{"id":"ab58eff2-bf9b-446f-8c41-7818fd9ce3dc","arxiv_id":"2607.19473","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":7,"one_line_summary":"At z≈0.43, massive star-forming galaxies host 3x more O VI in their inner halos than quiescent galaxies of the same stellar mass, with an O VI mass offset of ~1.5 dex, contradicting the virial-thermometer prediction.","lead":"GOLIATH is a new space-telescope survey of extremely rare massive galaxies that are still actively forming stars, measuring the hot oxygen gas in their outskirts. The measurements show these starburst galaxies hold about three times more O VI than equally massive dead galaxies, pointing to star formation—not just halo size—as the engine of this gas.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"SF–quiescent contrast at log M* ≈ 11 hinges on sSFR classification of four borderline GOLIATH sightlines; excluding them lowers the SF covering fraction from ~62.5% toward ~57% and may erase the nominal 3× contrast.","rationale":"The reader's weakest-assumption identified the same point: the GOLIATH systems are classified as star-forming from Prospector sSFR fits, but at least four high-mass sightlines have sSFR upper limits or values at/below the adopted threshold. This is precisely the condition that must hold for the central claim—that star-forming activity, not halo mass, regulates the warm CGM—because the high-mass SF sample is almost entirely GOLIATH. The concern is concrete and testable from existing data: the same MODS/KCWI spectra used for redshifts can provide emission-line classifications or Dn4000, and the Prospector posteriors can yield P(log sSFR > −11). The paper's own GOLIATH-only covering fraction does not resolve the issue because it mixes all GOLIATH systems regardless of sSFR. Other issues (J1319 column-density ambiguity, statistical inconsistencies between sections, turbulence-unconstrained cooling interpretation) are real but secondary; they affect the physical interpretation or individual sightlines, whereas the sSFR classification directly controls the magnitude and significance of the headline SF–quiescent contrast. A sensitivity test with the four borderline systems reclassified is the single most decisive check. If the contrast survives that test, the reader's CONDITIONAL verdict stands; if not, the central claim would need to be substantially weakened. I therefore keep the verdict unchanged at CONDITIONAL, with the same core concern the reader flagged.","tokens_in":34800,"tokens_out":18947,"duration_ms":183568,"concrete_test":"Re-run the Section 4.4 and 4.2 analysis with the four borderline systems (J1319+2728, J0956+2515, J0912+2450, J0958+3224) moved to the quiescent bin, using the Table 1 sSFR upper limits as the classification criterion. Then recompute the high-mass inner-CGM SF covering fraction at log N ≥ 14 and the SF–Q mean column-density offset. If the covering fraction drops below ~50% or the SF–Q offset falls below ~0.3 dex (or the Fisher exact p-value exceeds 0.05), the central contrast is not robust to the classification ambiguity.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The most load-bearing step for the paper's central claim is the binary sSFR cut used to define the high-mass star-forming sample. In the log M* = [11,12) inner-CGM bin, the SF covering fraction is quoted as 62.5% (5/8) and the ~0.5 dex mean column-density offset depends on which GOLIATH sightlines are counted as SF. Table 1 lists four systems with sSFR at or below the adopted threshold: J1319+2728 (log sSFR < −12.8), J0956+2515 (< −16.3), J0912+2450 (< −12.2), and J0958+3224 (−11.1). The first three are formally quiescent by the paper's own criterion; J0958 sits at the green-valley boundary and is the only one with an error bar that reaches the SF threshold. If all four are excluded, the high-mass SF set shrinks from 11 to 7 clear-SF sightlines, and the N≥14 covering fraction becomes 4/7 ≈ 57% (4/8 = 50% if J0958 is also excluded), with correspondingly wider confidence intervals and a weaker Fisher exact test. The excluded systems include two detections (J1319 at 14.01, J0912 at 14.31) and one marginal detection (J0958 at 13.97); moving them to the quiescent side also raises the quiescent mean, so the quoted 'factor ~3' in column density and '~1.5 dex' in OVI mass could shrink substantially. Because the paper does not report Prospector posterior probabilities or independent emission-line classifications for these systems, the robustness of the central SF–Q contrast is currently unresolved. The GOLIATH-only covering fraction of 64.7% includes all GOLIATH systems, so it does not by itself settle the SF–Q comparison.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents the GOLIATH survey, an HST/COS program targeting the inner CGM of 18 massive (log M*/M_sun ≈ 11), ostensibly blue, starburst/post-starburst galaxies at z ≈ 0.43, and compares them with archival star-forming and quiescent samples. It reports that in the log M*/M_sun = [11,12) inner-CGM bin, star-forming galaxies have OVI column densities a factor of ~3 higher, CGM OVI masses ~1.5 dex higher, and covering fractions ~2.6 times higher (62.5% vs 24%) than quiescent galaxies of similar mass. The paper interprets this as evidence that feedback-driven radiative cooling, not virial temperature, regulates the warm CGM at high mass, and that OVI is a short-timescale tracer of quenching.","tokens_in":35292,"tokens_out":10099,"duration_ms":95110,"significance":"If the central contrast is robust, the survey fills a genuinely unoccupied region of the M*–SFR plane and directly discriminates between the virial-thermometer and feedback scenarios. The paper's strengths include new COS spectroscopy, joint Voigt-profile fitting with MCMC-derived uncertainties, treatment of non-detections as censored upper limits, and two independent OVI-mass estimators that agree with each other and with literature values at overlapping mass. However, the headline SF–quiescent contrast rests on a small number of GOLIATH sightlines whose star-forming classification is not robust, and on one sightline whose reported column density is internally inconsistent. These issues must be resolved before the strong abstract-level claims can be accepted.","major_comments":[{"comment":"The central 62.5% vs 24% covering fraction and ~1.5 dex mass-offset claims depend on the binary sSFR cut at 10^-11 yr^-1 (Section 3.2). In the high-mass bin, four GOLIATH sightlines have sSFR upper limits at or below that cut: J1319+2728 (<−12.8), J0956+2515 (<−16.3), J0912+2450 (<−12.2), and J0958+3224 (−11.1+0.3−0.6). By the paper's own criterion the first three are quiescent and J0958 is at best marginal. Reclassifying them changes the high-mass SF covering fraction from 62.5% (5/8) to ~57% (4/7) or ~50% (4/8) and raises the quiescent mean, directly weakening the claimed factor-of-three contrast. Please report Prospector posterior probabilities and independent emission-line classifications for these systems, and rerun the Fisher exact test, covering fractions, and mass estimates under alternative classification choices.","section":"Table 1; §4.4"},{"comment":"The most massive sightline J1319+2728 is reported inconsistently. The text assigns log N = 14.01 to the system associated with the primary galaxy, Table 3 lists the integrated column as 14.01 but tabulates five components including two at v ≈ −1765 and −1619 km/s whose linear sum is 14.65, and Fig. 7 labels log N = 14.65 integrated over ±2000 km/s. The Table 3 note says only components within −300 to +200 km/s are used, but this does not match the listed rows, and the figure caption quotes −200 to +300 km/s. Because J1319 is the most massive object and would fall in the quiescent high-mass bin, this inconsistency affects the quiescent covering fraction and mass estimates. Clarify the adopted velocity window and propagate a single value through all analysis.","section":"Table 3; §4.5; Fig. 7"},{"comment":"The stated survey selection requires extremely blue galaxies with u−r < 1.65, but several Table 1 entries exceed this value (J1126+1204: 1.86; J1342−0053: 1.66; J1305+5301: 2.53; J1244+0755: 1.68; J1405+4704: 1.74), and others are upper limits (J1319+2728 <2.64; J0912+2450 <2.83) that do not by themselves prove blue colors. Because the sample definition underpins the claim that these are starburst/post-starburst systems, please clarify the color definition/redshift correction and discuss which sightlines actually satisfy the 'extremely blue' criterion.","section":"§2; Table 1"},{"comment":"The feedback-driven radiative cooling conclusion is currently supported by a visual comparison of component column densities and line widths with cooling curves. No quantitative test distinguishes the T≈5.5 K curve from the T≈6.0 K curve or from photoionization models, and the b-parameter uncertainties in Table 3 are large. Since the radiative-cooling interpretation is a headline claim (Abstract, Conclusion 3), add a statistical comparison that accounts for b uncertainties and upper limits, or soften the claim to 'consistent with, but not discriminating among, these scenarios.'","section":"§5.1; Fig. 10"}],"minor_comments":[{"comment":"The text states that all GOLIATH >3σ detections reside above log N_Th = 14.0, but J0958+3224 has log N = 13.97. Specify the threshold rounding convention or the exact counting rule used for the 62.5% covering fraction.","section":"§4.4"},{"comment":"The generalized Kendall's tau value is reported as τ = 0.191 in Section 4.2 and τ = 0.193 in the Conclusions. Unify the values and quote the same p-value throughout.","section":"§4.2 vs Conclusions"},{"comment":"The full-sample Kendall test is described as showing 'a small monotonic decrease' with τ = −0.068, which is confusing because the subsequent inner-CGM test gives a strongly positive τ = 0.393. Clarify that the sign reflects the R/Rvir selection effect and report both tests with consistent wording.","section":"§4.3"},{"comment":"Figure numbering is duplicated: the main-text Fig. 10 (OVI column density vs line width) and the Appendix spectra are all labeled 'Figure 10'. Renumber the appendix figures.","section":"Appendix"},{"comment":"The text refers to SDSS redshifts for 'J0906+015'; from the table this appears to be a typo for J0909+0121. Check all target names for consistency.","section":"§3.1"},{"comment":"Section 2 says the sample comprises 20 sightlines with 12 new and 8 archival, while Table 2 lists N=18 for GOLIATH. Clarify that two archival sightlines (J1126+1204 and J1240+0949) lack OVI coverage and are excluded from the absorption analysis.","section":"§2; Table 2"},{"comment":"The quiescent power-law fit is quoted as α = −0.28 with no uncertainty and appears to be driven by very few points. Report the number of sightlines/bins and the fit uncertainties, or present the quiescent relation as an upper envelope rather than a fitted power law.","section":"§4.6.3"}],"recommendation":"major_revision","confidential_remarks":"The observational dataset is valuable and the paper addresses a real gap, but the headline SF–quiescent contrast is sensitive to a small number of classification choices and to the internally inconsistent J1319 column density. I would ask for a robustness table showing all central statistics under alternate sSFR classifications before considering this for publication. The fact that the two theoretical models compared are from co-authors is not by itself problematic, but an independent simulation comparison or a more quantitative model test would substantially raise confidence in the feedback-driven cooling interpretation."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Good observational paper with a real gap to fill: nobody had measured OVI around massive, actively star-forming galaxies at z~0.4. GOLIATH gives you 18 sightlines, careful joint Voigt fitting, MCMC, and upper limits handled as censored data. That part is solid. The claim that massive SF galaxies have a higher OVI covering fraction than quiescent galaxies at fixed stellar mass is probably true—the clear-SF subset already shows 5 detections out of 8, and the archival quiescent samples sit low.\n\nBut the paper's own classification rules are applied loosely. Table 1 lists J1319+2728, J0956+2515, and J0912+2450 with sSFR upper limits below the adopted -11 cut; J0958 sits right at the boundary. By the paper's own criterion these are quiescent, yet they are treated as part of the SF story. Two of them have detections (14.01 and 14.31), so moving them to the quiescent bin would eat into the 3x column-density contrast and the 1.5 dex mass offset. The 62.5% covering fraction is really 5/8 in the clear-SF subset; it may survive, but the paper should show the numbers with and without the borderline systems. Right now the SF–Q contrast is less robust than the abstract implies.\n\nThere's also a concrete inconsistency on J1319: the text and Table 3 give log N = 14.01 for the primary system, while the Figure 7 caption says 14.65 integrated over ±2000 km/s. That has to be reconciled. And the GOLIATH-only covering fraction of 64.7% doesn't match the 11/18 detections I count at log N >= 14; the paper should clarify exactly which sightlines enter the denominator.\n\nThe interpretation part is the weakest. The consistency with feedback-driven radiative cooling rests on a visual match to model curves in Figure 10, with no constraint on turbulence. It's a plausible story, not a test. The simulation comparison is honest about the fact that GOLIATH-like systems are not modeled, so the 'inconsistent with virial thermometer' claim is an extrapolation.\n\nOverall: the data are a contribution and deserve a serious referee. The paper needs a careful revision on the classification, the J1319 reporting, the covering fraction arithmetic, and a less confident tone on the cooling model. I'd send it out.","headline":"Solid new dataset, but the SF–Q contrast and the cooling-flow interpretation are both softer than the abstract suggests.","tokens_in":35888,"tokens_out":11500,"would_cite":true,"duration_ms":100568,"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":"Massive star-forming galaxies carry roughly three times more warm OVI gas in their inner halos than quiescent galaxies of the same stellar mass, pointing to star formation—not halo mass—as the regulator of the warm circumgalactic medium at","keywords":["circumgalactic medium","OVI absorption","starburst galaxies","galaxy quenching","massive galaxies","galaxy evolution","absorption-line spectroscopy","warm-hot gas"],"falsifier":"Obtain deep rest-optical emission-line spectroscopy for the four ambiguous GOLIATH systems (J1319+2728, J0956+2515, J0912+2450, J0958+3224) to measure firm sSFRs. If two or more classify as quiescent, the star-forming covering fraction at log N_OVI ≥ 14 drops from 62.5% toward the quiescent 24%, directly weakening the central contrast. A complementary test: measure OVI around a sample of spectroscopically confirmed massive starbursts with independent SFR indicators (e.g., H-alpha or mid-infrared) to see whether the ~3x enhancement persists.","tokens_in":34679,"feed_emoji":"🌌","tokens_out":5389,"duration_ms":49942,"temperature":0.7,"pith_summary":"The GOLIATH survey targets a rare population: massive, blue, starburst and post-starburst galaxies at an average redshift of 0.43, caught on the eve of quenching. Using ultraviolet spectra of background quasars, it measures OVI absorption from the warm-hot circumgalactic medium inside 0.6 times the virial radius. The central result is that these massive star-forming galaxies have OVI column densities about three times higher, and CGM OVI masses about 1.5 dex higher, than quiescent galaxies of the same stellar mass. The paper argues this breaks a long-standing degeneracy: at stellar masses above 10^11 solar masses, star-formation activity—not halo virial temperature—controls the warm oxygen reservoir. Because the OVI cooling time is only 10–100 million years, the reservoir must be continuously replenished by feedback, making OVI a short-timescale tracer of the starburst-to-quiescent transition.","feed_headline":"Warm gas around massive galaxies tracks star formation, not mass","feed_subtitle":"GOLIATH survey finds 3x more OVI in starbursts than quiescent giants, rewriting how quenching is traced.","key_machinery":"The central observable is the OVI doublet (1031, 1037 Å) absorption in far-ultraviolet quasar spectra, which traces collisionally ionized gas at T ≈ 10^5.5 K. The analysis rests on comparing inner-halo sightlines (R/Rvir ≤ 0.6) of mass-matched star-forming and quiescent galaxies, using two independent estimators of total CGM OVI mass: an empirical annular integration weighted by covering fraction, and a radial-profile beta-function fit with censored upper limits. Together these turn sparse pencil-beam sightlines into a census of warm oxygen per halo.","core_discovery":"The survey fills a previously empty corner of the mass–star-formation plane: galaxies that are simultaneously massive (median log M*/M_sun ≈ 11.1) and actively forming stars. In the inner CGM (R/Rvir ≤ 0.6), star-forming galaxies in the log M*/M_sun = [11,12) bin exceed quiescent galaxies by a factor of ~3 in mean OVI column density and ~1.5 dex in total CGM OVI mass (log M_OVI/M_sun ≈ 7.3 versus ≈ 5.8), with covering fractions of 62.5% versus 24% at log N_OVI ≥ 14. OVI column density rises with both stellar mass and specific star-formation rate across the star-forming population, and the detected kinematic components remain gravitationally bound with velocity spreads up to ~1200 km/s. The a","pith_inferences":["The large scatter in GOLIATH column densities hints that some sightlines are already partially depleted; a larger sample of post-starburst galaxies could map the depletion curve and directly time the quenching event.","If the same trend holds at even higher masses, the most massive star-forming galaxies might show OVI masses approaching 10^7.5 M_sun, a testable prediction for future ultraviolet spectroscopy of rare hyper-luminous starbursts.","The bound kinematics suggest that outflows at this mass scale recycle gas within the halo rather than ejecting it; if so, the high OVI mass may be connected to metal retention, which could be tested by measuring the metallicity of the cool CGM phase.","Green-valley galaxies should show intermediate OVI covering fractions; targeted observations of this population could provide an independent check of the rapid-depletion picture."],"forward_implications":["Massive star-forming galaxies at log M*/M_sun ≈ 11 keep a warm oxygen reservoir despite virial temperatures that should ionize OVI away, so halo temperature alone cannot explain the observed OVI bimodality.","The CGM OVI mass rises with stellar mass for star-forming galaxies (power-law slope ≈ 0.5) while declining for quiescent galaxies, so the star-forming—quiescent gap widens toward higher mass.","The ~10–100 Myr cooling time means OVI is a short-timescale feedback tracer; it should disappear quickly after quenching, before the galaxy's colors fully transition.","Current cosmological simulations underpredict the OVI budget in massive halos; reproducing these measurements may require higher oxygen yields or more efficient metal transport by feedback.","OVI absorption can be used to catch galaxies in the act of quenching, since the warm gas depletes much faster than the ~1 Gyr star-forming-to-quiescent transition."],"fun_headline_variants":["OVI traces star formation, not stellar mass, in massive halos","Starburst galaxies hold 3x more OVI in their halo than quiescent ones","CGM OVI spikes in starbursts, not quiescent giants, at same mass","Feedback not mass sets hot gas in massive galaxies, OVI shows","Warm-hot CGM gas tracks quenching: OVI 3x higher in starbursts"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The GOLIATH galaxies are classified as star-forming on the basis of SED-derived specific star-formation rates, but for four of the 18 OVI sightlines the sSFR is an upper limit below the adopted star-forming threshold, and two have color upper limits that do not prove blue colors; if those galaxies are actually quiescent or post-starburst contaminants, the high-mass covering fraction and the 1.5 dex OVI mass offset would shrink.","fun_headline_variants_meta":{"raw":{"variants":["OVI traces star formation, not stellar mass, in massive halos","Starburst galaxies hold 3x more OVI in their halo than quiescent ones","CGM OVI spikes in starbursts, not quiescent giants, at same mass","Feedback not mass sets hot gas in massive galaxies, OVI shows","Warm-hot CGM gas tracks quenching: OVI 3x higher in starbursts"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000862,"raw_usage":{"total_tokens":3735,"prompt_tokens":1060,"completion_tokens":2675,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":804,"completion_tokens_details":{"reasoning_tokens":2563}},"tokens_in":804,"tokens_out":2675,"duration_ms":17516,"temperature":1.0,"reasoning_tokens":2563,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T12:38:10.417379+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Obtain deep rest-optical emission-line spectroscopy for the four ambiguous GOLIATH systems (J1319+2728, J0956+2515, J0912+2450, J0958+3224) to measure firm sSFRs. If two or more classify as quiescent, the star-forming covering fraction at log N_OVI ≥ 14 drops from 62.5% toward the quiescent 24%, directly weakening the central contrast. A complementary test: measure OVI around a sample of spectroscopically confirmed massive starbursts with independent SFR indicators (e.g., H-alpha or mid-infrared) to see whether the ~3x enhancement persists.","supporting_citations":[],"review_version":1}