{"id":"f16d7566-2743-46c8-8e2c-111d1234634f","arxiv_id":"2507.20687","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"UDGs and NUDGE galaxies with independent halo masses follow the globular cluster number-halo mass relation but scatter off the stellar mass-halo mass relation in proportion to their GC counts.","lead":"The authors compare five ultra-diffuse galaxies with independently measured halo masses against two standard scaling relations. The objects sit on the globular cluster count versus halo mass relation while drifting off the stellar mass versus halo mass relation in a way that scales with their globular cluster counts.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Halo mass definitions are heterogeneous; DF44's virial mass is an extrapolated cored-model value and Sgr's is a pre-infall reconstruction, so the all-five-within-scatter result may reflect mass-scale offsets rather than confirming that UDGs follow the GC-halo relation.","rationale":"The reader's verdict is CONDITIONAL with MODERATE confidence, and the weakest assumption identified is that the literature halo masses are unbiased estimates of the same virial mass used in the reference relation. My independent read agrees: this is the most load-bearing premise, and it is not secure. The heterogeneous methods - inner-profile extrapolation for DF44, pre-infall reconstruction for Sgr, rotation-curve fits for the rest - mean that the quoted masses are not operationally identical, and the paper provides no sensitivity analysis showing that the conclusion is invariant to a consistent halo-mass definition. In addition, the statistical test is weak: five points lying within a ~0.3 dex band is consistent with the relation but does not strongly confirm it, especially since the paper itself notes that scatter increases at the low-mass end. I do not see fraud, misrepresentation, or internal logical contradiction; the paper is transparent about the Sgr uncertainties and explicitly calls for larger samples. The concern lowers confidence but does not overturn the conditional acceptance. The reader already set CONDITIONAL, so I recommend no change to that verdict. The proposed concrete test - re-deriving DF44's halo mass under an NFW profile and re-testing Sgr with current rather than pre-infall values - would settle whether the central claim survives a uniform mass definition.","tokens_in":8655,"tokens_out":9316,"duration_ms":121938,"concrete_test":"Re-analyse Fig. 1 under a single, uniformly applied halo-mass definition. For DF44, re-derive M_200 from the published enclosed mass at 5.1 kpc using an NFW profile with the standard concentration-mass relation instead of a freely cored profile, and also using the cored Burkert profile from the original modelling, recording the change in residual to the Burkert & Forbes line. For Sgr, recompute the residual using the current, tidally stripped halo mass and current GC count rather than the pre-infall values, and separately using the upper and lower pre-infall mass limits. If either galaxy's residual shifts by more than ~0.3 dex under these changes, the conclusion that all five UDGs follow the relation is method-dependent; if both remain within the scatter for all combinations, the concern is retired.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing premise, acknowledged by the authors in Section 2 for Sgr, is that the five quoted halo masses are unbiased measurements of the same present-day virial halo mass used to calibrate the Burkert & Forbes (2020) relation. This is not established. DF44's log M_Halo = 11.2 +/- 0.6 is a total mass obtained by assuming a cored halo profile and extrapolating from kinematic data inside 5.1 kpc; the 0.6 dex uncertainty is larger than the ~0.3 dex scatter of the reference relation, so DF44's support for the relation is weak. Sgr's mass is a pre-infall N-body reconstruction in the range 1-6 x 10^10 Msun, and its GC count is also pre-infall, so the point tests a different epoch and mass definition than the other galaxies. WLM, IC2574, and DDO52 come from HI kinematic fits with different model assumptions, while no correction is applied for halo concentration priors, enclosed-mass radius, or tidal stripping. Because the conclusion that UDGs follow the GC-halo relation rests on all five points sitting inside the scatter, a systematic halo-mass offset of only 0.2-0.3 dex - the plausible size of these method differences - would move DF44 and possibly Sgr outside the relation. The paper's own 'average scatter of 0.19 dex' appears to be the mean absolute residual of the five points rather than a significance test, and with n=5 this is not statistically discriminating.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper compiles a small sample of five low-mass galaxies (three ultra-diffuse galaxies, DF44, WLM, and the Sgr dwarf; and two 'NUDGE' galaxies, IC2574 and DDO52) that have both globular cluster (GC) counts and halo masses determined independently from GC counts. The authors place these objects on the GC number-halo mass relation of Burkert & Forbes (2020) and on the stellar mass-halo mass relation, reporting that all five fall within the scatter of the GC-halo relation (average scatter 0.19 dex) while deviating systematically from the stellar mass-halo mass relation in a GC-count-dependent way. They interpret the most GC-rich objects as 'failed galaxies' with massive halos and suppressed star formation.","tokens_in":8952,"tokens_out":5529,"duration_ms":60548,"significance":"The paper's principal strength is that it uses halo masses derived from kinematic tracers rather than from GC counts, thereby avoiding the circularity that has affected some earlier UDG halo-mass estimates. If correct, the result would provide empirical support for the failed-galaxy scenario and would extend the GC number-halo mass relation to the lowest-mass galaxies with independent mass determinations. The five-object sample is, however, too small and too heterogeneous to bear the weight of the headline claim as it stands, and the statistical treatment is insufficient to distinguish genuine agreement from systematic offsets of order 0.2-0.3 dex.","major_comments":[{"comment":"The statement that 'All five lie well within the scatter of the relation for normal galaxies with an average scatter of 0.19 dex' is not a statistical significance test: the cited 0.19 dex is reported without definition, and with n=5 an average absolute residual of this size cannot distinguish consistency with the ~0.3 dex intrinsic scatter of the Burkert & Forbes (2020) relation from a systematic offset of similar magnitude. Please provide a quantitative assessment, such as residuals in units of the combined (intrinsic-plus-measurement) uncertainty, a chi-square statistic, or a bootstrap probability, and state the assumed intrinsic scatter separately.","section":"Section 3 and Figure 1"},{"comment":"DF44's adopted total halo mass log M_Halo = 11.2 ± 0.6 is an extrapolation from kinematic data inside 5.1 kpc under a cored-halo assumption, rather than a virial mass measured in the same way as the masses used to calibrate the reference relation. Because the quoted uncertainty (0.6 dex) is twice the reference relation's ~0.3 dex scatter, DF44 alone cannot discriminate 'follows the relation' from 'does not follow the relation.' The authors should quantify how the conclusion changes when DF44 is excluded or when its mass is re-derived under alternative profile assumptions.","section":"Section 2, DF44"},{"comment":"The Sgr dwarf is included using a pre-infall halo mass of 1-6 x 10^10 M_sun and a pre-infall GC count of 8-9, both of which refer to an earlier epoch than the present-day virial masses used for the other galaxies. Although the paper acknowledges these parameters are 'highly uncertain,' the central claim that all five points lie within the scatter depends critically on this pre-infall reconstruction. Please either exclude Sgr from the primary sample, demonstrate that the conclusion holds across the full allowed mass range, or justify explicitly why a pre-infall mass is the correct comparison quantity for a present-day scaling relation.","section":"Section 2, Sgr dwarf"},{"comment":"The inclusion of IC2574 and DDO52 as 'reasonable proxies' for UDGs relies on the assumption that both will fade into the UDG regime within ~1 Gyr with no new star formation and no dramatic change in effective radius. This is a plausible but unverified assumption, and it means the sample is not a pure UDG sample. The authors should present these two objects as a separate class (as the 'NUDGE' label suggests) or explicitly discuss how a failure of the fading assumption would alter the conclusions.","section":"Section 2, NUDGE galaxies"},{"comment":"The claim that DF44 and IC2574 'deviate by ~1 sigma and 2 sigma from the scatter of the normal galaxy relation' is not supportable without a defined measure of the intrinsic scatter of the comparison stellar mass-halo mass relation at these masses. Since this deviation is the basis for the failed-galaxy interpretation, the analysis should specify the adopted relation, its scatter, and how the sigma values are computed, so that the reader can reproduce the claim.","section":"Section 3, stellar mass-halo mass relation"}],"minor_comments":[{"comment":"The text contains numerous typographical errors and missing spaces, including 'Aremarkablenear-linear...' in the Introduction, 'IC 2574 ... has R_e = 2.7 kpc and central g band surface brightness of μ ∼ 23 mag/arcsec2. has R_e' (missing 'It') in Section 2, 'higher higher halo masses' in Section 3, and 'massses' in the Conclusions.","section":"Throughout"},{"comment":"The caption states that 'Most uncertainties are smaller than the symbol size,' but DF44's 0.6 dex uncertainty is larger than a typical symbol size; error bars should be shown for all objects or the statement should be qualified to avoid misleading the reader.","section":"Figure 1"},{"comment":"Several citations refer to works listed as 'submitted' or 'in preparation' (e.g., Gannon et al. 2025, submitted; Gannon et al. 2025, in prep). These should be updated to published or accepted versions, or clearly identified as private communications, before final publication.","section":"References"},{"comment":"The sentence beginning 'This fading on a relatively short timescale, and assuming no dramatic reduction in effective radius, suggests both galaxies are a reasonable proxy for a UDG' is grammatically awkward; consider rewriting to make the subject and verb clearer.","section":"Section 2"}],"recommendation":"major_revision","confidential_remarks":"The paper is well-written in substance but the central claim is stated too strongly for a sample of five heterogeneous objects. I recommend major revision as outlined, with particular attention to the statistical treatment and the treatment of DF44 and the Sgr dwarf. If the authors can provide a quantitative significance test and clearly separate the present-day and pre-infall mass definitions, the paper could become a useful contribution, though as a Letter it may remain at the level of a tentative result."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a five-object empirical letter arguing that UDGs follow the GC number–halo mass relation. It's a fair compilation and not circular, but the sample is too small and the halo mass definitions too heterogeneous to establish that conclusion strongly.\n\nThe new bit is the compilation: DF44, WLM, Sgr, IC2574, and DDO52, each with a halo mass from kinematics rather than from GC counts. That avoids the circularity in earlier work that used GC counts to infer masses. Including the MW and Sombrero as anchors is a nice touch, and the authors are upfront about the Sgr dwarf uncertainties. The figures are clear and the writing is honest.\n\nThe soft spots are real. DF44's mass has a 0.6 dex uncertainty, larger than the ~0.3 dex scatter of the reference relation, so that point cannot be doing much work. Sgr's mass and GC count are pre-infall, so it's a different epoch and mass definition. The HI-based masses for the other three come from different model assumptions, and there is no treatment of tidal stripping or concentration priors. The \"average scatter of 0.19 dex\" is the mean absolute residual of five points, not a significance test. With n=5, selection effects alone could produce the same appearance.\n\nThe stress-test note is on the mark: a systematic offset of a few tenths of a dex between mass definitions would move DF44 and possibly Sgr outside the relation. I don't think this kills the paper, but it means the conclusion is suggestive, not demonstrated.\n\nThe paper is still worth a serious referee: the question matters, the method is a step up from earlier work, and the caveats are acknowledged. My own verdict: read it if you work on UDGs, but don't treat it as the final word. I wouldn't cite it as a standalone result, but it could appear in a review alongside other UDG samples.","headline":"Five points with heterogeneous halo masses make a suggestive but not decisive case that UDGs follow the GC number–halo mass relation.","tokens_in":9509,"tokens_out":3173,"would_cite":false,"duration_ms":39066,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Ultra-diffuse galaxies follow the globular cluster–halo mass relation, and the GC-rich among them are likely failed galaxies with massive halos and suppressed star formation.","keywords":["ultra-diffuse galaxies","globular clusters","halo mass","scaling relations","failed galaxies","dwarf galaxies","stellar mass–halo mass relation","GC number–halo mass relation"],"falsifier":"The claim would be undermined by a reanalysis of the Sgr dwarf's pre-infall orbit that places its halo mass outside the $1$–$6\\times10^{10}$ M$_\\odot$ range, moving Sgr off the GC number–halo mass relation; it would also be falsified if a larger sample of UDGs with halo masses from kinematics at large radii (e.g., beyond $3R_e$) showed a systematic offset from the Burkert & Forbes relation rather than scattering about it.","tokens_in":8426,"feed_emoji":"🌌","tokens_out":12010,"duration_ms":105178,"temperature":0.7,"pith_summary":"The paper asks whether ultra-diffuse galaxies (UDGs) obey the same near-linear relation between globular cluster (GC) number and dark matter halo mass that normal galaxies follow. Using five galaxies with halo masses measured independently from kinematic tracers at large radii, it finds that all five lie within the scatter of the GC number–halo mass relation (average scatter 0.19 dex). The same galaxies deviate systematically from the stellar mass–halo mass relation, with the deviation growing with GC count: DF44 and IC2574 scatter to higher halo masses (or lower stellar masses) by about 1$\\sigma$ and 2$\\sigma$. If correct, this means GC-rich UDGs are 'failed galaxies' whose halos assembled but whose star formation was suppressed, and it makes GC count a practical predictor of a UDG's halo mass.","feed_headline":"Ultra-diffuse galaxies follow the globular cluster–halo mass relation","feed_subtitle":"GC-rich ultra-diffuse galaxies follow the GC–halo relation but deviate from the stellar mass–halo relation.","key_machinery":"The load-bearing object is the Burkert & Forbes (2020) GC number–halo mass relation, a near-linear scaling of about $5\\times10^9$ M$_\\odot$ of halo mass per globular cluster that holds from dwarfs to galaxy clusters with roughly 0.3 dex scatter. The paper tests this relation against the stellar mass–halo mass relation using a sample of five galaxies (DF44, WLM, the Sgr dwarf, IC2574, and DDO52) whose halo masses come from independent kinematic tracers reaching large radii, rather than from radial extrapolations within one effective radius. The GC count is the quantity that predicts halo mass and hence the expected position on the stellar mass–halo mass relation; deviation from that relation is measured as a function of GC richness.","core_discovery":"The central discovery is that UDGs and NUDGE galaxies with independent halo mass estimates lie within the scatter of the GC number–halo mass relation for normal galaxies, while their offset from the stellar mass–halo mass relation increases with GC count. DF44 (74 GCs) and IC2574 (27 GCs) deviate by roughly 1$\\sigma$ and 2$\\sigma$ from the stellar mass–halo mass relation, toward higher halo masses or equivalently lower stellar masses at a given halo mass. The paper argues that this supports the 'failed galaxy' scenario for GC-rich UDGs: they occupy massive dark matter halos that quenched star formation early, rather than being puffed-up classical dwarfs.","pith_inferences":["The paper's reasoning implies that a UDG's GC count is a statistical proxy for its halo mass, so a wide-field survey of GC counts around UDGs could map the halo mass function of the UDG population without expensive spectroscopy.","The failed-galaxy scenario predicts that GC-rich UDGs should have old stellar populations and little or no recent star formation; deep photometry or spectroscopy of DF44 and analogues could test this.","Because the two NUDGE galaxies are expected to fade into the UDG regime within about 1 Gyr while keeping their GC counts, the paper implicitly predicts that some blue, star-forming dwarfs today will evolve into GC-rich UDGs, preserving the relation.","If the offset from the stellar mass–halo mass relation grows with GC count, the most GC-rich UDGs (e.g., more than 50 GCs) should occupy halos of roughly $10^{11}$ M$_\\odot$ or more, which is testable with future integral-field kinematic observations."],"forward_implications":["If UDGs follow the GC number–halo mass relation, then GC counts can be used to assign halo masses to UDGs that lack kinematic data, extending the relation to a regime where direct halo mass measurements are scarce.","GC-rich UDGs (e.g., more than 20 GCs) are inferred to have massive halos and low stellar masses, implying that they are 'failed galaxies' whose star formation was quenched early, with later assembly times.","The systematic offset of IC2574 and DF44 from the stellar mass–halo mass relation, at roughly 1$\\sigma$ and 2$\\sigma$, predicts that a larger sample of GC-rich UDGs will show a similar trend of increasing deviation with GC count.","The Milky Way and Sombrero, which sit on the same relation with independent halo masses, show that the GC number–halo mass relation is not restricted to UDGs and that UDGs are not exceptional in their GC–halo scaling."],"supporting_citations":[{"why":"supplies the reference GC number–halo mass relation and its scatter used throughout the analysis.","marker":"Burkert & Forbes (2020)"},{"why":"provides DF44's rising velocity dispersion profile and total halo mass of log M_halo = 11.2 ± 0.6 from kinematics to 5.1 kpc.","marker":"van Dokkum et al. (2019)"},{"why":"models DF44's kinematics to support a cored halo and the same total halo mass.","marker":"Wasserman et al. (2019)"},{"why":"gives WLM's halo mass of 8.3 ± 2.2 × 10^9 M_sun from HI rotation.","marker":"Read et al. (2016)"},{"why":"gives DDO52's halo mass of log M_halo = 10.08 ± 0.10 M_sun.","marker":"Read et al. (2017)"},{"why":"provides the Sgr dwarf's pre-infall halo mass range of 1–6 × 10^10 M_sun from N-body reconstruction.","marker":"Dierickx & Loeb (2017)"},{"why":"gives the Sgr dwarf's pre-infall GC count of 8 or 9.","marker":"Forbes (2020)"},{"why":"supplies IC2574's GC count (27 ± 5), halo mass (log M_halo = 10.93 ± 0.08), and stellar mass (log M* = 8.39).","marker":"Karim et al. (2024)"},{"why":"provides the favoured DF44 GC count of 74 ± 18 and introduces the NUDGE classification used here.","marker":"Forbes & Gannon (2024)"},{"why":"catalogues Local Group galaxies meeting the UDG definition, including WLM and the Sgr dwarf.","marker":"Gannon et al. (2024)"}],"fun_headline_variants":["UDGs follow globular cluster–halo relation, not stellar mass–halo","GC-rich UDGs obey halo mass relation, not stellar mass","Ultra-diffuse galaxies trace globular clusters, not stars","GC-rich UDGs may be failed galaxies, following halo mass"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the halo masses taken from the literature for the five galaxies are unbiased estimates of the same virial halo mass used in the Burkert & Forbes (2020) relation, even though the estimates come from different methods (stellar kinematics within 5.1 kpc, HI rotation, kinematic fitting, models, and a pre-infall N-body reconstruction) and the Sgr dwarf's pre-infall parameters are acknowledged as highly uncertain.","fun_headline_variants_meta":{"raw":{"variants":["UDGs follow globular cluster–halo relation, not stellar mass–halo","GC-rich UDGs obey halo mass relation, not stellar mass","Ultra-diffuse galaxies trace globular clusters, not stars","GC-rich UDGs may be failed galaxies, following halo mass"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000793,"raw_usage":{"total_tokens":3451,"prompt_tokens":864,"completion_tokens":2587,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":480,"completion_tokens_details":{"reasoning_tokens":2508}},"tokens_in":480,"tokens_out":2587,"duration_ms":22425,"temperature":1.0,"reasoning_tokens":2508,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T13:20:54.777980+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"The claim would be undermined by a reanalysis of the Sgr dwarf's pre-infall orbit that places its halo mass outside the $1$–$6\\times10^{10}$ M$_\\odot$ range, moving Sgr off the GC number–halo mass relation; it would also be falsified if a larger sample of UDGs with halo masses from kinematics at large radii (e.g., beyond $3R_e$) showed a systematic offset from the Burkert & Forbes relation rather than scattering about it.","supporting_citations":[],"review_version":1}