{"id":"dcd75f02-bb70-4130-8528-19dfc8033114","arxiv_id":"2411.19009","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"A wide Subaru HSC survey finds 7 candidate radio galaxies at z~4 with stellar masses of a few 10^11 solar masses whose rest-frame UVJ colors indicate fast quenching in about 0.1 Gyr.","lead":"Astronomers matched a new g-dropout Lyman-break galaxy catalog from Subaru HSC with VLA FIRST radio data and identified 7 likely radio galaxies at z~4 that are already massive and quenched. The result suggests that the most massive galaxies in the early universe assembled and stopped forming stars quickly, though the sample is small and lacks spectroscopic confirmation.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified beyond the reader's conditional on spectroscopic confirmation; the analysis is internally coherent and its own simulations support the claimed selection mass floor.","rationale":"I read the paper as a survey paper whose central claims are: (1) a g-dropout+FIRST selection yields 146 HzRG candidates, with 7 having near/mid-IR photometry and photo-z 3.3–4.5; (2) these 7 have large stellar masses ~4.2e11 Msun; (3) their rest-frame UVJ colors indicate fast ~0.1 Gyr quenching. The load-bearing assumption is that the 7 objects are genuinely at z~4. The reader identified this same concern, and the paper itself acknowledges the z~2 degeneracy via Kubo et al. (2024). I agree with the conditional verdict: the analysis is coherent, the SED fits have acceptable reduced chi2, and the authors run useful robustness checks (sfh2exp, Monte Carlo UVJ uncertainties, and a completeness simulation). However, none of these checks can break the photo-z degeneracy because they all use the same photometric data and the same stellar population models; the degeneracy is external to the fitting procedure. The selection simulation in Section 4.1 is helpful for interpreting the sample as biased to Mstar>1e11, and the KS test versus USS-selected samples is suggestive but not independent of the same photo-z issue. Since the paper does not overclaim spectroscopic confirmation and explicitly notes the limitation, the appropriate verdict remains conditional rather than accept or reject. A concrete test that would settle the concern is spectroscopy; without it, the mass and quenching claims should not be treated as established.","tokens_in":23544,"tokens_out":1573,"duration_ms":49705,"concrete_test":"Obtain optical/NIR spectroscopy (e.g., VLT/X-shooter or JWST/NIRSpec) for the seven candidates, or search archival data, to measure redshifts from Ly-alpha, rest-UV absorption lines, or the 4000 A break. If any object yields z<3, verify whether the z~2 fast-quenching template from Kubo et al. (2024) fits its HSC+VIKING/UKIDSS+WISE SED and recompute its mass and UVJ colors at the spectroscopic redshift to assess how much of the central conclusion survives.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The reader's conditional verdict identifies the central risk: all seven 'final HzRG candidates' lack spectroscopic redshifts and were chosen by re-running X-CIGALE with the redshift prior restricted to 3.3–4.5 after the free photo-z run placed 21 of 28 candidates elsewhere, with the paper itself citing Kubo et al. (2024) that z~2 fast-quenching galaxies can mimic the near-infrared signature of z~4 galaxies. This is a genuine and paper-acknowledged degeneracy, not an internal inconsistency. The claim most at risk is not the existence of massive radio-loud LBGs but the specific assertion that these objects are z~4 galaxies with Mstar~4.2e11 Msun and fast (~0.1 Gyr) quenching; if the objects are z~2 interlopers, the masses and UVJ quenching interpretation are not correct. The paper also engages the selection concern through a 17,856-model completeness simulation that supports the Mstar>1e11 floor, and tests the fast-quenching result against sfh2exp, which keeps the conclusion for 5 of 7 objects. The remaining concern is thus purely the need for external confirmation, which the paper itself does not claim to have.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper presents a search for z~4 radio galaxies by matching g-dropout Lyman-break galaxies selected from ~560 deg^2 of Subaru HSC-SSP data to VLA FIRST radio sources, producing 146 HzRG candidates. Restricting to 28 objects with VIKING or UKIDSS near-infrared and unWISE mid-infrared detections, the authors run X-CIGALE SED fits and select 7 objects with photometric redshifts in 3.31e11 selection floor, a Monte Carlo uncertainty analysis for the UVJ colors, and an sfh2exp robustness check that broadly reproduces the fast-quenching pattern for 5 of 7 objects. The main astrophysical conclusions, however, remain conditional on photometric redshifts and on model-generated UVJ colors; the paper is transparent about both limitations, which is a credit to the authors.","major_comments":[{"comment":"The final seven-object sample rests entirely on photometric redshifts. In Section 3.1 the unrestricted X-CIGALE run places the median best-fit redshift of the 28 NIR/MIR-bright candidates at z=2.30, and the text itself cites Kubo et al. (2024) to note that z~2 fast-quenching galaxies show a spectral break near rest 1600 Angstrom that can be misidentified as the near-infrared signature of z~4 galaxies. Since none of the seven objects has a spectroscopic redshift, the restricted 3.31e11 Msun, with completeness roughly 0.4-0.6 for the massive range and near zero below it. Given this floor, the conclusion in Section 5 that 'the massive nature is a general characteristic of HzRGs independent of the selection method' is stronger than the data support. The KS comparison with USS-selected samples (p=1.2e-3) also mixes different SED-fitting codes and excludes objects with mass upper limits. The authors are transparent about the bias in Section 4.1, but the wording of the abstract and conclusions should be tempered to say that the massive nature applies to the NIR/MIR-selected subset.","section":"Sec. 4.1, Fig. 7, Sec. 5"},{"comment":"The Monte Carlo realizations are said to show that fast quenching is preferred, but no quantitative fraction is given. Only ID3 and ID6 formally satisfy the Muzzin et al. (2013) quiescent criteria, while ID2, ID4, and ID7 are near the boundary; without a stated percentage of realizations that fall within the tau=0.1 Gyr track rather than the tau=1.0 Gyr track, the 'rough timescale of 0.1 Gyr' is not statistically supported. Please report the classification probability or a posterior distribution on the effective quenching timescale.","section":"Sec. 4.3, Figs. 10 and 12"}],"minor_comments":[{"comment":"The caption contains the typo 'Furtehr' and should read 'Further'; please proofread the figure captions.","section":"Figure 9 caption"},{"comment":"The text contains 'soruces' where 'sources' is intended; the same typo appears in the VLASS discussion and should be corrected.","section":"Sec. 4.4"},{"comment":"The near-infrared columns labeled Ks and K are easy to confuse because the survey (VIKING or UKIDSS) is not repeated in the column headers; please add explicit survey labels or a clearer footnote.","section":"Table 4"},{"comment":"The abstract and conclusion state the ~0.1 Gyr quenching timescale without mentioning that the UVJ colors are model-derived; please add a caveat or soften the wording.","section":"Abstract and Sec. 5"},{"comment":"The phrase 'convolvedflux 0 20 mag' is not defined in the text; please explain this HSC-SSP magnitude type or cite the relevant documentation.","section":"Sec. 2.1"},{"comment":"No data availability statement is included; providing the final photometric catalog and the X-CIGALE configuration would improve reproducibility.","section":"Data availability"}],"recommendation":"major_revision","confidential_remarks":"This is a workmanlike candidate-sample paper, and the analysis pipeline is internally consistent. The main risk is not methodology but the absence of spectroscopy: the seven objects are photometric candidates, and the paper's own citation of Kubo et al. (2024) makes the z~2 degeneracy concrete. I would not reject the paper for this alone, because the authors are explicit about the caveat, but the conclusions must be softened or supported by follow-up spectroscopy before the strong claims about z~4 masses and fast quenching can stand. If the journal has a format for candidate papers with clearly labeled caveats, this would fit after revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a genuinely useful paper, and it deserves referee time, but the headline claims about mass and fast quenching rest on photometric redshifts in a degeneracy the authors themselves flag, plus a UVJ analysis that is partly circular. Treat the 7 objects as strong candidates, not established z~4 massive galaxies.\n\nWhat is actually new: this is the first systematic wide-area application of the g-dropout LBG + FIRST matching approach to find z~4 radio galaxies, extending the single-object work of Yamashita et al. (2020) to 560 deg^2. The authors assemble 146 photometric HzRG candidates, focus on 28 with NIR/MIR coverage, and end with 7 objects whose SEDs are consistent with z~3.3-4.5. They also show that only 4 of the 7 satisfy the USS criterion, which is a useful concrete demonstration that Lyman-break selection catches radio galaxies that steep-spectrum searches miss. The comparison with USS-selected HzRGs is accompanied by an honest discussion of the Mstar > 10^11 Msun selection floor, backed by a 17,856-model completeness simulation. That simulation is real work and it supports the claim that the sample is biased toward high masses.\n\nThe photometry is also handled carefully: they redo aperture photometry for sources missing from the VIKING/UKIDSS catalogs, validate against HSC y-band, and estimate errors from sky fluctuations. The SED fits are internally consistent, with reduced chi-squared from 0.6 to 1.6, and Bayesian mass errors of 20-30%. The sfh2exp robustness check is a good-faith attempt to show the quenching result is not an artifact of one SFH parameterization.\n\nSoft spots, in proportion: the main one is spectroscopic confirmation. The free photo-z run placed 21 of the 28 candidates elsewhere, and the final 7 were selected after restricting the redshift prior to 3.3-4.5. The paper itself cites Kubo et al. (2024) showing z~2 fast-quenching galaxies can mimic the near-IR signature of z~4 galaxies. If any or all of these 7 are actually at z~2, the masses and the UVJ interpretation collapse. The authors do not overclaim in the text, but the abstract and conclusion state the masses and quenching timescale as results, and that framing needs a prominent 'pending spectroscopic confirmation' caveat.\n\nSecond, less severe: the UVJ colors are not measured; they are generated from the best-fit X-CIGALE models whose SFH parameters already encode the quenching history. The simulation and sfh2exp test mitigate this, but they do not fully break the circularity. This is a moderate caveat, not a fatal flaw, because the paper is transparent about the method.\n\nThe citation pattern looks appropriate, with self-citations limited to the relevant WERGS series and methodological precedents.\n\nBottom line: readers working on high-z massive galaxies, radio galaxies, or photo-z selection will get value from this. I would send it to a serious referee, and I would cite it as a candidate sample, not as a confirmed measurement. If I were the editor, I would ask for one more revision round emphasizing the confirmation status and softening the abstract's tone.","headline":"A careful, well-caveated first systematic sample of g-dropout-selected z~4 radio galaxy candidates with plausible masses, but the absence of spectroscopy and the model-derived UVJ colors mean the headline numbers should be read as candidates, not measurements.","tokens_in":24408,"tokens_out":2172,"would_cite":true,"duration_ms":23892,"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":"Seven g-dropout radio galaxies at $z\\sim4$ are as massive as ultra-steep-spectrum-selected ones and show fast quenching.","keywords":["high-redshift radio galaxies","Lyman break galaxies","photometric redshifts","stellar mass assembly","galaxy quenching","UVJ diagram","radio spectral index","z~4 galaxies"],"falsifier":"Spectroscopic redshifts for the seven candidates would settle the claim: if their Lyman-$\\alpha$ or rest-frame UV features place them at $z\\sim2$ rather than $z\\sim3.3{-}4.5$, the claimed stellar masses and fast-quenching timescales at $z\\sim4$ would collapse. A definitive test would be deep near-infrared spectroscopy targeting the 4000 Angstrom/Balmer break region expected at $z\\sim4$, or detection of a confirmed Lyman-$\\alpha$ line at $z\\sim3.8{-}4.0$.","tokens_in":23375,"feed_emoji":"🌌","tokens_out":7608,"duration_ms":63260,"temperature":0.7,"pith_summary":"The paper argues that radio galaxies selected through the Lyman-break technique, rather than by ultra-steep radio spectra, are already massive and passive by $z\\sim4$. Cross-matching g-dropout Lyman-break galaxies from the Subaru HSC survey with VLA FIRST radio sources yields 146 candidates, of which seven have optical-to-mid-infrared SEDs consistent with $z\\sim4$. SED fitting gives stellar masses between $2.6\\times10^{11}\\,M_\\odot$ and $5.6\\times10^{11}\\,M_\\odot$, with an average of $4.2\\times10^{11}\\,M_\\odot$, similar to previously known USS-selected high-redshift radio galaxies. The rest-frame UVJ diagram places these objects near or in the quiescent region and favors a fast quenching timescale of roughly 0.1 Gyr. If true, this means massive, already-quenched galaxies were present among radio-loud systems at $z\\sim4$ even when selection did not require a steep radio spectrum.","feed_headline":"Seven radio galaxies at z~4 are born massive and quenched fast","feed_subtitle":"A Lyman-break search finds galaxies as heavy as ultra-steep-spectrum ones, hinting early massive galaxies formed fast.","key_machinery":"The argument is carried by the g-dropout Lyman-break selection, which isolates galaxies at $z\\sim3.3{-}4.5$ through the color cuts $g-r>1.0$, $r-i<1.0$, and $g-r>1.5(r-i)+0.8$, followed by a 1-arcsec positional match to FIRST radio sources. Among 146 candidates, the 28 with VIKING/UKIDSS near-infrared and unWISE mid-infrared photometry were fitted with X-CIGALE using BC03 stellar population models and delayed star-formation histories; the best-fit SEDs were then used to construct rest-frame UVJ colors. The UVJ diagram with BC03 evolutionary tracks for $\\tau=0.1$, $0.5$, and $1.0$ Gyr is the diagnostic that supports the fast-quenching claim.","core_discovery":"On the paper's own terms, the discovery is that a Lyman-break-selected sample of radio galaxies at $z\\sim4$, chosen without requiring an ultra-steep radio spectrum, contains objects with stellar masses of $(2.6{-}5.6)\\times10^{11}\\,M_\\odot$ (average $4.2\\times10^{11}\\,M_\\odot$), placing them at the massive end of the $z\\sim4$ galaxy mass function. Their rest-frame UVJ colors, derived from the best-fit SED models, are consistent with a fast past quenching of roughly 0.1 Gyr. The paper also finds that the radio spectral indices among the seven are diverse: four satisfy the USS criterion and three do not, so Lyman-break selection catches radio galaxies that USS surveys miss.","pith_inferences":["A direct extension would be deep near-infrared spectroscopy of the seven candidates to confirm their redshifts; if the photo-$z$ redshifts hold, these objects become prime targets for proto-cluster searches around powerful radio jets.","ALMA or submillimeter observations could check whether the modest $E(B-V)$ values from SED fitting hide obscured star formation, which would change the inferred quenching completeness.","Comparing the radio luminosity function of Lyman-break-selected high-redshift radio galaxies with the quasar luminosity function at $z\\sim4$ could test whether the fast quenching is synchronized with luminous AGN phases.","If confirmed, the mass similarity between Lyman-break and USS-selected high-redshift radio galaxies would strengthen the view that such galaxies trace the most massive halos at $z\\sim4$."],"forward_implications":["The $z\\sim4$ radio galaxy population is not confined to ultra-steep-spectrum sources; Lyman-break selection reveals comparable massive galaxies with flatter radio spectra.","These objects sit near the massive end of the $z\\sim4$ stellar mass function, implying substantial early assembly before cosmic noon.","The fast-quenching signal favors a quenching mechanism acting on roughly 0.1 Gyr timescales, such as quasar-mode feedback, over slower radio-mode feedback.","Because the sample requires near-infrared and mid-infrared detection, it is biased toward $M_\\star>10^{11}\\,M_\\odot$; the true average mass of Lyman-break-selected high-redshift radio galaxies could be lower.","Future deeper infrared surveys can measure the full stellar-mass distribution and test whether Lyman-break-selected and USS-selected high-redshift radio galaxies form one population."],"supporting_citations":[{"why":"Defines the g-dropout LBG color criteria and the expected $z\\sim3.3{-}4.5$ redshift range used to build the parent sample.","marker":"Ono et al. (2018)"},{"why":"Provides the HSC-SSP photometry and aperture-matched magnitudes adopted for color selection and total flux measurement.","marker":"Harikane et al. (2022)"},{"why":"Supplies the X-CIGALE code used for all SED fitting and photometric redshift estimation.","marker":"Yang et al. (2020)"},{"why":"BC03 stellar population models provide the SED templates from which stellar masses and ages are derived.","marker":"Bruzual & Charlot (2003)"},{"why":"Provides the rest-frame UVJ quiescent-galaxy selection criteria used to classify the candidates.","marker":"Muzzin et al. (2013)"},{"why":"Supplies the fast-quenching evolutionary tracks and timescale interpretation used in the UVJ diagram.","marker":"Belli et al. (2019)"},{"why":"Demonstrates that $z\\sim2$ fast-quenching galaxies can mimic the near-infrared signature of $z\\sim4$ galaxies, motivating the caution in interpreting the photometric redshifts.","marker":"Kubo et al. (2024)"},{"why":"Provides comparison stellar masses of USS-selected high-redshift radio galaxies at $z\\sim4$.","marker":"Seymour et al. (2007)"},{"why":"Provides comparison masses for USS-selected high-redshift radio galaxies in the $3