{"id":"6d9d0e45-5693-409e-bd3e-59bf8b611f3c","arxiv_id":"2608.06463","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Post-starburst galaxies drive the low-mass upturn in the quiescent galaxy mass function out to z~2.25, and a simple model in which all post-starbursts become passive overpredicts low-mass passive galaxies at later times.","lead":"Using JWST imaging of the UDS field, the authors identify over 800 post-starburst galaxies at redshifts 0.5 to 3 and show that recently quenched galaxies, not older passive ones, drive the low-mass upturn in the quenched galaxy mass function out to z~2. The result points to a fast, possibly environment-driven quenching route for low-mass galaxies at cosmic noon, and it questions simple one-way evolution from starburst to passive.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The z~2 PSB-dominance claim is not yet secured: the high-redshift comparison uses a passive sample that is incomplete at low mass and a PSB sample with ~35% estimated contamination, so the ratio may be an artifact.","rationale":"I agree with the reader's conditional verdict. The PCA classification at low mass and high redshift is a real weak point, but the more specific flaw is that the dominance ratio at z~2 is not measurable from the current data because passive incompleteness and PSB contamination both bias toward the conclusion. The lower-redshift result is on much firmer ground, and the independent UDS comparison, UVJ consistency, and corrected-upturn test are genuine supporting evidence. However, none of these directly measures the high-z passive population, so the 'since z~2' claim is overreaching. This does not invalidate the paper; it means the title and abstract should be qualified until the low-mass passive number density is recovered or the PSB fraction is conservatively estimated. The reader's weakest_assumption highlighted classification reliability but not the asymmetric completeness of the passive comparison, so my agreement is partial.","tokens_in":23653,"tokens_out":7213,"duration_ms":65519,"concrete_test":"Recompute phi_PSB/(phi_PSB+phi_passive) for log(M*/Msun)<10 in each redshift bin using only mass bins above the larger of the PSB and passive 90% completeness limits, after subtracting the Section 4.3 contaminant mass function and, as a conservative floor, reducing the PSB counts by the 1-sigma lower bound on strict-spectroscopy purity (44%). Propagate Poisson and PSB-number bootstrap uncertainties. If the 68% confidence lower limit on this fraction in the 1.75<z<2.25 bin falls below 0.5, the title/abstract claim should be softened to 'out to at least z~1.5'.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim is a ratio claim: at log(M*/Msun)<10, phi_PSB/(phi_PSB+phi_passive) exceeds 1/2 out to z~2, and therefore the low-mass quenched upturn is 'primarily driven by post-starburst galaxies'. In the two highest-redshift bins the passive mass function is only defined above its 90% completeness limit, which is higher than the PSB limit; Fig. 7 marks the low-mass passive bins as lower limits, and its caption states the dominance is shown 'out to at least z~1.5', not z~2. The title and abstract claim 'since z~2' is therefore stronger than the plotted comparison supports. The same comparison is also affected by PSB contamination: the strict spectroscopic purity is 57+/-13% (8/14), and the Section 4.3 resampling test estimates ~35% of low-mass PSBs may be scattered star-forming galaxies. Both systematics bias the ratio in the same direction in the redshift/mass regime of interest: star-forming contaminants inflate the numerator while passive incompleteness suppresses the denominator. The corrected PSB mass functions in Fig. 9 still show an upturn, which is encouraging, but an upturn in PSBs alone does not establish that PSBs dominate the quenched population unless the passive low-mass counts are measured or conservatively bounded. In short, the lower-redshift result is credible, but the 'since z~2' framing is not yet demonstrated.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper uses JWST PRIMER imaging of the UDS field, combined with HST/ACS and ground-based UDS data, to measure stellar mass functions for photometrically selected star-forming, passive, and post-starburst (PSB) galaxies over 0.5 < z < 3.0. Applying the Wild et al. (2014) PCA super-colour technique, the authors identify 807 PSBs and report a low-mass upturn in the quenched mass function below log(M*/M_sun) ~ 10 out to z ~ 2, which they attribute primarily to the PSB population. They then use a simple evolutionary model (Eq. 2) to predict the passive mass function at later times from the passive plus visibility-time-scaled PSB mass functions, finding an overprediction of low-mass passive galaxies that persists even with longer assumed PSB visibility times, and discuss mergers, rejuvenation, and stochastic star formation as possible resolutions.","tokens_in":23814,"tokens_out":5121,"duration_ms":43365,"significance":"If the central claim holds, the paper provides important evidence that a rapid, likely environment-linked quenching channel for low-mass galaxies was already active by cosmic noon, and that the low-mass quiescent population cannot be understood as the product of slow, one-way passive evolution alone. The analysis has several genuine strengths: the two independent selection methods (SC and UVJ) show the same qualitative upturn pattern; binned and KDE mass functions agree; the star-forming contamination test leaves the PSB upturn largely intact; and the comparison with UDS DR11 classifications provides an independent photometric cross-check. The evolutionary model in Eq. 2 is predictive rather than circular, since the visibility time is adopted from earlier work and the prediction is compared with, not fitted to, the measured passive mass function. The main weakness is that the headline 'since z~2' claim is stronger than the plotted passive/PSB comparison supports, because at z > 1.75 the passive mass function is incomplete exactly in the low-mass regime where the ratio is claimed to be PSB-dominated, and the PSB sample purity is only modest.","major_comments":[{"comment":"The headline claim that low-mass PSBs dominate the quenched mass function 'since z~2' is stronger than the plotted comparison supports. Fig. 7's caption states the dominance is demonstrated 'out to at least z~1.5', and in the two highest-redshift bins (1.75 < z < 2.25 and 2.25 < z < 3.0) the passive mass function is only defined above its 90 per cent completeness limit, which lies above the PSB completeness limit; low-mass passive bins are marked as lower limits. Because the claimed dominance is a ratio phi_PSB/(phi_PSB + phi_passive), passive incompleteness suppresses the denominator in exactly the regime where the claim is made. The authors should either provide a conservative upper bound on phi_passive below the passive completeness limit in these bins and recompute the ratio, or restrict the dominance claim to z < 1.75 in the abstract, title, and conclusions.","section":"Abstract/§4.1, Fig. 7"},{"comment":"The PSB classification purity is only 57 ± 13 per cent (8/14) for the strict W_Hd > 5 Å criterion and 79 ± 11 per cent (11/14) for the relaxed criterion, and the Section 4.3 resampling test estimates that roughly 35 per cent of low-mass PSBs (M* < 10^10 M_sun) may be scattered star-forming contaminants. Both effects bias the PSB-dominance ratio in the same direction at high redshift: contaminants inflate the numerator and passive incompleteness suppresses the denominator. The authors need to quantify the net effect on phi_PSB/(phi_PSB + phi_passive) in the 1.75 < z < 2.25 and 2.25 < z < 3.0 bins, for example by propagating the 35 per cent contamination estimate and the passive upper limits through the ratio. The corrected PSB mass functions in Fig. 9 establish an upturn in the PSB population alone, but they do not by themselves establish dominance over the passively evolving population unless the passive counts are measured or conservatively bounded.","section":"§3.3 and §4.3"},{"comment":"The statement that 'we do not see this trend in the older passive population at any redshift' is vulnerable to completeness: the passive sample is shown only above its own 90 per cent mass completeness curve (Fig. 6), which is higher than the PSB curve at the same redshift, particularly at z > 1.75. A non-detection of a low-mass upturn in older passive galaxies in the highest-redshift bins is therefore expected from the selection, not an observed physical absence. The text should explicitly separate the redshift/mass regime where the passive mass function is measured from the regime where it is not, and adjust the wording in Section 6 accordingly.","section":"§4.1, Fig. 6"},{"comment":"The evolutionary model in Eq. (2) is a useful predictive tool and is not circular, since t_PSB = 0.5 Gyr is adopted a priori from Wild et al. (2016) and the predicted passive mass function is compared with, rather than fitted to, the measured one. However, the quantitative overprediction factors quoted in Section 5.4 (25x, 13x, 3.1x, etc.) assume that all growth of the passive mass function comes from PSBs and neglect mergers and rejuvenation, and the cluster correction in Section 5.3 changes the ratios by large factors (e.g. the low-mass ratio in the 0.5 < z < 0.75 bin increases from 2.4 to 10.7 after correction). The conclusions drawn from these ratios are plausible but should be presented as upper-limit-style estimates, and the sensitivity to the adopted visibility timescale should be stated more explicitly in the abstract and conclusions.","section":"§5.1–§5.4, Eq. (2), Table 2"}],"minor_comments":[{"comment":"The name 'Schecter' is misspelled; it should be 'Schechter'.","section":"Throughout (e.g. §1, footnote 1 in §4.1)"},{"comment":"The subsection heading 'Auxillary HST Imaging' contains a typo; it should be 'Auxiliary HST Imaging'.","section":"§2.1.1"},{"comment":"The caption refers to 'The combined passive population is shown above the PSB 90 per cent mass completeness limit', but the panel shows the combined quenched population (passive + PSB); the wording should say 'combined quenched population'.","section":"Fig. 7 caption"},{"comment":"The phrase 'we estimate that approximately ~35 per cent' uses both 'approximately' and '~'; please use one form.","section":"§4.3"},{"comment":"The statement that 'our mass functions do not extend as low in stellar mass due to our conservative magnitude limit' would benefit from a quantitative statement of the limiting mass in the 1.75 < z < 2.25 bin, to clarify the comparison with Hamadouche et al. (2025).","section":"§4.2"}],"recommendation":"major_revision","confidential_remarks":"For the editor: the analysis is carefully done and the lower-redshift results are likely robust; I see no circularity in the evolutionary model. The principal issue is that the abstract and title overstate the high-redshift dominance claim relative to what Fig. 7 and the completeness limits actually show. The needed fixes (conservative upper limits on the passive mass function at z > 1.75, propagation of the contamination estimate through the ratio, and correspondingly softened wording) are straightforward and within the scope of a revision, so I recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know two things up front. First, the observational core here is solid: the PSB mass functions from JWST PRIMER are a genuine step forward, and the cross-checks are done carefully. Second, the headline claim as written—that post-starbursts dominate the low-mass quiescent population since z~2—is stronger than the plotted data justify. The paper itself admits in the Fig. 7 caption that the dominance is shown 'out to at least z~1.5', not z~2.\n\nWhat is actually new: this is the first JWST-based PSB stellar mass function reaching log(M*/Msun)<10 above z=2. They use two independent selection methods (PCA supercolours and UVJ), and the KDE and binned mass functions agree. The comparison with ground-based UDS DR11 over 6618 overlapping galaxies is a good consistency check. The contamination resampling test is a reasonable attempt to remove star-forming scatter into the PSB region, and the corrected PSB mass functions still show an upturn. That is real evidence, and it should be credited.\n\nThe soft spots are concentrated in the high-redshift bins. In the 1.75<z<2.25 and 2.25<z<3.0 bins, the passive mass function is only defined above a 90% completeness limit that sits higher than the PSB limit; those low-mass passive bins are marked as lower limits. The ratio PSB/(PSB+passive) therefore has an inflated numerator and a suppressed denominator. Add the photometric purity issue—strict spectroscopic purity is 57±13% on 14 objects, and the resampling test suggests ~35% of low-mass PSBs could be star-forming contaminants—and both systematics push the ratio in the same direction. An upturn in the PSB mass function alone does not establish dominance unless the passive counts are measured or conservatively bounded. The paper also gives no formal significance for the high-z upturn.\n\nNone of this is fatal. The lower-redshift result is credible, and the overprediction model in Section 5 is transparent and not circular: the visibility time is adopted from Wild et al. (2016), not tuned to force agreement. The discussion of mass-dependent visibility times, mergers, rejuvenation, and stochastic star formation is speculative but clearly flagged.\n\nThis paper is for anyone working on quiescent galaxy assembly or PSB selection in JWST-era surveys. It deserves a serious peer review; the high-z dominance claim should be recalibrated, and the authors should present a bounded estimate of the passive low-mass counts above z~1.75. I would not cite the numbers until the catalogue is public, but I would cite the PSB mass functions once they are released.\n\nRecommendation: send it to review, but the editor should insist the 'since z~2' framing be revised to match what the data actually show.","headline":"The low-redshift result that post-starbursts drive the low-mass quenched upturn holds up, but the 'since z~2' framing overstates what the data can support; the paper deserves review with a required revision of that claim.","tokens_in":24565,"tokens_out":2684,"would_cite":true,"duration_ms":24645,"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":"Post-starburst galaxies, not older passive galaxies, drive the low-mass upturn in the quenched galaxy mass function out to $z\\sim2$.","keywords":["galaxies: evolution","galaxies: high-redshift","galaxies: formation","galaxies: starburst","galaxies: star formation","galaxies: fundamental parameters","post-starburst galaxies","galaxy stellar mass functions"],"falsifier":"A decisive test is a large spectroscopic survey of the photometric post-starburst candidates below $10^{10}\\,M_\\odot$ in the redshift bins above $z=1.75$. If the majority show $W_{\\rm H\\delta}>5$ Å with weak or absent emission lines, the claim survives. If instead most are older passives with $4<W_{\\rm H\\delta}<5$ Å or dusty or star-forming interlopers, then the low-mass upturn would not be primarily post-starburst in origin, and the paper's central conclusion would be refuted.","tokens_in":23315,"feed_emoji":"🔭","tokens_out":12294,"duration_ms":88969,"temperature":0.7,"pith_summary":"The paper uses JWST near-infrared imaging of the UDS field to construct stellar mass functions for star-forming, passive, and post-starburst galaxies across $0.5<z<3.0$, identifying more than 800 photometrically selected post-starbursts. It confirms a low-mass upturn in the quenched galaxy mass function at $M_\\ast\\lesssim10^{10}\\,M_\\odot$ out to $z\\sim2$ and shows that this upturn is primarily driven by post-starburst galaxies; the older passive population alone shows no upturn. This places a rapid, likely environment-linked quenching channel at cosmic noon. The paper then tests the one-way evolutionary picture by assuming every post-starburst fades into a passive galaxy with a uniform visibility time, and finds that this overpredicts the number density of low-mass passive galaxies at $z<1.75$; even the longest plausible visibility times do not fully close the gap. The authors conclude that mergers, tidal disruption, rejuvenation, or stochastic star formation must remove or revive a substantial fraction of low-mass post-starbursts.","feed_headline":"Post-starbursts explain low-mass quiescent galaxy upturn","feed_subtitle":"JWST counts show recently quenched galaxies, not old passive ones, drive the excess by z~2.","key_machinery":"The engine of the analysis is the PCA super-colour method: each galaxy's rest-frame $2500\\,$–$\\,15000$ Å SED is projected onto three eigenvectors built from stellar population synthesis models, producing amplitudes SC1, SC2, and SC3. SC1 tracks the $R$-band light-weighted stellar age and dust content, SC2 traces the fraction of stellar mass formed in the last Gyr, and SC3 breaks the degeneracy between metallicity and burst fraction; galaxies are classified as star-forming, passive, dusty star-forming, or post-starburst from their position in SC space. The same fit supplies stellar masses. The mass functions are then fed into a continuity formula, $\\phi_{\\rm pred}(M_\\ast)=\\phi_{\\rm PAS,hi-z}(M_\\ast)+(\\Delta t_{\\rm cos}/t_{\\rm PSB})\\,\\phi_{\\rm PSB,hi-z}(M_\\ast)$, which converts the post-starburst population in a higher redshift bin into a predicted contribution to the passive population at lower redshift, with $t_{\\rm PSB}$ the visibility time of the post-starburst phase.","core_discovery":"The central claim is that the excess of low-mass quenched galaxies below $M_\\ast\\sim10^{10}\\,M_\\odot$ out to $z\\sim2$ consists mostly of post-starburst galaxies, not older passive galaxies. In the PCA super-colour classification, the passive population shows no low-mass upturn in any redshift bin, while the post-starburst mass function rises steeply below $10^{10}\\,M_\\odot$ out to $z=2.25$. This means a fast quenching pathway for low-mass galaxies was already active by cosmic noon. The paper further shows that a simple continuity model in which all post-starbursts transition into passives, scaled by a 500 Myr visibility time, overproduces low-mass passive galaxies at $z<1.75$; adopting longer visibility times for low-mass systems reduces the discrepancy but does not remove it. The conclusion is that low-mass post-starbursts do not follow a simple one-way path to the passive population: they may quench more slowly, rejuvenate back to star formation, or be removed through mergers and tidal disruption.","pith_inferences":["Inference: if the post-starburst-driven upturn is real, the instantaneous post-starburst abundance at $z\\sim2$ is not a reliable predictor of the future passive population; continuity models such as the one in equation (2) need an explicit removal or rejuvenation term at low masses, and simulations that assume strict one-way quenching will tend to overproduce low-mass quiescent galaxies.","Inference: the factor by which low-mass post-starbursts are overproduced relative to high-mass ones (roughly 3–25 times, depending on redshift and visibility assumptions) can be recast as an empirical disruption-plus-rejuvenation rate that could be compared directly with merger and tidal-disruption rates from simulations in the same volume.","Inference: a testable extension is that deeper spectroscopy of the same field will find that a substantial fraction of low-mass post-starburst candidates at $z>1.5$ retain detectable nebular emission or show tidal or merger features, because the paper's own accounting implies many of these systems do not simply fade into the passive population."],"forward_implications":["The low-mass upturn in the total quenched mass function out to $z\\sim2$ is caused by recently quenched galaxies; older passive galaxies contribute little to it.","A rapid, likely environment-driven quenching route for low-mass galaxies was already active by cosmic noon ($z\\sim2$), not just at lower redshifts.","Assuming every post-starburst fades into a passive galaxy with a uniform 500 Myr visibility time overpredicts the number density of low-mass passive galaxies at $z<1.75$.","Mass-dependent post-starburst visibility times, up to about 1.5 Gyr at low masses, reduce but do not eliminate this overprediction, so additional processes are required.","Mergers, tidal disruption, rejuvenation, and stochastic star formation must suppress the build-up of the low-mass passive population, affecting low-mass post-starbursts roughly 3–25 times more often than high-mass post-starbursts depending on redshift and visibility assumptions."],"supporting_citations":[{"why":"Supplies the PCA super-colour method and the eigenbasis definitions that the paper applies to JWST photometry.","marker":"Wild et al. 2014"},{"why":"Establishes the earlier post-starburst mass functions and the 500 Myr visibility-time model that the paper adapts to predict passive mass functions.","marker":"Wild et al. 2016"},{"why":"Provides the original spectroscopic verification of photometric post-starburst selection, setting the purity and completeness expectations quoted in this work.","marker":"Maltby et al. 2016"},{"why":"Defines the UDS DR11 super-colour classifications and the boundary-transformation procedure used to translate selection boundaries into the JWST filter set.","marker":"Wilkinson et al. 2021"},{"why":"Found the low-mass upturn in post-starburst abundance in dense environments out to $z=1.5$, the prior result this paper extends to higher redshift.","marker":"Taylor et al. 2023"},{"why":"Reported the low-mass upturn in the quenched galaxy mass function out to $z\\sim2$, which this paper confirms and attributes to post-starbursts.","marker":"Santini et al. 2022"},{"why":"Detected the low-mass upturn in the $1.75<z<2.25$ bin that provides the comparison point for the paper's quenched samples.","marker":"Hamadouche et al. 2025"},{"why":"Supplies the method used to compute the 90 per cent stellar mass completeness limits that define where each mass function is trusted.","marker":"Pozzetti et al. 2010"},{"why":"Supplies the stellar population synthesis models used to build the eigenbasis and to estimate stellar masses.","marker":"Bruzual & Charlot 2003"},{"why":"Offers tentative evidence that post-starburst visibility timescales increase toward lower masses, the key correction explored when testing whether longer visibility closes the overprediction.","marker":"Skarbinski et al. 2026"}],"fun_headline_variants":["Post-starbursts dominate low-mass quiescent upturn","Low-mass quiet galaxies are mostly post-starburst","JWST reveals fast quench path for small galaxies","Rapid shutdown of small galaxies starts by z~2","Post-starbursts rule the quiet low-mass population"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The result stands on the assumption that the photometric PCA super-colour classification reliably separates post-starburst galaxies from older passive and dusty star-forming galaxies at stellar masses below $10^{10}\\,M_\\odot$ and redshifts out to $z\\sim2$; the spectroscopic validation is based on 8–14 objects, with roughly 57 per cent purity and 62 per cent completeness for the strict criterion, so a large contamination fraction among the 807 photometric post-starbursts would weaken the claim that post-starbursts drive the low-mass upturn.","fun_headline_variants_meta":{"raw":{"variants":["Post-starbursts dominate low-mass quiescent upturn","Low-mass quiet galaxies are mostly post-starburst","JWST reveals fast quench path for small galaxies","Rapid shutdown of small galaxies starts by z~2","Post-starbursts rule the quiet low-mass population"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000323,"raw_usage":{"total_tokens":1897,"prompt_tokens":1110,"completion_tokens":787,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":726,"completion_tokens_details":{"reasoning_tokens":706}},"tokens_in":726,"tokens_out":787,"duration_ms":6411,"temperature":1.0,"reasoning_tokens":706,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T14:33:30.479423+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A decisive test is a large spectroscopic survey of the photometric post-starburst candidates below $10^{10}\\,M_\\odot$ in the redshift bins above $z=1.75$. If the majority show $W_{\\rm H\\delta}>5$ Å with weak or absent emission lines, the claim survives. If instead most are older passives with $4<W_{\\rm H\\delta}<5$ Å or dusty or star-forming interlopers, then the low-mass upturn would not be primarily post-starburst in origin, and the paper's central conclusion would be refuted.","supporting_citations":[],"review_version":2}