{"id":"bdae24be-9044-4698-a0b8-8faf2624d891","arxiv_id":"2412.04553","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"The scatter in black hole mass at fixed stellar mass sets the low-mass slope of the quenched central galaxy mass function, and the observed slope requires a scatter of about 0.5 to 0.8 dex.","lead":"This paper shows that the spread in black hole masses within galaxies controls the low-mass slope of the quenched central stellar mass function, with larger spread producing shallower slopes. Matching the observed slope requires a black hole mass scatter of at least 0.5 dex at fixed stellar mass, a new constraint on galaxy and black hole coevolution.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The sigma_BH >= 0.5 dex inference hinges on comparing hand-chosen local power-law slopes from models to P12's global Schechter slope; the two need not agree, and the paper gives no error bars or fit-range sensitivity.","rationale":"The paper has real strengths: the Dark Sage controlled experiments vary sigma_BH at fixed physics and show a clear monotonic change in the local QCSMF slope; the trend is physically intuitive and the comparison includes multiple independent models. The reader's conditional verdict is therefore appropriate. The load-bearing weak point is not the internal logic of the experiments but the quantitative bridge to observations. Section 5 explicitly states the model slopes are hand-chosen local power laws, and footnote 3 acknowledges the observed P12 population deviates from a single Schechter function at low masses. Because no error bars or robustness tests are presented for alpha, and because the observed constraint is a global fit rather than the same local quantity, the lower bound sigma_BH >= 0.5 is a model-dependent inference rather than a secure measurement. The proposed check—computing the local P12 slope with the same procedure—would directly settle whether the bridge holds. No fatal flaw was found; the correct disposition is the same conditional acceptance the reader recommended.","tokens_in":12527,"tokens_out":7232,"duration_ms":71183,"concrete_test":"Digitize or retrieve the P12 quenched-central SMF and apply exactly the Section 5 fitting procedure used for the models: restrict to the mass interval just below the QCSMF peak over which the data are plausibly a single power law, fit log Phi vs log M*, and bootstrap over the published error bars. Compare this local alpha_P12 with the global Schechter value used in Figure 4. If the two differ by more than the scatter among model alpha values in Figure 4, the comparison underlying sigma_BH >= 0.5 is invalid; if they agree, the inference is supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central quantitative claim—that matching the observed QCSMF slope requires sigma(log10 MBH | M*) >= 0.5 dex—depends on comparing two different slope definitions. In Section 5, alpha for each model is obtained by a hand-chosen local power-law fit over 'the largest values of M* below the peak in the QCSMF where the slope is plausibly represented by a single powerlaw value.' The observational constraint, by contrast, is the single Schechter-function slope from P12, fit over the full quenched-central population. Footnote 3 admits that P12's data below the completeness limit rise above the best-fit Schechter function, so the observed QCSMF may also be non-Schechter; in that case the global Schechter slope is not necessarily the local low-mass slope the models are tuned to reproduce. The paper does not report uncertainties on the measured alpha values or test the sensitivity of alpha to the chosen fitting range. If the local observed slope differs from P12's global value by even ~0.1-0.2, the intersection with the model trend in Figure 4 moves substantially, and the sigma >= 0.5 lower bound is not secure. Compounding this, the toy-model relation in Eq. (5) linearizes the normal CDF about z=0; for the fiducial parameters used to derive Eq. (6), z=(log M* - y)/sigma ~ -1.25, which is not in the linear regime.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes that the low-mass slope of the quenched central stellar mass function (QCSMF) is a sensitive diagnostic of the scatter in black hole mass at fixed stellar mass, σ(log10 MBH | M*). Using the Dark Sage semi-analytic model with controlled modifications to the black hole population and to AGN feedback, the authors find that increasing the imposed BH mass scatter systematically flattens the QCSMF low-mass slope. This trend is reported across several independent models (TNG, EAGLE, SAGE, SHARK, UniverseMachine). Comparing the model slopes with the Peng et al. (2012) single-Schechter fit to observed quenched centrals, the paper infers σBH ≳ 0.5 dex and argues this is consistent with direct observational estimates of BH–galaxy scaling relation scatter. A toy model in which quenching occurs above a critical BH mass is used to interpret the relation between slope and scatter.","tokens_in":12930,"tokens_out":7202,"duration_ms":74185,"significance":"If the central claim is correct, the QCSMF slope provides a novel, observationally cheap probe of BH–galaxy co-evolution and AGN feedback, complementary to direct measurements of the MBH–M* relation. The paper's clear strength is the controlled Dark Sage experiment: within a fixed feedback prescription, varying only the imposed BH scatter yields a monotonic trend, and the qualitative agreement across independent simulation and SAM families is encouraging. The authors are also candid about limitations, noting in footnote 3 that the observed P12 data deviate from a single Schechter function below the completeness limit. However, the quantitative lower bound σBH ≳ 0.5 dex is not yet securely established: it rests on comparing hand-fitted local model slopes with a global observed Schechter slope, and the toy-model inversion used to interpret the comparison fits rather than independently predicts σBH. These issues are load-bearing for the abstract's main quantitative claim.","major_comments":[{"comment":"The model slopes α are obtained from hand-chosen fitting ranges with no reported uncertainties, and the observational anchor is P12's single-Schechter slope for the entire quenched-central population. Footnote 3 concedes that the P12 data rise above their best-fit Schechter function below the completeness limit, so the observed QCSMF may itself be non-Schechter; in that case a global Schechter slope is not necessarily equal to the local low-mass slope measured in the models. Because the σBH ≳ 0.5 dex lower bound is read off the intersection of the model trend with the P12 line in Figure 4, the authors should provide uncertainties on α, test the sensitivity of α to the chosen fitting range, and ideally measure a local slope from the P12 data over the same mass range used for the models.","section":"Section 5 and Figure 4"},{"comment":"The value σBH ≈ 0.78 dex is obtained by solving Eq. (6), which is derived by requiring the toy-model slope to equal the P10 slope 1+β at a chosen M*. It is therefore a fitted value, not an independent prediction, and the abstract's σBH ≳ 0.5 dex should be attributed to the simulation trend under an assumed observed slope rather than to the toy model. In addition, Eq. (5) linearizes the normal CDF about z=0, but for the fiducial parameters (γ=1, A=MBH,crit=10^8 Msun, M*=10^10 Msun, σBH≈0.78) one has z≈−1.25, well outside the regime where the expansion is accurate; using the full CDF derivative would change the inferred σBH.","section":"Section 6, Eq. (6)"},{"comment":"The fixed conditional distribution models impose a lognormal scatter at fixed halo mass, whereas the abstract and Section 6 state the constraint as σ(log10 MBH | M*), scatter at fixed stellar mass. Halo mass at fixed stellar mass is itself scattered, so the two quantities are not identical. Figure 4's horizontal error bars, described as 'the span of scatters within the stellar mass range used to estimate α,' do not establish that the model quantity is the same as the observational quantity. The authors should either compute and report the effective scatter at fixed stellar mass in the models, or reframe the observational comparison to match the quantity actually varied in the simulations.","section":"Sections 2 and 6"},{"comment":"The toy-model curves in Figure 4 are labeled as predictions for α, but Eq. (5) actually gives d log fQ/d log M*, and the paper's own relation d log fQ/d log M* = α − αblue implies that the plotted quantity differs from α by the slope of the star-forming SMF. Since αblue is not specified for the toy model or for each simulation, the direct overlay of Eq. (5) on the model α values in Figure 4 is not well-defined. The authors should plot the same quantity on both axes, or explain how the offset is accounted for.","section":"Section 7 and Figure 4"}],"minor_comments":[{"comment":"The caption sentence listing observational data is garbled, with '(Wright et al. 2017) (grey diamonds dashed line)' appearing after the list and the P12 label in the right-hand panel not being explained.","section":"Figure 1 caption"},{"comment":"The chosen fitting ranges for α should be reported explicitly, for example in a table, since they are currently visible only as short overplotted lines in Figures 3 and 6.","section":"Section 5"},{"comment":"The definition y=11γ−log A+log MBH,crit implicitly assumes that masses are in solar masses; this should be stated, and log10 notation should be used consistently.","section":"Equation (5)"},{"comment":"The text says the P10 argument 'only applies near M*' and then immediately says that P10 point out the formula 'if extended to lower masses, predicts Equation 2'; this is confusing and should be rephrased.","section":"Section 7"},{"comment":"The statement that no artificial neural networks were used is irrelevant to the scientific content and could be removed.","section":"Footnote 1"},{"comment":"Several LaTeX artifacts remain, such as 'Porras-V alverde' and 'M¯ u' in the affiliation block; these should be corrected in the production version.","section":"Author affiliation and text"}],"recommendation":"major_revision","confidential_remarks":"The qualitative trend is solid and worth publishing, but the quantitative lower bound σBH ≳ 0.5 dex is currently overstated relative to the support provided. The revision should focus on the slope-definition comparison, error bars and fit-range sensitivity for α, and the toy-model derivation; no new simulations are necessarily required, but re-analysis of existing outputs and a corrected comparison with P12 data are needed."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The one thing to know: the paper shows a clean, physically intuitive trend — larger scatter in black hole mass at fixed stellar mass makes the low-mass slope of the quenched central stellar mass function shallower — and it demonstrates this with controlled SAM experiments plus a broad model comparison. The quantitative claim that sigma_BH >= 0.5 dex is plausible but not as secure as the abstract implies; the slope comparison it rests on has an apples-to-oranges problem that the paper itself half acknowledges.\n\nWhat's genuinely new: the connection between sigma_BH and QCSMF slope, and the use of a tunable SAM to show causality rather than just correlation. The trend holds across TNG, EAGLE, SAGE, SHARK, and the authors' Dark Sage variants, which is real evidence. The toy model in Section 6 gives a simple explanation for why the P10 slope could arise from a black-hole-mass threshold with scatter. That's worth having.\n\nThe soft spots are in the measurement of alpha and its comparison to P12. The authors fit a powerlaw by hand in a mass range chosen by eye, report no uncertainties, and don't test sensitivity to range choice. They compare these local slopes to P12's single Schechter slope, which is a global fit over the whole quenched-central population. Footnote 3 notes P12's data actually rise above the fit below the completeness limit, which means the observed QCSMF may also be non-Schechter. If the local observed slope differs from the global one by 0.1-0.2, the inferred sigma_BH moves substantially. That's a real fragility, not a manufactured one.\n\nThe toy model has a related issue: Equation 6 is derived by setting the toy model slope equal to the P10 slope at a chosen M*, so the resulting sigma ~ 0.78 dex is a fitted value rather than an independent prediction. The Taylor expansion in Eq. 5 is also outside its linear regime for the fiducial parameters (z ~ -1.25), though the quadratic formula seems to give a reasonable answer anyway. The simulation trend is forward-modeled and independent, so the paper's core observation stands; only the calibration of the lower bound is softer.\n\nBottom line: this deserves peer review. The trend is likely to be useful regardless of whether the exact sigma_BH bound survives closer scrutiny. A good referee should push for error bars on alpha, a sensitivity analysis of the fit ranges, and a direct comparison of local observed slopes (or a discussion of why the global Schechter slope is the right target). I'd cite it for the trend and take it to reading group.","headline":"A clean, well-tested trend linking black hole mass scatter to the quenched central galaxy mass function slope, with a plausible but not fully secure lower bound of sigma_BH >= 0.5 dex.","tokens_in":13422,"tokens_out":2068,"would_cite":true,"duration_ms":59806,"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 argues that the low-mass slope of the quenched central stellar mass function is controlled by the scatter in black hole mass at fixed stellar mass, and that matching the observed slope requires a scatter of at least 0.5 dex.","keywords":["black hole mass scatter","quenched central stellar mass function","AGN feedback","mass quenching","semi-analytic models","cosmological simulations","galaxy formation","black hole–galaxy co-evolution"],"falsifier":"Measure the intrinsic scatter of the $M_{\\mathrm{BH}}\\text{--}M_*$ relation from a large sample with accurate black hole masses and a well-characterized selection function; if the true scatter comes out below about $0.4$ dex, the paper’s required floor of $0.5$ dex would be contradicted unless AGN feedback is considerably more gradual than the implemented models. Alternatively, construct a model with $\\sigma_{\\mathrm{BH}} < 0.3$ dex and a gradual feedback ramp that still reproduces the P12 slope, which would weaken the claimed causal link to scatter.","tokens_in":1801,"feed_emoji":"🕳️","tokens_out":2356,"duration_ms":50479,"temperature":0.7,"pith_summary":"The paper argues that the low-mass slope of the quenched central stellar mass function (QCSMF) encodes the scatter in supermassive black hole masses at a fixed stellar mass. By modifying the semi-analytic model Dark Sage and comparing other semi-analytic models and cosmological hydrodynamic simulations, the authors show that larger scatter in black hole mass spreads quenching over a wider range of stellar masses and flattens the low-mass slope. Reproducing the observed slope from Peng et al. (2012) requires $\\sigma(\\log_{10} M_{\\mathrm{BH}} | M_*) \\gtrsim 0.5$ dex. If true, the QCSMF low-mass slope becomes an indirect diagnostic of black hole–galaxy co-evolution and of AGN feedback physics.","feed_headline":"Black hole scatter sets quenched galaxy counts","feed_subtitle":"Larger spread in black hole masses at fixed stellar mass flattens the quenched mass function; matching data needs 0.5 dex.","key_machinery":"The central object is the quenched central stellar mass function (QCSMF) and its low-mass powerlaw slope $\\alpha \\equiv d\\log_{10}\\Phi/d\\log_{10} M_*$ measured just below the QCSMF peak near $10^{11}\\,M_\\odot$. The relationship between $\\alpha$ and the black hole mass scatter $\\sigma_{\\mathrm{BH}}$ is quantified through a toy model in which quenching occurs when a black hole exceeds a critical mass $M_{\\mathrm{BH,crit}}$; expanding the quenched fraction $f_Q$ gives $d\\log f_Q/d\\log M_*$ decreasing with increasing $\\sigma_{\\mathrm{BH}}$, and requiring this slope to approach the Peng et al. (2010) limit $1+\\beta$ at a particular stellar mass yields a quadratic equation for $\\sigma_{\\mathrm{BH}}$, giving roughly $0.78$ dex for fiducial parameters.","core_discovery":"The central claim is that the scatter in black hole mass at a fixed stellar mass, $\\sigma_{\\mathrm{BH}}$, is the dominant factor setting the low-mass slope $\\alpha = d\\log_{10}\\Phi/d\\log_{10} M_*$ of the quenched central stellar mass function. Across Dark Sage variants and in TNG100, EAGLE, SAGE, SHARK, and UniverseMachine, higher $\\sigma_{\\mathrm{BH}}$ yields shallower slopes, because quenching then occurs over a broader range of stellar masses. A comparison with the observed QCSMF slope implies $\\sigma(\\log_{10} M_{\\mathrm{BH}} | M_*) \\gtrsim 0.5$ dex, consistent with direct measurements of the $M_{\\mathrm{BH}}\\text{--}M_*$ scatter that include all galaxy types. The paper further shows that a sudden black-hole-threshold quenching law with large scatter can mimic the single Schechter function of quenched centrals that Peng et al. (2010) attributed to star-formation-rate-proportional quenching, suggesting the success of that ansatz may be coincidental.","pith_inferences":["If the relation holds at higher redshift, the low-mass QCSMF slope should evolve as the scatter in the $M_{\\mathrm{BH}}\\text{--}M_*$ relation evolves, offering a cross-check that existing surveys could attempt.","A direct, selection-function-corrected measurement of the intrinsic $M_{\\mathrm{BH}}\\text{--}M_*$ scatter in a mass-complete local sample to better than $0.1$ dex precision would either confirm or contradict the inferred floor of $0.5$ dex.","The paper’s toy model suggests a degeneracy between $\\sigma_{\\mathrm{BH}}$ and the gradualness of feedback; a testable extension is to fit both parameters jointly using the full QCSMF shape rather than a single slope."],"forward_implications":["Higher black hole mass scatter produces shallower QCSMF low-mass slopes, and the observed shallow slope implies $\\sigma_{\\mathrm{BH}} \\gtrsim 0.5$ dex.","Direct scatter measurements that include all morphological types and active galaxies (around 0.5–0.8 dex) are consistent with this bound, while the tighter scatter seen for classical bulges alone would require more gradual AGN feedback.","The Peng et al. (2010) quenching law with $\\eta \\propto$ SFR may be a coincidence: since more massive galaxies have higher star formation rates and more likely host massive black holes, a black-hole-threshold law with large scatter reproduces the same QCSMF shape.","Successful galaxy formation models must produce black hole growth decoupled from stellar mass growth; adjusting the AGN feedback recipe alone is unlikely to yield the required scatter."],"supporting_citations":[{"why":"Establishes the mass-quenching framework and the predicted relation between the quenched and star-forming mass function slopes that the paper compares against.","marker":"Peng et al. 2010"},{"why":"Provides the observed single Schechter fit to the quenched central stellar mass function whose slope is the observational constraint requiring $\\sigma_{\\mathrm{BH}} \\gtrsim 0.5$ dex.","marker":"Peng et al. 2012"},{"why":"Supplies the black hole seeding and feedback modifications to Dark Sage that approximate IllustrisTNG and generate the model variants used here.","marker":"Porras-Valverde et al. 2024"},{"why":"Describes Dark Sage, the semi-analytic model that is modified to control the black hole mass scatter and feedback prescriptions.","marker":"Stevens et al. 2016"},{"why":"Provides the IllustrisTNG simulation whose TNG100 run is compared as a cosmological hydrodynamic simulation in the slope versus scatter plane.","marker":"Pillepich et al. 2018"},{"why":"Gives the UniverseMachine empirical model, an additional comparison point for the QCSMF slope.","marker":"Behroozi et al. 2019"},{"why":"Describes the SHARK semi-analytic model included as another independent SAM comparison.","marker":"Lagos et al. 2024"},{"why":"Provides the EAGLE simulation data point used in Figure 4 to extend the trend to a second hydrodynamic simulation.","marker":"McAlpine et al. 2016"}],"fun_headline_variants":["Quenched galaxy slope points to broad black hole mass scatter","Black hole mass spread sets quenched stellar mass function slope","Large BH scatter yields shallower quenched galaxy slopes","Quenching in galaxies linked to black hole mass scatter","Quenched galaxy data imply 0.5 dex black hole scatter"],"cache_read_input_tokens":15488,"weakest_assumption_plain":"The inference that $\\sigma_{\\mathrm{BH}} \\gtrsim 0.5$ dex depends on comparing a hand-chosen local powerlaw slope fitted just below the QCSMF peak in each model to the single Schechter slope that Peng et al. (2012) fitted over a broader mass range, and the paper itself notes that few model QCSMFs are single Schechter functions and that the P12 data show an upturn below the completeness limit.","fun_headline_variants_meta":{"raw":{"variants":["Quenched galaxy slope points to broad black hole mass scatter","Black hole mass spread sets quenched stellar mass function slope","Large BH scatter yields shallower quenched galaxy slopes","Quenching in galaxies linked to black hole mass scatter","Quenched galaxy data imply 0.5 dex black hole scatter"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00028,"raw_usage":{"total_tokens":1674,"prompt_tokens":975,"completion_tokens":699,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":591,"completion_tokens_details":{"reasoning_tokens":618}},"tokens_in":591,"tokens_out":699,"duration_ms":7071,"temperature":1.0,"reasoning_tokens":618,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T21:24:01.241945+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the intrinsic scatter of the $M_{\\mathrm{BH}}\\text{--}M_*$ relation from a large sample with accurate black hole masses and a well-characterized selection function; if the true scatter comes out below about $0.4$ dex, the paper’s required floor of $0.5$ dex would be contradicted unless AGN feedback is considerably more gradual than the implemented models. Alternatively, construct a model with $\\sigma_{\\mathrm{BH}} < 0.3$ dex and a gradual feedback ramp that still reproduces the P12 slope, which would weaken the claimed causal link to scatter.","supporting_citations":[],"review_version":1}