{"id":"c9d36e0f-e71c-4729-9d1f-3412a1c96c69","arxiv_id":"2608.04371","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":9,"one_line_summary":"The FUDS0 pilot HI sample yields a baryonic Tully-Fisher slope k=3.32 and shows no significant evolution at z<0.42, though the highest-redshift bin constrains it only indirectly.","lead":"Using 74 neutral hydrogen galaxies from the FAST Ultra-Deep Survey pilot field, this paper measures the baryonic Tully-Fisher relation and finds a slope of about 3.3 with no significant evolution below redshift 0.42. The work is a check on whether the tight galaxy mass-rotation relation stays constant over the last four billion years.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The no-evolution claim at z=0.12–0.42 rests on the untested assumption that the outlier component's slope and vertical offset are redshift-independent; Section 7.2.2 fixes both from the full IGS sample before inferring the high-z BTFR zero point, so a redshift-evolving δ_Outlier would masquerade…","rationale":"The paper's strongest contribution is a careful local BTFR measurement: the HQS and IGS fits are consistent, the Gaussian mixture method is validated on mocks, and the random forest analysis plausibly identifies SNR and inclination uncertainty as the main drivers of the outlier population. Those parts of the argument are credible and deserve credit. The load-bearing weakness is confined to the high-redshift leg of the no-evolution claim. The reader's weakest_assumption identifies exactly this point: the high-z inference assumes that the outlier component shares the full-sample slope and vertical offset. My reading of Table 1 and Sections 7.2.1–7.2.2 confirms this is the pivotal untested assumption. The paper itself states the limitation honestly in Section 7.2.2, noting that genuine evolution might be masked, but the abstract and summary present the indirect high-z result as supporting the no-evolution conclusion. Because the high-z bin is the only evidence beyond z≈0.12, the headline claim 'no significant evolution across z<0.42' is not fully demonstrated. The appropriate response is a conditional acceptance: require either a demonstration that δ_Outlier is redshift-independent (e.g. through the injected-evolution simulation described above) or a softening of the claim to the directly measured z<0.12 range. This does not invalidate the local measurement or the outlier-origin analysis, so a conditional verdict rather than rejection is warranted.","tokens_in":27701,"tokens_out":4421,"duration_ms":44940,"concrete_test":"Use the Appendix C mock pipeline, which already includes measurement uncertainties, selection criteria, and redshift distributions, but inject a redshift-dependent outlier offset, e.g. δ_Outlier(z)=0.47+0.2×(z/0.42), while keeping the true C_BTFR fixed. Then apply the exact Section 7.2.2 procedure to the high-z mock bin: fix k=3.32, fix δ_Outlier to the full-sample value, and record the recovered b. If the recovered b shifts by more than the quoted 1σ uncertainty (~0.20) or the procedure reports 'no evolution' despite the injected evolution, the high-z inference is not robust. As a secondary check, refit the 17 real high-z galaxies with δ_Outlier as a free parameter, using a prior derived only from the z<0.12 sample, and test whether b remains consistent with the low-z value.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim is that the BTFR does not evolve significantly out to z=0.42. Direct support comes only from the two low-redshift bins (z≤0.08 and 0.08<z≤0.12), with 25 and 32 galaxies respectively. In the highest bin (z=0.12–0.42, 17 galaxies) the C_BTFR component vanishes; Section 7.2.2 therefore infers the BTFR zero point by fitting the C_Outlier component with k fixed to 3.32 and δ_Outlier fixed to 0.47, both taken from the full IGS sample. This is not an independent evolutionary test: the full-sample δ_Outlier is dominated by the same low-z galaxies that already define the no-evolution baseline, so fixing it removes the very redshift dependence the test is meant to detect. If δ_Outlier changes with redshift—for example because high-z detections are systematically lower in SNR or more affected by inclination errors, or because genuine evolution moves galaxies out of C_BTFR—the inferred b is biased by the same amount. Table 1 shows the sensitivity: in the high-z bin, allowing δ_Outlier to float changes the inferred zero point from b=10.21 (fixed δ) to b=10.46 (fixed k, free δ), a shift comparable to or larger than the quoted 1σ uncertainties. The paper itself acknowledges in Section 7.2.2 that a transformation from C_BTFR to C_Outlier caused by evolution cannot be ruled out and that genuine evolution might be masked. Consequently, the statement that the high-z outlier component 'aligns with the conclusion' of no evolution is an overreach; the no-evolution claim at z>0.12 is not established by the current data.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper uses 74 HI-selected galaxies from the FUDS0 pilot field (with a 24-galaxy high-quality subsample) to measure the baryonic Tully-Fisher relation at redshifts up to z=0.42. A two-component Gaussian mixture model yields slope k=3.32(+0.12,-0.11), zero point b=10.07(+0.03,-0.03), and intrinsic scatter sigma_BTFR=0.036(+0.010,-0.009) for the tight component, together with a dispersed outlier component. A random forest analysis attributes outliers primarily to low signal significance and inclination uncertainties. The sample is divided into three redshift bins, and the paper claims no significant evolution of slope and zero point out to z=0.42, with the highest-redshift bin constrained indirectly through the outlier component.","tokens_in":28152,"tokens_out":4912,"duration_ms":50261,"significance":"If the low-scatter baryonic Tully-Fisher relation indeed persists to z~0.4 in an HI-selected sample, this is a valuable addition to the sparse HI-based constraints on BTFR evolution and has implications for using the BTFR as a distance indicator at moderate redshift. The paper has genuine strengths: the MCMC fitting is carefully described, the mock verification in Section 5.3 and Appendix C is a serious attempt to test recovery of the model parameters under selection effects and measurement errors, and the authors explicitly acknowledge in Section 7.2.2 that genuine evolution might be masked by observational effects. However, the highest-redshift bin contains no detected C_BTFR component, so the no-evolution conclusion at z=0.12-0.42 is not an independent measurement; it rests on an untested assumption that the outlier component's vertical offset is redshift-independent.","major_comments":[{"comment":"The statement that the highest-redshift bin (z=0.12-0.42) 'aligns with the conclusion' of no zero-point evolution is not supported as an independent measurement. The quoted indirect zero point b=10.21(+0.20,-0.19) is obtained by fixing both k=3.32 and delta_Outlier=0.47 to the full IGS sample values. When delta_Outlier is instead allowed to float with k fixed (Table 1, row 6), the inferred zero point shifts to b=10.46(+0.27,-0.68), a change comparable to or larger than the quoted 1-sigma uncertainty. Because the full-sample delta_Outlier is dominated by the same low-redshift galaxies that already define the no-evolution baseline, fixing it removes the very redshift dependence the test is meant to detect. The paper itself concedes in Section 7.2.2 that a transformation from C_BTFR to C_Outlier caused by evolution cannot be ruled out and that genuine evolution might be masked, so the abstract's 'aligns with the conclusion' wording overreaches.","section":"7.2.2, Table 1"},{"comment":"The mock verification does not validate the specific procedure used for the high-redshift bin. The step-function evolution test injected in Section 5.3 varies the BTFR slope k and zero point b between redshift bins but keeps the outlier offset delta_Outlier fixed at the input value. It therefore cannot detect a bias in the indirect zero-point reconstruction if delta_Outlier itself evolves with redshift. Since the random forest analysis in Section 7.1 identifies low SNR and inclination uncertainties as the main drivers of C_Outlier, and both effects worsen with redshift, a redshift-dependent delta_Outlier is a plausible alternative that would masquerade as no evolution under the fixed-offset procedure. The manuscript should either fit a model in which delta_Outlier is allowed to vary with redshift, or explicitly restrict the no-evolution claim to z<=0.12.","section":"5.3, Appendix C"},{"comment":"The slope-evolution claim at high redshift is also weaker than the text suggests. The z=0.12-0.42 bin contains only 17 galaxies and no detected C_BTFR component; the free-slope fit in that bin gives k=3.97(+0.72,-1.00), which is consistent both with the low-redshift slopes and with substantially different values. The statement that the shared slope derived from C_Outlier 'implies' a consistent C_BTFR slope relies on the model assumption that both components share the same slope and that the outlier population is described by the same relation at all redshifts. The data in this bin are too sparse to confirm no slope evolution, and the text should present this bin only as a weak consistency check, not as independent confirmation.","section":"6.1, 7.2.1, Table 1"}],"minor_comments":[{"comment":"There are several typographical issues: 'fomation' in the abstract, 'Higalaxies' should be 'HI galaxies' throughout, and 'plat part' in Section 3.1 should be 'flat part'.","section":"Abstract, Section 1"},{"comment":"The HQS sample size is stated inconsistently: Section 2.2 says the final HQS sample has 24 galaxies, Section 6.1 says 26 galaxies, and Table 1 lists N=24. The text and figures should be harmonized.","section":"Section 2.2, Section 6.1, Table 1"},{"comment":"The histogram label 'Nubmer' is a typo for 'Number', and the figure caption should identify the panels more explicitly, since the current caption lists many variables without clear visual markers for each panel.","section":"Figure 1"},{"comment":"The random forest labels are derived from the same two-component model's posterior probabilities (Equation 20), so the importance ranking partly encodes the model's own definition of C_Outlier. This does not invalidate the conclusion, because SNR and inclination uncertainty are not inputs to the mixture fit, but the text should acknowledge that the classification is not fully independent of the model being interpreted.","section":"Section 7.1"}],"recommendation":"major_revision","confidential_remarks":"The low-redshift analysis and the outlier-origin study are well executed and likely publishable after revision. The main issue is that the abstract and summary overstate the evidentiary weight of the highest-redshift bin: the no-evolution claim at z=0.12-0.42 depends on fixing delta_Outlier to the full-sample value, which is not an independent test. I would encourage the editors to require that the authors either present a joint fit with redshift-dependent delta_Outlier or explicitly reframe the high-z result as a consistency check under an assumed constant offset. The paper's own caveats in Section 7.2.2 are a good starting point, but they need to be reflected in the abstract and summary."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: the paper delivers a clean, well-documented local BTFR measurement from a new FAST pilot field, and the low-redshift no-evolution test (z≤0.12) is fine. But the headline extension to z=0.42 is not actually tested. The high-z bin has only 17 galaxies and no C_BTFR component, so the authors fix the outlier offset δ_Outlier from the full sample and then read off the zero point. That is circular in exactly the way that matters: if δ_Outlier evolves with redshift, the inferred b is biased by the same amount. Table 1 shows the sensitivity: letting δ_Outlier float changes b from 10.21 to 10.46 (W_m,20), a shift comparable to the 1σ uncertainties. So the no-evolution conclusion at z>0.12 is conditional on an untested assumption.\n\nWhat the paper does well: the FUDS0 sample is genuinely new, from the FAST telescope, and the MCMC setup with the two-component Gaussian mixture model is described in detail. The mock verification—though not a formal reproducibility artifact—is a real strength, showing the method recovers static parameters and can detect strong evolution, while also flagging that detection is harder in the high-z bin. The random forest analysis attributing outliers to low SNR and inclination errors is sensible, and the completeness treatment is careful. The local fit (k=3.32±0.12, b=10.07±0.03, σ=0.036) is consistent with the HQS fit and with the ALFALFA slope, so the paper earns its local BTFR numbers.\n\nThe soft spot is the high-z inference, and I want to be fair: the authors do not hide it. Section 7.2.2 explicitly says that a transformation from C_BTFR to C_Outlier caused by evolution cannot be ruled out and that genuine evolution might be masked. Given that, the abstract's phrase \"aligns with the conclusion\" overstates what the data show. The direct evidence for no evolution covers z≤0.12; beyond that, the paper should present the result as a non-detection under a stated assumption, or marginalize over the redshift dependence of δ_Outlier. The HQS reference sample is also strongly selected and limited to z<0.15, which is fine for a local anchor but should not be leaned on for the evolution claim.\n\nWho is this for? Astronomers working on HI-selected scaling relations and BTFR evolution. It is a solid pilot result that deserves serious refereeing. The right path is to send it out, with a clear request to reframe the high-z claim in the abstract and conclusions to match the actually tested redshift range.\n\nRecommendation: accept for peer review, but the referee should push on the δ_Outlier assumption before the no-evolution headline is published.","headline":"Solid local BTFR from a new FAST field, but the no-evolution claim at z>0.12 is inference dressed as a detection, and the paper itself admits the assumption.","tokens_in":28707,"tokens_out":3361,"would_cite":true,"duration_ms":33696,"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":"Using 74 HI-selected galaxies from the FUDS0 pilot field, this paper finds the baryonic Tully-Fisher relation has slope k=3.32, zero point b=10.07, and no significant evolution out to z=0.42, with outliers dominated by measurement effects.","keywords":["baryonic Tully-Fisher relation","HI line emission","FAST Ultra-Deep Survey","galaxy kinematics","galaxy evolution","Gaussian mixture model","random forest classifier","rotation curves"],"falsifier":"Measure the BTFR directly at $0.12<z<0.42$ with a sample of a few hundred high-significance HI galaxies with accurate inclinations, fit the two-component model with $k$ and $\\delta_{\\rm Outlier}$ free, and test whether the recovered $C_{\\rm BTFR}$ zero point and slope agree with $k=3.32$, $b=10.07$; a departure beyond the quoted $1\\sigma$ uncertainties would falsify the no-evolution claim.","tokens_in":27494,"feed_emoji":"🔭","tokens_out":8711,"duration_ms":70417,"temperature":0.7,"pith_summary":"The paper's goal is to measure the baryonic Tully-Fisher relation (BTFR) at redshifts up to $z=0.42$ using 74 HI-selected galaxies from the pilot FUDS0 field, and to test whether the relation evolves. The central finding is a tight power law with slope $k=3.32_{-0.11}^{+0.12}$, zero point $b=10.07_{-0.03}^{+0.03}$, and intrinsic scatter $\\sigma_{\\rm BTFR}=0.036_{-0.009}^{+0.010}$ dex, consistent with the local relation. A second, dispersed component of outliers is shown by a random forest classifier to be driven mainly by low signal significance and poor inclination estimates, meaning the scatter is observational rather than evolutionary. Splitting the sample into redshift bins, the paper finds the slope and zero point stay consistent within $1\\sigma$ up to $z\\le0.12$, and infer the same for $z=0.12-0.42$ by extrapolating the outlier component. If correct, the BTFR remains a usable distance indicator and a stable reference at intermediate redshift.","feed_headline":"Baryonic Tully-Fisher relation holds steady to z=0.42","feed_subtitle":"74 HI galaxies from FUDS0 give a constant BTFR out to redshift 0.42","key_machinery":"The load-bearing mechanism is a two-component Gaussian mixture model fitted to the BTFR plane, written as $y=k(x-2.4)+b$ for the tight component and a parallel component offset by $\\delta_{\\rm Outlier}$ with its own scatter; a shared slope $k$ lets the model recover the main relation even when a large outlier population is present. The companion machinery is a random forest classifier that ranks galaxy properties by their ability to predict whether a galaxy belongs to $C_{\\rm BTFR}$ or $C_{\\rm Outlier}$, which is what links the outlier population to observational errors. The velocity measure is the rest-frame $W^{\\rm rest}_{\\rm rot,20}$ width from reprocessed Busy-function profiles, corrected for turbulence and inclination, and baryonic mass includes stellar mass plus 1.4 times the atomic plus predicted molecular gas.","core_discovery":"The paper argues that, in the $M_{\\rm bary}$--$W^{\\rm rest}_{\\rm rot,20}$ plane, the FUDS0 IGS sample is described by two Gaussian components that share one slope. The tight component, $C_{\\rm BTFR}$, has $k=3.32_{-0.11}^{+0.12}$, $b=10.07_{-0.03}^{+0.03}$, and $\\sigma_{\\rm BTFR}=0.036_{-0.009}^{+0.010}$, matching the high-quality HQS fit and the local ALFALFA relation in slope; the dispersed component, $C_{\\rm Outlier}$, sits at a vertical offset $\\delta_{\\rm Outlier}=0.47_{-0.13}^{+0.14}$ with larger scatter. The paper claims the tight component's slope and zero point do not change between the $z\\le0.08$ and $0.08<z\\le0.12$ bins, and that at $z=0.12-0.42$, where only the outlier component is detectable, fixing the slope and using the full-sample offset recovers a zero point consistent with no evolution. The random forest analysis is the evidence for the observational origin of the outliers: line-significance and inclination uncertainty outrank morphology, confusion, or redshift as predictors of membership.","pith_inferences":["An implication the paper leaves implicit is that a non-evolving BTFR could serve as a redshift-independent distance rung connecting local HI galaxies to $z\\sim0.4$ systems that lack other distance anchors.","If the outliers are truly measurement-driven, earlier claims of BTFR evolution based on optical rotation curves may need to revisit inclination and line-significance systematics before interpreting offset as cosmic evolution.","A testable extension is to apply the same two-component plus random-forest pipeline to the full FUDS survey and check whether $\\delta_{\\rm Outlier}$ is truly constant with redshift rather than an artifact of the pilot sample.","Since molecular masses are predicted from an optical scaling relation rather than measured, a redshift-dependent molecular fraction would shift the baryonic masses and could alter the inferred zero point; the no-evolution result inherits this uncertainty."],"forward_implications":["At $z<0.42$, baryonic masses of disk galaxies can be predicted from rotation widths with about $0.036$ dex intrinsic scatter once observational outliers are removed.","BTFR-based distance estimates, previously trusted only locally, remain a valid tool out to $z\\sim0.4$.","Simulation predictions of a flatter slope and higher intercept by $z\\sim0.5$ are not supported by this sample.","Targeting high signal significance and reliable inclinations is sufficient to obtain a clean BTFR, without invoking galaxy evolution to explain the scatter.","The planned full FUDS survey, with roughly six times the FUDS0 sample, should sharpen or challenge these constraints."],"supporting_citations":[{"why":"Supplies the FUDS0 catalog of 128 HI detections and the completeness function used to define and select the IGS/HQS samples.","marker":"Xi et al. 2024"},{"why":"Provides the multiwavelength counterparts and SED-derived stellar masses that enter $M_{\\rm bary}$.","marker":"Xi et al. 2025"},{"why":"Introduces the two-component Gaussian mixture model for the BTFR and the ALFALFA comparison sample.","marker":"Ball et al. 2023"},{"why":"Defines the $W_{m,20}$ and $W_{m,50}$ linewidth measures adopted as rotation-velocity proxies.","marker":"Courtois et al. 2009"},{"why":"Provides the empirical scaling relation used to estimate molecular gas mass for FUDS0 galaxies.","marker":"Tacconi et al. 2018"},{"why":"The MIGHTEE-HI study with which the paper's no-evolution result is directly compared at $z<0.081$.","marker":"Ponomareva et al. 2021"},{"why":"The BUDHIES $z\\sim0.2$ measurement that supplies the previous higher-redshift BTFR comparison.","marker":"Gogate et al. 2023"}],"fun_headline_variants":["BTFR holds firm out to z=0.42","74 galaxies keep Tully-Fisher relation constant to z=0.42","No evolution in baryonic Tully-Fisher relation to z=0.42","Tully-Fisher relation intact at z=0.42","Baryonic Tully-Fisher relation: no change to z=0.42"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The no-evolution conclusion at $z>0.12$ depends on assuming the outlier population at those redshifts sits at the same vertical offset from the main relation, and follows the same slope, as the outlier population in the full sample; if either of those changes with redshift, the inferred high-redshift zero point would be biased.","fun_headline_variants_meta":{"raw":{"variants":["BTFR holds firm out to z=0.42","74 galaxies keep Tully-Fisher relation constant to z=0.42","No evolution in baryonic Tully-Fisher relation to z=0.42","Tully-Fisher relation intact at z=0.42","Baryonic Tully-Fisher relation: no change to z=0.42"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000361,"raw_usage":{"total_tokens":2062,"prompt_tokens":1170,"completion_tokens":892,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":786,"completion_tokens_details":{"reasoning_tokens":795}},"tokens_in":786,"tokens_out":892,"duration_ms":7515,"temperature":1.0,"reasoning_tokens":795,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-08T18:49:36.786044+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the BTFR directly at $0.12<z<0.42$ with a sample of a few hundred high-significance HI galaxies with accurate inclinations, fit the two-component model with $k$ and $\\delta_{\\rm Outlier}$ free, and test whether the recovered $C_{\\rm BTFR}$ zero point and slope agree with $k=3.32$, $b=10.07$; a departure beyond the quoted $1\\sigma$ uncertainties would falsify the no-evolution claim.","supporting_citations":[],"review_version":1}