{"id":"deb7817f-a3dd-488e-b34b-19eb1e7131ed","arxiv_id":"2601.17113","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Bursty and time-extended star formation histories make simulated dwarf galaxies more likely to develop flat dark matter cores, and adding a star-formation-history ratio improves the fit to the inner-slope versus stellar-mass relation.","lead":"The paper compares two galaxy simulation suites to test whether the timing of star formation changes dark matter density in galaxy centers. It finds that bursty or extended star formation tends to flatten inner dark matter profiles, and offers a new formula using star-formation history to improve slope predictions.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Inner slopes measured at 1–2% Rvir lack resolution checks and per-galaxy uncertainties; if unresolved systems bias the slopes, both the SFH correlations and the Eq. 6 improvement could be numerical artifacts.","rationale":"The reader's weakest assumption is exactly that the 1–2% Rvir slope is converged and resolution-independent. This is the most load-bearing point because it underlies all of the paper's conclusions: the homogeneous comparison between NIHAO and FIRE-2, the correlations with burstiness and SFH deviation, and the quantitative improvement from Eq. 6. The paper provides no explicit test of this assumption for the main slope measurement, and Appendix B shows that resolution limits do affect at least one related measurement (150 pc densities in some FIRE-2 galaxies). The proposed test directly checks whether the trends survive when only clearly resolved galaxies are used, and would therefore settle whether the central claim is robust or a numerical artifact. The verdict remains CONDITIONAL, as the paper would need to include such a test to support its claims fully.","tokens_in":15615,"tokens_out":5919,"duration_ms":66120,"concrete_test":"Recompute the full analysis (Figs. 1,3,5,7 and §3.3 MSE values) after excluding every galaxy for which 0.01 Rvir is smaller than max(r_200, 3×gravitational softening length), where r_200 is the radius enclosing 200 dark-matter particles (defined in Appendix B). If the burstiness/SFH-deviation trends and the Eq. 6 improvement persist with similar magnitude and scatter on the retained sample, the results are robust to resolution; if they weaken or disappear, the central claim is not supported. As a secondary check, bootstrap the density profile of each galaxy to assign a slope uncertainty and test whether the split-sample offsets are significant.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 3 defines the inner slope as a linear fit over 0.01–0.02 Rvir, using 35 bins out to Rvir. The paper provides no convergence test, no requirement that this radial window is resolved (e.g., >r_200 or >3×softening) for each galaxy, and no slope uncertainties. This matters because the sample spans M* ~1e5–1e11 Msun, so 0.01 Rvir ranges from tens of pc in dwarfs to kpc in massive halos, while resolution also varies. Appendix B admits that for some FIRE-2 galaxies even 150 pc lies below r_200 and requires extrapolation; the analogous risk for the 1–2% Rvir window is never assessed. If unresolved or marginally resolved slopes are biased, and if the resolved fraction correlates with M*/Mvir or with SFH (low-mass dwarfs tend to be burstier), then the split-sample trends in Figs. 3 and 5 and the MSE gains in §3.3 could reflect numerical resolution rather than physics. This is the single load-bearing concern because every claim—homogeneous NIHAO/FIRE-2 agreement, burstiness/SFH effects, and Eq. 6—passes through this slope measurement.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper analyzes 93 NIHAO and 109 FIRE-2 zoom-in galaxies, measuring inner dark-matter logarithmic slopes between 1% and 2% of R_vir and characterizing star formation histories via a bursty mass fraction (defined by SFR(50 Myr) > 1.5 SFR(500 Myr)) and the post-to-pre reionisation stellar mass ratio M_post/M_pre. It argues that, after homogenizing profile construction and overdensity definition, the two simulation suites agree much better than previously reported, and that residual burstiness and SFH concentration at fixed stellar mass correlate with cored versus cuspy inner profiles. It presents Eq. (6), which adds an M_post/M_pre correction to the standard inner-slope fitting formula, and reports lower mean squared errors in both suites. It also reproduces Muni et al.'s 150 pc density test for FIRE-2 and offers a falsifiable prediction for M* ~ 10^8 M_sun galaxies.","tokens_in":15958,"tokens_out":5237,"duration_ms":51342,"significance":"If the central claims hold, this is a valuable contribution: it demonstrates that previous NIHAO/FIRE-2 differences were partly due to differing overdensity definitions and binning, it identifies a plausible second parameter (burstiness / SFH timing) that reduces scatter in the inner-slope relation, and it gives a concrete observational prediction tied to resolved stellar populations. The homogeneous re-analysis and the explicit reproduction of Muni et al.'s EDGE test are strengths. However, the quantitative claims are not yet established: the inner-slope measurement lacks resolution validation, and the reported improvement of Eq. (6) is in-sample and, for NIHAO, effectively a constant offset rather than an SFH-dependent term. The qualitative correlation between bursty/extended SFHs and cores is plausible, but the paper currently overstates the predictive gain.","major_comments":[{"comment":"The inner slope is a linear fit to the logarithmic density profile between 0.01 and 0.02 R_vir, using 35 bins down to 0.005 R_vir. The paper never establishes that this window is resolved for every galaxy: there is no requirement such as r_inner > r_200 or > 3× softening, no convergence test, and no per-galaxy slope uncertainties. The sample spans M* ~ 10^5–10^11 M_sun, so 0.01 R_vir ranges from tens of pc in dwarfs to kpc in massive halos. Appendix B admits that for some FIRE-2 galaxies even 150 pc lies below the r_200 resolution limit and requires extrapolation; the analogous risk for the 1–2% R_vir window is not assessed. If unresolved or marginally resolved systems are biased toward shallower slopes, and if the resolved fraction correlates with M*/Mvir or with SFH (low-mass dwarfs being the burstiest), then the split-sample trends in Figs. 3 and 5 and the MSE gains in §3.3 could be n","section":"§3, Figs. 1/3/5, Appendix B"},{"comment":"The claimed improvement in prediction is in-sample: Eq. (6) is fitted and evaluated on the same galaxies, so a lower MSE is expected merely from adding parameters (n2, α) and removing an outlier. No cross-validation or held-out sample is presented. More seriously, the NIHAO fit gives α = 2.06×10^-3; over any plausible M_post/M_pre range, y^α ≈ 1, so the correction term is nearly a constant offset. The NIHAO MSE reduction therefore cannot be attributed to an SFH dependence of the slope. The FIRE-2 value α = 0.14 is also weak, and neither fit reports parameter uncertainties or covariance. Please add out-of-sample validation, report the actual dynamic range of y^α, and show the sensitivity of the results to the removed outlier.","section":"§3.3, Eq. (6), Table 4"},{"comment":"The 'burst deviation' and 'SFH deviation' are residuals from second-order polynomial fits to the same galaxies that are then split into positive/negative deviation groups for slope fits. This construction guarantees that the two groups differ by construction; it can create apparent secondary correlations when the underlying relation already has mass-dependent scatter. To support the claim that burstiness and SFH concentration explain scatter at fixed M*/Mvir, the authors should either use independent variables (e.g., direct SFR metrics not de-trended against M*) or demonstrate with cross-validation that the split-sample fits predict slopes of galaxies not used in the polynomial fit. As written, the explanatory claim is partly in-sample.","section":"§3.1–3.2, Figs. 3 and 5"}],"minor_comments":[{"comment":"Duplicate word in 'the temporal temporal concentration of SFHs'.","section":"§4"},{"comment":"'median relative deviation over 0.15%' conflicts with the earlier 'values above ~0.15'; if 0.15 is a fraction, it should be written as 0.15 or 15%.","section":"Appendix B"},{"comment":"No indication of the distribution of y=M_post/M_pre is given. Please show the range and percentiles for each suite, otherwise the plotted curves may suggest extrapolation beyond the data.","section":"Fig. 7"},{"comment":"The FIRE-2 (<0) fit has β=348, clearly an unconstrained/degenerate parameter. Report parameter uncertainties or use a more robust fitting procedure for the split samples.","section":"Table 2"},{"comment":"The equation display does not define x and y; x=M*/Mvir and y=M_post/M_pre are given in the text but should be repeated in the equation caption or below the display.","section":"Eq. (6)"},{"comment":"The sentence about the NIHAO peak value rounding to the same value at one significant digit is confusing; report more significant digits.","section":"Appendix A"}],"recommendation":"major_revision","confidential_remarks":"This is a worthwhile paper for A&A after a major revision. The main concern is the unresolved/marginally resolved inner-slope measurement; a resolution test and per-galaxy uncertainties are essential. The in-sample MSE comparison and the near-constant NIHAO correction term also need to be addressed. I see no grounds for rejection, but the current quantitative claims overstate what is demonstrated."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Hi [colleague],\n\nThe short version: I buy the qualitative story—galaxies with burstier, more temporally extended star formation tend to have shallower inner dark matter slopes at fixed stellar-to-halo mass ratio—but I don't buy the specific quantitative claim that Eq. 6 improves predictions. The improvement is measured in-sample on the same galaxies used to fit it, and the NIHAO correction term is nearly constant, so for that suite Eq. 6 is effectively just a renormalized version of Eq. 2.\n\nWhat the paper does well is the homogeneous reanalysis of NIHAO and FIRE-2. By using the same virial definition, binning, and slope measurement, they show the two suites agree much better than the literature comparison suggested (Tollet et al. 2016 vs Lazar et al. 2020). That is a real contribution and worth citing. The split-sample fits in Figs. 3 and 5 are suggestive: galaxies with above-average burstiness or higher post/pre-reionisation mass ratios sit on shallower relations. The effect is in the direction you'd expect from Di Cintio et al. (2017) and Muni et al. (2024), so the paper is extending an existing idea rather than introducing a new mechanism.\n\nThe soft spots are real. First, no cross-validation. The MSE drop from 0.073 to 0.055 (FIRE-2) and 0.063 to 0.048 (NIHAO) is computed on the training set. With 6-7 free parameters in Eq. 6, that drop is expected. Second, the NIHAO fit has alpha=2e-3, making y^alpha essentially constant over the plotted range; the correction is a constant offset, not a secondary predictor. That parameterization looks degenerate. Third, the slope measurement at 1-2% Rvir has no convergence test and no per-galaxy uncertainty. Appendix B admits that for some FIRE-2 galaxies even 150 pc is below the r_200 resolution limit and requires extrapolation. If the 1-2% Rvir window is unresolved for the lowest-mass galaxies, and those galaxies also tend to be burstier, then the correlations in Figs. 3 and 5 could be partly numerical. The paper needs to show the slopes are converged, or at least that the correlation survives when only resolved galaxies are used.\n\nThere are also some unstable split-sample fits (e.g., beta=348 for FIRE-2 below-average SFH) with no error bars on the parameters. That's minor if the qualitative trend in Fig. 5 is robust, but it makes me wary of overinterpreting the shape of the relation.\n\nSo: send to peer review. A good referee can push for out-of-sample or cross-validated error estimates, resolution checks, and parameter uncertainties. The core idea is sound and the homogenized comparison is valuable; the quantitative claims need to be downsized or properly validated.\n\nBest","headline":"Plausible qualitative result, but the quantitative improvement claim needs out-of-sample validation and resolution checks.","tokens_in":16480,"tokens_out":3308,"would_cite":true,"duration_ms":34725,"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":"Star formation histories, not just galaxy mass, set whether dark matter centers are cored or cuspy.","keywords":["star formation histories","dark matter cusp/core","inner density slope","galaxy simulations","stellar feedback","burstiness","reionization","stellar-to-halo mass ratio"],"falsifier":"A concrete test would be to recompute the same inner slopes from the same simulated galaxies at two or more resolution levels (or with different binning and centering choices) and check whether the burstiness and SFH-deviation trends survive; alternatively, an observational survey of dwarf galaxies with well-measured star formation histories and dark matter cores could check whether early-concentrated star formation systematically correlates with cusps, as the paper predicts. If the correlation vanishes or reverses, the claim is falsified.","tokens_in":15530,"feed_emoji":"🌌","tokens_out":3943,"duration_ms":42207,"temperature":0.7,"pith_summary":"This paper argues that the shape of a galaxy's inner dark matter profile—whether it has a steep cusp or a flat core—depends on more than how much stellar mass it has relative to its halo. By comparing two large sets of simulated galaxies analysed with identical methods, the authors show that galaxies with burstier and more temporally extended star formation histories develop shallower dark matter slopes at fixed stellar-to-halo mass ratio. They distill this into a new two-variable fitting formula that predicts the inner slope more accurately than the standard one-parameter relation. If true, this resolves apparent discrepancies between simulation suites and gives observers a direct prediction: galaxies with old, early-peaked stellar populations should retain cusps, while those with extended, late star formation should host cores.","feed_headline":"Bursty star formation flattens dark matter cusps into cores","feed_subtitle":"Galaxies with extended, clumpy star formation develop shallower dark matter profiles at fixed mass, a new two-variable fit shows.","key_machinery":"The central objects are the burstiness metric—the fraction of stellar mass formed in starburst phases, defined by comparing star formation rates averaged over 50 and 500 Myr—and the post-to-pre reionization stellar mass ratio M_post/M_pre, which measures how temporally concentrated star formation is relative to the reionization epoch at z~6.5. Armed with these, the paper defines 'burst deviation' and 'SFH deviation' (residuals from the mass trends) and shows they order the scatter in the inner slope relation. The final fitting formula (Eq. 6) combines the stellar-to-halo mass ratio x with a power of y=M_post/M_pre to predict the inner slope.","core_discovery":"The central discovery is that the inner logarithmic slope of the dark matter density profile, measured between 1% and 2% of the virial radius, correlates with the star formation history in a way that is independent of the stellar-to-halo mass ratio. Galaxies whose star formation rate is burstier than average for their mass—quantified by a bursty mass fraction—and galaxies that formed a larger fraction of their stars after reionization (high M_post/M_pre) systematically have shallower (more cored) profiles. The paper packages this in a modified fitting formula (their Eq. 6) that adds a term proportional to a power of M_post/M_pre to the existing mass-ratio fit, cutting the mean squared predic","pith_inferences":["The burstiness metric relies on a specific threshold (1.5 times the ratio of SFR averaged over 50 Myr vs 500 Myr); the paper does not test sensitivity to this choice, so a natural extension is to verify with other burst definitions or timescales.","If the trend holds, the scatter in observed core/cusp properties of dwarf galaxies could be used as a probe of their star formation histories, even where direct SFH measurement is difficult.","The correlation between burstiness and M_post/M_pre may mean the two metrics are not independent; disentangling their relative importance (e.g., with galaxies that are bursty but early-forming) could sharpen the causal claim.","The paper's exclusion of a small subsample lacking pre-reionization stars and one outlier in the fitting leaves open how universal the formula is for extremely low-mass galaxies."],"forward_implications":["At fixed stellar-to-halo mass ratio, galaxies with burstier star formation develop shallower dark matter cores.","Galaxies with more extended SFHs (higher post-to-pre reionization ratio) are more efficient at creating cores; earlier-concentrated star formation preserves cusps.","The new two-variable fitting formula (Eq. 6) reduces the mean squared error of inner-slope prediction by roughly one-quarter in both simulation suites (FIRE-2: 0.073 to 0.055; NIHAO: 0.063 to 0.048).","The apparent disagreement between previous NIHAO and FIRE-2 inner-slope trends largely disappears when analysis is homogenised (same halo definition, binning), implying that much of the scatter is physical rather than numerical.","Observational prediction: among galaxies near the peak core-formation mass (M_star~10^8 Msun), those with early-peaked stellar age distributions should retain cusps, while extended, late star formation should correlate with large cores; finding large cores in early-forming dwarfs would challenge feedback-driven core formation."],"fun_headline_variants":["Bursty star formation flattens dark matter cusps","Star formation history shapes dark matter cores","Post-reionization stars turn cusps into cores","Clumpy star growth predicts dark matter slopes","New fit ties star burstiness to halo profiles"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The paper's central claim rests on the assumption that the inner slope measured by a linear fit to the density profile between 1% and 2% of the virial radius is a converged, resolution-independent measure of dark matter cusp/core structure; no convergence test is given, and Appendix B notes that for some FIRE-2 galaxies the analogous 150 pc radius lies below the resolution limit, requiring extrapolation.","fun_headline_variants_meta":{"raw":{"variants":["Bursty star formation flattens dark matter cusps","Star formation history shapes dark matter cores","Post-reionization stars turn cusps into cores","Clumpy star growth predicts dark matter slopes","New fit ties star burstiness to halo profiles"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000257,"raw_usage":{"total_tokens":1494,"prompt_tokens":900,"completion_tokens":594,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":644,"completion_tokens_details":{"reasoning_tokens":520}},"tokens_in":644,"tokens_out":594,"duration_ms":6776,"temperature":1.0,"reasoning_tokens":520,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T08:24:14.440279+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A concrete test would be to recompute the same inner slopes from the same simulated galaxies at two or more resolution levels (or with different binning and centering choices) and check whether the burstiness and SFH-deviation trends survive; alternatively, an observational survey of dwarf galaxies with well-measured star formation histories and dark matter cores could check whether early-concentrated star formation systematically correlates with cusps, as the paper predicts. If the correlation vanishes or reverses, the claim is falsified.","supporting_citations":[],"review_version":1}