{"id":"d902f9ff-c874-4acf-b5af-9f28d16b3de0","arxiv_id":"2607.08604","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":4.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":3,"one_line_summary":"Using ZTF DR2 data, AGN presence in Type Ia supernova host galaxies at z<0.15 shows no significant systematic effect on Hubble residuals.","lead":"This paper tests whether active galactic nuclei (AGN) in host galaxies systematically bias the measured properties of Type Ia supernovae used for cosmology. It finds no significant effect, ruling out AGN as a source of systematic error in current low-redshift supernova surveys.","discovery_kind":"unclear","skeptic_critique":{"model":"glm-5.2","headline":"No significant objection identified. The null result is adequately powered for the claim being made: with 58 AGN hosts, effects larger than ~0.06 mag are detectable at 2σ, well below the ~0.1 mag mass step the paper rules out AGN as an explanation for.","rationale":"The reader correctly identified the MILLIQUAS reliability question as the weakest assumption, and it is a fair concern. However, I assess it as not load-bearing because the claim is a null result at a specific effect-size scale (~0.1 mag mass step), and the sample of 58 AGN hosts provides sufficient statistical power to detect effects of that magnitude even with moderate classification impurity. The paper is transparent about its limitations (small MIR-AGN sample, ZTF DR2 calibration caveats, differing selection criteria from Ginolin et al.). The mass-matched comparison design is sound and directly addresses the dominant confounder. The ACCEPT verdict with HIGH confidence is appropriate. The one improvement I would suggest is explicitly stating the minimum detectable effect size, which would strengthen the reader's confidence that the null result is meaningful rather than merely underpowered. But this is a presentation gap, not a logical flaw.","tokens_in":4914,"tokens_out":2745,"duration_ms":158486,"concrete_test":"Quantify the detection floor explicitly: compute the minimum detectable effect size at 2σ and 3σ given N=58 AGN hosts and the observed σ_∆µ, then verify that a simulated AGN effect of 0.1 mag (the mass step magnitude) injected into 58 randomly drawn SNe from the mass-matched non-AGN sample is recovered with >95% probability. This would confirm the sample has adequate statistical power for the specific claim being made.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim is that AGN presence cannot account for the observed brightness differences in high-mass galaxies (the mass step, ~0.1 mag). With 58 AGN-hosting SNe and per-mean uncertainties of ±0.02 mag, the combined uncertainty on the difference is ~0.028 mag, giving a 2σ sensitivity floor of roughly 0.06 mag. This is sufficient to rule out AGN as the cause of the mass step. The reader's concern about MILLIQUAS purity is legitimate but not load-bearing: even with a 20% false-positive rate (~12 misclassified galaxies), a real AGN effect large enough to explain the mass step would remain detectable after dilution. The MIR diagnostics, while too few for robust statistics (2–10 AGN), are consistent with the MILLIQUAS-based null result, providing a partial cross-check. The mass-matching strategy (nearest non-AGN galaxy above and below each AGN host in mass) appropriately controls for the dominant confounder. The standardization parameters (α, β, M₀ from Ginolin et al. 2025) are applied identically to both samples, so any calibration mismatch cancels in the comparison. The claim is modest and well-scoped: AGN are ruled out as the origin of the mass step at z < 0.15, not as a source of any conceivable sub-threshold systematic.","agreement_with_reader":"partial"},"referee_report":{"model":"glm-5.2","summary":"This short paper investigates whether the presence of active galactic nuclei (AGN) in Type Ia supernova (SN Ia) host galaxies introduces a systematic bias in Hubble residuals, using the ZTF DR2 spectroscopically-classified SN Ia sample. The author identifies AGN-hosting galaxies via mid-infrared (MIR) color selection and cross-matching with the MILLIQUAS catalog, finding that MIR selection yields too few AGN for useful statistics but MILLIQUAS identifies 58 AGN hosts (3.5% of the sample). By comparing Hubble residuals of SNe Ia in AGN-hosting galaxies to a mass-matched non-AGN sample, the author finds no statistically significant difference (⟨∆µ⟩ = -0.07±0.02 vs. -0.08±0.02), concluding that AGN presence cannot account for the observed mass step in SN Ia Hubble residuals at z < 0.15. The experimental design—mass-matching to control for the dominant confounder—is appropriate, and the null result is adequately powered for the modest claim being made.","tokens_in":5641,"tokens_out":1203,"duration_ms":133423,"significance":"The paper addresses a well-motivated question: as SN Ia cosmology pushes toward per-millimeter systematic control, any environmental variable correlated with host properties warrants scrutiny. The mass-matched comparison is a clean experimental design that isolates the AGN effect from the dominant host-mass confounder. The conclusion is modest and well-scoped: AGN are ruled out as the origin of the mass step at low redshift, not as a source of any conceivable sub-threshold systematic. The result is a useful negative finding that narrows the space of environmental systematics. The use of two independent AGN diagnostics (MIR and MILLIQUAS), even though one yields too few objects, provides a partial cross-check. The standardization parameters adopted from Ginolin et al. (2025) are applied identically to both samples, so calibration choices cancel in the comparison.","major_comments":[{"comment":"§2, mass-matching procedure: The mass-matched sample is described as 'the unique galaxies closest in mass both above and below each AGN hosting galaxy.' It is unclear whether this means two control galaxies per AGN host (one above, one below) or some other scheme. With 58 AGN hosts, this could yield 116 controls or some other number. The text should specify the exact number of galaxies in the mass-matched control sample and confirm that no galaxy is used as a control for more than one AGN host (or if reuse is permitted, justify it). The uncertainties on ⟨∆µ⟩ for the control sample (±0.02) should be derived from the actual sample size and scatter, and this should be stated explicitly.","section":null},{"comment":"§3: The uncertainties on ⟨∆µ⟩ are quoted as ±0.02 for both the AGN and mass-matched samples. For 58 AGN hosts, the standard error on the mean should be σ/√58. If the intrinsic scatter in Hubble residuals is ~0.12–0.15 mag (typical for SNe Ia after standardization), this gives σ_mean ≈ 0.016–0.020 mag, which is consistent. However, the paper should state the intrinsic scatter and confirm whether the quoted uncertainties are standard errors of the mean or include any additional systematic component. If the mass-matched sample is larger (e.g., 116 galaxies), its uncertainty should be smaller; the fact that both are ±0.02 suggests either similar sample sizes or that the uncertainties are dominated by a common systematic floor. This should be clarified.","section":null}],"minor_comments":[{"comment":"§2: The redshift cut is stated as '0.01 ≤ z ≤ 0.15' in the methods but the abstract says 'z < 0.15'. Make these consistent.","section":null},{"comment":"§2: The WISE apparent magnitude cuts (W1 < 14.70, W2 < 14.45) are mentioned without explanation of how they were chosen or their completeness implications. A brief justification or reference would help.","section":null},{"comment":"§3, Eq. (2): The sign convention for the color term (−βc) should be briefly defined or referenced to the Tripp relation for readers unfamiliar with the specific Ginolin et al. (2025) formulation, to improve self-containedness.","section":null},{"comment":"Figure 1: The caption mentions 'circles and hexagons' for full-sample means but these symbols are not clearly distinguishable in the figure as described. Consider clarifying what each symbol represents or simplifying the legend.","section":null},{"comment":"§4: The sentence beginning 'However, too few reliably classifiable MIR-AGN hosts within it to perform robust analysis' is grammatically incomplete. Consider revising to 'However, there are too few reliably classifiable MIR-AGN hosts within it to perform robust analysis.'","section":null},{"comment":"§1: The phrase 'the largest of which being their astrophysics' is slightly awkward. Consider 'the largest of which relates to their astrophysics' or similar.","section":null}],"recommendation":"minor_revision","confidential_remarks":"The paper is short and the analysis is straightforward. The null result is credible and the mass-matching design is sound. The main concern is presentation clarity around the mass-matched sample construction and uncertainty estimation, not a fundamental problem with the analysis. I see no reason this should not be published after minor revisions. The reader's concern about MILLIQUAS purity is legitimate but not load-bearing: even with moderate contamination, a real AGN effect large enough to explain the ~0.1 mag mass step would remain detectable after dilution, so the null conclusion is robust to this concern."},"author_rebuttal":null,"desk_editor":{"model":"glm-5.2","letter":"This is a straightforward null result that closes a specific question for SN Ia cosmology: whether AGN in host galaxies systematically bias Hubble residuals. Using ZTF DR2, Clark cross-matches SN Ia hosts against MILLIQUAS, finds 58 AGN-hosting galaxies (3.5% of the sample), and compares their Hubble residuals to a mass-matched non-AGN control sample. The result is unambiguous: ⟨∆µ⟩ = -0.07±0.02 for AGN hosts versus -0.08±0.02 for the mass-matched controls. They agree within 0.01 mag. AGN are not driving the mass step. The mass-matching is the right experimental design here—host mass is the dominant confounder, and controlling for it directly isolates the AGN question. The standardization parameters (α, β, M₀) are adopted from Ginolin et al. 2025 rather than fit to this data, which is fine because they're applied identically to both samples and cancel in the comparison. The paper is also honest about its limitations, which I appreciate. The MIR AGN selection essentially fails—only 2 to 10 AGN found depending on the diagnostic, too few for any meaningful statistics. So the entire result rests on the MILLIQUAS cross-match. The reader flagged this as the weakest assumption, and I agree it's the main soft spot, but it's not load-bearing. MILLIQUAS is a reasonably robust catalog, and even with a 20% false-positive rate, a real AGN effect large enough to explain the ~0.1 mag mass step would still be detectable after dilution. With 58 AGN hosts and per-mean uncertainties of ±0.02 mag, the 2σ sensitivity floor is roughly 0.06 mag—well below the mass step. So the null is adequately powered for the claim being made. One thing I'd push on if I were refereeing: the paper doesn't validate MILLIQUAS purity or completeness against its own MIR diagnostics, even though the MIR sample is too small for statistics. A sentence acknowledging this gap would help. Also, the title is slightly stronger than the result—it's a null at z < 0.15 with one AGN selection method, not a universal exclusion. But the body is appropriately scoped. This is for SN Ia cosmologists who need to tick off environmental systematics one by one. It's incremental but useful, and the mass-matched design is clean. Deserves a serious referee, though the fixes are minor.","headline":"Short, clean null result: AGN in SN Ia hosts don't drive the mass step","tokens_in":5710,"tokens_out":599,"would_cite":false,"duration_ms":74773,"reading_group":"no","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"glm-5.2","headline":"AGNs cleared as source of supernova measurement bias","keywords":["Type Ia supernovae","active galactic nuclei","Hubble residuals","cosmological systematics","host galaxy mass step","Zwicky Transient Facility"],"falsifier":"A future sample with independently verified AGN classifications (e.g., from X-ray or optical spectroscopic diagnostics) that shows a statistically significant offset in Hubble residuals between AGN and non-AGN hosts at matched stellar mass would falsify the null result.","tokens_in":4936,"feed_emoji":"🔭","tokens_out":3974,"duration_ms":133820,"temperature":0.7,"pith_summary":"The paper tests whether active galactic nuclei — supermassive black holes actively feeding on gas in the centers of some galaxies — could be systematically biasing the properties measured for Type Ia supernovae hosted in those galaxies. Type Ia supernovae are the workhorses of cosmological distance measurement, but their brightness after standardization still varies systematically with host galaxy properties: supernovae in more massive galaxies appear brighter. If AGN feedback altered gas content or temperature in ways that affected supernova progenitors, or if AGN light biased measurements of host galaxy stellar mass, this would create a hidden systematic that could distort cosmological results. Using 2,425 spectroscopically confirmed Type Ia supernovae from the Zwicky Transient Facility at redshifts below 0.15, the author identifies 58 host galaxies containing AGNs (via the MILLIQUAS catalog) and compares their supernova Hubble residuals — the difference between observed and cosmologically predicted brightness — against a mass-matched sample of non-AGN hosts. The two samples show statistically identical residuals (-0.07±0.02 versus -0.08±0.02), ruling out AGN presence as the driver of the well-known environmental differences in standardized supernova brightness at low redshift.","feed_headline":"AGNs cleared as source of supernova measurement bias","feed_subtitle":"Supernovae in AGN-hosting galaxies show identical brightness residuals to matched non-AGN hosts — one cosmological worry ruled out.","key_machinery":"The Hubble residual, defined as the difference between the observed brightness of a standardized Type Ia supernova and the brightness predicted by a cosmological model, serves as the diagnostic. By comparing the mean Hubble residual of supernovae in AGN-hosting galaxies to that of a mass-matched control sample of non-AGN hosts, any AGN-specific effect would appear as a statistically significant offset between the two distributions. The absence of such an offset is the null result.","core_discovery":"The Hubble residuals of Type Ia supernovae in AGN-hosting galaxies are statistically indistinguishable from those in mass-matched non-AGN hosts (-0.07±0.02 versus -0.08±0.02), demonstrating that AGN activity does not account for the environmental systematics — particularly the brightness difference in high-mass galaxies — observed in standardized supernova brightnesses at low redshift.","pith_inferences":["If the mass step is not driven by AGN presence, the remaining candidates are properties that correlate with galaxy mass but are not caused by central black hole activity — stellar population age, metallicity, or gas-to-stellar mass ratio.","The 3.5% AGN fraction in the sample is small enough that even a real AGN effect would need to be large to be detectable; the null result places an upper bound on the magnitude of any such effect, but does not rule out subtle effects below the detection threshold.","At higher redshifts near cosmic noon (z ~ 2), where AGN activity was substantially more common, the fraction of AGN-hosting supernova systems could be high enough that the null result may not hold."],"forward_implications":["One candidate explanation for the mass step in supernova brightness is eliminated at low redshift, narrowing the search to stellar population age, metallicity, or gas content as drivers.","Future cosmological surveys using Type Ia supernovae need not treat AGN-hosting galaxies as a class requiring separate calibration.","The very low mid-infrared AGN identification rate (2–10 galaxies, consistent with prior ~0.4% fractions) suggests MIR color selection is too conservative for this purpose; broader multi-diagnostic catalogs like MILLIQUAS are needed.","Planned larger samples from upcoming facilities could test whether the null result extends to higher redshifts where AGN activity was more common."],"fun_headline_variants":["AGN activity does not bias Type Ia supernova Hubble residuals","Type Ia supernovae in AGN hosts show no added Hubble residual bias","AGN presence leaves Type Ia supernova brightness residuals unaffected","No evidence of AGN-induced bias in Type Ia supernova measurements","Host galaxy AGNs do not introduce Type Ia supernova measurement bias"],"cache_read_input_tokens":0,"weakest_assumption_plain":"The paper relies on the MILLIQUAS catalog, cross-matched within 3 arcseconds, to correctly identify which host galaxies contain AGNs. The mid-infrared method, which the author also tried, found too few AGN to be useful. If MILLIQUAS misclassifies galaxies — either missing real AGNs or flagging non-AGN galaxies as AGN hosts — the null result could reflect classification error rather than a true absence of effect.","fun_headline_variants_meta":{"raw":{"variants":["AGN activity does not bias Type Ia supernova Hubble residuals","Type Ia supernovae in AGN hosts show no added Hubble residual bias","AGN presence leaves Type Ia supernova brightness residuals unaffected","No evidence of AGN-induced bias in Type Ia supernova measurements","Host galaxy AGNs do not introduce Type Ia supernova measurement bias","Type Ia supernova standardization unaffected by host galaxy AGNs"]},"model":"glm-5.2","effort":"low","cost_usd":0.0,"raw_usage":{"total_tokens":2169,"prompt_tokens":405,"completion_tokens":1764,"prompt_tokens_details":null},"tokens_in":405,"tokens_out":1764,"duration_ms":38276,"temperature":1.0,"reasoning_tokens":1666,"cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-10T04:34:26.790176+00:00","model_set":{"reader":"glm-5.2"},"falsifier":"A future sample with independently verified AGN classifications (e.g., from X-ray or optical spectroscopic diagnostics) that shows a statistically significant offset in Hubble residuals between AGN and non-AGN hosts at matched stellar mass would falsify the null result.","supporting_citations":[],"review_version":1}