{"id":"a99edd19-3cbb-4fcd-a1f6-80d17008fc07","arxiv_id":"2501.04788","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Rest-frame near-infrared half-light radii of quiescent galaxies trace their half-mass radii, and young and old quiescent galaxies have consistent sizes at intermediate stellar masses from z=3 to z=0.5.","lead":"Using JWST images from the PRIMER survey, researchers measured the sizes of 455 quiet, non-star-forming galaxies and compared how light and stellar mass are distributed within them. They find near-infrared sizes match mass-weighted sizes, and that recently quieted and older quiet galaxies are similarly sized at intermediate masses, which helps explain how galaxy sizes grow over cosmic time.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Rest-frame infrared sizes are claimed to trace mass-weighted sizes, but the half-mass radii derive from F444W light scaled by an unvalidated radial M*/L assumption; resolved SED fitting is needed.","rationale":"Agree with the reader: the weakest link is the assumption that the integrated color-M*/L relation (Eq. 2) applies at all radii. The paper's methodology is standard and carefully executed, and the residual checks in the appendix partially address dependence on mass and age, but they do not validate the radial extrapolation. The claim that F444W sizes trace mass is particularly sensitive because the F444W profile is the basis for the half-mass radius; the agreement between half-light and half-mass in F444W is expected if the color gradient is small. External validation is essential to break this circularity. I also note minor internal inconsistencies (sample size: 426+39=465 vs 455 in abstract and Sec 4.1; mass-weighted size growth: 0.30/0.35 dex in Sec 4.1 vs 0.45/0.40 dex in Sec 6) that do not affect the verdict but signal that the manuscript needs careful revision. The proposed test - resolved SED fitting on a subset - is straightforward with existing PRIMER data and would directly determine whether the color-M*/L relation holds radially. Given the acknowledged limitations and lack of external validation, the conditional verdict remains appropriate; the paper is worth publishing with the caveat clearly stated.","tokens_in":19041,"tokens_out":14000,"duration_ms":119594,"concrete_test":"Select ~20 galaxies spanning the mass and redshift range, including both young and old quiescent galaxies. Perform spatially resolved SED fitting in 3-5 annuli using the full NIRCam photometry (F090W-F444W) plus any available HST data, deriving M*/L radial profiles directly from the SEDs. Recompute half-mass radii and re,mass/re,light ratios from these independent mass profiles. If the half-mass radii differ from the paper's color-based values by more than ~0.1 dex, or if the young/old agreement disappears, the central claims are not secure; if they agree, the assumption is validated.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim that rest-frame infrared sizes accurately trace mass-weighted sizes is not empirically demonstrated. Half-mass radii are derived by taking the F444W light profile and multiplying by an M*/L profile obtained from the F277W-F444W color via a relation fitted to integrated colors and SED masses (Eq. 2, Table 1), with the explicit assumption that this relation is valid at every radius (Sec. 3.2). The appendix (Fig. 8) checks residuals of the integrated relation against total mass, age, and flux, but never tests whether the relation holds as a function of radius. Because the F444W profile is the backbone of the half-mass radius, the observed closeness of re,mass and re,light(F444W) is partly a consequence of the method: a shallow color gradient implies a nearly constant M*/L profile, forcing the two radii to agree. It does not prove that the F444W light distribution traces the stellar mass distribution. The authors themselves cite Bernardi et al. (2023), who show that IMF or metallicity gradients can change half-mass sizes by up to 0.3 dex. Without external validation - spatially resolved SED fitting, dynamical mass profiles, or lensing - the headline claims about IR sizes tracing mass and about the young/old size agreement in half-mass radii are based on an unverified assumption rather than a measurement.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript uses JWST/NIRCam PRIMER imaging to measure half-light and half-mass radii of 455 quiescent galaxies at 0.5 < z < 3, split into 426 old and 39 young quiescent galaxies. Half-light radii are measured in six filters with GALFIT and residual-corrected surface brightness profiles. Half-mass radii are derived by converting F277W-F444W color profiles into M*/L444 profiles using a redshift-binned linear relation (Eq. 2) calibrated on the same galaxies' integrated colors and SED masses, then multiplying the F444W luminosity profiles by these M*/L profiles and integrating. The paper reports that rest-frame infrared sizes trace mass-weighted sizes while rest-frame optical sizes are 0.1-0.2 dex larger, that young and old quiescent galaxies have consistent sizes at intermediate masses, and that the changing young/old mix can mimic apparent evolution in re,mass/re,light. These results are interpreted as evidence for a combination of progenitor bias and minor mergers driving size growth.","tokens_in":19305,"tokens_out":4798,"duration_ms":47526,"significance":"If the half-mass radii are robust, the paper provides one of the first direct light- versus mass-weighted size comparisons for age-selected quiescent galaxies at 0.5 < z < 3 with JWST, and it has practical relevance for using infrared sizes as proxies for mass-weighted sizes. The strengths include careful Monte Carlo uncertainty propagation, residual-corrected SBPs, validation of 1D versus 2D radii, comparisons with literature measurements, and explicit discussion of caveats such as IMF-gradient effects. However, the central empirical claim depends on an unvalidated assumption about the radial applicability of the integrated color-M*/L relation, so the key results are currently conditional on that assumption rather than independently demonstrated.","major_comments":[{"comment":"The color-M*/L relation is calibrated on the integrated colors and SED masses of the same sample and then applied at every radius inside each galaxy. Because the mass profile is the F444W luminosity profile scaled by the resulting M*/L profile, the re,mass/re,light ratio at F444W is partly built in: a shallow F277W-F444W color gradient yields a nearly constant M*/L profile, forcing re,mass to track re,light(F444W). The agreement between F444W sizes and half-mass sizes is therefore not an independent test of the claim that infrared light traces stellar mass. The appendix (Fig. 8) checks residuals against total mass, age, and flux, but not against radius. I request validation with spatially resolved SED fitting in annuli, or comparison with dynamical or lensing mass profiles; absent that, the headline claims should be reframed as conditional on the radial M*/L assumption.","section":"Section 3.2, Eq. (2), and Section 3.3"},{"comment":"The young quiescent sub-sample comprises only 39 galaxies, of which 9 are above log(M*/M_sun) > 11. The claim that young and old quiescent galaxies agree in size at intermediate masses is based on small-N bins and large galaxy-to-galaxy scatter. Although the paper notes this limitation, I would like a quantitative statement of the statistical power: what fractional size difference could be detected at 1-sigma with N=39 in the 10 < log(M*/M_sun) < 11 range? Without this, the conclusion of 'a combination of progenitor bias and minor mergers' is not strongly constrained by the data.","section":"Sections 4.1 and 5"},{"comment":"Thirty-five percent of the parent sample is excluded because galaxies cannot be securely fit in all six available filters, and no test is presented of how the excluded systems differ in mass, redshift, concentration, or merging state. Since fitting failures in compact or irregular quiescent galaxies are plausibly correlated with the size measurements, this selection could bias the derived size distributions and the young/old comparison. I request a comparison of the included and excluded samples in mass-redshift space and in at least one measured property such as the F444W half-light radius from a forced fit.","section":"Section 2"}],"minor_comments":[{"comment":"The expression '(10<log(M*/M_sun)> 11)' should read '(10<log(M*/M_sun)<11)'.","section":"Conclusions, bullet 3"},{"comment":"The label 'F227W' is a typo and should be 'F277W'.","section":"Figure 2 caption"},{"comment":"The exponent reported as '(-1.85 ± -0.27)' should be written with a positive uncertainty, e.g., '(-1.85 ± 0.27)', and the sign convention for R ∝ (1+z)^alpha should be stated explicitly.","section":"Section 4.1"},{"comment":"The phrase 'old and young quiescent galaxies quiescent galaxies are, on average, constant' contains a duplicated word and should be revised.","section":"Section 4.4"},{"comment":"The caption contains the typo 'Depency'; it should be 'Dependence'.","section":"Appendix, Figure 8 caption"},{"comment":"The quoted uncertainties on A and C are highly disparate across redshift bins, which likely indicates strong covariance between the two fitted parameters; presenting the full covariance matrix or a corner plot would help readers assess whether the bin-to-bin variations in A are meaningful.","section":"Table 1"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is transparent and the analysis is careful, but the central claim that infrared sizes trace mass-weighted sizes is not empirically established because the half-mass radii inherit the radial M*/L assumption. I believe the paper can be made acceptable by either adding an external validation (resolved SED fitting, dynamics, or lensing) or by softening the language to make the conditional nature of the result explicit. The small young-quiescent sample and the 35% exclusion rate are serious limitations, though they are acknowledged; the selection-bias test would help. The paper is within scope for the journal and represents useful observational progress if these issues are addressed."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe thing to know: this paper has a genuinely new age-split comparison of half-mass and half-light radii for quiescent galaxies at 0.5<z<3 using PRIMER, and the point that the changing young/old mixture can mimic redshift evolution of color gradients is worth taking seriously. But the headline claim—that rest-frame infrared sizes accurately trace mass-weighted sizes—is weaker than it looks, because the half-mass radii are derived from the F444W light profiles multiplied by an M*/L profile obtained from a color-M*/L relation fitted to the same sample. The agreement between re,mass and F444W re,light is partly built into the method.\n\nWhat it does well: the analysis is careful. Monte Carlo uncertainties, residual tests, and comparison with Suess et al. and van der Wel et al. are all there. The Appendix checks residuals of the integrated relation against mass, age, and flux and finds no strong dependence. The age-split comparison across the full redshift range is new relative to the cited literature. The conclusion that intermediate-mass young and old quiescent galaxies agree in size, and that the size growth is a combination of progenitor bias and minor mergers, is cautiously stated.\n\nThe soft spots: the central validation rests on the assumption that the color-M*/L relation fitted to integrated colors holds at every radius inside each galaxy (Sec. 3.2). That is not tested. The Appendix checks residuals of integrated quantities, not radial gradients. The authors themselves cite Bernardi et al. (2023), who show that IMF or metallicity gradients can change half-mass sizes by up to 0.3 dex. Without spatially resolved SED fitting, dynamical mass profiles, or lensing, the claim that F444W sizes trace mass sizes is an assumption, not a measurement. This is a moderate weakness, not a fatal one, because the paper's other results—the age-split size comparison and the mimicry effect—do not depend critically on the exact half-mass radii. The small young sample (39 galaxies, 9 above 1e11 Msun) and the 35% fitting exclusion are limitations, but they are acknowledged and handled reasonably. The private photometric catalog is a practical obstacle for reproducibility, though the paper does point to the observation data DOI.\n\nWho this is for: anyone working on the evolution of quiescent galaxy sizes. The paper deserves a serious referee; the subject is important and the authors are transparent about their assumptions. The referee should push for external validation of the mass profiles, or for a softened statement of the IR-trace-mass claim.\n\nRecommendation: send it to peer review. With appropriate revisions it can be a useful contribution.","headline":"A careful age-split size study with a new mimicry result, but the headline claim that F444W sizes trace mass sizes is partly built into the method and needs external validation.","tokens_in":19969,"tokens_out":2739,"would_cite":true,"duration_ms":26583,"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":"Rest-frame infrared sizes faithfully trace the mass-weighted sizes of quiescent galaxies, and young and old quiescent galaxies have the same average size at intermediate masses across $0.5<z<3$.","keywords":["quiescent galaxies","half-mass radius","size evolution","color gradients","JWST/NIRCam","M*/L gradients","progenitor bias","minor mergers"],"falsifier":"Compare the half-mass radii obtained here with radii derived from spatially resolved stellar-population fitting (fitting the SED in concentric annuli or pixels) for the same galaxies; if the radial $M_*/L$ gradients differ, the half-mass radii and the $r_{e,\\mathrm{mass}}/r_{e,\\mathrm{light}}$ ratios would shift by more than the quoted uncertainties.","tokens_in":16,"feed_emoji":"🔭","tokens_out":6817,"duration_ms":120755,"temperature":0.7,"pith_summary":"Using JWST/NIRCam imaging from the PRIMER survey, this paper asks whether the growth of quiescent galaxy sizes with cosmic time is real or a measuring artifact. It derives both half-light radii (from six filters) and half-mass radii (from the F444W profile plus a color-to-mass-to-light relation) for 455 quiescent galaxies at $0.5<z<3$, splitting the sample into recently quenched and older galaxies. The paper finds that rest-frame infrared sizes closely track mass-weighted sizes, that optical sizes are 0.1–0.2 dex larger, and that young and old quiescent galaxies agree in size at intermediate masses. The authors conclude that the observed size growth of intermediate-mass quiescent galaxies is driven by a combination of progenitor bias and minor mergers, and that the changing mix of young and old galaxies can mimic evolution in color gradients.","feed_headline":"Young and old quiescent galaxies have the same size, JWST finds","feed_subtitle":"Rest-frame infrared half-light radii track half-mass radii; size growth reflects mergers and progenitor bias.","key_machinery":"The load-bearing object is the empirical color-to-mass-to-light relation, $\\log(M_*/L_{\\mathrm{F444W}}) = -2.5\\,\\log_{10}(F277W/F444W)\\,A + C$, fitted in redshift bins to integrated colors and SED-based masses (Table 1). The paper uses the observed $F277W-F444W$ color to predict the stellar mass-to-light ratio at every radius, multiplies the F444W luminosity profile by this radial $M_*/L$ to get a mass profile, and integrates to obtain the half-mass radius. This is the step that turns light profiles into mass profiles, and it is where the assumption of universal radial $M_*/L$ enters.","core_discovery":"The central claim is that the size growth of the red sequence between $z=3$ and $z=0.5$ is intrinsic to the population and not an artifact of measuring light rather than mass. The paper demonstrates that half-light radii measured at rest-frame infrared wavelengths ($\\sim$4.4 µm) agree with half-mass radii, while rest-frame optical sizes ($0.5$–$0.7$ µm) run 0.1–0.2 dex larger. For intermediate-mass galaxies ($10<\\log(M_*/M_\\odot)<11$), young quiescent galaxies have the same average half-light and half-mass radii as old quiescent galaxies within uncertainties; the previously reported size difference appears only at $\\log(M_*/M_\\odot)>11$, where the present sample has only nine young objects. The paper argues that these results favor a mix of progenitor bias and minor mergers driving size growth, and that the increasing share of young quiescent galaxies at high redshift can mimic a redshift evolution of the $r_{e,\\mathrm{mass}}/r_{e,\\mathrm{light}}$ ratio.","pith_inferences":["The same color-to-$M_*/L$ method could be extended to star-forming galaxies, where stronger dust and star-formation gradients would test whether the universality of radial $M_*/L$ holds more generally.","If the relation is truly universal in radius, then the similarity of young and old color gradients implies that intermediate-mass galaxies share similar radial stellar population structure regardless of quenching age; resolved spectroscopy could verify this directly.","The 0.1–0.2 dex offset between optical and infrared sizes implies that single-band optical size measurements systematically overestimate the mass scale of quiescent galaxies; the method applied to wider-area JWST surveys could confirm this at the high-mass end, where the current sample has only nine young galaxies.","Because IMF gradients could shift half-mass radii by up to 0.3 dex, the absolute sizes reported here rest on the assumed IMF; a spatially resolved dynamical or lensing mass calibration would set the scale."],"forward_implications":["Half-mass radii of quiescent galaxies grow by about 0.4–0.45 dex from $z=3$ to $z=0.5$, so the size growth of the red sequence is not an artifact of measuring light.","Rest-frame infrared (F444W) half-light radii can be used as a direct proxy for half-mass radii in quiescent galaxies, simplifying future size studies.","At intermediate masses, the agreement between young and old quiescent galaxies implies that both progenitor bias and minor mergers contribute to size growth, so simulations of quenching must include both channels.","The apparent redshift evolution of the $r_{e,\\mathrm{mass}}/r_{e,\\mathrm{light}}$ ratio in the general quiescent population can be explained by the growing fraction of young quiescent galaxies at high redshift."],"supporting_citations":[{"why":"Establishes the color-profile method for deriving half-mass radii from imaging, which this paper adapts and extends.","marker":"Szomoru et al. (2013)"},{"why":"Provides the approach to extracting radial color gradients and deriving $M_*/L$ profiles; serves as the methodological baseline.","marker":"Miller et al. (2023)"},{"why":"Supplies the multi-color $M_*/L$ relation approach and the comparison for mass-weighted sizes and size evolution.","marker":"van der Wel et al. (2024)"},{"why":"Offers the alternative half-mass radius method the paper benchmarks against and the color-gradient evolution it reinterprets.","marker":"Suess et al. (2019a)"},{"why":"Shows that rest-frame infrared (4.4 µm) sizes are smaller than optical sizes and can act as a proxy for mass-weighted sizes; the paper validates this claim directly.","marker":"Suess et al. (2022)"},{"why":"Prior study of size differences between young and old quiescent galaxies at $1.5<z<3$; the paper extends and tests its conclusions at intermediate masses.","marker":"Clausen et al. (2024)"}],"fun_headline_variants":["JWST: Young and old quiescent galaxies equal size","Infrared sizes trace mass radii in galaxies","Galaxies grow by mergers, not light measurement bias","Optical sizes overstate galaxy mass radii, JWST shows"],"cache_read_input_tokens":22016,"weakest_assumption_plain":"The load-bearing premise is that the color-to-mass-to-light relation fitted to each galaxy's integrated colors also holds at every radius inside the galaxy, so that radial color profiles can be converted into radial mass profiles.","fun_headline_variants_meta":{"raw":{"variants":["JWST: Young and old quiescent galaxies equal size","Infrared sizes trace mass radii in galaxies","Galaxies grow by mergers, not light measurement bias","Optical sizes overstate galaxy mass radii, JWST shows"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001421,"raw_usage":{"total_tokens":5828,"prompt_tokens":1132,"completion_tokens":4696,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":748,"completion_tokens_details":{"reasoning_tokens":4631}},"tokens_in":748,"tokens_out":4696,"duration_ms":35401,"temperature":1.0,"reasoning_tokens":4631,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T21:28:16.621992+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compare the half-mass radii obtained here with radii derived from spatially resolved stellar-population fitting (fitting the SED in concentric annuli or pixels) for the same galaxies; if the radial $M_*/L$ gradients differ, the half-mass radii and the $r_{e,\\mathrm{mass}}/r_{e,\\mathrm{light}}$ ratios would shift by more than the quoted uncertainties.","supporting_citations":[{"cited_title":"Rest-frame near-infrared sizes of galaxies at cosmic noon: objects in JWST's mirror are smaller than they appeared","cited_arxiv_id":"2207.10655","evidence_quote":"Shows that rest-frame infrared (4.4 µm) sizes are smaller than optical sizes and can act as a proxy for mass-weighted sizes; the paper validates this claim directly."},{"cited_title":"3D-DASH: The Evolution of Size, Shape, and Intrinsic Scatter in Populations of Young and Old Quiescent Galaxies at 0.5 < z < 3","cited_arxiv_id":"2405.09354","evidence_quote":"Prior study of size differences between young and old quiescent galaxies at $1.5<z<3$; the paper extends and tests its conclusions at intermediate masses."}],"review_version":1}