{"id":"5e80f3bb-2926-4b2e-acde-0c030342c1c9","arxiv_id":"2411.11965","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":7,"one_line_summary":"The gravitational wave bias of mock binary black holes increases with the pivot mass of the assumed host galaxy mass selection, and can exceed the galaxy bias by up to about 30 percent only for steep host probability slopes.","lead":"This paper models how the clustering of gravitational wave sources from binary black hole mergers depends on the galaxies that host them, by populating two all-sky galaxy surveys with mock mergers. It offers a new way to turn future gravitational wave clustering measurements into constraints on black hole formation physics.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Qualitative trend is robust, but the ~30% headline amplitude is only reached at the extreme δl=0.5 corner, outside the physically motivated range; the physically motivated δl~1 case (~10%) is comparable to the paper's own unresolved systematics (photo-z scheme, App.","rationale":"The paper is an honest forward-modeling study with a sensible pipeline: the uniform-limit validation (App. B), the KS-tested ℓmin selection (App. C), and the kmax robustness check (Sec. 6.4) give real independent support to the mechanics, and the qualitative direction of the central claim follows by construction from the mass-bias relation. The reader's concern (host-galaxy probability restricted to a broken power law in M* alone; no SFR or metallicity) is real, and the paper concedes it explicitly in Sec. 4.2 and Sec. 7, pointing to a companion paper. I do not treat that concession as fatal, since the paper is explicitly a first-step phenomenological framework. My stress-test identifies a sharper, text-internal reason to keep the verdict conditional: the quantitative envelope of the claim is not yet robust. The 30% headline is an edge-of-parameter-space value (δl=0.5), and the paper's own comparisons to Artale+20 and Santoliquido+22 suggest the physically relevant region sits nearer δl≈1–2, where the effect is ~10%. That 10% is of the same order as the spread induced by the photo-z convolution scheme (App. D, Fig. 21) and is entangled with the WISC 0.2<z<0.3 bin that Sec. 6.2 itself calls 'unphysical'. Additionally, the printed Eq. (4.7) has an ambiguous leading M* factor that, if literal, changes all quoted slopes by ±1 and voids the App. B uniform test; this ambiguity should be resolved before the numbers are taken at face value. The proposed test is designed to settle both the amplitude robustness and the prefactor question. Because none of these issues overturn the qualitative trend or the framework, I do not move the verdict; CONDITIONAL remains the right call.","tokens_in":40079,"tokens_out":22478,"duration_ms":209814,"concrete_test":"Reproduce the physically motivated scenario (δl=1, δh=5, M_K=10^12; red curves in Fig. 13) for 2MPZ and WISC using each of the six photo-z convolution schemes defined in App. D, and additionally recompute the WISC 0.2<z<0.3 bin with galaxies below M*=10^10 M_sun removed (the regime Sec. 6.2 calls 'anomalously high' and 'unphysical'). If the b_GW/b_g enhancement at this point varies by more than ~5 percentage points across schemes or shrinks below 5% when the low-mass WISC regime is excised, the abstract's amplitude claim and the inferred constraining power on M_K and slopes should be downgraded to a trend-only statement. Separately, a minimal implementation check: run the uniform limit δl,δh→∞ of Eq. (4.7) through the pipeline; if the recovered b_GW disagrees with b_g (which App. B claims they do agree), a spurious M* prefactor is present in the code or in the printed equation.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim has a secure qualitative core: raising M_K or lowering δl by construction shifts host selection toward higher stellar mass and therefore higher bias, and the uniform-selection validation (App. B), scale-robustness checks (Sec. 6.4), and KS-tested scale cuts (App. C) support the pipeline. What is load-bearing is the amplitude and the derived constraining power, and the paper's own text weakens those. (1) The abstract's 'maximum change of about 30%' is obtained only for δl=0.5 (Sec. 7 requires a 'rapid growth rate, around 1/δl ≳ 2'); the literature comparison in Sec. 4.2 puts physically motivated low-mass slopes at δl≈1.2 (Artale+20) and δl≈1.7–2.1 (Santoliquido+22), and Sec. 6.5 notes that the 30% enhancement in Libanore+21 requires such rapid growth. The headline therefore quantifies an extreme corner, not a representative prediction. (2) For the physically motivated δl≈1 case the claimed effect is ~10% (Sec. 7), which is the same order as the systematic spread the paper documents: App. D and Fig. 21 show b_GW shifting by up to ~10% between redshift-convolution schemes, with the zshift/σ corrections fit from a mass-dependent subset of galaxies (Fig. 22); Sec. 6.2 flags the WISC 0.2<z<0.3 bin as 'anomalously high' and 'unphysical', yet that bin enters Figs. 10, 13, and 14. (3) Eq. (4.7) as printed carries a leading M* factor; taken literally it makes the stated slopes (1/δl=0.25, −1/δh=−2) equal to 1+1/δl=1.25 and 1−1/δh=−1, and makes the δl,δh→∞ 'uniform' limit of App. B non-uniform (P∝M*). If this is a typesetting artifact the equation misstates the model; if not, every slope interpretation shifts. None of these reverse the trend's sign, but together they mean the quantitative claims, including the promised constraints on M_K and slopes, are not yet established at the stated precision.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents a forward-modeling framework for the gravitational-wave (GW) bias parameter of binary black hole mergers. Mock BBH catalogs are generated by assigning hosts from the 2MPZ and WISC galaxy catalogs according to a broken power-law host-galaxy probability P(GW|M*) (Eq. 4.7), with pivot mass M_K and slopes 1/δ_l, 1/δ_h. The angular power spectra of the mock sirens are measured and fit to a linear bias model (Eq. 2.4), and b_GW is compared to the galaxy bias b_g measured from the same catalogs. The central finding is that b_GW increases with M_K and with 1/δ_l, with a maximum enhancement of about 30% relative to b_g over the explored parameter space, while the fiducial and physically motivated cases give enhancements of order 5–10%. The paper is framed as a first phenomenological bridge between observed galaxy catalogs and GW clustering, to be tested with future GW observations.","tokens_in":40554,"tokens_out":7559,"duration_ms":73546,"significance":"If the central mapping is correct, this paper provides a practical way to translate assumptions about BBH host-galaxy properties into a prediction for the GW bias, which is useful for forecasts and for interpreting future dark-siren cross-correlations. The pipeline is carefully validated in several respects: a uniform host-galaxy probability recovers the galaxy bias (Appendix B), the results are robust to the adopted kmax/ℓmax over a broad range (Section 6.4), and the ℓmin choices are tested with a Kolmogorov–Smirnov statistic on the χ² distribution (Appendix C). The treatment of mass-dependent photometric-redshift systematics in Appendix D is more thorough than is common in this type of study. The paper is also appropriately honest that this is a forward model rather than a measurement, and that incompleteness affects absolute bias values. However, the predictive content is largely a mapping from a mass-selection rule to the bias of a mass-selected subset, so the qualitative trend is expected; the value added is in the quantitative calibration and in the explicit exploration of the parameter space, not in an observational detection.","major_comments":[{"comment":"The printed host-galaxy probability contains an explicit leading factor M*, so log10 P = const + (1 + 1/δ_l) log10 M* on the low-mass side and log10 P = const + (1 − 1/δ_h) log10 M* on the high-mass side. The text immediately below Eq. (4.7) states that δ_l = 4 gives a 'positive slope of ~0.25' and δ_h = 0.5 a 'negative slope of ~−2', but the actual slopes are 1.25 and −1 if the equation is taken literally. This same issue propagates into the comparison with Artale et al. (2020): their α1 ~ 0.8 is translated into 1/δ_l = 0.8, yielding δ_l ~ 1.2, but under the printed equation one would obtain 1 + 1/δ_l = α1, i.e. δ_l ~ 5. Please either remove the leading M* factor or revise the slope definitions, and re-derive the literature mapping and fiducial parameter choices consistently.","section":"Sec. 4.2, Eq. (4.7)"},{"comment":"The 'maximum change of about 30%' is reached only for the extreme δ_l = 0.5 corner of the parameter space (blue curves in Fig. 13), while the physically motivated δ_l ~ 1 case yields about 10%, as stated in Sec. 7 and Sec. 6.5 ('would require a rapid growth rate, around 1/δ_l ≳ 2'). At the same time, Appendix D and Fig. 21 show that b_GW shifts by up to ~10% between the different photometric-redshift convolution schemes, so the representative signal is of the same order as a documented systematic. In addition, Sec. 6.2 flags the WISC 0.2 < z < 0.3 bin as 'anomalously high' and 'unphysical' due to contamination or photo-z problems, yet this bin enters Figs. 10, 13, and 14. I recommend re-scoping the abstract and presentation so that the headline amplitude reflects the physically motivated regime, and quantifying a systematic floor before claiming future constraining power.","section":"Abstract; Secs. 6.2, 6.5, 7"},{"comment":"The central trend b_GW increasing with M_K is, as the paper itself explains in Sec. 6.1, a consequence of selecting higher-stellar-mass subsets of the same galaxy catalogs whose bias b_g is measured, together with the fact that 'galaxy bias is related to stellar mass'. This does not invalidate the forward model, but it means the comparison b_GW vs. b_g is not an independent test of the host-galaxy probability; it is a calibration of the mass-selection mapping. The paper should state this limitation more prominently in the conclusions, and where possible validate the mapping against an external stellar-mass–bias relation or a separate galaxy sample.","section":"Secs. 2 and 6.1"}],"minor_comments":[{"comment":"The domain log10 M* < 7 is not covered by either branch of Eq. (4.7); please state explicitly that the probability is zero there, or extend the low-mass branch to lower masses.","section":"Sec. 4.2, Eq. (4.7)"},{"comment":"The sentence 'The GW bias at M_K = 10^12 M⊙ exceeds the galaxy bias by at most ~5% and exceeds the galaxy bias at M_K = 10^11 M⊙ by ~10%' appears to invert the trend shown in Fig. 10, where the enhancement grows with M_K; please clarify the intended comparison.","section":"Sec. 7"},{"comment":"References [115] and [136] are the same HEALPix paper, and references [117] and [142] are the same Balaguera-Antolinez et al. paper; please consolidate the duplicates.","section":"References"},{"comment":"Figure 21 omits error bars for clarity, but the spread among the convolution schemes is central to the systematic assessment; please include representative error bars or provide a numerical summary of the spread in the caption or text.","section":"Appendix D, Fig. 21"}],"recommendation":"major_revision","confidential_remarks":"The core pipeline is sound and the paper is a useful contribution, but the inconsistency in Eq. (4.7) is load-bearing because it affects the fiducial parameter interpretation and the literature comparison. The amplitude-versus-systematics issue also needs to be addressed before the abstract can be taken at face value. I do not see a novelty problem, and the authors are appropriately cautious about the forward-model interpretation."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The new thing here is not another simulation-based estimate of b_GW; it is the first attempt I know of to populate observed all-sky photometric galaxy catalogs (2MPZ, WISC) with mock BBHs using a flexible broken power-law host probability, and then map the host parameters onto clustering bias. That is a useful complement to the Libanore/Peron simulation approach, because it uses real survey selection and a much larger volume, at the price of no SFR or metallicity information. The pipeline is careful: uniform-host validation, KS-tested scale cuts, and a dedicated appendix on photometric redshift convolution sensitivity. The authors are honest that b_GW/bg is the meaningful quantity rather than b_GW itself, given catalog incompleteness. That deserves credit.\n\nThe soft spots are real but mostly fixable. First, the abstract's 'maximum change of about 30%' is reached only for δl=0.5, which the paper's own literature comparison places outside the physically motivated range (δl≈1.2–2.1). For δl≈1, the effect is ~10%, the same order as the systematic spread from photo-z convolution choices (Appendix D) and the anomalous WISC 0.2<z<0.3 bin that the authors themselves flag as unphysical. So the headline number is an extreme corner, not a representative prediction. Second, Eq. (4.7) as printed contains a leading M* factor. Taken literally, the slopes are 1+1/δl and 1−1/δh; the stated fiducial slopes 0.25 and −2 become 1.25 and −1, and the δl,δh→∞ uniform limit is not uniform. The surrounding text clearly intends a pure power law, so this is probably a rendering artifact, but the equation must be corrected before the quantitative claims are used. Third, no code or data are provided; for a paper whose output is a mapping others will want to use in forecasts, that is a gap, though not fatal.\n\nThe qualitative core is secure: raising M_K or steepening the low-mass slope selects higher-mass hosts and therefore higher bias, and the pipeline checks out. I would send this to a serious referee, expecting major revision on the equation, the amplitude framing, and a more direct comparison of the ~10% signal to the documented systematics. If you work on GW×LSS systematics it is worth reading now; if you want a citable mapping for forecasts, wait for the revision.","headline":"A useful forward-modeling framework for the GW bias—qualitative trends are secure, but the headline 30% amplitude is an extreme corner and Eq. (4.7) as printed misstates the slopes.","tokens_in":41207,"tokens_out":3859,"would_cite":true,"duration_ms":37849,"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":"This paper models the gravitational-wave bias of binary black hole mergers as a consequence of the stellar-mass-dependent host galaxy probability, and shows that measuring this bias can constrain how BBHs populate galaxies.","keywords":["gravitational wave bias","binary black hole mergers","angular power spectrum","host galaxy probability","stellar mass","2MPZ","WISC","large-scale structure"],"falsifier":"Cross-correlate a future dark-siren catalog from next-generation detectors (O5 or Einstein Telescope) with 2MPZ/WISC-like galaxy maps at $z<0.4$, measure $b_{\\rm GW}$ from the angular power spectrum, and compare it to the predicted $b_{\\rm GW}(M_K)$ curves. If the recovered bias does not rise with the mean stellar mass of host galaxies, or falls outside the ~30% envelope around the galaxy bias, the broken power-law host-galaxy probability is ruled out.","tokens_in":39889,"feed_emoji":"🌌","tokens_out":7945,"duration_ms":66558,"temperature":0.7,"pith_summary":"This paper builds a forward model that connects the clustering bias of binary black hole mergers, $b_{\\rm GW}$, to the stellar mass of their host galaxies. The authors populate the 2MPZ and WISC galaxy catalogs with mock BBH mergers according to a broken power-law host-galaxy probability $P({\\rm GW}|M_*)$ with a pivot mass $M_K$, measure the angular power spectra of the resulting siren catalogs, and fit $b_{\\rm GW}$ relative to the dark matter clustering. Their central finding is that $b_{\\rm GW}$ increases with both $M_K$ and the low-mass slope $1/\\delta_l$, with a maximum shift of about 30% relative to the galaxy bias across the parameter range explored. If this model is right, future measurements of $b_{\\rm GW}$ from GW-galaxy cross-correlations can constrain $M_K$ and the slopes of the host-galaxy probability, offering a new observational handle on the astrophysics of BBH formation.","feed_headline":"Black hole mergers cluster up to 30% differently than galaxies","feed_subtitle":"New model ties merger hosts' stellar mass to the gravitational-wave bias future surveys can measure.","key_machinery":"The load-bearing object is the broken power-law host-galaxy probability $P({\\rm GW}|M_*)$ of Eq. (4.7): $P\\propto M_*\\,10^{(\\log_{10}M_*-\\log_{10}M_K)/\\delta_l}$ below $M_K$ and $\\propto M_*\\,10^{(\\log_{10}M_K-\\log_{10}M_*)/\\delta_h}$ above it. This function decides which galaxies in 2MPZ and WISC host the mock BBH sirens, shifting the mean stellar mass of hosts and therefore their large-scale clustering. The measured $b_{\\rm GW}$ comes from fitting $b_{\\rm GW}^2 C_\\ell^{\\rm Mod}$ to the siren angular power spectrum, where $C_\\ell^{\\rm Mod}$ is computed from the siren redshift distribution and the dark matter transfer function. The merger-rate normalization and delay-time distribution set the number of sirens but are shown not to change the clustering, so the host-galaxy probability alone carries the parameter dependence of the bias.","core_discovery":"The central claim is that the gravitational wave bias parameter is a diagnostic of the host-galaxy probability function, not a fixed number of the GW source population. For a broken power law $P({\\rm GW}|M_*)$ peaking at $M_K$, the paper finds generically that increasing $M_K$ shifts the mock siren hosts toward more massive, more clustered galaxies and raises $b_{\\rm GW}$, while decreasing $M_K$ pushes hosts toward low-mass galaxies and suppresses $b_{\\rm GW}$ below the galaxy bias. Across the explored parameter space ($M_K\\in[10^9,10^{12}]\\,M_\\odot$, $\\delta_l\\in[0.5,10]$, $\\delta_h\\in[0.1,5]$), the maximum deviation of $b_{\\rm GW}$ from the galaxy bias is about 30%. The authors further claim that the ratio $b_{\\rm GW}/b_g$ is the robust observable given catalog incompleteness, varying by less than 15% with redshift up to $z=0.4$ for the fiducial model, and that different formation scenarios (short vs. long delay times) leave distinguishable signatures in this ratio.","pith_inferences":["The paper's restriction to stellar mass may mask a degeneracy: if real BBH hosts are selected by star formation rate or metallicity at fixed stellar mass, an observed $b_{\\rm GW}$ would be attributed to the wrong $M_K$ and slopes; testing with spectroscopic samples that measure SFR and metallicity would break this degeneracy.","The ~30% maximum bias shift is derived from photometric catalogs with $z<0.4$; extrapolating the broken power-law model to higher redshift would require adding redshift evolution of $M_K$, which the current framework does not include.","A practical shortcut suggested by the logistic fits is that a single measurement of $b_{\\rm GW}/b_g$ near $z\\sim0.15$ could already place a lower or upper bound on $M_K$ relative to the mean galaxy mass, without needing the full parameter scan.","The method could be validated before GW detections by applying the same host-galaxy probability to a complete spectroscopic sample and checking whether the predicted $b_{\\rm GW}$ from 3D clustering matches the angular-spectrum result."],"forward_implications":["A measurement of $b_{\\rm GW}$ in the local universe can be inverted to constrain the pivot mass $M_K$ and the slopes $\\delta_l$, $\\delta_h$ of the BBH host-galaxy probability.","The ratio $b_{\\rm GW}/b_g$, rather than the absolute bias, should be used for cosmological inference because it is robust to galaxy catalog incompleteness.","Scenarios with short delay times (strong suppression at high mass, small $\\delta_h$) and long delay times (mild or no suppression, $\\delta_l\\sim1$) predict distinguishable $b_{\\rm GW}$ values, up to ~30% apart.","The method transfers directly to binary neutron star and neutron star-black hole mergers, and to deeper future surveys such as DESI, Euclid, Rubin, and Roman.","The logistic fit of $b_{\\rm GW}$ versus $\\log_{10}M_K$ implies a steep transition near the mean galaxy mass of each redshift bin, so the bias measurement locates the characteristic stellar mass scale of the hosts."],"supporting_citations":[{"why":"Provides the population-synthesis host galaxy mass and star formation rate distributions that motivate the $\\delta_l\\sim1$, no-suppression limit of the broken power-law model.","marker":"[43]"},{"why":"Supplies observational scaling relations whose fitted parameters translate to the $\\delta_l\\sim1.7$-$2.1$ slopes used to set the explored low-mass slope range.","marker":"[40]"},{"why":"Gives host-galaxy probabilities from semi-analytic galaxy formation in the Millennium-II simulation, providing the $\\delta_l\\sim1$, $\\delta_h\\sim0.5$-$2$ reference values for short delay times.","marker":"[38]"},{"why":"Presents the 2MPZ photometric redshift catalog that serves as the low-redshift host galaxy sample for mock sirens.","marker":"[97]"},{"why":"Presents the WISExSCOS (WISC) photometric redshift catalog that serves as the deeper host galaxy sample out to $z\\sim0.4$.","marker":"[111]"},{"why":"Supplies the W1-band mass-to-light ratios used to estimate stellar masses for active and passive galaxies in both catalogs.","marker":"[119]"},{"why":"Supplies the photometric redshift convolution method and the redshift-dependent scatter relation for 2MPZ used to build the model angular power spectrum.","marker":"[142]"},{"why":"Provides the theoretical angular power spectrum formula that connects the siren redshift distribution to the model $C_\\ell$ used to fit $b_{\\rm GW}$.","marker":"[89]"},{"why":"Provides the local binary black hole merger rate used to normalize the number of mock sirens as a function of redshift.","marker":"[124]"}],"fun_headline_variants":["GW clustering deviates from galaxies by up to 30%","Black hole merger clustering reveals host galaxy mass via bias","GW bias: a 30% signature of merger host galaxies","How black hole mergers cluster maps their galaxy hosts' mass"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The central premise is that the probability a galaxy hosts a binary black hole merger is a broken power law of stellar mass alone, with the same functional shape in every redshift bin and no dependence on star formation rate, metallicity, or environment.","fun_headline_variants_meta":{"raw":{"variants":["GW clustering deviates from galaxies by up to 30%","Black hole merger clustering reveals host galaxy mass via bias","GW bias: a 30% signature of merger host galaxies","How black hole mergers cluster maps their galaxy hosts' mass"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001445,"raw_usage":{"total_tokens":5865,"prompt_tokens":1034,"completion_tokens":4831,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":650,"completion_tokens_details":{"reasoning_tokens":4763}},"tokens_in":650,"tokens_out":4831,"duration_ms":33067,"temperature":1.0,"reasoning_tokens":4763,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T18:04:10.159575+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Cross-correlate a future dark-siren catalog from next-generation detectors (O5 or Einstein Telescope) with 2MPZ/WISC-like galaxy maps at $z<0.4$, measure $b_{\\rm GW}$ from the angular power spectrum, and compare it to the predicted $b_{\\rm GW}(M_K)$ curves. If the recovered bias does not rise with the mean stellar mass of host galaxies, or falls outside the ~30% envelope around the galaxy bias, the broken power-law host-galaxy probability is ruled out.","supporting_citations":[{"cited_title":"Wide-area tomography of CMB lensing and the growth of cosmological density fluctuations","cited_arxiv_id":"1805.11525","evidence_quote":"Supplies the photometric redshift convolution method and the redshift-dependent scatter relation for 2MPZ used to build the model angular power spectrum."}],"review_version":1}