicated diagnostic (e.g., enthalpy residual or stellar mode check). The GPL code release, promised upon acceptance, would let a referee check this directly.\n\nThe remaining gaps are in proportion: single-resolution evolution, no statistical error bars on the PN comparison, and a one-example eccentricity test are all acknowledged. None of them undermines the core algorithmic contribution. The citation pattern is appropriate and the paper engages with the relevant methods literature.\n\nRecommendation: send it to peer review. A referee who works with FUKA, SGRID, or SpECTRE will get value from it. The revision should quantify the neglected spin-curl term and add a convergence check on the evolution, even if the full phase-error study remains deferred.","headline":"A genuinely useful spectral-initial-data methods paper whose main caveat is an unquantified spin-curl approximation in the tilted-neutron-star hydrostatic equilibrium.","tokens_in":23198,"tokens_out":3739,"would_cite":true,"duration_ms":36414,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["83-08","65N35","65F50","83C05"],"pacs":["04.25.Dm","04.30.-w","04.40.Dg"],"model":"deepseek-v4-flash","headline":"Celephais constructs spectrally accurate compact-binary initial data with arbitrary spin orientations, using a sparse, adaptively refined spectral solve that cuts unknowns by about three in high-mass-ratio benchmarks.","keywords":["spectral methods","initial data","compact binaries","precessing spins","neutron stars","black hole-neutron star binaries","adaptive hp-refinement","sparse Jacobian"],"falsifier":"Recompute one of the precessing binary-neutron-star configurations keeping the spin-curl term in the fluid Euler equation instead of dropping it, and compare the enthalpy or rest-mass-density distribution plus the measured eccentricity and waveform phase against the current data; a visible drift or a systematic tilt-dependent offset in the stellar profile would falsify the claim that arbitrary spin orientations are handled to the stated accuracy.","tokens_in":22305,"feed_emoji":"🌌","tokens_out":9262,"duration_ms":82788,"temperature":0.7,"pith_summary":"The paper introduces Celephais, a spectral initial-data code that solves the Einstein constraint equations for binary neutron stars and black-hole–neutron-star binaries without assuming the spins are aligned with the orbital axis. The aim is to make spectrally accurate, precessing compact-binary initial data cheap enough to produce in bulk for numerical-relativity surveys. To do so it assembles only the nonzero entries of the global multi-domain Jacobian, factors the sparse matrix once, and reuses that factorisation as a preconditioner in a matrix-free Newton–Krylov solve, then adaptively refines only the spectral directions that are underresolved. For a mass-ratio-20 black-hole–neutron-star benchmark, the adaptive scheme reaches the uniform-grid constraint accuracy with about three times fewer unknowns. The paper validates the data by binding-energy comparisons with post-Newtonian theory, an eccentricity-reduction test, and a full evolution whose waveform-reconstructed precession axis follows a post-Newtonian simple-precession model.","feed_headline":"Spectral solver cuts unknowns threefold for unequal-mass binaries","feed_subtitle":"A new code builds neutron-star and black-hole–neutron-star initial data with tilted spins, tested through merger.","key_machinery":"The load-bearing object is the sparse, globally coupled multi-domain Jacobian of the extended conformal thin-sandwich residual system. The code builds it column-wise by scanning each residual's symbolic expression through a dependency graph, evaluating only residual rows whose supports contain the varied field–domain pair, and propagates several compatible Jacobian columns per forward-mode automatic-differentiation traversal. The resulting matrix is factored once by a sparse direct solver and reused as a right preconditioner in a matrix-free Newton–Krylov solve, with the factors refreshed periodically or when residual growth demands it; when the binary admits a reflection symmetry about the orbital plane with spins in that plane, the Jacobian block-diagonalises into even and odd parity sectors that are factored separately. Adaptive hp-refinement then inspects spectral-tail ratios in each domain and coordinate direction, marking pairs whose tail ratios exceed tolerances and choosing between radial h moves and anisotropic p moves. The final piece is the spin-aware post-Newtonian force balance: the orbital frequency is solved from the instantaneous radial balance of the nonspinning circular baseline plus the leading spin–orbit and spin–spin accelerations treated as vectors, with the nonradial part left to precessional dynamics.","core_discovery":"The central claim is that the cost barrier to spectral compact-binary initial data with generic spin orientations can be broken by explicitly exploiting the sparsity of the Newton Jacobian rather than treating it as dense. The paper shows that a structurally filtered, batched automatic-differentiation assembly yields a sparse matrix with only a few percent nonzero entries for a representative precessing binary neutron star; factoring that matrix once with a sparse direct solver and reusing the factors as a refreshed right preconditioner keeps the memory bounded while the matrix-free Newton–Krylov iteration retains the full Jacobian action at every step. On top of this, an hp-refinement scheme driven by spectral-tail ratios concentrates resolution in the domains and directions that limit convergence, recovering the uniform-grid constraint accuracy with about three times fewer unknowns for a mass-ratio-20 black-hole–neutron-star system and about 1.6 times fewer for mass ratio 10. The author further extends the post-Newtonian force-balance condition to arbitrary spin orientations and shows that, for a precessing binary neutron star with one star tilted at 81 degrees, the post-Newtonian-initialised data have eccentricity below $10^{-2}$ and one evolution-based correction removes most of the orbital-frequency oscillation; the evolved waveform's reconstructed radiation axis tracks the 3.5PN simple-precession model through most of the inspiral.","pith_inferences":["If the approximate Bernoulli relation is the binding limitation, then retaining the spin-curl term in the fluid Euler equation would let the same machinery claim true quasiequilibrium for arbitrary spin tilts; the paper's own diagnostics (virial error, eccentricity, waveform phase) provide the template for testing that upgrade.","The sparse-Jacobian-plus-refreshed-preconditioner strategy is not tied to binary-neutron-star physics; the same dependency-filtered batched differentiation and parity-sector splitting could lower memory for other global spectral elliptic solvers in numerical relativity, such as puncture black-hole initial data or scalar-tensor field equations.","The aligned-spin approximation in the radial-infall estimate is the weakest part of the eccentricity-control pipeline for strongly precessing systems; a testable extension is a fully vectorial 3.5PN infall-rate estimate with the spin projection evolved along the precession cone.","The spectral-tail ratios used for adaptive refinement could double as an automated resolution-selection criterion in production waveform campaigns, since the paper's results suggest the final accuracy is largely independent of the refinement path once the limiting direction reaches sufficient order."],"forward_implications":["Precessing binary-neutron-star and black-hole–neutron-star initial data can be produced spectrally at laptop scale: for the production precessing binary-neutron-star case the first Jacobian assembly takes under 10 minutes and the full job uses under 17 GB of memory.","High-mass-ratio black-hole–neutron-star data can be brought to uniform-grid constraint accuracy with about three times fewer unknowns, and mass-ratio-10 data with about 1.6 times fewer, easing the cost of large parameter surveys.","Post-Newtonian force-balance initialisation gives precessing binary-neutron-star eccentricity below 10^-2 before correction and a usable starting point after one evolution-based iteration.","A full evolution of the precessing binary-neutron-star system is stable through merger, and the waveform-reconstructed precession axis follows the 3.5PN simple-precession model through most of the inspiral, linking prescribed spin tilts to observed dynamics.","The code is planned for public release under the GNU General Public License, which would let other groups reproduce the configurations and build extensions such as puncture black-hole data or alternative theories of gravity."],"supporting_citations":[{"why":"Supplies the multi-domain spectral library on which Celephais is built.","marker":"[9]"},{"why":"Provides the aligned-spin solver whose notation, construction, and compatibility test are extended to arbitrary spin orientations.","marker":"[10]"},{"why":"Provides the constant-rotational-velocity construction for generic neutron-star spin that Celephais adopts.","marker":"[11]"},{"why":"Describes a pseudospectral initial-data code for generic spins whose methods are compared and extended.","marker":"[13]"},{"why":"Supports the quasi-local spin construction and binary-neutron-star initial data with arbitrary spins.","marker":"[15]"},{"why":"Supplies the adjustable-eccentricity binary-neutron-star procedure used for evolution-based eccentricity correction.","marker":"[32]"},{"why":"Provides the post-Newtonian spin–orbit and spin–spin accelerations used in the spin-aware force-balance estimate.","marker":"[73]"},{"why":"Gives the 3.5PN precession frequency used in the simple-precession model for the waveform-reconstructed precession axis.","marker":"[74]"},{"why":"Gives the principal-radiation-axis method used to reconstruct the precession axis from waveform multipoles.","marker":"[90]"}],"fun_headline_variants":["Sparse spectral method tames generic-spin binaries","Spectral initial data with threefold savings for tilted spins","Faster spectral setup for unequal-mass precessing binaries","Sparse Jacobian enables cheap spectral binary data"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The neutron-star equilibrium in tilted-spin configurations is computed with an approximate Bernoulli relation that neglects the spin-curl term in the Euler equation; if spin-driven fluid currents significantly change the enthalpy balance, the stellar models are not truly quasiequilibrium.","fun_headline_variants_meta":{"raw":{"variants":["Sparse spectral method tames generic-spin binaries","Spectral initial data with threefold savings for tilted spins","Faster spectral setup for unequal-mass precessing binaries","Sparse Jacobian enables cheap spectral binary data"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000492,"raw_usage":{"total_tokens":2486,"prompt_tokens":1084,"completion_tokens":1402,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":700,"completion_tokens_details":{"reasoning_tokens":1339}},"tokens_in":700,"tokens_out":1402,"duration_ms":11231,"temperature":1.0,"reasoning_tokens":1339,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T00:26:39.816007+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute one of the precessing binary-neutron-star configurations keeping the spin-curl term in the fluid Euler equation instead of dropping it, and compare the enthalpy or rest-mass-density distribution plus the measured eccentricity and waveform phase against the current data; a visible drift or a systematic tilt-dependent offset in the stellar profile would falsify the claim that arbitrary spin orientations are handled to the stated accuracy.","supporting_citations":[],"review_version":1}