{"id":"9a5ee657-b6b8-4e64-be20-c9c4cf618376","arxiv_id":"2507.08179","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":4,"one_line_summary":"Using 341 X-ray-selected AGN, this study calibrates the [Ne v] lambda3427 line as a tracer of black-hole accretion power, finding a 43% detection rate and no strong dependence on obscuration, black-hole mass, or Eddington ratio.","lead":"This paper measures a specific ultraviolet emission line, [Ne v], in more than 340 nearby active galaxies to test whether it can reliably signal the presence and power of a supermassive black hole, even when the galaxy is heavily obscured. It finds the line in about 43% of cases and shows that, when detected, its brightness tracks the hard X-ray output with a scatter as good as or better than the commonly used [O III] tracer.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No single load-bearing flaw found; the key caveat of uncorrected host dust is acknowledged, quantified, and partially mitigated by the concordant dust-insensitive MIR [Ne v] study.","rationale":"The reader's weakest_assumption (host-scale dust attenuation correlated with AGN properties) is plausible and well-identified, so I agree partially. However, my reading gives more weight to the mitigating evidence: (1) the MIR [Ne v] 14.3/24.3 micron measurements from the same BASS parent sample (Bierschenk et al. 2024) show comparable scatter and the same absence of trends with M_BH, lambda_Edd, and N_H; MIR lines are insensitive to dust attenuation, so this independent probe weakens the concern that host dust created the null results; (2) the paper explicitly quantifies the required dust (E(B-V)~0.8 for a 2 dex range), which is high for a narrow-line AGN sample; (3) the sample is hard-X-ray selected, so line-detection-dependent selection effects are limited. The reader also correctly notes that host dust could hide correlations, but such a correlation would need to exist between intrinsic line strength and N_H, M_BH, or lambda_Edd—and the concordant MIR results plus the lack of any physical expectation for such a dust correlation make this unlikely. The paper's main claims are modest, well-supported, and internally consistent. No machine-checked proofs exist, but none are expected for an observational paper; the analysis uses standard statistical methods and is reproducible from BASS public data. The parameter count is small and the conclusion is robust to reasonable choices in the analysis. Therefore I recommend UNCHANGED (ACCEPT).","tokens_in":30131,"tokens_out":1881,"duration_ms":20282,"concrete_test":"One verification worth running: repeat the analysis of Section 4.3 including the 14 narrow-line, low-N_H (log[N_H/cm^-2]=20) Seyfert 1.9 sources that are currently included in the sample, and also re-run the correlation tests after applying a simple host-extinction correction to [Ne v] fluxes using the Balmer decrement (H-alpha/H-beta) measurements available from BASS/DR2, where available. If the null trends survive and the scatter in log L[Nev]/L14-150 remains ~0.45 dex after Balmer-decrement correction, the central claim would be robust to the host-dust concern.","verdict_should_be":"UNCHANGED","load_bearing_attack":"No significant objection identified. The central claim—that no strong trends exist between narrow [Ne v] lambda3427 detection rate or relative strength and L14-150, M_BH, lambda_Edd, or N_H—is supported by the data and by the stated correlation tests (Appendix C, Table 2). The most plausible alternative explanation for the null results is host-galaxy dust attenuation affecting the blue [Ne v] line. The authors acknowledge this in Section 3.1 and Section 4.5, and they quantify that explaining a 1 dex (2 dex) spread requires E(B-V) ~0.4 (0.8) for a foreground screen, or ~0.8 (1.6) for mixed dust. This is a legitimate caveat, but it is not a fatal flaw: the uniform selection of BASS depends on hard X-rays, not optical line flux, and the independent, dust-insensitive MIR [Ne v] lines studied in Bierschenk et al. (2024) show the same scatter (~0.5 dex) and the same lack of trends with N_H, M_BH, and lambda_Edd. That independent concordance is strong evidence that the null results are not artifacts of host dust. The main remaining uncertainty is that the sample is restricted to narrow-line AGN (type 1.8-2), so the conclusion is not directly tested for type 1 AGN; the paper appropriately restricts its claims accordingly. The conclusion that the L[Nev]/L14-150 scatter of 0.45 dex is 'smaller than' the [O III] scatter is somewhat fragile because it compares different samples and fitting procedures, but the qualitative conclusion (comparable scatter) is robust. Overall, the paper's claims are appropriately scoped and the weakest assumption is explicitly quantified and bounded.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper measures the narrow [Ne V] λ3427 emission line in 341 ultra-hard X-ray (14-195 keV) selected, narrow-line AGN from the BASS survey, using individual fits and spectral stacking. It reports a detection fraction of 42.8% (146/341), a median ratio log L[NeV]/L(14-150 keV) = -3.75 with ~0.45 dex scatter, and finds no strong trends of detection rate or relative line strength with X-ray luminosity, black hole mass, Eddington ratio, or column density, except for a weak anti-correlation with L(14-150). The paper concludes that [Ne V] λ3427 is a robust AGN accretion tracer even in heavily obscured systems, and discusses implications for high-redshift JWST studies and dual AGN searches.","tokens_in":30666,"tokens_out":8417,"duration_ms":91439,"significance":"If the results hold, this is a valuable calibration of an AGN tracer that can penetrate heavy obscuration and is accessible with JWST at high redshift. The analysis is careful in several respects: Monte-Carlo-based uncertainties, conservative multi-criteria detection thresholds, explicit correlation tests including upper limits (Kendall τ), a stacking analysis, and comparison with the dust-insensitive MIR [Ne V] lines from the companion BASS study (Bierschenk et al. 2024). The concordance between the optical and MIR [Ne V] results is a strong argument that the null trends are not primarily artifacts of host-galaxy dust. The sample is homogeneous in selection (ultra-hard X-rays) and the paper appropriately restricts its conclusions to narrow-line AGN.","major_comments":[{"comment":"The abstract states that \"we find no significant relations between the (relative) strength of [Ne V] and the basic AGN/SMBH properties under study,\" but Table 2 reports a statistically significant anti-correlation between log(F[Nev]/F14-150) and log L14-150 (Spearman ρ=-0.34, P=3.1e-5; Kendall τ=-0.11, P=0.008). The body of the paper correctly identifies this as the only statistically significant relation, but the abstract's blanket claim is too strong. Please soften the abstract to \"no strong trends\" or explicitly state the exception, otherwise readers will incorrectly infer that no correlation with X-ray luminosity exists.","section":"Abstract; §4.3; Table 2 (Appendix C)"},{"comment":"There is a direct contradiction about which spectra enter the stacking analysis. §3.2 says that stacks were constructed from either all AGN or only non-detected AGN, and that §4.4 focuses on the \"former set\" (all AGN). §4.4 then says \"These stacked spectra, and the analysis that follows, focus only on those AGN where [Ne V] was not individually detected.\" This inconsistency makes it impossible to interpret Figure 8 and Table 1 precisely. Please correct the text and specify in the figure/table caption which subset is shown.","section":"§3.2 vs. §4.4"},{"comment":"The scaling ratios (median log L[NeV]/L14-150 = -3.75, scatter 0.45 dex) and the linear fit in Eq. (1) are derived exclusively from the 146 [Ne V]-detected sources, while 195 sources (57% of the sample) provide only upper limits. As the paper itself cautions, the intrinsic scatter is larger, but the median itself may also be biased if non-detections preferentially have lower L[NeV]/L_X. A quantitative treatment of the censoring (e.g., a Kaplan-Meier or survival-regression estimate of the median ratio, or at least a bound on the bias) is needed to support the use of these values as a calibration. Absent such an analysis, the paper should explicitly state that the reported ratios apply only to sources in which [Ne V] is detected, not to the full AGN population.","section":"§4.2; Eq. (1); §5"}],"minor_comments":[{"comment":"The detection fraction at high column density is quoted inconsistently: the abstract says \">70% for log[N_H/cm^-2] > 23\", §4.1 says \"f_det ≈ 60%\" at log(N_H/cm^-2) ≳ 23, and §4.3 says \"exceeding f_det ≈ 50%\" at ≳23.5 and \"reaching f_det ≳ 80%\" toward >24. Please harmonize the numbers and specify the exact bin edges.","section":"Abstract; §4.1; §4.3"},{"comment":"The statement that the [Ne V] scatter (0.45 dex) is \"smaller than\" the [O III] scatter is based on comparisons with different samples and fitting procedures (Heckman et al. 2005; Berney et al. 2015). The qualitative conclusion that the scatter is comparable is well supported, but the quantitative claim of being \"smaller\" should be caveated accordingly.","section":"§4.2"},{"comment":"The host-dust discussion is well done, but the paper could strengthen it by testing sensitivity to host extinction using available Balmer-decrement measurements (or excluding known edge-on/merger hosts) rather than only a foreground-screen estimate. This would further diminish the concern that the null trends are dust-induced.","section":"§4.5"},{"comment":"The text says the detection fraction \"saturates near f_det ≃ 75% at S/N ≳ 30\", but the plotted bins in Figure 2 reach at most 69% (11/16). Please adjust the wording to match the actual binned values.","section":"Figure 2; §4.1"},{"comment":"Minor language issues: §2.1 \"the line is very close to the noisy, blue end\" could be rephrased; §4.4 contains \"arrow [Nev]\" which appears to be a typo for \"[Nev]\"; also the phrase \"we adopt a the aforementioned criteria\" in §3.2 should be corrected.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The paper is a solid BASS legacy product with a clean, well-documented analysis. The main issues are an overstatement in the abstract, an internal contradiction about the stacking subsets, and a missing censoring analysis for the quoted scaling ratios. These are all fixable without changing the scientific message, but the censoring point is load-bearing for the calibration claim, so I recommend major revision rather than minor. There is no concern about novelty or scope; the paper fits the journal well."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's my read of the Reiss et al. paper. The headline result is that [Ne v] lambda3427 is detected in ~43% of 341 hard-X-ray-selected narrow-line AGN, with a median L[NeV]/L14-150 of 10^-3.75 and ~0.45 dex scatter, and no statistically meaningful trends with NH, MBH, or Eddington ratio. That is a genuinely useful calibration, and it's the largest sample of this line in a homogeneous, ultra-hard-X-ray-selected parent sample by a long shot.\n\nWhat's new: dedicated fits with conservative detection criteria, Monte Carlo uncertainties, including upper limits in correlation tests, and a stacking analysis that supports the null results. The consistency with the dust-insensitive MIR [Ne v] study (Bierschenk et al. 2024) is the strongest argument that the nulls are not artifacts of host dust. The paper is also honest about the main caveat: no host-galaxy extinction correction is applied, and they quantify that explaining the observed spread would require E(B-V) ~0.4-1.6 depending on geometry.\n\nSoft spots, in order: (1) The comparison of scatter to [O III] from other studies is not apples-to-apples because samples and fitting procedures differ. The qualitative claim that [Ne v] scatter is comparable or smaller is fine, but I wouldn't lean hard on \"smaller.\" (2) NH estimates exist for only 234 of 341 sources, which weakens the NH null at the highest columns. (3) The visual inspection step is subjective, though it seems to be conservative rather than optimistic. (4) The sample is narrow-line AGN only, so the scaling relation should not be assumed for type 1 without further work. None of these are fatal; the paper's claims are appropriately scoped. The minor note about a forthcoming Trakhtenbrot et al. JWST paper is a bit self-referential but harmless.\n\nMy bottom line: this is a solid, useful observational paper. The scaling relation will be cited as a reference calibration for [Ne v] work at high redshift, and the null results are worth having even if they are not surprising. It deserves a serious referee. I'd bring it to reading group and would cite it.","headline":"A careful, well-scoped calibration of [Ne v] lambda3427 as a tracer of obscured AGN power; the null trends hold up and the main caveat is quantified, not fatal.","tokens_in":31149,"tokens_out":2136,"would_cite":true,"duration_ms":24339,"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":"The paper shows that the narrow [Ne v] $\\lambda3427$ line, detected in 43% of a complete ultra-hard-X-ray-selected AGN sample, calibrates accretion power with median $\\log(L_{\\rm [Ne\\,v]}/L_{14-150}) = -3.75$ and about 0.45 dex scatter…","keywords":["active galactic nuclei","[Ne v] lambda3427","ultra-hard X-ray selection","obscured AGN","emission-line diagnostics","black hole mass","Eddington ratio","spectral stacking"],"falsifier":"Measure, for each of the 341 AGN, a host-scale dust indicator such as the Balmer decrement or the ratio of [Ne v] $\\lambda3427$ to the dust-insensitive mid-infrared [Ne v] 14.3 $\\mu$m line; if $F_{\\rm [Ne\\,v]}/F_{14-150}$ declines systematically with increasing host $E(B-V)$, especially among high-column sources, the reported null trends and the high detection fraction at $\\log N_{\\rm H} \\gtrsim 23$ would be partly artifacts of attenuation.","tokens_in":29981,"feed_emoji":"🔭","tokens_out":8585,"duration_ms":88661,"temperature":0.7,"pith_summary":"This paper asks whether the narrow [Ne v] $\\lambda3427$ emission line, a spectral signature of gas ionized by photons with energies above 97 eV and essentially only produced by accreting supermassive black holes, can serve as a reliable measure of how hard an active galactic nucleus (AGN) is accreting. Using 341 narrow-line AGN selected by their ultra-hard X-rays (14-195 keV), it reports that the line is detected in 43% of the sample, and that detection is achieved even among the most heavily obscured systems. The median ratio of [Ne v] luminosity to intrinsic 14-150 keV X-ray luminosity is $\\log(L_{\\rm [Ne\\,v]}/L_{14-150}) = -3.75$, with a scatter of about 0.45 dex, comparable to or smaller than the scatter of the commonly used [O III] $\\lambda5007$ tracer. Neither the detection fraction nor the relative line strength shows a significant trend with X-ray luminosity, black-hole mass, Eddington ratio, or line-of-sight column density. If the claim holds, [Ne v] becomes a calibrated, nearly star-formation-free probe of accretion power in hidden AGN, including dual and high-redshift sources.","feed_headline":"[Ne v] exposes hidden AGN in 43% of ultra-hard X-ray sources","feed_subtitle":"Line strength tracks X-ray luminosity with 0.45 dex scatter, independent of obscuration, black-hole mass, and Eddington ratio.","key_machinery":"The central object is the narrow [Ne v] $\\lambda3427$ emission line, a coronal line requiring ionizing photons with $h\\nu > 97$ eV, which in practice are produced almost exclusively by the AGN accretion flow rather than by star formation. The argument is carried by measuring this line in 341 narrow-line AGN selected through their ultra-hard X-ray emission, using Gaussian fits to individual VLT/X-Shooter and Palomar/DBSP spectra plus continuum-normalized stacks in bins of luminosity, black-hole mass, Eddington ratio, and column density. The key quantitative machinery is the luminosity ratio $L_{\\rm [Ne\\,v]}/L_{14-150}$ and its distribution: a median of $-3.75$ dex with ~0.45 dex scatter, which is what converts a line flux into an accretion-power estimate and what sets the claim that [Ne v] works as a tracer.","core_discovery":"The central claim is that narrow [Ne v] $\\lambda3427$ emission is a reliable tracer of intrinsic AGN radiative output across the full range of obscuration, not a line that only appears in unobscured or low-column systems. The paper's dedicated fits and stacking of 341 ultra-hard-X-ray-selected, narrow-line AGN yield a detection fraction $f_{\\rm det} = 42.8\\%$ (146/341), which saturates near ~75% even at extremely high signal-to-noise rather than reaching 100%, so the absence of [Ne v] in an individual spectrum does not mean the AGN is weak. Among detections, the median scaling is $\\log(L_{\\rm [Ne\\,v]}/L_{14-150}) = -3.75$ with 0.45 dex standard deviation (and $-3.36$ relative to 2-10 keV), and the relative strength shows no significant correlation with $L_{14-150}$ (except a weak sub-linear luminosity dependence), $M_{\\rm BH}$, Eddington ratio, or $N_{\\rm H}$, with the detection fraction remaining above ~60% at $\\log(N_{\\rm H}/{\\rm cm^{-2}}) \\gtrsim 23$. Stacking spectra of individually non-detected AGN fails to produce [Ne v] in most bins and shows no clear trend with any property. The authors conclude that [Ne v] can be used as-is to estimate AGN bolometric luminosity ($L_{\\rm bol} \\approx 45{,}000 \\times L_{\\rm [Ne\\,v]}$) and to identify obscured and dual AGN, including at high redshift with JWST.","pith_inferences":["If host-scale dust is later found to dim [Ne v] more in certain merging or high-column hosts, the apparent independence from $N_{\\rm H}$ and the detection fraction at high obscuration could be partly artifacts; the paper itself flags this as its main caveat.","Because the line is almost immune to star-formation contamination, the measured scaling could be applied to large optical surveys to estimate obscured accretion luminosity even without X-ray data, provided the signal-to-noise-dependent detection fraction is modeled.","The saturation of the detection fraction near 75% at high $S/N$ suggests an intrinsic spread in narrow-line-region properties, so stacking non-detections is a promising route to recover average line strength for faint populations.","A direct comparison of [Ne v] $\\lambda3427$ with the dust-unaffected mid-infrared [Ne v] lines on the same objects would test whether the null trends with $N_{\\rm H}$ are real or an attenuation artifact."],"forward_implications":["[Ne v] $\\lambda3427$ can be used to estimate AGN bolometric luminosity through $L_{\\rm bol} \\approx 45{,}000 \\times L_{\\rm [Ne\\,v]}$, with scatter comparable to or below that of [O III] $\\lambda5007$ at ~0.5 dex.","The line remains detectable in a large fraction of Compton-thick and heavily obscured AGN, so it can flag buried accretion that X-ray surveys miss.","For $z \\gtrsim 6$ sources observed with JWST/NIRSpec, a detected [Ne v] line directly implies AGN-powered ionization, and the measured scaling converts line flux into an accretion-power estimate.","The absence of strong trends with black-hole mass and Eddington ratio means simple thin-disk predictions, that high-mass or low-$\\lambda_{\\rm Edd}$ flows should produce little extreme-UV radiation, are not reflected in the relative strength of [Ne v] emission.","Narrow-band [Ne v] imaging can reveal sub-kpc dual AGN in obscured mergers, where [O III] is contaminated by star formation."],"supporting_citations":[{"why":"Supplies the BASS/DR2 parent sample of ultra-hard X-ray selected AGN and the optical spectra used here.","marker":"Koss et al. 2022a"},{"why":"Provides the intrinsic 14-150 and 2-10 keV luminosities and line-of-sight $N_{\\rm H}$ from multi-mission X-ray spectral decomposition.","marker":"Ricci et al. 2017b"},{"why":"Supplies the stellar-velocity-dispersion-based black-hole masses used for the $M_{\\rm BH}$ and Eddington-ratio trends.","marker":"Koss et al. 2022b"},{"why":"Provides the earlier BASS/DR2 spectral catalog and redshifts whose line measurements this work refines.","marker":"Oh et al. 2022"},{"why":"Established earlier [Ne v]-to-X-ray scaling and the idea that low ratios flag heavily obscured AGN, the baseline this study compares against.","marker":"Gilli et al. 2010"},{"why":"Found a mild correlation and ~0.5 dex scatter between [Ne v] and ultra-hard X-ray emission in ~50 local AGN, the direct predecessor scaling.","marker":"Berney et al. 2015"},{"why":"The companion BASS study of mid-infrared [Ne v] lines, providing the dust-unaffected comparison and similar null trends.","marker":"Bierschenk et al. 2024"},{"why":"Demonstrated tight mid-infrared [Ne v] line correlations with ultra-hard X-ray luminosity, establishing the tracer's credentials.","marker":"Weaver et al. 2010"},{"why":"Provides the $M_{\\rm BH}$-$\\sigma_*$ scaling relation used to convert stellar velocity dispersions into black-hole masses.","marker":"Kormendy & Ho 2013"},{"why":"Supplies the [O III]-to-X-ray scaling with ~0.5 dex scatter that the paper compares against for [Ne v].","marker":"Heckman et al. 2005"}],"fun_headline_variants":["Ne v line finds hidden AGN even in thick obscuration","[Ne v] reveals AGN through thick gas: 43% detection","Neon line probes AGN power despite heavy X-ray absorption","Hard X-ray AGN: [Ne v] light persists in 43% of sources","Obscured AGN: [Ne v] still detected in 43%"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The analysis assumes that dust in the host galaxy does not dim the [Ne v] line more for some AGN than for others in a way that tracks the same properties being studied, because no correction for host-galaxy extinction is applied.","fun_headline_variants_meta":{"raw":{"variants":["Ne v line finds hidden AGN even in thick obscuration","[Ne v] reveals AGN through thick gas: 43% detection","Neon line probes AGN power despite heavy X-ray absorption","Hard X-ray AGN: [Ne v] light persists in 43% of sources","Obscured AGN: [Ne v] still detected in 43%"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000718,"raw_usage":{"total_tokens":3395,"prompt_tokens":1287,"completion_tokens":2108,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":903,"completion_tokens_details":{"reasoning_tokens":2010}},"tokens_in":903,"tokens_out":2108,"duration_ms":17516,"temperature":1.0,"reasoning_tokens":2010,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T18:24:33.866869+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure, for each of the 341 AGN, a host-scale dust indicator such as the Balmer decrement or the ratio of [Ne v] $\\lambda3427$ to the dust-insensitive mid-infrared [Ne v] 14.3 $\\mu$m line; if $F_{\\rm [Ne\\,v]}/F_{14-150}$ declines systematically with increasing host $E(B-V)$, especially among high-column sources, the reported null trends and the high detection fraction at $\\log N_{\\rm H} \\gtrsim 23$ would be partly artifacts of attenuation.","supporting_citations":[{"cited_title":"J., et al","cited_arxiv_id":null,"evidence_quote":"The companion BASS study of mid-infrared [Ne v] lines, providing the dust-unaffected comparison and similar null trends."}],"review_version":1}