{"id":"0c37e87b-05e8-4c4a-afeb-eb34a5bc82fc","arxiv_id":"1908.03596","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Using numerical relativity thresholds, cosmological perturbation theory, and peak theory, the authors derive updated upper limits on the small-scale primordial curvature power spectrum from primordial black hole abundance, roughly an order of magnitude tighter than earlier constraints.","lead":"The paper computes new upper limits on the size of early-universe density ripples, derived from how many primordial black holes could exist. It finds the limits are about ten times stronger than earlier estimates, and that even one such black hole would require a dramatic departure from the nearly flat ripple spectrum seen by Planck.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Concern: the claimed one-order-of-magnitude tightening rests on the crude TNL/TLIN estimate in Section 5.1, which directly multiplies the reconstructed P_zeta in Equation 6.1 and is not given an error bar.","rationale":"The reader's verdict (CONDITIONAL) is appropriate, and the reader correctly flags the Gaussian peak-theory and non-linear transfer-function approximations. I do not find a fully fatal objection, because the qualitative single-PBH scale-invariant conclusion in Fig. 7 is robust to the T_NL/T_LIN uncertainty, and the paper is transparent about its approximations. However, I disagree slightly with the emphasis: the most load-bearing quantitative concern is the non-linear transfer function ratio, because it enters directly and quadratically in the central Equation 6.1 and thus in the claimed factor-of-~10 improvement in Fig. 6. The Gaussian peak-theory assumption mainly affects the variance reconstruction via Eq. 6.4, and the authors show that Press-Schechter overestimates the amplitude by 20-70%, which is comparable but is at least partially diagnosed internally. By contrast, the T_NL/T_LIN ratio is estimated by a single crude numerical comparison and no error bar or sensitivity test is reported. A rigorous re-derivation or a robustness test of that ratio would either confirm or undermine the main quantitative claim. I therefore keep the verdict at CONDITIONAL rather than ACCEPT, and the concrete test above would settle the concern.","tokens_in":47456,"tokens_out":1605,"duration_ms":16365,"concrete_test":"Recompute the Fig. 6 upper limits using the linear transfer function T_LIN (i.e., set T_NL/T_LIN = 1) and using the upper and lower ends of the quoted T_NL/T_LIN range, keeping every other step identical. If the resulting difference in P_zeta(kt) is above ~50% of the separation between the new and previous constraints, the claimed order-of-magnitude tightening is not robust to the transfer-function uncertainty; if the difference is smaller than ~20%, the headline quantitative claim is stable. As a complement, run a second-order perturbation-theory computation of T_NL for the same profile family and compare its ratio to Equation 5.6.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central quantitative claim, that the new pipeline yields constraints roughly an order of magnitude tighter than previous estimates (Fig. 6 and Section 7 point 5), is directly proportional to the non-linear transfer function ratio introduced in Section 5.1. In Equation 6.1 the reconstructed P_zeta scales as 1/TNL^2, and the paper estimates TNL/TLIN ~ 1.2-1.8 by comparing numerical and linear density profiles (Eq. 5.6). The authors call this 'crude' and defer the full calculation, yet no uncertainty or robustness test is attached to it. Since TNL appears squared in the key amplitude equation, even a 30% error in TNL/TLIN changes the headline upper limits by roughly a factor 2, which is a substantial fraction of the claimed order-of-magnitude improvement. The same sensitivity applies to the 'one single PBH' conclusion for a scale-invariant spectrum in Fig. 7, where the amplitude gap is several orders of magnitude and therefore robust; the load-bearing fragility is specific to the tighter quantitative bound in Fig. 6. The authors state in Section 5.1 that pressure effects are inefficient and assume nu' ~ nu, but this does not bound the uncertainty in the numerical ratio itself. A concrete propagation of this uncertainty, or a direct second-order perturbative computation of TNL, would settle whether the claimed tightening survives.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"Using numerical-relativity collapse thresholds, a physically motivated smoothing prescription, and Gaussian peak theory, this paper constructs an end-to-end pipeline that maps observational upper limits on the primordial black hole abundance f_PBH into upper limits on the primordial curvature power spectrum P_zeta(k) over k ~ 10^5 - 10^15 Mpc^-1, and demonstrates how to invert the mapping from a spectrum to the abundance. The central quantitative claims are (i) that the maximal allowed amplitude of a spike-type P_zeta is about one order of magnitude tighter than previous estimates (Fig. 6 and Section 7, point 5), and (ii) that even a single PBH in the observable universe is incompatible with a scale-invariant spectrum (Fig. 7). The paper also reconstructs the shape of the spike on scales k_t to 5 k_t (Fig. 8), provides a step-by-step recipe for using the reconstructed shape to determine the profile parameter alpha and the abundance (Section 6, items 1-6), and derives carefully the relativistic degree-of-freedom conversion between beta and f_PBH (Appendix B).","tokens_in":47664,"tokens_out":20470,"duration_ms":204806,"significance":"If the central quantitative claim survives scrutiny, this is a genuinely useful methodological step forward: it is the first PBH-to-spectrum mapping that combines simulation-calibrated shape-dependent thresholds, a physical filtering prescription, and peak theory in one self-consistent pipeline. The paper is transparent and unusually complete in its exposition: the reconstruction steps are enumerated (Section 6, items 1-6), the g* factors are derived rather than approximated (Appendix B), and the Section 4.3 treatment of how non-linearities mix all n-point functions of zeta into the density power spectrum is a valuable formal result. The qualitative conclusions are robust and falsifiable: the incompatibility of one single CPBH with a scale-invariant spectrum, the sensitivity of the bounds to the modelling of collapse, and the predicted SKA/LISA gravitational-wave signatures.","major_comments":[{"comment":"The headline claim of Section 7 (point 5) and Figure 6 — that the new pipeline gives constraints about one order of magnitude tighter than previous estimates — is controlled by the numerical factors T_NL/T_LIN ~ 1.2-1.8 quoted in Section 5.1, because the reconstructed amplitude scales as P_zeta proportional to 1/T_NL^2 in Eq. (6.1). The paper explicitly calls this estimate 'crude' and defers the full calculation to Ref. [188], but no uncertainty or robustness test is attached to the quoted factors, and the estimator in Eq. (5.6) is not uniquely defined (the radial range over which the profiles are averaged, the weighting, and the normalization convention are not specified). Since T_NL enters squared, a 30% uncertainty in T_NL/T_LIN changes the reconstructed P_zeta by roughly a factor of two, which is a substantial fraction of the claimed order-of-magnitude improvement; the red band in Fig. 6 therefore does not yet carry the precision implied by the abstract and the conclusions. The companion assumption nu' ~ nu in Section 5.1 is an uncontrolled approximation of similar size. I stress that this concern is specific to the quantitative bound in Fig. 6: the 'one single PBH' conclusion of Fig. 7 rests on a gap of several orders of magnitude and is robust. A concrete remedy is to propagate an explicit uncertainty band around T_NL (or to show the conservative case T_NL = T_LIN) in Fig. 6 and to rescale the 'order of magnitude' claim accordingly.","section":"Sec. 5.1, Eq. (5.6); Sec. 6, Eq. (6.1) and Fig. 6"},{"comment":"The abundance mapping that fixes sigma_0 in Eq. (6.4) assumes that the smoothed overdensity field is Gaussian for the purposes of peak theory (Section 5, first paragraphs), and Section 6 states that the reconstruction is performed 'assuming Gaussian initial conditions'. However, Section 4.3 shows that the non-linear relation (3.7) makes the smoothed density field non-Gaussian even when zeta is Gaussian: Eqs. (4.13)-(4.14) give a non-zero one-point function and two-point contributions from all higher n-point functions of zeta, including the <zeta zeta zeta zeta> terms that survive at Gaussian zeta. The Gaussian assumption is admitted in the text as 'too strict', and the filtering of the density field in the presence of curvature is likewise acknowledged to be approximate (Section 4.2, Eqs. 4.8-4.10), but no bounds are placed on how these approximations shift the reconstructed sigma_0 and hence P_zeta. This is consequential because the inference of sigma_0 from f_PBH runs through the exponential tail e^{-nu^2/2} of the peak-height distribution, which is precisely the part of the distribution modified by non-Gaussianity. The manuscript cites contemporaneous work on the non-linear density-curvature relation and PBH abundance (Refs. [160-162, 179]) but does not import any of these results into the quantitative constraints. I recommend either a quantitative evaluation of the leading non-Gaussian correction to the peak-height distribution (e.g., the one-loop terms in Eq. 4.14 at Gaussian zeta), or an explicit caveat in the abstract and conclusions that the tightened bound of Fig. 6 applies only under the Gaussian smoothed-field assumption.","section":"Sec. 4.3, Eqs. (4.13)-(4.15); Sec. 5 (first paragraphs); Sec. 6, Eq. (6.4)"}],"minor_comments":[{"comment":"The sentence introducing the filter functions says equation (4.14) is multiplied by W'_s(k1)W'_s(k1); the second factor should presumably read W'_s(k2), since the two-point function depends on two wavevectors.","section":"Sec. 4.3, text after Eq. (4.14)"},{"comment":"The claim that the Gaussian and mixed higher-order terms are 'highly suppressed' because they contain four transfer functions is not compelling in the regime used later, where the transfer function is evaluated at horizon crossing with T_LIN(k tau ~ 1) ~ 0.9 (Section 5.1); the transfer-function counting does not by itself bound the loop integrals. A sentence with a dimensional estimate, or an explicit statement that these terms are dropped in the Gaussian limit, would avoid overstating the case.","section":"Sec. 4.3, discussion after Eq. (4.15)"},{"comment":"The assertion that different choices of (delta_I - delta_I,c) 'do not have any impact on the constraint itself' is stronger than what the preceding discussion shows: Eq. (6.5) fixes the mass-scale calibration, and a factor-3 shift in M_PBH for a given k_t changes which f_PBH upper limit is applied at that scale, which can shift the vertical position of the bound in regions where f_max varies steeply. Please state the resulting calibration uncertainty explicitly.","section":"Sec. 6, paragraph after Eq. (6.4)"},{"comment":"'We prove that none of these quantities is accurately computed using linear theory' overstates the demonstration, which is performed for the one-parameter family of profiles of Eq. (3.5); 'show' would be more accurate than 'prove'.","section":"Sec. 2, Executive Summary; Sec. 3.3"},{"comment":"References [188] and [211] are listed as 'arXiv:ToAppear' without identifiers; since the argument of Section 5.1 explicitly defers the non-linear transfer function to [188], these placeholders should be completed or removed before publication.","section":"References [188], [211]"}],"recommendation":"major_revision","confidential_remarks":"The paper is competent and likely to be influential, but the contemporaneous literature (Refs. [160-162, 179]) already covers parts of the non-linear density-curvature abundance mapping, and the authors cite these works without incorporating them into the quantitative constraints. I would advise the editor that the 'one order of magnitude' tightening claim, which is the most citable result, needs to be made conditional on the T_NL and Gaussianity assumptions or accompanied by a robustness band; otherwise the paper risks being cited for a crisp number that the internal analysis does not fully support. The qualitative results (Fig. 7 and the modelling-sensitivity conclusions) are solid and should not be delayed."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Let me give you the short version first. This is the most complete PBH-abundance-to-curvature-spectrum pipeline I have seen: numerical relativity thresholds that depend on profile shape, a physically motivated smoothing prescription, a non-linear density-curvature relation, and peak theory instead of Press-Schechter. The qualitative headline—one single CPBH would be incompatible with a scale-invariant spectrum extrapolated from CMB scales—is robust, because the gap is orders of magnitude. The quantitative claim, that the updated upper limits on P_zeta are about an order of magnitude tighter than previous estimates, is plausible but not yet nailed down.\n\nThe genuinely new pieces are the combined pipeline and the shape reconstruction for k > k_t. Section 4.3's exact expansion of the non-linear overdensity field is a useful contribution in its own right, and the paper is transparent about where shortcuts remain. It explicitly calls the Gaussian peak-theory assumption \"too strict\" and gives the T_NL/T_LIN estimate as crude.\n\nThe soft spots. The stress-test note is right: T_NL/T_LIN enters squared in Eq. (6.1), and the estimate from Eq. (5.6) is an average profile comparison with no error bar. A 30% uncertainty there shifts the headline amplitude by about a factor of two, which is a large fraction of the claimed order-of-magnitude improvement. It is not fatal for the qualitative single-PBH statement, but it means the tight quantitative bounds in Fig. 6 are conditional on that number. Second, peak theory is applied to a field the authors themselves show is non-Gaussian; no quantification of the bias on the abundance tail is given. Third, the monochromatic, single-shape population is an admitted simplification; it is standard in the field, but again unquantified. These are not reasons to dismiss the paper—they are reasons to ask for error bands before calling the tightening definitive.\n\nWho is this for? Anyone working on PBH constraints, inflationary small-scale power spectrum, or forecasting LISA/SKA synergies will read it. I would cite it for the methodology and for the qualitative exclusion. It deserves a serious referee, not a desk rejection; I'd send it out with a request for a sensitivity analysis on T_NL/T_LIN and at least a parametric estimate of non-Gaussian corrections to the peak height distribution. That is a major revision, not a fundamental flaw.","headline":"The most careful PBH-to-spectrum pipeline to date, with a robust qualitative exclusion of scale-invariant spectra, but the headline order-of-magnitude tightening depends on a crude transfer-function estimate that needs an error bar before it is definitive.","tokens_in":48264,"tokens_out":2095,"would_cite":true,"duration_ms":21838,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Primordial black hole abundance, translated through non-linear collapse physics and peak theory, caps the small-scale curvature power spectrum about an order of magnitude more tightly than previously thought, and a single such black hole…","keywords":["primordial black holes","primordial curvature power spectrum","peak theory","numerical relativity","gravitational collapse threshold","small-scale perturbations","dark matter","non-Gaussianity"],"falsifier":"Re-derive the PBH abundance at fixed $\\mathcal{P}_\\zeta$ using the same non-linear $\\zeta\\to\\delta$ map but with a non-Gaussian peak distribution, or with numerical simulations that include the full mode coupling, and compare the resulting $f_{\\rm PBH}$ to the Gaussian peak-theory prediction; a difference of more than about an order of magnitude at fixed amplitude would shift the central constraint by more than the claimed tightening. A direct check is also possible observationally: a future measurement of the stochastic gravitational-wave background from the same scalar perturbations on the scales of Figure 6 would confirm or rule out the amplitude ceiling.","tokens_in":47203,"feed_emoji":"🕳️","tokens_out":14709,"duration_ms":133328,"temperature":0.7,"pith_summary":"Primordial black holes (PBHs) forming from large curvature perturbations are one of the few probes of the primordial curvature power spectrum on scales far smaller than the cosmic microwave background can reach. This paper builds a sharper bridge between PBH abundance and that power spectrum, replacing simplified collapse criteria with numerically simulated thresholds, an exact non-linear density–curvature relation, and peak theory for the smoothed overdensity field. The main quantitative claim is that the maximum allowed amplitude of the curvature power spectrum on scales $k \\sim 10^5$–$10^{15}\\,\\mathrm{Mpc}^{-1}$ is about an order of magnitude lower than earlier estimates. The paper also concludes that even a single PBH formed this way would be incompatible with a scale-invariant spectrum extrapolated from CMB scales, so detecting just one would signal a distinctly non-slow-roll inflationary epoch.","feed_headline":"Black-hole census tightens early-universe ripple limits tenfold","feed_subtitle":"Full non-linear collapse physics turns black-hole counts into a stricter small-scale test of inflation.","key_machinery":"The load-bearing object is the compaction function, twice the mass excess inside a sphere divided by its areal radius, $C = 2\\delta M/R$; the paper adopts the criterion that a PBH forms when the value of $C$ at its maximum radius and horizon-crossing time exceeds the shape-dependent threshold $\\delta_{I,c}(\\alpha)$. The second piece is the exact non-linear map from the curvature perturbation $\\zeta$ to the density contrast, $\\delta \\propto e^{2\\zeta}[\\nabla^2\\zeta - \\tfrac{1}{2}(\\nabla\\zeta)^2]$, which replaces the usual linearized Laplacian relation and changes the inferred peak scale, horizon-crossing mass, and threshold. The third piece is standard peak theory for Gaussian random fields, whose spectral moments $\\sigma_0,\\sigma_1,\\sigma_2$ are fed through the abundance equation so that the variance $\\sigma_0$ is fixed by the observed PBH fraction, and the mean peak profile is converted via equation (6.1) into the curvature power spectrum $\\mathcal{P}_\\zeta(k)$ on the modes relevant to collapse.","core_discovery":"The paper's central claim is that, with the collapse physics treated properly, PBH abundance constraints become an order of magnitude stronger: on scales $k \\simeq 10^5$–$10^{15}\\,\\mathrm{Mpc}^{-1}$ the allowed amplitude of $\\mathcal{P}_\\zeta(k)$ drops by about a factor of ten relative to earlier analyses. The improvement comes from three corrections that all push in the same direction: the collapse threshold $\\delta_{I,c}(\\alpha)$ ranges from $0.4135$ to $2/3$ across the profile family instead of being a single number; the full non-linear relation between density and curvature damps and shrinks the overdensity profile, with linear theory underestimating the perturbation size and horizon mass by up to a factor of about six; and standard peak theory gives a variance $\\sigma_0$ that is $10$–$30\\%$ smaller than the usual simplified abundance estimate, so previous work overestimated the power-spectrum amplitude by $20$–$70\\%$. Combining these elements, the reconstruction formula yields both the peak amplitude and the shape of $\\mathcal{P}_\\zeta(k)$ on modes $k_t$ to $5k_t$, and the same machinery shows that generating even one PBH in the observable Universe would require $\\mathcal{P}_\\zeta \\sim 10^{-3}$–$10^{-2}$, orders of magnitude above the CMB-scale value.","pith_inferences":["A non-Gaussian extension of peak theory could move the constraints either way; because the abundance is set by the tail of the peak-height distribution, even mild small-scale non-Gaussianity could shift $\\mathcal{P}_\\zeta$ by as much as the order-of-magnitude tightening claimed here.","The filtering prescription — smooth on scales far below the perturbation rather than at the horizon scale — transfers directly to other rare-object statistics, such as heavy halos or cosmic-string loops, where window-function choice is a known source of systematic uncertainty.","Running the same pipeline backwards with a full critical-collapse mass function would predict a specific PBH mass distribution for each spike shape, letting a measured mass function break the degeneracy between $\\mathcal{P}_\\zeta$ peak amplitude and profile parameter $\\alpha$.","For broad or plateau-like power spectra, the black-hole-in-black-hole problem becomes relevant; implementing an excursion-set peak theory would likely change the derived constraints relative to the monochromatic-spike case considered here."],"forward_implications":["The maximum allowed amplitude of the primordial curvature power spectrum on scales $k \\sim 10^5$–$10^{15}\\,\\mathrm{Mpc}^{-1}$ is roughly an order of magnitude lower than previous PBH-based estimates.","A single curvature-origin PBH in the observable Universe would force $\\mathcal{P}_\\zeta$ to rise to $\\sim 10^{-3}$–$10^{-2}$, so a scale-invariant spectrum at the CMB amplitude is incompatible with even one such object.","The reconstructed power-spectrum shape on modes $k_t \\lesssim k \\lesssim 5k_t$ steepens as the profile parameter $\\alpha$ grows, so measuring the high-$k$ slope of a spike constrains the typical perturbation profile.","Because $\\sigma_0$ depends on the abundance only through $(-\\log f_{\\rm PBH})^{-1/2}$, the resulting amplitude limits are nearly insensitive to which observational $f_{\\rm PBH}$ bound is used; the collapse modelling, not the abundance data, dominates the uncertainty.","If PBHs are detected but future gravitational-wave searches see no scalar-induced stochastic background at the corresponding scales, the curvature-perturbation formation mechanism would be disfavoured in favour of alternatives such as topological defects."],"supporting_citations":[{"why":"Provides the compaction-function collapse criterion the paper adopts for deciding when a perturbation forms a black hole.","marker":"[132]"},{"why":"Supplies the numerically computed shape-dependent collapse thresholds and mass-scaling coefficients for the profile family.","marker":"[137]"},{"why":"Supplies the peak-theory statistics (peak density, mean profile, spectral parameters) used to convert collapse conditions into abundance.","marker":"[166]"},{"why":"Gives the earlier PBH-based constraints on the curvature power spectrum that the new calculation tightens by roughly an order of magnitude.","marker":"[121]"},{"why":"Gives the most recent prior PBH abundance constraints used as the comparison baseline in Figure 6.","marker":"[126]"},{"why":"Establishes the shape-dependence of PBH abundance that the reconstruction pipeline builds on.","marker":"[189]"},{"why":"Provides the current and forecast bounds from spectral distortions and gravitational-wave backgrounds used to show where the new PBH limits sit.","marker":"[108]"}],"fun_headline_variants":["Black-hole census slashes early-universe ripple limits tenfold","Non-linear PBH collapse tightens spectrum constraints tenfold","PBH abundance now tests primordial curvature ten times better","Full collapse physics sharpens black-hole cosmology bounds","Numerical relativity plus peak theory tighten PBH spectrum limits"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The calculation assumes the smoothed overdensity field is Gaussian so that peak theory can be applied, an assumption the authors themselves call too strict; the non-linearities they quantify make the true field non-Gaussian, and because PBH abundance is set by the tail of the peak-height distribution, relaxing this assumption could move the reconstructed power-spectrum amplitude substantially.","fun_headline_variants_meta":{"raw":{"variants":["Black-hole census slashes early-universe ripple limits tenfold","Non-linear PBH collapse tightens spectrum constraints tenfold","PBH abundance now tests primordial curvature ten times better","Full collapse physics sharpens black-hole cosmology bounds","Numerical relativity plus peak theory tighten PBH spectrum limits"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000445,"raw_usage":{"total_tokens":2290,"prompt_tokens":1025,"completion_tokens":1265,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":641,"completion_tokens_details":{"reasoning_tokens":1186}},"tokens_in":641,"tokens_out":1265,"duration_ms":11371,"temperature":1.0,"reasoning_tokens":1186,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:09:03.884782+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Re-derive the PBH abundance at fixed $\\mathcal{P}_\\zeta$ using the same non-linear $\\zeta\\to\\delta$ map but with a non-Gaussian peak distribution, or with numerical simulations that include the full mode coupling, and compare the resulting $f_{\\rm PBH}$ to the Gaussian peak-theory prediction; a difference of more than about an order of magnitude at fixed amplitude would shift the central constraint by more than the claimed tightening. A direct check is also possible observationally: a future measurement of the stochastic gravitational-wave background from the same scalar perturbations on the scales of Figure 6 would confirm or rule out the amplitude ceiling.","supporting_citations":[],"review_version":1}