{"id":"e519a452-f322-4c8e-92fb-7b2ad4e8c2bc","arxiv_id":"1908.04990","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Using two high-quality quasar samples, the authors confirm two dominant principal components and propose a new two-axis plane based on distance along a fitted decay curve.","lead":"This paper tests whether a few key numbers can describe quasar diversity, using newer spectra. It confirms two main patterns in the data and proposes a curved two-axis map that may be linked to black hole mass and luminosity.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The claimed plane rests on removing ten endpoint quasars; without independent evidence that they are artifacts, the axes remain significantly correlated and the central claim is unsupported.","rationale":"The reader's weakest assumption identifies exactly the load-bearing concern: post-hoc removal of ten outliers is required to make the two new axes linearly independent. I agree with that diagnosis. The PCA replication in Sections 2.1 and 2.2 is credible and not the issue; the new contribution is the curved-coordinate plane, and that contribution is conditional on the outlier exclusion. The paper itself flags the uncertainty in Section 4.1, but it does not test the exclusion independently, and it reports no variance fraction for the new plane or fitted parameter values. A CONDITIONAL verdict is therefore appropriate rather than ACCEPT or REJECT: the claim could be rescued if independent spectral re-measurement shows the ten objects are artifacts, but as it stands the linear-independence result is not established. The concrete test above would settle the concern directly.","tokens_in":11340,"tokens_out":3005,"duration_ms":32961,"concrete_test":"Independently re-measure RFeII and H-beta FWHM for all ten excluded quasars using the higher-fidelity spectral fitting method of Sniegowska et al. (2018), then refit y = 1/(a + x^b) on the full 175-object reduced Shen sample with corrected values. If the Spearman correlation between arc-length and residual remains significant at p < 0.05 in the corrected full sample, the axes are not independent and the central claim fails. If the correlation becomes insignificant without excluding any objects, the post-hoc exclusion is vindicated.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim in Section 5 is that the quasar main sequence can be described by 'two linearly independent parameters': distance along a decay curve and distance from the curve. Section 3 shows that this independence is not present in the full reduced Shen sample: a Spearman test gives p = 0.03 for an anti-correlation between the two new axes. The authors then exclude two groups of five quasars at the ends of the curve, and only after this post-hoc removal does the correlation disappear. Section 4.1 offers plausible reasons the outliers may be measurement artifacts: high-RFeII values may be overestimated, citing Sniegowska et al. (2018) where 21/27 such objects did not retain high RFeII, and double-peaked H-beta profiles may bias FWHM measurements. But these are possibilities, not independent tests; the paper itself says it is 'not certain' that RFeII > 1.3 values are accurate and calls the width explanation 'plausible'. The offset-envelope statistic (60% versus 37%) is suggestive but weak. If the ten objects are genuine members of the main sequence, the Spearman p = 0.03 remains, the new axes are not linearly independent across the sample, and the claimed plane collapses. Because the exclusion is decided after seeing the correlation and is never validated on independent data, it is the load-bearing assumption of the paper.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper applies PCA to two quasar samples — a 30-object XSHOOTER/Capellupo sample and a quality-selected 175-object subset of the Shen et al. (2011) catalog — and finds that two dominant eigenvectors reproduce the classical Boroson & Green result. It then fits a two-parameter decay curve y = 1/(a + x^b) to the FWHM(Hβ)–RFeII plane of the reduced Shen sample, defines two new coordinates (distance along the curve and perpendicular distance from the curve), removes two groups of five outliers at the curve ends to make these coordinates linearly independent, and interprets the main axis as an anti-correlation with black hole mass and the secondary axis as luminosity. The abstract and Section 5 conclude that the quasar main sequence can be described by a plane spanned by these two nonlinear coordinates.","tokens_in":11605,"tokens_out":4542,"duration_ms":47277,"significance":"If the central claim were fully established, the paper would provide a useful nonlinear generalization of the quasar main sequence and a concrete mapping of its two coordinates onto physical drivers. The PCA replication on newer high-quality samples (Table 2) is a genuine contribution, and the paper is clearly written with explicit parameter lists and reproducible enough to be checked. However, the new-plane result rests on a post-hoc exclusion of ten objects and on a variance claim that is not independently quantified, so the significance of the headline claim is currently limited.","major_comments":[{"comment":"The central claim that the two new parameters are linearly independent rests on the post-hoc removal of ten quasars. For the full reduced Shen sample, the Spearman test on the two new axes gives p = 0.03, i.e., a significant anti-correlation. The removal of the two endpoint groups is decided after observing this correlation, and Section 4.1 supports it only with suggestive evidence: the Sniegowska et al. (2018) result concerns 21/27 objects, the offset-envelope statistic is 60% versus 37% on a sample of ten, and the paper itself says it is \"not certain\" that RFeII > 1.3 values are accurate and calls the double-peaked Hβ explanation \"plausible.\" No independent test is provided, such as applying the same procedure to the Capellupo sample, to a withheld subset, or to bootstrapped samples with random rejection. The abstract and Section 5 claim that the plane is spanned by linearly independent axes, but this is not established for the full sample and needs either independent validation or an explicit conditional statement.","section":"Section 3 and Fig. 4"},{"comment":"The statement that the new plane \"accounts for the majority of the variance\" is not quantified and is partly circular. The main axis is the arc length along the decay curve whose parameters a and b in Eq. (1) are fitted to the same reduced Shen sample, and the secondary axis is the residual distance from that fitted curve. A two-parameter least-squares curve will by construction reduce the residuals of the training data; what is missing is a variance fraction, a comparison with PCA on the same two variables, and an assessment of out-of-sample behavior. No uncertainties or bootstrap estimates for a, b, or the variance fraction are provided. As written, the variance claim is a property of the fit rather than an independent test.","section":"Section 3 and Section 5"},{"comment":"The physical interpretation — main axis anti-correlating with black hole mass, secondary axis with luminosity — is not quantitatively tested. The supporting evidence is a color-coded figure (Fig. 4) and qualitative statements about correlations; no regression coefficients, Spearman rho values for the new coordinates against MBH and L3000, or partial correlations are given, and no attempt is made to separate the proposed mass/luminosity drivers from the inclination effect discussed in Section 4. Since the abstract and conclusions present these identifications as the main physical result, the authors should either provide the quantitative correlations or weaken the conclusion to a conjecture.","section":"Sections 4 and 5"}],"minor_comments":[{"comment":"The outlier threshold is RFeII > 1.2 in Section 3, but Section 4.1 later says \"not certain that values of RFeII greater than approximately 1.3\" are accurate; please make the threshold consistent.","section":"Section 3 vs. Section 4.1"},{"comment":"The vector label \"fwhmmg\" in Fig. 2 does not match the parameter label \"fwhmmg2\" used in the text.","section":"Fig. 2"},{"comment":"The column headers \"Capellupo 17\" and \"Capellupo 13\" should explicitly state that the numbers refer to the number of input parameters, and the caption should indicate that the Boroson & Green column is reproduced from their Table 4 rather than newly computed.","section":"Table 2"},{"comment":"Please state the allowed domain (x > 0) and note that y diverges as x approaches zero, which motivates the chosen starting point in Section 3.","section":"Eq. (1)"},{"comment":"The caption says the envelope is calculated from a \"five-point moving magnitude average\"; this should be clarified (mean or median), and the text \"mean envelop\" should be corrected to \"mean envelope.\"","section":"Fig. 7 caption and Section 4.1"}],"recommendation":"major_revision","confidential_remarks":"The PCA replication is solid and useful, but the headline nonlinear-plane claim is substantially stronger than the evidence presented. The post-hoc exclusion of ten objects and the unquantified variance claim are the two load-bearing points; both can be addressed by independent validation or by reframing the conclusions as conditional, so major revision rather than rejection seems appropriate."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper gives you a clean confirmation that the two-eigenvector structure of the quasar main sequence survives in newer, high-quality samples. That part is well done and useful. The second half, proposing a curved coordinate plane, is a genuinely new way to parameterize the FWHM(Hβ)-RFeII plane, but the evidence for it is weaker than the abstract suggests.\n\nWhat is actually new: the decay-curve coordinate system, arc length along y = 1/(a+x^b) and perpendicular distance from it, has not appeared in the cited literature. That is a legitimate contribution. The paper also deserves credit for being upfront about the outliers, showing example spectra and citing Sniegowska et al. for possible RFeII overestimation. That is honest reporting.\n\nThe soft spot is load-bearing. The two new axes are significantly correlated (Spearman p=0.03) in the full reduced Shen sample. The authors then remove two groups of five quasars at the ends of the curve, and the correlation disappears. Since the curve and the coordinate system are fitted to the same data used for the projection, and the outlier removal is decided after seeing the correlation, the claim that the plane is spanned by linearly independent axes is not independently tested. The paper's own caveats in Section 4.1, that it is 'not certain' RFeII>1.3 values are accurate and that the double-peaked Hβ explanation is 'plausible', confirm that this is a hypothesis, not a demonstrated result.\n\nOther concerns are minor by comparison: the Capellupo analysis uses only 30 objects with 13 parameters and no bootstrap or error analysis; no variance fraction is reported for the new plane; and no code or fitted a,b values are given. The physical interpretation in terms of black hole mass and luminosity is plausible but speculative, and the paper does hedge it appropriately.\n\nWho this is for: AGN observers who treat the main sequence as a diagnostic. They will get a useful confirmation of PCA stability and a new coordinate idea worth testing on independent samples. The central plane claim should not be taken at face value.\n\nRecommendation: send it to peer review. A serious referee can ask for out-of-sample validation, bootstrap uncertainties, and the variance fraction captured by the new axes. Those are answerable requests, not fatal flaws.","headline":"The PCA replication is solid and worth knowing; the new curved-plane claim rests on removing the same outliers that create the correlation, so it is conditional until validated independently.","tokens_in":12163,"tokens_out":2547,"would_cite":false,"duration_ms":24786,"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":"The paper argues the quasar main sequence is a nonlinear curve; distance along it and from it give two linearly independent coordinates that explain most of the variance, with black hole mass and luminosity as likely drivers.","keywords":["quasar main sequence","principal component analysis","Fe II to Hβ ratio","Hβ FWHM","black hole mass","Eddington ratio","active galactic nuclei","nonlinear decay curve"],"falsifier":"Re-measure Hβ FWHM and RFeII for the ten excluded quasars from high signal-to-noise spectra with a physically motivated Fe II model; if the measurements confirm the catalog values, or if the Spearman correlation between the two new coordinates is still significant (p < 0.05) when the full sample is included, the claimed plane is falsified.","tokens_in":11088,"feed_emoji":"🔭","tokens_out":13893,"duration_ms":106864,"temperature":0.7,"pith_summary":"The paper asks whether the quasar main sequence is really one line or a two-dimensional plane, and argues that a nonlinear curve gives the better picture. Two high-quality quasar samples—the XSHOOTER-based Capellupo sample and a clean subsample of the Shen catalog—both show the same two dominant principal components found in earlier work. Fitting a decay curve $y = 1/(a + x^b)$ to the plane of Hβ FWHM versus $R_{\\mathrm{FeII}}$, the authors define new coordinates: arc length along the curve and perpendicular distance from it. In the reduced Shen sample these coordinates are linearly independent once ten end-of-curve outliers are removed, and the plane they span accounts for most of the variance. If this is right, quasar spectral diversity in this plane is governed by just two physical parameters, most plausibly black hole mass along the curve and luminosity off it.","feed_headline":"Quasar diversity fits a curve, not a straight line","feed_subtitle":"Distance along and from that curve explains most of the variance, with black hole mass and luminosity as likely drivers","key_machinery":"The load-bearing object is the nonlinear decay curve fitted to the mean-normalized FWHM(Hβ)–$R_{\\mathrm{FeII}}$ plane, with $y$ the Hβ FWHM, $x$ the $R_{\\mathrm{FeII}}$, and $a,b$ the fitted parameters. Its 'nearpoint' projection assigns each quasar two new coordinates: the arc length along the curve and the perpendicular offset from it. These coordinates serve as a nonlinear analogue of the PCA eigenvectors, and the curve's steep-to-shallow shape is what lets the projection absorb variance that a straight line cannot.","core_discovery":"The central discovery is a re-parameterization of the quasar main sequence in the FWHM(Hβ)–$R_{\\mathrm{FeII}}$ plane. Rather than taking the two dominant PCA eigenvectors as straight orthogonal axes, the paper fits a simple two-parameter decay curve to the data and projects each quasar onto its nearest point on that curve. The new principal coordinate is the distance along the curve from a fixed origin; the new secondary coordinate is the Euclidean distance of the data point from the curve. After excluding two groups of five quasars at the curve's ends, these coordinates are linearly independent, and the plane they define accounts for the majority of the variance. The authors identify the main axis as an anti-correlation with black hole mass, with Eddington ratio rising along it, and the secondary axis as an anti-correlation with luminosity, while noting that inclination likely still shapes Hβ width.","pith_inferences":["A natural next test would be to fit the same decay curve to samples with independently measured virial black hole masses and Eddington ratios, rather than catalog estimates, to see whether arc length tracks mass cleanly.","The same curved-coordinate projection could be applied to rest-frame UV planes such as Mg II or C IV based diagnostics, to test whether the two-coordinate structure persists at higher redshift.","If the ten excluded quasars are reobserved and turn out to be genuine extreme members of the sequence, the plane would reduce to a line with residual correlation, so their re-measurement is the decisive check.","The functional form of the fitted curve gives photoionization and disk-structure models a concrete shape to reproduce, turning a geometric claim into a physical prediction."],"forward_implications":["The two leading eigenvectors remain dominant in newer high-quality quasar samples, so the main-sequence idea is not an artifact of the original 1992 sample.","Distance along the decay curve anti-correlates with black hole mass in the reduced Shen sample, making mass the best-supported candidate for the main axis.","Distance from the curve anti-correlates with luminosity, consistent with the broad-line-region radius–luminosity relation for the secondary axis.","If the mass interpretation holds, Eddington ratio declines at higher black hole mass along the sequence, possibly tied to a sub-Eddington limit and gas depletion in massive galaxies.","Inclination is not a primary driver of either axis but likely contaminates Hβ FWHM, so orientation should remain in the modeling."],"supporting_citations":[{"why":"Introduced PCA to quasar spectra and defined Eigenvector 1 as the anti-correlation between RFeII and Hβ FWHM; this is the baseline the paper replicates and extends.","marker":"Boroson & Green (1992)"},{"why":"Provides the reduced Shen sample's measurements of FWHM, RFeII, luminosities, Eddington ratios, and black hole masses.","marker":"Shen et al. (2011)"},{"why":"Earlier unified explanation of EV1 whose triangular distribution on the FWHM–RFeII plane is the shape comparison the decay curve improves on.","marker":"Shen & Ho (2014)"},{"why":"Finds that most quasars cataloged with very high RFeII do not actually have such high values, supporting the removal of the five high-axis outliers as artifacts.","marker":"Sniegowska et al. (2018)"},{"why":"Supplies the 30-object XSHOOTER sample whose measured parameters are used for the first PCA analysis.","marker":"Capellupo et al. (2015)"},{"why":"Established the quasar main sequence concept and the FWHM(Hβ)–RFeII plane used throughout the paper.","marker":"Sulentic et al. (2000)"},{"why":"Proposes a slim-disk/shielding picture connecting high Eddington ratio to stronger Fe II, supporting the black-hole-mass interpretation of the main axis.","marker":"Marziani et al. (2018)"},{"why":"Independent multivariate classification concluding black hole mass drives the optical plane, the main prior support for identifying the main axis with mass.","marker":"Fraix-Burnet et al. (2017)"},{"why":"Supplies the broad-line-region radius–luminosity relation used to argue that the secondary axis tracks luminosity.","marker":"Kaspi et al. (2005)"}],"fun_headline_variants":["Quasar main sequence is a curve, not a plane","Distance along and off the curve explains quasar variance","Black hole mass and luminosity shape quasar sequence","Curve projection beats PCA for quasar main sequence","New quasar axes: along and perpendicular to the sequence"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The linear independence of the two new axes rests on removing ten quasars at the ends of the fitted curve after the fact; if those objects are real members of the main sequence rather than measurement artifacts, the axes stay correlated and the claimed plane does not exist.","fun_headline_variants_meta":{"raw":{"variants":["Quasar main sequence is a curve, not a plane","Distance along and off the curve explains quasar variance","Black hole mass and luminosity shape quasar sequence","Curve projection beats PCA for quasar main sequence","New quasar axes: along and perpendicular to the sequence"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000709,"raw_usage":{"total_tokens":3248,"prompt_tokens":1055,"completion_tokens":2193,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":671,"completion_tokens_details":{"reasoning_tokens":2116}},"tokens_in":671,"tokens_out":2193,"duration_ms":17462,"temperature":1.0,"reasoning_tokens":2116,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:27:04.428365+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Re-measure Hβ FWHM and RFeII for the ten excluded quasars from high signal-to-noise spectra with a physically motivated Fe II model; if the measurements confirm the catalog values, or if the Spearman correlation between the two new coordinates is still significant (p < 0.05) when the full sample is included, the claimed plane is falsified.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduced PCA to quasar spectra and defined Eigenvector 1 as the anti-correlation between RFeII and Hβ FWHM; this is the baseline the paper replicates and extends."},{"cited_title":"T., Strauss, M","cited_arxiv_id":null,"evidence_quote":"Provides the reduced Shen sample's measurements of FWHM, RFeII, luminosities, Eddington ratios, and black hole masses."},{"cited_title":"M., Netzer, H., Lira, P., Trakhtenbrot, B., & Mej ´ıa-Restrepo, J","cited_arxiv_id":null,"evidence_quote":"Supplies the 30-object XSHOOTER sample whose measured parameters are used for the first PCA analysis."},{"cited_title":"2018, Frontiers in Astronomy and Space Sciences, 5, 28","cited_arxiv_id":null,"evidence_quote":"Proposes a slim-disk/shielding picture connecting high Eddington ratio to stronger Fe II, supporting the black-hole-mass interpretation of the main axis."},{"cited_title":"2017, Frontiers in Astronomy and Space Sciences, 4, 1","cited_arxiv_id":null,"evidence_quote":"Independent multivariate classification concluding black hole mass drives the optical plane, the main prior support for identifying the main axis with mass."}],"review_version":1}