{"id":"80e255e6-f942-4465-a94d-30c33ebc4e55","arxiv_id":"2512.04156","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Kinematics of three ultra-faint Milky Way satellites—Segue 1, Triangulum II, Tucana V—rule out warm dark matter lighter than ~5.8 keV at 95% confidence.","lead":"Warm dark matter is expected to leave heavy, dense 'prompt cusps' at the centers of tiny galaxies. Comparing simulated cusps to the observed motions of three ultra-faint dwarfs puts a new, independent lower limit on the dark matter particle mass of about 5.8 keV.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The m_chi>5.8 keV exclusion depends on the prompt-cusp amplitude A of infalling subhalos: a 20% assembly-bias offset plus a lognormal scatter, calibrated only on power-law simulations and not natively produced by Galacticus. An error here directly shifts the predicted v_circ(r_h) tails and the resul","rationale":"I read the paper in good faith. The central claim is plausible and the implementation is a real step: prompt cusps are physically motivated, the cusp-NFW profile is explicit, the analysis is transparent about its limitations, and several known omissions (tidal heating of stars, subhalo concentration assembly bias) would strengthen rather than weaken the limit. The paper also provides code/data references and a calibration appendix. However, the single most load-bearing element is the prompt-cusp amplitude A for subhalos: v_circ(r_h) is directly controlled by A at these radii, and the two corrections that set the shape and tail of the A distribution — the 20% assembly-bias offset and the scatter — are calibrated on the lead author's power-law simulations, not independently verified for WDM. The paper itself flags that Galacticus does not naturally produce the assembly bias, so the median A is not securely anchored. If this calibration is off, the p-values underlying the m_chi>5.8 keV limit will shift. The conditional verdict is appropriate: the constraint should not be treated as a firm exclusion until the A calibration is checked, e.g., by evaluating A at infall and/or measuring subhalo cusp scatter in a WDM simulation. I do not see a reason to reject or upgrade; the concern is a calibration/verification issue, not an internal inconsistency.","tokens_in":14720,"tokens_out":12685,"duration_ms":125600,"concrete_test":"Re-run the full Section 5 analysis with prompt-cusp coefficients drawn from the field-halo cusp-halo relation at (M_ifl,z_ifl) — the dotted distributions in Fig. 5 — instead of the Galacticus (M_res,z_res) assignment, keeping all galaxy-formation, tidal, and observational inputs fixed. If the resulting combined 95% lower limit on m_chi drops by more than ~0.5 keV or falls below 5 keV, the claimed exclusion is not robust to the subhalo assembly-bias calibration; if it remains consistent, that correction is not the deciding factor.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central predicted observable, v_circ(r_h), is set at radii where the cusp-NFW density profile is dominated by the prompt cusp A r^-1.5. The model assigns A from the Delos (2025) cusp-halo relation at the resolution-crossing mass M_res=1e7 M_sun, then relies on two simulation-based corrections: newly infalling subhalos have median A ~20% higher than same-mass field halos, and A is lognormally scattered with sigma given by Eq. A2 (Sections 3 and 4.1, Appendix A). Figure 5 shows this raises the Galacticus A distribution by ~20% and narrows it relative to using (M_ifl,z_ifl). The paper itself notes (footnote 9) that assembly bias does not emerge naturally in Galacticus, so the median A being correct 'is not clear.' Appendix A calibrates both the 20% offset and the scatter on self-similar power-law simulations (n=-2.67,-2,1), not on WDM power spectra, using limited mass bins and no independent check. The 95% limit is set by how improbable the low observed v_circ of Tri II and Tuc V are in the WDM tail; a 20% median A shift or a larger true scatter would raise those p_i values and lower the m_chi limit. This is the load-bearing assumption of the central claim.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a new test of warm dark matter (WDM) using the central density cusps predicted to form in all dark matter halos. The authors implement the cusp-halo relation of Delos (2025) in the semi-analytic galaxy formation model Galacticus, generate Milky Way-like satellite populations for WDM masses of 3–40 keV and CDM, and compare the predicted circular velocities at the half-light radius with the observed kinematics of the three smallest/faintest confirmed Milky Way satellites: Segue 1, Triangulum II, and Tucana V. They find that Tri II and Tuc V have observed v_circ(r_h) values that are unusually low compared to WDM predictions, while Segue 1 is consistent with WDM. Combining the three systems with a Fisher-type chi^2 statistic, they derive m_chi > 5.8 keV at 95% confidence and m_chi > 9.4 keV at 90% confidence for thermal-relic WDM, competitive with existing limits from satellite abundances and lensing. The central claim is that the prompt-cusp enhancement of small-scale structure provides a new, powerful probe of the dark matter particle mass.","tokens_in":15052,"tokens_out":6773,"duration_ms":60226,"significance":"If the result holds, this is a novel and competitive WDM constraint based on an enhancement (prompt cusps) rather than a suppression of small-scale structure. The method is conceptually distinct from abundance and lensing probes and could be sharpened with better kinematics of the same or additional ultrafaint dwarfs. The paper is unusually transparent: it explicitly acknowledges that Galacticus half-light radii are about half the observed values, that tidal heating is not applied to stars, that concentration assembly bias is omitted, and that the subhalo assembly-bias calibration in Appendix A rests on power-law simulations. These acknowledged limitations are mostly in the conservative direction (they would strengthen the limit), which lends credibility. The code and model version are publicly referenced, and the analysis uses empirical posterior distributions for the key kinematic inputs rather than simple Gaussian approximations. However, the central exclusion depends sensitively on the adopted prompt-cusp amplitude and scatter for subhalos, and the statistical combination needs more justification. With robustness tests, the paper could provide a landmark constraint; in its curr","major_comments":[{"comment":"The central exclusion m_chi > 5.8 keV is driven by the predicted lower tail of v_circ(r_h) for Tri II and Tuc V, which in turn is set by the prompt-cusp amplitude A assigned to subhalos. The model uses a ~20% assembly-bias boost (Fig. 9) and a lognormal scatter with sigma given by Eq. (A2), calibrated on self-similar power-law simulations (n = -2.67, -2, 1) rather than on WDM power spectra. Footnote 9 concedes that assembly bias does not emerge naturally in Galacticus and that the median A 'is not clear' to be physically correct. Because a 20% shift in the median A or a 50% increase in scatter would directly move the predicted v_circ distributions and hence the 95% limit, I ask for a sensitivity analysis (e.g., vary the boost from 0% to 40% and vary the scatter by a factor of 2) or a direct validation on a WDM simulation. Without such a test, the robustness of the headline limit to the m","section":"Secs. 3, 4.1, and Appendix A"},{"comment":"The combined probability uses Fisher's method, chi^2 = -2 sum ln p_i with 6 degrees of freedom, which assumes the p_i are independent and uniformly distributed under the null. Here the p_i are posterior predictive p-values computed after conditioning on the observed M_V, r_h, and r_p, and using a model distribution estimated from only 10 Milky Way-like Galacticus realizations with resampling (about 100–2000 unique analogues per galaxy). These p-values are not guaranteed to be uniform: the limited effective sample size and the conditioning on observed properties can produce non-uniformity. Since the 95% and 90% limits are read directly from the combined curve in Fig. 7, the authors should calibrate the null distribution of the combined statistic (e.g., by drawing mock galaxies from the model and repeating the full procedure) or adopt a more conservative combination. This is load-bearing f","section":"Sec. 5, Eq. (3)"}],"minor_comments":[{"comment":"The statement that subhalo cusps are 'about 20 percent higher' is an oversimplification of Fig. 9, where the bias varies substantially with mass and time (roughly 1.0–1.8). Please specify the mass/redshift range relevant to the Segue 1, Tri II, and Tuc V analogues when quoting this number.","section":"Sec. 4.1, Fig. 5"},{"comment":"The caveat that assembly bias does not emerge naturally from Galacticus and that the median A 'is not clear' to be physically correct is a significant modeling limitation. This should be stated more prominently in Section 6 rather than in a footnote, given its direct impact on the result.","section":"Sec. 6 / Footnote 9"},{"comment":"The adopted log-normal distribution for Tuc V's sigma_los is said to closely match Hansen et al. (2024). Please cite the specific figure or table in that work, since different summary statistics (e.g., median vs. mode) can shift the inferred v_circ.","section":"Sec. 2, Tuc V"},{"comment":"The abstract says 'three faintest Milky Way satellites,' but the selection is based on both faintness and small size; Segue 1, Tri II, and Tuc V are not strictly the three faintest in absolute magnitude. Consider phrasing such as 'three smallest and faintest confirmed satellites' for precision.","section":"Abstract and Sec. 1"},{"comment":"The label 'P(> sigma_los)' is a survival function; it would be clearer to write 'P(sigma_los > x)' or to state that it is the complementary CDF.","section":"Fig. 3 caption"}],"recommendation":"major_revision","confidential_remarks":"The paper relies heavily on the lead author's prior work (Delos 2023, 2025) for the cusp-halo relation and prompt-cusp physics. This is natural in a fast-moving subfield, but the Appendix A calibration is central and should be treated as a key claim requiring direct validation. The statistical combination is also simpler than ideal; a calibration test would materially increase confidence. The authors are transparent and the overall direction is promising. I recommend major revision rather than rejection because the issues are addressable with sensitivity tests and a null-distribution check."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The new thing here is real: the first implementation of prompt cusps in a full semi-analytic galaxy formation model, and an observational comparison of the three smallest Milky Way satellites that yields a competitive WDM limit (m_chi > 5.8 keV at 95%) from a different physical effect than abundance or lensing. The paper is honest about its limitations, and several of the known corrections would strengthen rather than weaken the limit. That alone makes it worth reading and citing.\n\nWhat the paper does well: it clearly explains the physical mechanism, uses a well-established model (Galacticus) with a published galaxy formation prescription, and is transparent about the mismatch between modeled and observed half-light radii, the neglect of tidal heating on stars, and the absence of concentration assembly bias. The code and data references are given, and the authors flag places where their own modeling is uncertain, including footnote 9 where they admit it is not clear that the median cusp amplitude is physically correct.\n\nWhere I would push back: the load-bearing input is the prompt-cusp coefficient A, assigned to subhalos at the resolution threshold using a cusp-halo relation from the lead author's prior work, then boosted by ~20% and scattered using calibrations from power-law simulations in Appendix A. That is not independently verified, and the scatter correction is tuned to match a simulated distribution rather than predicted from Galacticus. The stress-test note is right that an error in this step directly shifts the predicted v_circ tails and the resulting limit. That said, the likely direction of the correction—if subhalos are actually more concentrated than modeled—is to raise v_circ and strengthen the exclusion, so the central claim is probably conservative. The statistical combination via a chi^2 with 6 degrees of freedom is oversimplified—the p_i values are not independent Gaussians, and there is a look-elsewhere element in picking the three faintest galaxies. But that mostly affects the exact confidence level, not the qualitative conclusion that Tri II and Tuc V are outliers in light WDM.\n\nOverall: a solid, careful paper with one calibrated-from-the-author's-own-simulations ingredient that deserves scrutiny. It is not a desk reject. Send it to a referee who knows both semi-analytic merger trees and cusp formation, and ask for a sensitivity analysis of the A calibration and a more defensible combined statistic. The paper is a serious contribution and I would cite it if my work touches WDM limits.","headline":"A genuinely new and plausible WDM limit from prompt cusps in the faintest dwarfs, but the calibration of the subhalo cusp amplitude is the main thing to stress-test before trusting the number.","tokens_in":15576,"tokens_out":1292,"would_cite":true,"duration_ms":14752,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["95.35.+d"],"model":"deepseek-v4-flash","headline":"The kinematics of the three faintest Milky Way satellite galaxies require thermal-relic warm dark matter to be heavier than 5.8 keV at 95 percent confidence; Triangulum II and Tucana V are the decisive outliers.","keywords":["warm dark matter","prompt cusps","Milky Way satellite galaxies","dwarf galaxy kinematics","velocity dispersion","thermal relic","cusp-halo relation","dark matter mass limit"],"falsifier":"Calculate the prompt-cusp coefficients of subhalos at their time of infall in a high-resolution warm-dark-matter cosmological simulation; if the median coefficient exceeds the field-halo value by substantially less than 20 percent at the relevant masses and redshifts, the predicted v_circ(r_h) distributions would be lower and the derived mass limit would be overstated.","tokens_in":1628,"feed_emoji":"🔭","tokens_out":1557,"duration_ms":64059,"temperature":0.7,"pith_summary":"This paper argues that the measured orbital speeds of stars in the three faintest known Milky Way satellite galaxies—Segue 1, Triangulum II, and Tucana V—are direct evidence against light warm dark matter. In warm dark matter models, every dark matter halo forms with a dense central cusp whose strength is set by the particle's free-streaming scale; heavier particles produce weaker cusps. Using a semi-analytic galaxy-formation model that implements these 'prompt cusps', the authors generate theoretical counterparts of the three galaxies and compare predicted circular velocities at the half-light radius with observed kinematics. The measured velocities of Triangulum II and Tucana V fall far below the warm-dark-matter predictions, yielding a 95 percent confidence lower bound of 5.8 keV on the thermal-relic particle mass (and 9.4 keV at 90 percent confidence). The result matters because it constrains dark matter using a predicted enhancement of small-scale structure, rather than the usual suppression of galaxy abundance, and it is already competitive with the strongest existing limits.","feed_headline":"Kinematics of three tiny galaxies put warm dark matter above 5.8 keV","feed_subtitle":"Three ultrafaint satellite galaxies orbit too slowly for warm dark matter lighter than 5.8 keV; more precise stellar velocities would raise","key_machinery":"The central object is the prompt cusp: a density profile rho(r) = A r^-1.5 that forms in every dark matter halo at the free-streaming scale. The cusp coefficient A is set by the cusp-halo relation, which links A to the halo's mass and formation time; the paper assigns A to each simulated subhalo at the moment its mass crosses the resolution threshold, includes a lognormal scatter calibrated to simulations, and applies a ~20 percent upward bias for subhalos about to fall into a larger host. The density profile is then evolved using a cusp-NFW form with tidal heating, and the key observable is the circular velocity at the half-light radius, v_circ(r_h), estimated from observed line-of-sight ve","core_discovery":"The central claim is that the central density cusps—'prompt cusps'—that warm dark matter produces in every small halo are massive enough to leave a detectable kinematic signature in the smallest galaxies, and that the measured velocity dispersions of Triangulum II and Tucana V are serious outliers for warm dark matter with particle masses below roughly 6–9 keV. Embedding a cusp-halo relation that ties cusp strength to halo mass and formation time into a full model of Milky Way satellite populations, the paper finds that for a 10 keV thermal-relic particle the predicted circular velocity at the half-light radius is substantially higher than in cold dark matter, and that the observed low value","pith_inferences":["We infer that the same prompt-cusp test could be applied to other dark matter models that suppress small-scale power, such as sterile neutrinos or interacting dark matter, provided the cusp-halo relation can be recalibrated for those models.","We infer that the 20 percent assembly-bias correction to subhalo cusp coefficients is the single most important calibration uncertainty; a direct simulation resolving prompt cusps in infalling subhalos would be the most straightforward way to test whether the derived limit is overstated.","We infer that the contrast between Segue 1 (orbiting faster than cold-dark-matter predictions) and the other two galaxies could, if the cold-dark-matter baseline is correct, provide a separate probe of tidal disruption and orbital histories of ultrafaint dwarfs.","We infer that the statistical power of this method will grow faster than linearly with sample size, since each new compact ultrafaint galaxy adds an independent draw from a distribution that is sharply peaked in warm-dark-matter models."],"forward_implications":["If the limit holds, it joins satellite-abundance and strong-lensing constraints as one of the strongest existing bounds on thermal-relic warm dark matter, while resting on a fundamentally different observable signature.","Improving velocity dispersion measurements of Segue 1, Triangulum II, and Tucana V could substantially sharpen the mass bound, because the predicted difference from cold dark matter remains visible even at particle masses around 20 keV for these compact galaxies.","Discovering and kinematically characterizing more galaxies as compact and faint as these would allow the constraint to be pushed toward higher masses with a modest number of additional systems.","The model predicts that tidal stripping suppresses circular velocities most strongly for low-pericenter systems like Triangulum II, making such galaxies especially discriminating between warm and cold dark matter.","Accounting for assembly bias in subhalo concentrations would raise the predicted circular velocities and therefore strengthen the warm-dark-matter limit, as the paper explicitly notes."],"fun_headline_variants":["Ultrafaint galaxies push warm dark matter past 5.8 keV","Tiny galaxies' slow stars rule out light warm dark matter","Dwarf galaxies set lower bound on warm dark matter mass","Faintest galaxies tighten warm dark matter mass limit"],"cache_read_input_tokens":16768,"weakest_assumption_plain":"The limit rests on the calibration that newly infalling subhalos carry prompt cusps about 20 percent denser than typical field halos of the same mass, with a lognormal scatter correctly captured by the model; if this subhalo cusp bias or scatter is overestimated, the predicted circular velocities would be too high and the derived mass limit would be too strong.","fun_headline_variants_meta":{"raw":{"variants":["Ultrafaint galaxies push warm dark matter past 5.8 keV","Tiny galaxies' slow stars rule out light warm dark matter","Dwarf galaxies set lower bound on warm dark matter mass","Faintest galaxies tighten warm dark matter mass limit"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000167,"raw_usage":{"total_tokens":1062,"prompt_tokens":677,"completion_tokens":385,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":421,"completion_tokens_details":{"reasoning_tokens":315}},"tokens_in":421,"tokens_out":385,"duration_ms":3882,"temperature":1.0,"reasoning_tokens":315,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T18:37:41.364495+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Calculate the prompt-cusp coefficients of subhalos at their time of infall in a high-resolution warm-dark-matter cosmological simulation; if the median coefficient exceeds the field-halo value by substantially less than 20 percent at the relevant masses and redshifts, the predicted v_circ(r_h) distributions would be lower and the derived mass limit would be overstated.","supporting_citations":[],"review_version":1}