{"id":"b996b166-20bb-4aec-9606-7e38f436ce98","arxiv_id":"2412.05662","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"In a dark-matter-only simulation forecast, combining power spectrum multipoles with Minkowski functionals and tensors tightens constraints on neutrino mass and f(R) gravity parameters by factors of 1.5 to 3.4.","lead":"This paper uses the shapes and orientations of cosmic structure in redshift space to tell apart two competing effects: modified gravity that speeds up structure growth and massive neutrinos that slow it down. The authors show that adding Minkowski morphology statistics to standard power spectrum multipoles improves forecasted constraints on neutrino mass and the f(R) gravity parameter by factors of 1.5 to 3.4.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Headline improvement factors for M_nu and sigma_8 rest on a secant derivative whose noise is shown to be non-convergent at the 10% level; the central claim is CONDITIONAL, not UNCHANGED.","rationale":"The reader's weakest_assumption identifies exactly the concern that I consider most load-bearing: the numerical derivatives and covariance matrix may not be converged, and the specific M_nu derivative (Eq. 4.8, a secant from M_nu=0 to 0.4 eV) plus the Appendix B evidence of >10% fluctuations for M_nu and sigma_8 at Nderi=500 directly undermine the precision of the headline improvement factors. I agree with the CONDITIONAL verdict and with the judgment that this is not fatal to the paper's broad conclusion (complementary non-Gaussian/anisotropic information does help), but it is fatal to quoting 3.3-3.4x factors as point values without error bars or a conservative alternative in the main text. The paper is honest: Appendix B explicitly states the fluctuation, Appendix C concedes a possible slight overestimate, and Appendix D provides a more conservative result. Those internal admissions are evidence of good faith, not fraud; my critique is about the central quantitative claim's robustness. I do not see an independent, stronger concern: the Gaussianity test (Appendix A) is an appropriate check, the k-cut test (Appendix E) shows modest sensitivity, and the use of dark-matter-only forecasts is clearly stated. The single most load-bearing issue remains the noise/non-convergence of the derivatives that feed the Fisher matrix, because the headline numbers are exactly ratios of those Fisher errors. My concrete test targets that issue directly by re-estimating the headline ratios with both a more conservative fR0 derivative and a higher-order M_nu derivative, and by attaching bootstrap error bars to the Nderi=500 values. If that test shows shifts below ~15%, the CONDITIONAL verdict could be upgraded; if it shows larger shifts, the headline should be revised downward to the conservative Appendix D numbers.","tokens_in":36399,"tokens_out":2197,"duration_ms":18252,"concrete_test":"Recompute Tables 2 and 3 for the large-scale configuration (kmax=0.25 h/Mpc, RG=10 Mpc/h) using the more conservative fR0 estimator fR04 from Eq. C.10 and a higher-order M_nu derivative estimator (e.g., a Richardson or 4-point stencil through M_nu=0, 0.1, 0.2, 0.4 eV) with 500 realizations per node, re-running the Figure 15 convergence test at Nderi=500 with 10 independent resamplings to attach error bars to each marginalized sigma. If the ratio sigma(P0,2,4)/sigma(combined) for M_nu or sigma_8 shifts by more than 15% relative to Table 3, the headline improvement factors should be reported with uncertainties or replaced by the conservative Appendix D values.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The paper's headline claim is that combining P0,2,4 with MTs improves power-spectrum constraints on M_nu and sigma_8 by factors of about 3.3-3.4 (Table 3). This claim is load-bearing for the abstract, yet it rests on the Fisher-matrix combination of a derivative estimator that the paper itself shows is not converged. In Eq. 4.8, the M_nu derivative is a two-point secant between M_nu=0 and M_nu=0.4 eV, both with Zel'dovich ICs, chosen specifically because it is the lowest-noise option. Appendix B (Figure 15) shows that for the MTs at large scales (kmax=0.25 h/Mpc, RG=10 Mpc/h), the combined-estimator marginalized constraints on M_nu and sigma_8 still fluctuate by more than 10% even when Nderi approaches 500; the authors state that a large number of simulations would be required for these constraints to converge below the 5% level. The same large-scale panel is exactly the configuration quoted for the headline '2.8/2.2/3.4/3.4/1.5' improvement factors, and the small-scale panel also shows visible residual scatter for M_nu and sigma_8. Because the constraint improvements are ratios of two noisy quantities, a 10% fluctuation in the denominator can plausibly translate into a 15-20% shift in the headline ratio, and if the residual fluctuations indicate a biased derivative rather than pure sampling noise, the quoted 3.3-3.4x factors on M_nu and sigma_8 would be systematically overestimated. The paper's own Appendix C further concedes that the fR0 estimator (fR03a) used in the main text may slightly overestimate the improvement in the power-spectrum constraints, although it argues the effect is not significant. Since the strongest claim is stated as point values without error bars, the central quantitative claim is not yet fully secured.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper forecasts cosmological constraints from redshift-space power spectrum multipoles (P0, P2, P4), 3D scalar Minkowski functionals (MFs), and tensor Minkowski functionals (MTs) using the Quijote and Quijote-MG simulations. The authors measure derivatives and covariance from 500 and 5000 realizations, respectively, use a combined Fisher estimator, and report that adding MFs+MTs improves P0,2,4 constraints on Omega_m, h, sigma_8, M_nu, and |f_R0| by factors of 3.4, 3.0, 3.3, 3.3, and 1.9 on small scales (kmax=0.5 h/Mpc, RG=5 h^-1 Mpc) and 2.8, 2.2, 3.4, 3.4, and 1.5 on large scales (kmax=0.25 h/Mpc, RG=10 h^-1 Mpc). The paper also analyzes the information in perpendicular and parallel MT components and in low- versus high-density thresholds, finding complementary information. The analysis includes extensive convergence tests, Gaussianity checks, and a more conservative forecast in the appendices.","tokens_in":36752,"tokens_out":11169,"duration_ms":88057,"significance":"If the headline improvements are robust, this work establishes Minkowski tensors as a valuable addition to the cosmological probe toolbox for breaking the f(R)-massive-neutrino degeneracy using non-Gaussian and anisotropic information. The analysis is notably careful: covariance from 5000 fiducial simulations with Hartlap debiasing, a combined standard/compressed Fisher estimator to mitigate derivative noise, likelihood Gaussianity checks, and convergence tests with respect to both derivative and covariance noise. The physical interpretation of the perpendicular/parallel MT elements and the low- versus high-density threshold separation is a strength. The main caveat is that the MT constraints on M_nu and sigma_8 are not fully converged at the 10% level, which directly affects the headline improvement factors for those parameters.","major_comments":[{"comment":"The headline improvement factors for M_nu and sigma_8 (3.3–3.4 on both small and large scales) rest on the combined Fisher estimator whose convergence is not established for these parameters. In Figure 15, for the MTs on large scales (kmax=0.25 h/Mpc, RG=10 h^-1 Mpc), sigma_theta(Nderi)/sigma_theta(Nderi=500) fluctuates by more than 10% for M_nu and sigma_8 even when Nderi approaches 500, and the authors state that 'a large number of simulations would be required for these constraints to converge below the 5% level.' Since the power-spectrum constraints are well converged, the quoted ratios are a converged quantity divided by a non-converged one; a 10% fluctuation in the denominator can plausibly shift the headline ratios by 15–20%, and if the fluctuation reflects bias rather than sampling noise, the improvement would be systematically overestimated. The paper should provide more converged forecasts for M_nu and sigma_8 (e.g., with more realizations or a different derivative estimator) before claiming these factors.","section":"Appendix B, Fig. 15, and Table 3"},{"comment":"The M_nu derivative is estimated as a two-point secant between M_nu=0 and M_nu=0.4 eV using Zel'dovich ICs, and the paper argues that IC differences cancel in derivatives. However, this cancellation is only tested for the MFs (Figure 2), not for the MTs, which are exactly the statistics exhibiting the worst non-convergence. The assumption that the MT derivatives are unbiased is load-bearing for the M_nu forecasts. A concrete test would be to compute the M_nu derivative with 2LPT ICs for the fiducial and M_nu=0.4 eV models, or to use a smaller step with more realizations, and compare the resulting Fisher forecasts.","section":"Eq. 4.8 and Section 2"},{"comment":"The abstract reports constraints on five parameters (Omega_m, h, sigma_8, M_nu, f_R0), but the analysis fixes Omega_b and n_s in the main forecast (Appendix B, Figure 11 and surrounding text). The headline improvement factors are therefore not marginalized over the full cosmological parameter set. The paper should either state this restriction explicitly in the abstract or extend the forecast to include Omega_b and n_s, because degeneracies with these parameters could weaken the claimed improvements, as the paper itself notes in Section 6 when comparing with [79].","section":"Abstract and Appendix B"}],"minor_comments":[{"comment":"The transformation from f_R0 to |f_R0|lg2 is not clearly defined; the listed mapping from f_R0 = 0, -5e-7, -5e-6, -5e-5, -5e-4 to 0, 0.0127, 0.0254, 0.0507, 0.101 does not correspond to a standard logarithmic base-10 or base-2 transformation. Please clarify the exact definition.","section":"Section 4.1"},{"comment":"In the captions, the MT data vector is repeatedly written as 'W0 + W^{0,2}_1 + W^{0,2}_1 + W3'; the second term should be W^{0,2}_2.","section":"Captions of Figures 12 and 15"},{"comment":"The statement that Omega_b and n_s are fixed appears only in Appendix B; please reiterate this restriction in Section 5.3 or the abstract so that readers do not misinterpret the marginalized constraints as being over the full seven-parameter space.","section":"Section 5.3"}],"recommendation":"major_revision","confidential_remarks":"The paper is technically solid and unusually transparent about its convergence limitations, but the abstract's improvement factors for M_nu and sigma_8 are stronger than the convergence tests in Appendix B support. A major revision with additional realizations or a recalibrated derivative estimator for the MTs would place the central claim on firmer ground. I agree with the reader's conditional verdict."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This is a careful simulation-based Fisher forecast, and it probably has the right broad conclusion: adding Minkowski functionals and tensors to power spectrum multipoles in redshift space tightens joint constraints on f(R) gravity and neutrino mass. The headline improvement factors are plausible, but the point values on M_nu and sigma_8 are not fully secured by the paper's own convergence tests.\n\nWhat is genuinely new: the first joint use of 3D Minkowski tensors for the f(R)-neutrino degeneracy, and a useful interpretation of how perpendicular and parallel MT components respond to Kaiser versus Fingers-of-God effects on nonlinear scales. The pipeline is exemplary—5000 fiducial simulations for the covariance, Hartlap debiasing, three Fisher estimators, Gaussianity checks, and a serious convergence appendix. The paper is honest about most of its limitations.\n\nThe soft spots are real but not fatal. Appendix B, Figure 15, shows that for the MTs at large scales the combined-estimator marginalized constraints on M_nu and sigma_8 fluctuate by more than 10% even when Nderi approaches 500. That is exactly the configuration quoted for the headline '2.8/2.2/3.4/3.4/1.5' improvements. Because the improvements are ratios of two noisy quantities, a 10% fluctuation in the denominator can shift the headline ratio by 15-20%. The M_nu derivative is a secant between 0 and 0.4 eV using Zeldovich initial conditions; it was chosen for low noise, but that does not rule out bias. Appendix C also concedes that the fR0 derivative estimator used in the main text may slightly overestimate the improvement from combining with the power spectrum. None of this kills the central argument, but it means the abstract's point values overstate the current precision.\n\nOne presentation issue: the abstract says 'MTs' but the data vector is W0 + W^{0,2}_1 + W^{0,2}_2 + W3, which includes the scalar MFs W0 and W3. The paper's own Section 5.2 shows that the tensors add only modest information on top of the MFs; the scalar MFs carry most of the gain over P0,2,4. The 'first application on fully non-linear scales' phrasing in the conclusion is also broader than the more careful claim in the introduction.\n\nWho is this for? Anyone forecasting MG and neutrino constraints with non-Gaussian statistics. It deserves a serious referee. I would send it out, but ask the authors to quote ranges or produce more converged derivatives for the headline numbers, and to make the scalar-MF dominance explicit in the abstract.","headline":"Careful Fisher forecast showing Minkowski statistics tighten MG+neutrino constraints; headline improvements plausible but the M_nu and sigma_8 factors are not fully converged.","tokens_in":37361,"tokens_out":3859,"would_cite":true,"duration_ms":32885,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper argues that the 3D scalar and tensor Minkowski functionals of the redshift-space density field carry non-Gaussian and velocity information that breaks the f(R)-gravity versus neutrino-mass degeneracy, improving Fisher-forecast…","keywords":["large-scale structure","Minkowski functionals","Minkowski tensors","redshift-space distortions","massive neutrinos","modified gravity","f(R) gravity","Fisher forecast"],"falsifier":"Recompute the neutrino-mass derivative with a smaller step or central difference (for example using the $M_\\nu=0.2$ and $0.4$ eV runs against the fiducial) and repeat the combined-estimator Fisher forecast for $P_{0,2,4}+$MFs+MTs; if the claimed $\\sim3.3$-fold improvement on $M_\\nu$ does not survive, the headline result is dominated by derivative noise, which the paper's own Appendix B shows still produces more than 10% fluctuations on the large-scale $M_\\nu$ and $\\sigma_8$ constraints.","tokens_in":36156,"feed_emoji":"🌌","tokens_out":11067,"duration_ms":92109,"temperature":0.7,"pith_summary":"The paper is a forecasting study about whether the morphology of the cosmic web in redshift space can separate the growth-enhancing effects of $f(R)$ modified gravity from the growth-suppressing effects of massive neutrinos, two ingredients that are strongly degenerate in standard statistics like the power spectrum and halo counts. The authors claim that the non-Gaussian information captured by the scalar Minkowski functionals and the anisotropic, velocity-sensitive information captured by the tensor Minkowski functionals break this degeneracy. On simulation-based Fisher forecasts, adding these morphological statistics to the power-spectrum multipoles tightens the constraints on $\\Omega_m$, $h$, $\\sigma_8$, $M_\\nu$, and $f_{R0}$ by factors of 3.4, 3.0, 3.3, 3.3, and 1.9 on small scales, and 2.8, 2.2, 3.4, 3.4, and 1.5 on large scales. If correct, the result matters because it identifies a relatively untapped source of non-Gaussian and velocity information that could help Stage-IV surveys jointly measure neutrino mass and test gravity.","feed_headline":"Minkowski tensors tighten neutrino and gravity constraints by 3.4x","feed_subtitle":"Adding shape and velocity statistics to the power spectrum could break the f(R)-versus-neutrino degeneracy in surveys.","key_machinery":"The load-bearing object is the redshift-space morphological data vector: the four scalar Minkowski functionals (volume, surface area, integrated mean curvature, and Euler characteristic) together with the two translation-invariant rank-2 Minkowski tensors $W^{0,2}_1$ and $W^{0,2}_2$, computed on isodensity surfaces of the density field smoothed with a Gaussian filter of radius $R_G=5$ or $10\\,h^{-1}\\mathrm{Mpc}$. The tensors are split into perpendicular and parallel components relative to the line of sight; these components separately encode the anisotropies produced by the Kaiser effect and the Fingers-of-God effect, which is how velocity information enters. The forecast itself is carried by a combined Fisher estimator, the geometric mean of the standard and compressed estimators, chosen to avoid the noise-tightening bias that simulation-derived derivatives can introduce.","core_discovery":"The central claim is that the 3D scalar Minkowski functionals and rank-2 tensor Minkowski functionals, measured on the redshift-space CDM density field, contain non-Gaussian and anisotropic information that breaks the degeneracy between $f(R)$ modified gravity and massive neutrinos. The authors find distinct imprints of $f_{R0}$ and $M_\\nu$ in these statistics, particularly at low density thresholds, and show that the perpendicular and parallel components of the Minkowski tensors respond differently to the Kaiser and Fingers-of-God regimes of redshift-space distortion. In a combined-estimator Fisher forecast, adding the Minkowski statistics to the monopole, quadrupole, and hexadecapole of the power spectrum improves the marginalized constraints on $\\Omega_m$, $h$, $\\sigma_8$, $M_\\nu$, and $|f_{R0}|_{\\mathrm{lg}2}$ by factors of 3.4, 3.0, 3.3, 3.3, and 1.9 on small scales ($k_{\\max}=0.5\\,h\\,\\mathrm{Mpc}^{-1}$, $R_G=5\\,h^{-1}\\mathrm{Mpc}$) and 2.8, 2.2, 3.4, 3.4, and 1.5 on large scales ($k_{\\max}=0.25\\,h\\,\\mathrm{Mpc}^{-1}$, $R_G=10\\,h^{-1}\\mathrm{Mpc}$). This is presented as the first application of 3D Minkowski tensors on fully nonlinear scales.","pith_inferences":["The improvements reported here are for the dark-matter field; for galaxy catalogs the gains may shrink, since bias and shot noise dilute the non-Gaussian signal and the authors note that Fisher forecasts for biased tracers face convergence problems.","Because the paper treats $R_G=5$ and $10\\,h^{-1}\\mathrm{Mpc}$ separately, a multi-scale combination across both smoothing scales would likely tighten constraints further than either scale alone.","The same machinery should generalize to other modified-gravity models or to higher-rank Minkowski tensors, which the authors suggest could capture more of the anisotropic information; those extensions are untested here.","A practical consequence: the grid-based Minkowski functional measurement is fast enough for survey-scale analyses, so the expensive tensor measurement may be targeted only where its perpendicular-parallel anisotropy pays off."],"forward_implications":["If the forecast is right, the power-spectrum multipole constraints on $\\Omega_m$, $h$, $\\sigma_8$, and $M_\\nu$ improve by factors of roughly 2 to 3.4 when Minkowski functionals and tensors are added, on both quasi-linear and nonlinear scales.","The $|f_{R0}|$--$M_\\nu$ degeneracy, which appears in the power spectrum, halo mass function, and halo bias, is reduced by the morphological statistics; the $f(R)$ parameter itself improves by factors of 1.9 on small scales and 1.5 on large scales.","The perpendicular and parallel elements of the Minkowski tensors carry complementary information, and low-density thresholds (voids) are particularly sensitive to $f_{R0}$ because the environment-dependent screening of the fifth force is weaker there.","The analysis implies that non-Gaussian and velocity information are effectively complementary to two-point statistics, motivating further development of efficient morphological estimators for galaxy surveys; grid-based Minkowski functionals give nearly identical constraints at more than 700 times lower computational cost.","The same statistics could be extended to smaller smoothing scales or to higher-rank Minkowski tensors to capture additional anisotropic information, though the paper notes that the computational cost of tensor measurements is currently limiting."],"supporting_citations":[{"why":"Supplies the fiducial N-body simulation suite, with 5000 realizations for covariance estimation and 500 for derivatives.","marker":"[55]"},{"why":"Supplies the modified-gravity simulation suite used to compute derivatives with respect to $f_{R0}$ and to identify f(R) imprints.","marker":"[56]"},{"why":"Shows that Minkowski functionals of large-scale structure carry constraining power on modified gravity, motivating their use here.","marker":"[39]"},{"why":"Establishes that Minkowski functionals can probe massive neutrinos and complement the power spectrum, the direct precursor of this analysis.","marker":"[48]"},{"why":"Extends Minkowski-functional constraints on neutrino mass to redshift space and supplies the threshold-binning scheme adopted here.","marker":"[49]"},{"why":"Provides the algorithm for measuring 3D Minkowski tensors on isodensity surfaces and quantifies RSD-induced anisotropy.","marker":"[51]"},{"why":"Supplies the analytic ensemble-average predictions for the perpendicular and parallel Minkowski-tensor components in redshift space.","marker":"[52]"},{"why":"Provides the combined Fisher estimator used to reduce the bias from noise in numerical derivatives.","marker":"[78]"}],"fun_headline_variants":["Minkowski tensors break f(R)-neutrino degeneracy in cosmic maps","Shapes and velocities in redshift space sharpen neutrino and gravity bounds","First 3D Minkowski tensors on nonlinear scales improve constraints 3x","Non-Gaussian shapes separate modified gravity from massive neutrinos","Shapes plus power spectrum: neutrino and f(R) constraints tightened up to 3.4x"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole forecast rests on the assumption that the simulation-estimated derivatives and covariance matrix approximate the true likelihood well enough that numerical noise does not tighten the Fisher errors; the neutrino-mass derivative in particular is a secant between $M_\\nu=0$ and $0.4$ eV from simulations with Zel'dovich initial conditions, and the paper's own Appendix B shows more than 10% fluctuations on the large-scale $M_\\nu$ and $\\sigma_8$ constraints even with all 500 derivative simulations.","fun_headline_variants_meta":{"raw":{"variants":["Minkowski tensors break f(R)-neutrino degeneracy in cosmic maps","Shapes and velocities in redshift space sharpen neutrino and gravity bounds","First 3D Minkowski tensors on nonlinear scales improve constraints 3x","Non-Gaussian shapes separate modified gravity from massive neutrinos","Shapes plus power spectrum: neutrino and f(R) constraints tightened up to 3.4x"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000941,"raw_usage":{"total_tokens":4207,"prompt_tokens":1319,"completion_tokens":2888,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":935,"completion_tokens_details":{"reasoning_tokens":2800}},"tokens_in":935,"tokens_out":2888,"duration_ms":17717,"temperature":1.0,"reasoning_tokens":2800,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T20:29:45.587569+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute the neutrino-mass derivative with a smaller step or central difference (for example using the $M_\\nu=0.2$ and $0.4$ eV runs against the fiducial) and repeat the combined-estimator Fisher forecast for $P_{0,2,4}+$MFs+MTs; if the claimed $\\sim3.3$-fold improvement on $M_\\nu$ does not survive, the headline result is dominated by derivative noise, which the paper's own Appendix B shows still produces more than 10% fluctuations on the large-scale $M_\\nu$ and $\\sigma_8$ constraints.","supporting_citations":[],"review_version":1}