{"id":"115bb3ec-1ba5-47d2-a214-195aa195a94d","arxiv_id":"2601.07917","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"Runaway stellar collisions in dense, metal-poor young clusters can build intermediate-mass black holes of roughly 300-6000 Msun, and the channel is suppressed by high metallicity or low cluster density, with about half of the cosmic IMBHs forming below z~1.5-3.","lead":"Using more than 1,400 star-cluster simulations, the authors map when runaway collisions of stars create intermediate-mass black holes, finding that cluster density and metallicity set the outcome. The result supplies fitting formulas for seeding galaxy-formation simulations and suggests many IMBHs form at moderate redshifts, not only in the early Universe.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Adopted Vink (2018) wind law extrapolated beyond 900 Msun with no cap is the pivotal assumption; capping at 1.5e-4 Msun/yr changes M_bullet from ~6000 to >25000 and likely shifts the Z threshold, so M_bullet(Sigma_h,Z) and cosmic rate are not robust.","rationale":"I read the paper as claiming a robust mapping from (Sigma_h,Z) to maximum IMBH mass and a cosmic rate. The N-body campaign is impressive: >1440 isolated runs, 30 hierarchical runs, cross-checks against Vergara, wind-boost tests, and explicit acknowledgement of wind sensitivity are real strengths. But the mapping is not parameter-free: it depends on a wind law extrapolated beyond its calibration range, and the authors themselves show that an equally common prescription with a cap changes the outcome qualitatively (SMS vs IMBH). The concern is therefore internal to the paper's own sensitivity analysis and is the single most load-bearing issue. It is not a disagreement with consensus; it is a correctness risk: the fitted formula and rate estimates inherit the choice. I would not reject: the paper is transparent, and the qualitative statement that dense low-Z clusters can produce massive BH seeds survives under both wind treatments. But the precise masses, the Z threshold, and the statement about local IMBH origins should be treated as conditional on wind physics. Hence UNCHANGED relative to the reader's CONDITIONAL verdict.","tokens_in":56757,"tokens_out":4751,"duration_ms":46466,"concrete_test":"Re-run the full isolated grid IM3D[1-5]Z[1-9] (at least one random realization per cell) and the HD9Z1/HD9Z2 hierarchical runs with BIFROST using the alternative wind model that produced Vergara et al. (2025a): Vink et al. (2001) line-driven winds plus the Humphreys-Davidson/LBV ceiling max(dm/dt)=1.5e-4 Msun/yr, with all other collision, radius, and integration settings unchanged. If the LOESS/fitted M_bullet(Sigma_h,Z) surface shifts by more than a factor of 2 in the dense low-Z region or the metallicity threshold moves above Z=0.3 Zsun, the published formula and the derived cosmic IMBH rate should be re-derived; if the surface is unchanged, the concern is resolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central quantitative claim is the fitting formula M_bullet(Sigma_h,Z) (Eqs. 9-10, Figs 7-8) and the associated metallicity threshold Z~0.2-0.3 Zsun. This is governed by Eq. (5), the Vink (2018) wind-loss rate, which the paper explicitly extends beyond its calibrated 900 Msun range with no maximum rate. At Z=0.01, Vink(2018) reaches 1.5e-4 Msun/yr near 1700 Msun; at Z=0.10 it already does so near 760 Msun. If, as in the commonly used NBODY6++GPU/MOCCA prescription (Vink 2001 plus the Humphreys-Davidson/LBV ceiling of 1.5e-4 Msun/yr), the wind rate is capped, the authors' own post-processing (Fig. 11) and direct test (Fig. C1) show monotonic growth past 20000-25000 Msun instead of the ~7000 Msun plateau, i.e. a factor ~4 change in final IMBH mass and entry into the SMS regime relevant for SMBH seeds. The same cap removes much of the metallicity dependence that drives the Z~0.2 threshold, so the suppression of IMBH formation above Z~0.2 Zsun, the peak redshift/rate of the cosmic formation model (Section 6), and the conclusion that local IMBHs are not necessarily failed high-z seeds all inherit the chosen wind extrapolation. Secondary uncertainties in collisional mass loss (Eq. 2, C=0.3 primary+secondary losses) and in EMS/SMS radii (Section 2.2.2, factor 100-1000 late-time differences with Vergara et al.) go in the same direction and are acknowledged in the text, but the wind prescription is the dominant load-bearing choice.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents a large suite of more than 1,440 direct N-body simulations of young massive star clusters with the BIFROST code, varying cluster mass, half-mass surface density, and metallicity, to study the formation of intermediate-mass black holes (IMBHs) via runaway stellar collisions. The central quantitative claim is that the maximum IMBH mass is determined almost entirely by cluster half-mass surface density and metallicity through a fitting formula M_bullet(Sigma_h, Z) (Eq. 9, with Eq. 10 for extrapolation), with IMBHs of roughly 300–6000 Msun forming at Z ≲ 0.2–0.3 Zsun and formation suppressed at higher metallicity and below Sigma_h ≲ 3e4 Msun/pc^2. The paper also presents 30 hierarchical cluster assembly simulations, estimates wind and collisional ejecta budgets, and constructs a cosmic IMBH formation rate model that peaks at z~2–4, with ~50% of IMBHs forming below z~1.5–3, thereby challenging the picture that all local IMBHs are failed high-redshift SMBH seeds. The authors explicitly test the sensitivity of their results to the adopted massive-star wind prescription in Section 4.2 and Appendices B and C.","tokens_in":57322,"tokens_out":3362,"duration_ms":37527,"significance":"If the central scaling holds, the fitting formulas provide a practical seeding prescription for galaxy-formation simulations and semi-analytic models, and the derived redshift-dependent IMBH formation rates would be an important constraint on black-hole seeding scenarios. The paper's main strengths are the exceptionally large and systematic simulation sample, the inclusion of hierarchical assembly alongside isolated clusters, and the unusually transparent treatment of the dominant stellar-wind uncertainty: the authors show in their own post-processing and dedicated test runs that an alternative widely used wind prescription (Vink 2001 with an LBV-like cap of 1.5e-4 Msun/yr) changes the maximum IMBH mass by a factor of about four and pushes collisionally grown stars into the supermassive-star regime. This honesty is commendable, but it also directly exposes the fragility of the headline numerical predictions and of the metallicity threshold that drives the cosmic rate model.","major_comments":[{"comment":"The load-bearing premise of the paper is the Vink (2018) wind prescription, Eq. (5), extrapolated beyond its stated 900 Msun calibration range with no upper rate limit. The paper's own analysis shows that imposing the commonly used LBV-like ceiling of 1.5e-4 Msun/yr, as in NBODY6++GPU/MOCCA, changes the mass growth from a plateau near ~7000 Msun to monotonic growth beyond 25000 Msun (Fig. 11 and Fig. C1), a factor of ~4 in the final IMBH mass, and removes much of the metallicity dependence that produces the Z~0.2 Zsun threshold. Since Eq. (9), the threshold quoted in the abstract, and the cosmic rate model in Section 6 all inherit this choice, the central quantitative claims are conditional on an extrapolated and uncapped wind law. The authors should either justify this extrapolation quantitatively or present the main fitting formulas and cosmic rates for both wind prescriptions, with a","section":"§2.2.3, Eq. (5); §4.2, Fig. 11; Appendix C"},{"comment":"The fitting formula Eq. (9) is a LOESS-plus-piecewise fit to the authors' own simulation data, and it is then used as input to the cosmic rate model in Section 6. The 'consistency' with JWST little-red-dot number densities reported in Section 6.6 is therefore a self-consistency check of the model chain, not an independent test of the fitting formula or of the IMBH channel. This is especially important because the rate model also depends on the cluster formation efficiency and the normalization of the cluster mass–radius relation (f_h), which are varied but not constrained by the N-body results. The text should be revised to state explicitly that the JWST comparison validates the assumed seeding efficiency only under the combined model assumptions, and should avoid presenting it as independent support for M_bullet(Sigma_h, Z).","section":"§3.4, Eq. (9); §6.1–6.6"},{"comment":"The collisional mass-loss model, Eq. (2) with C=0.3 and mass loss taken from both the primary and secondary, is another ad-hoc choice with a strong cumulative effect: Appendix A shows that for a star doubling its mass the cumulative fractional loss is ~0.214, about 2.85 times the single-collision maximum and 1–2 orders of magnitude larger than secondary-only loss prescriptions. Recent hydrodynamical work on extended, marginally bound extremely massive stars suggests that mass loss can be even more catastrophic in some collisions. Although the authors acknowledge this uncertainty, it is not varied in the main sample, so the error bars on the final IMBH masses and on the metallicity threshold reflect only a subset of plausible prescriptions. At minimum, the paper should state how Eq. (2) affects the peak IMBH masses relative to the wind uncertainty, and ideally provide a small number of te","section":"§2.2.1, Eq. (2); Appendix A"}],"minor_comments":[{"comment":"Typo in the heading: 'The faction of clusters' should read 'The fraction of clusters'.","section":"§3.2.1"},{"comment":"In the caption, 'IMBHs cannot from at Z=1.0 Zsun' should read 'cannot form'.","section":"Fig. 2 caption"},{"comment":"The sentence 'The maximum collision and TDE rates by IMBHs in models HD9Z2 is are' has a grammatical error; 'is are' should be 'are'.","section":"§4.6"},{"comment":"The piecewise definitions of A(Z), B(Z), C(Z), D(Z), E(Z) are terse; an explicit worked example of how to assemble, say, A(Z) and Sigma_crit(Z) from Table 4 would help users implement Eq. (9) without error.","section":"§3.4, Tables 4–5"},{"comment":"The cluster formation efficiency Gamma is set to 0.3 and treated as constant; the text notes the linear scaling, but a brief discussion of the observed environmental variation of Gamma and its likely redshift evolution would improve the robustness of the rate estimates.","section":"§6.1"}],"recommendation":"major_revision","confidential_remarks":"This is a technically impressive and useful paper: the simulation campaign is large, the numerical methods are state of the art, and the authors are unusually transparent about the sensitivity of their results. My main reservation is that the abstract, fitting formula, and cosmic-rate conclusions are presented as robust predictions, while the paper's own tests show that the adopted wind prescription—extrapolated beyond its calibrated range with no cap—is the single dominant factor controlling the final IMBH masses, the metallicity threshold, and the rate history. This is fixable in revision by reframing the results as conditional on the wind model and by providing quantitative versions of the main predictions for the capped-wind case. I would not reject the paper: the underlying simulations and the uncertainty analysis are valuable, and the central claims can be made defensible with appropriate conditions and caveats. However, in its current form the headline quantitative claims overstate what the simulations can establish."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The key thing to know: this is a large, systematic direct N-body study of runaway stellar collisions across a joint cluster-density and metallicity grid. The paper's qualitative claim—that IMBH formation is controlled mainly by half-mass surface density and metallicity, with a threshold near Z ~ 0.2–0.3 Zsun—is supported by a lot of simulation power and seems sturdy. The quantitative claim, M_bullet ~ 300–6000 Msun and the fitted formula, is more fragile because it inherits a specific choice for very massive star winds.\n\nWhat is actually new: the joint Sigma_h–Z grid in isolated clusters (three masses, five densities, nine metallicities, ten realizations each—over 1400 runs), the 30 hierarchical assembly simulations reaching extreme densities, and the first practical fitting formulas M_bullet(Sigma_h, Z) for seeding galaxy-formation models. The paper also includes a cosmic IMBH formation rate calculation that explicitly depends on cluster birth densities and SFRD metallicity. Credit where due: the authors test their weakest link. Section 4.2 and Appendix C show that capping the Vink (2018) wind at an LBV-like limit of 1.5e-4 Msun/yr makes the collisionally grown star grow monotonically past 25000 Msun instead of plateauing near 7000 Msun. That is an honest, important stress test. The radial and collisional-mass-loss uncertainties are also discussed, and the hierarchical runs are a step beyond the isolated-cluster literature.\n\nThe soft spots are real and proportional. The adopted wind law is extrapolated beyond 900 Msun with no cap, and the final IMBH masses, the metallicity threshold, and the cosmic rate all ride on that extrapolation. The fitting formula is a LOESS fit to the authors' own simulations, then fed into the rate model; the JWST consistency is an internal check, not an independent validation. The rate model also depends strongly on the poorly constrained birth-density normalization f_h, and the TDE accretion fraction of 0.5 is optimistic. The data are not public, only \"available on reasonable request,\" which limits reproducibility. None of these flaws undermines the central qualitative result—denser and more metal-poor clusters make heavier IMBHs—but they do mean the precise masses and rates should be treated as model-dependent.\n\nThis paper is for anyone modeling SMBH seeding, globular-cluster IMBH searches, or next-generation GW event rates. It deserves a serious referee: the simulation campaign is large, the analysis is careful, and the wind sensitivity is honestly flagged rather than hidden. I would send it to review, with the recommendation that the authors make public at least the fitting-formula coefficients and the wind-sensitivity runs. A revised version that presents the wind-capped case as an equally prominent companion scenario would be more useful than the current single-prescription headline.","headline":"A genuinely wide simulation grid mapping IMBH masses against cluster density and metallicity, but the headline masses and cosmic rates are anchored to an extrapolated wind law that the authors themselves show could change by a factor of four.","tokens_in":57786,"tokens_out":1795,"would_cite":true,"duration_ms":22098,"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":"In massive star clusters, the mass of an intermediate-mass black hole made by runaway stellar collisions is set almost entirely by the cluster's half-mass surface density and metallicity, through a simple fitting formula; the channel turns","keywords":["intermediate-mass black holes","runaway stellar collisions","star cluster surface density","metallicity","stellar winds","N-body simulations","black hole seeds","cosmic IMBH formation rate"],"falsifier":"Run one of the paper's dense, low-metallicity cluster setups (e.g., M_cl ~1.5e5 Msun, Z~0.1 Zsun, Sigma_h~1e5 Msun/pc2) with a wind-loss rate capped at 1.5e-4 Msun/yr: the paper itself reports that this produces monotonic growth past 20,000 Msun. If a single such run with a different but physically standard wind model reproduces the plateau near 7,000 Msun, the reported IMBH masses are robust; if it grows into the supermassive-star regime, the headline masses are an artifact of the extrapolated wind law. Observationally, detecting a >300 Msun black hole in a solar-metallicity massive cluster w","tokens_in":56665,"feed_emoji":"⚫","tokens_out":7301,"duration_ms":72030,"temperature":0.7,"pith_summary":"Runaway stellar collisions in young massive star clusters can build extremely massive stars that collapse into intermediate-mass black holes (IMBHs) of roughly 300–6000 solar masses, according to more than 1400 direct N-body simulations. The paper claims that for clusters heavier than about 10^4 solar masses, the maximum IMBH mass is set almost entirely by two host-cluster properties—half-mass surface density and metallicity—through a fitting formula. The channel works in low-metallicity, dense environments and is suppressed in the high-metallicity, low-density conditions typical of the local Universe. If correct, the fitting formulae let galaxy-formation models seed supermassive black holes from star clusters, and the derived cosmic IMBH formation rate—peaking at redshift 2–4, with half forming below z~1.5–3—would challenge the idea that all local IMBHs are failed high-redshift seeds.","feed_headline":"Dense, low-metal clusters forge 300-6000 solar-mass black holes","feed_subtitle":"New fits tie IMBH mass to cluster density and metallicity; half form below z~1.5-3.","key_machinery":"The load-bearing machinery is the competition between collisional mass gain and stellar wind mass loss for stars grown beyond about 600 solar masses. The paper uses an extrapolated wind-loss law with no cap, plus a mass-loss term in each collision (f_loss = C q/(q+1)^2, C=0.3), and metallicity-dependent radii, to produce the plateau in stellar mass that sets the IMBH mass. The central object that carries the argument is the fitting formula M_bullet = theta_H(Sigma_h - Sigma_crit(Z)) [A(Z) log10 Sigma_h + B(Z) (log10 Sigma_h)^2 + C(Z)], with coefficients tabulated in three metallicity ranges; a second linear formula handles extrapolation to higher Sigma_h. This converts a two-dimensional simu","core_discovery":"The paper establishes that in massive star clusters (M_cl ≳ 10^4 Msun) the maximum mass of an intermediate-mass black hole produced by runaway collisions is a function of only the cluster's half-mass surface density Sigma_h and its metallicity Z, stated as M_bullet = M_bullet(Sigma_h, Z) with a Heaviside threshold and piecewise metallicity-dependent coefficients. IMBHs with masses of roughly 300–6000 Msun form when Z is below about 0.2–0.3 Zsun and Sigma_h above about 3e4 Msun/pc^2; above that metallicity, strong stellar winds quench the runaway growth, and below that density collisions are too rare. The paper additionally gives a hierarchical-assembly extension showing that extremely dense","pith_inferences":[],"forward_implications":["The fitted M_bullet(Sigma_h,Z) can seed black holes in semi-analytic galaxy-formation models and high-resolution cosmological simulations without running star-by-star collision calculations.","The IMBH formation efficiency is strongly peaked below Z~0.1-0.2 Zsun; local young massive clusters with solar-like metallicity and low density should rarely produce IMBHs, explaining their absence.","The cosmic IMBH formation rate density peaks at z~2-4 with values up to ~1e-7 per year per cMpc^3; roughly half of IMBHs form below z~1.5-3, so a local IMBH need not be a leftover high-z SMBH seed.","The densest hierarchical models reach sustained tidal-disruption rates above 1e-5 per year per cluster, providing a growth channel that can push IMBHs beyond 1e4 Msun.","The sensitivity study shows that if a maximum wind-loss rate (a luminous-blue-variable-like cap) is used instead, collisionally grown stars pass 20,000-25,000 Msun, making the channel a viable supermassive-star seed route.","If the paper's picture holds, targeted searches for IMBHs should prioritize dense, low-metallicity old clusters rather than solar-metallicity clusters, where the suppression boundary predicts near-zero occupation.","If the formula is read as a predictive map, the same dense clusters that form IMBHs also produce 5-10% of their mass in wind and collisional ejecta, connecting IMBH formation to the chemical peculiarities seen in globular cluster multiple populations.","The wind-rate sensitivity suggests the dominant uncertainty is the physics of extremely massive stellar winds at low metallicity; if the cap is real, the channel may produce supermassive stars and heavier seeds than the paper's headline masses."],"fun_headline_variants":["IMBH mass set by cluster density and metallicity","Dense, metal-poor clusters forge 300-6000 Msun black holes","Runaway collisions need dense, low-metal clusters to make IMBHs","New fits predict IMBH mass from cluster density and metallicity","Low metallicity, high density: key to intermediate-mass black holes"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The entire result rests on the adopted wind mass-loss law for stars above about 600 solar masses—an extrapolation without an upper cap—because replacing it with a standard cap changes the final black-hole masses by a factor of about four and shifts the channel into the supermassive-star regime.","fun_headline_variants_meta":{"raw":{"variants":["IMBH mass set by cluster density and metallicity","Dense, metal-poor clusters forge 300-6000 Msun black holes","Runaway collisions need dense, low-metal clusters to make IMBHs","New fits predict IMBH mass from cluster density and metallicity","Low metallicity, high density: key to intermediate-mass black holes"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000273,"raw_usage":{"total_tokens":1581,"prompt_tokens":963,"completion_tokens":618,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":707,"completion_tokens_details":{"reasoning_tokens":525}},"tokens_in":707,"tokens_out":618,"duration_ms":6621,"temperature":1.0,"reasoning_tokens":525,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T10:58:37.960362+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run one of the paper's dense, low-metallicity cluster setups (e.g., M_cl ~1.5e5 Msun, Z~0.1 Zsun, Sigma_h~1e5 Msun/pc2) with a wind-loss rate capped at 1.5e-4 Msun/yr: the paper itself reports that this produces monotonic growth past 20,000 Msun. If a single such run with a different but physically standard wind model reproduces the plateau near 7,000 Msun, the reported IMBH masses are robust; if it grows into the supermassive-star regime, the headline masses are an artifact of the extrapolated wind law. Observationally, detecting a >300 Msun black hole in a solar-metallicity massive cluster w","supporting_citations":[],"review_version":1}