{"id":"5d9a3835-a57f-4562-9c55-7e59c452baac","arxiv_id":"2602.13578","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":8,"one_line_summary":"From 24 lensed quasars with new velocity-dispersion measurements, the mass-profile slope γ increases with surface density (γs=0.37) and decreases with redshift (γz=-0.35).","lead":"Using 24 gravitational lenses where a background quasar is the source, this paper measures the mass-density slopes of early-type galaxies and finds they steepen with surface density and flatten with redshift. It is one of the first systematic applications of joint lensing-and-dynamics to quasar lenses, offering an independent check on galaxy-structure results previously derived from galaxy-galaxy lenses.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The γs>0 result may be an artifact of using the observed velocity dispersion both as the data and to define the surface-density covariate; the fit can absorb noise and bias γs upward. A mock-injection test is needed.","rationale":"The reader's weakest assumption focused on sample selection, which is a valid concern. However, the more serious and load-bearing issue is the self-referential covariate in Eq. 11: the dependent variable (σ_obs) and the predictor (Σ̃) are derived from the same noisy measurement. This creates a feedback that can bias γs upward and inflate its apparent significance, directly threatening the paper's headline claim that γ increases with surface mass density. The reader listed 'self-referential covariate' among methodological weaknesses but did not make it the weakest assumption; hence partial agreement. The new measurements and framework may still be salvageable, but the central result must be validated against this bias. A mock-injection test under the null γs=0 is the minimal, decisive check. The verdict remains CONDITIONAL rather than ACCEPT or REJECT, because the test could confirm that the reported γs is robust despite the circularity.","tokens_in":20641,"tokens_out":6139,"duration_ms":57319,"concrete_test":"Generate mock catalogs from the 24 observed lenses using the true model with γs_true=0, γz_true=−0.35, γ0_true=1.62, and the same observational errors in σ (Table 1). Compute Σ̃ from the noisy σ and run the full PyMultiNest pipeline; if the 68% posterior of γs is biased upward by more than ~0.1, the real γs=0.37 is not trustworthy. Alternatively, rerun the inference with Σ̃ computed from an independent surface-density estimate not involving σ_obs (e.g., stellar mass or a half-split of the spectrum), and check whether γs remains positive.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"Section 4.1 defines the surface-density covariate in Eq. 11 as Σ̃ = (σ_e2/100 km s−1)^2 / (R_eff/10 h−1 kpc), where σ_e2 is the aperture-corrected observed velocity dispersion from Eq. 8. This same σ_obs,e2 appears as the dependent variable in the Gaussian likelihood, Eq. 9. For each galaxy, γ_i = γ0 + γz z_l + γs log Σ̃_i is therefore a function of the noisy σ_obs. A positive measurement error in σ_obs inflates Σ̃, raises γ_i, and raises the predicted σ_th (Eq. 6), moving it toward the observed value; a negative error does the opposite. The posterior can absorb the measurement noise in σ_obs by choosing a spuriously positive γs. This is a textbook endogenous-regressor problem—regressing a quantity on a function of itself—and it directly targets the paper's headline claim 'γ_s > 0' (abstract and Section 4.2). The reported uncertainty on γs (~0.08) reflects only the statistical width under the assumed model, not the bias from this feedback. The paper's 'robustly demonstrate' claim is therefore unsupported unless this circularity is shown to be negligible. The selection-effect concern is real but secondary; the self-referential covariate is the most load-bearing point for the central result.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper compiles a parent sample of 106 galaxy-scale strong gravitational lenses with quasar sources and selects a 24-system subset for joint lensing and stellar-kinematics analysis. Newly presented data include 11 stellar velocity dispersions measured from archival BOSS/DESI spectra with an iterative pPXF-based masking procedure, and 4 effective radii from DESI-LS imaging. Assuming a power-law total mass profile, the authors parameterize the slope as γ = γ0 + γz z_l + γs log Σ̃ and report γ0=1.62^{+0.11}_{-0.12}, γz=-0.35^{+0.08}_{-0.09}, γs=0.37^{+0.08}_{-0.07} (68%), interpreting γz<0 and γs>0 as robust evidence for redshift and surface-density trends consistent with earlier galaxy-galaxy strong-lensing studies.","tokens_in":21033,"tokens_out":5992,"duration_ms":61893,"significance":"If the central result holds, the paper would provide a valuable independent confirmation of mass-profile evolution trends using lensed quasars, a sample that is complementary to galaxy-galaxy lenses and that extends to higher lens redshifts. The new spectroscopic measurements for 11 lens galaxies and effective radii for 4 systems are useful observational contributions, and the joint lensing-dynamics machinery is applied in a relatively clean Bayesian framework. However, the headline claims of 'robustly demonstrate' are currently undermined by a self-referential covariate in the γ parameterization and by an incompletely characterized selection from the parent sample. The paper's contribution is therefore promising but not yet established.","major_comments":[{"comment":"The surface-density covariate Σ̃ in Eq. (11) is constructed from σ_e2, the aperture-corrected observed velocity dispersion defined in Eq. (8). This same σ_obs,e2 is the dependent variable in the Gaussian likelihood, Eq. (9). Through Eq. (10), the model slope γ_i becomes a function of the noisy observation. A positive measurement error in σ_obs inflates Σ̃, increases γ_i, and through Eq. (6) raises the predicted σ_th, moving it toward the observed value; a negative error acts oppositely. The posterior can therefore absorb measurement noise, producing a spuriously positive γ_s. The reported uncertainty on γ_s (~0.08) reflects only statistical scatter under the assumed model, not this feedback. This is an endogenous-regressor problem. I request a mock-injection test: generate synthetic samples with known γ_s and realistic σ_obs errors, then check recovery; and/or a reanalysis where Σ̃ is ba","section":"§4.1, Eqs. (10)–(11); §3.5, Eq. (9)"},{"comment":"The sample selection from 106 to 24 systems is based on data availability, but the selection process is not fully described and is not modeled. Section 2.1 states that 13 systems had all required values and that 11 new σ_obs and 4 new θ_eff measurements expanded the sample to 24. Section 2.2 then reports that a cross-match with BOSS/DESI yielded 70 matches, three spectra were discarded as lacking reliable absorption features, and 'this effort produced new velocity dispersion measurements for 11 systems'. This leaves about 56 matches unused without explanation. If the 11 were selected by S/N, visual quality, or other criteria, the analysis sample is not a random subset of the parent population. Differences in redshift, Einstein radius, velocity dispersion, or effective radius between the included and excluded systems could bias both γ_z and γ_s. At minimum, compare the distributions of th","section":"§2.1–2.2"},{"comment":"The statistical model is incompletely specified. The parameter vector is stated as Θ = {γ0, γz, γs, Ωm}, but the likelihood in Eq. (9) uses a total uncertainty Δσ_tot that 'incorporates the measurement error of σ_ap, the uncertainty propagated from the aperture correction (η), and a systematic term'. The prior on η is given in §3.4, but it is not stated whether η is sampled, analytically marginalized, or included as an added variance term. The 'systematic term' is never defined, and no prior or fixed value is provided for it. This prevents exact reproduction of the likelihood and leaves the amplitude of the systematic uncertainty unquantified. Please spell out the full generative model, including how η and σ_sys enter, and the priors used.","section":"§4.1 and §3.4–3.5"}],"minor_comments":[{"comment":"Please clarify the count discrepancy: 70 cross-matches, 3 discarded, but only 11 velocity dispersions reported. The selection among the remaining spectra is a key part of sample construction and should be explicit.","section":"Section 2.2"},{"comment":"There are naming inconsistencies: 'SDSSJ1640+1932' in Table 1 appears as 'SDSSJ1640-1932' in Appendix A; 'WFI2033-4723' in Table 1 appears as 'WFJ2033-4723' in Appendix A. Please unify.","section":"Table 1 / Appendix A"},{"comment":"For slit-based observations, the table lists a single θ_ap; please specify whether an equivalent circular aperture radius was adopted for the correction in Eq. (8), and how this was derived.","section":"§3.4 / Table 1"},{"comment":"The caption states that dashed lines show 'the mean value'. For skewed posteriors, the posterior median or mode would be more standard; please clarify which statistic is plotted.","section":"Figure 5"},{"comment":"The phrase 'robustly demonstrate' is stronger than the current evidence warrants given the issues above. I suggest tempering the wording unless the mock-injection test and selection checks support the claim.","section":"Abstract / §4.2"}],"recommendation":"major_revision","confidential_remarks":"The core concern raised in the stress-test note is, in my reading, well-founded: the covariate Σ̃ in Eq. (11) is built from the same observed velocity dispersion that the likelihood fits, creating a direct pathway for noise to bias γ_s. This must be addressed with a mock-injection test or a hierarchical reanalysis before the paper's central claim can be accepted. The sample-selection description in §2.2 also contains an unexplained 70→11 gap that needs to be resolved. If the bias test shows the effect is negligible, the paper could become acceptable after mostly local revisions; otherwise the conclusion may not survive."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a useful paper with real new measurements, but the central claim about surface-density dependence is undermined by an endogenous-regressor problem. I'd send it out, and require the authors to address that.\n\nThe genuine value is the data work: 11 new velocity dispersions for lensed quasar galaxies, obtained with a sensible iterative pPXF procedure to mask quasar emission lines, checked against BOSS pipeline values; plus 4 new effective radii. That fills a real gap and makes the GQSL population accessible to joint lensing-dynamics analysis. Applying the Chen et al. framework to 24 GQSL systems and getting trends consistent with earlier GGSL work is a legitimate first step, and the γz result (−0.35) is qualitatively in line with the literature.\n\nThe soft spot is in Section 4.1. Equation (11) defines Σ̃ from the aperture-corrected observed dispersion σ_e2, and Equation (9) fits that same observed dispersion. So the regression has the outcome on both sides. A positive noise on σ_obs inflates Σ̃, raises the predicted γ, and increases the predicted σ_th, pulling the model toward the data. The posterior can absorb measurement noise by adjusting γs upward. The reported uncertainty on γs (~0.08) does not include this feedback. The authors call the result 'robustly demonstrate'; I don't think that's supported until they run a mock-injection test or use an independent covariate (e.g., stellar mass from photometry).\n\nThe sample-selection issue is real but secondary: going from 106 to 24 by data availability and not modeling the selection could bias both trends, though the consistency with earlier work argues against a catastrophic effect.\n\nMinor points: the systematic term σsys is mentioned but not defined; no data product is released beyond Table 1.\n\nBottom line: the paper is a solid dataset contribution and worth a serious referee, but the quantitative surface-density slope should be treated as provisional. The reader's conditional verdict is about right.\n\nWho: strong-lensing and galaxy-evolution readers. Recommendation: accept for peer review, but the revision should address the endogeneity, ideally with mock injections using the same pipeline and noise model.","headline":"Useful new data and a genuine first application, but the headline γs>0 result is not robust because the covariate and the outcome come from the same measured velocity dispersion.","tokens_in":21514,"tokens_out":2937,"would_cite":true,"duration_ms":29786,"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":"This paper claims that the total mass-density slope of massive early-type lens galaxies evolves with both redshift and surface mass density, measured for the first time from a sample of lensed quasars.","keywords":["strong gravitational lensing","lensed quasars","galaxy mass density profiles","stellar velocity dispersion","early-type galaxies","galaxy evolution","joint lensing dynamics","power-law slope"],"falsifier":"Measure stellar velocity dispersions for all 106 parent-sample lenses from new deep spectroscopy and rerun the joint analysis; if the recovered γz and γs shift outside the quoted error bars, the reported trends are an artifact of which systems happened to have archival spectra.","tokens_in":20559,"feed_emoji":"🔭","tokens_out":5786,"duration_ms":52784,"temperature":0.7,"pith_summary":"The paper tries to establish that the internal mass structure of massive early-type galaxies changes with cosmic time and with galaxy density. By combining strong-lensing geometry with stellar kinematics for 24 lensed quasars, it measures the power-law slope of the total mass profile and finds it decreases with redshift and increases with surface mass density. This matters because it tests galaxy formation models and supports the inside-out growth picture, while validating lensed quasars as an independent probe of galaxy structure.","feed_headline":"Galaxy mass profiles steepen with density, flatten with redshift","feed_subtitle":"Denser lens galaxies have steeper mass profiles; higher redshift makes them shallower, per 24 lensed quasars.","key_machinery":"The central identity is the joint lensing–dynamics mass equality, M_grl = M_dyn, which equates the projected mass inside the Einstein radius from lens geometry with the dynamical mass from stellar kinematics. The analysis combines power-law density and luminosity profiles, the spherical Jeans equation, and an aperture correction to predict the velocity dispersion within a standardized aperture. The other key piece is a new iterative spectral-fitting procedure that masks contaminating quasar emission lines, allowing velocity dispersions to be measured for lens galaxies whose spectra are otherwise overwhelmed by the background quasar.","core_discovery":"Modeling each lens galaxy's total mass profile as a power law, ρ ∝ r^(−γ), the authors parameterize the slope as γ = γ0 + γz·z_l + γs·log Σ̃, where z_l is the lens redshift and Σ̃ is a normalized surface mass density. From joint lensing–dynamics analysis of 24 systems, they infer γ0 = 1.62(+0.11/−0.12), γz = −0.35(+0.08/−0.09), and γs = +0.37(+0.08/−0.07). The negative γz says that at fixed density, higher-redshift galaxies have shallower (less centrally concentrated) mass profiles; the positive γs says that at fixed redshift, denser galaxies have steeper profiles. The paper argues these trends are statistically significant and consistent with earlier galaxy–galaxy lensing results, and it pr","pith_inferences":["The authors do not model the selection of the 24 systems; because only systems with usable spectra enter the analysis, the reported γz and γs could be biased if measurable lenses are brighter or more massive. A complete kinematic census of the parent sample would settle this.","If the γs > 0 relation holds generally, it suggests a unified scaling between galaxy mass profile and stellar surface density that could be applied to non-lensed galaxies as a dynamical mass estimator.","The steeper γz compared with galaxy–galaxy compilations may reflect the higher redshift range of lensed quasars, but it could also indicate sample selection; targeted spectroscopy of the remaining parent lenses would test this directly.","The parameterized γ(z, Σ) provides a quantitative prediction that cosmological simulations of massive early-type galaxies should reproduce if the trends are physical."],"forward_implications":["If the measured trends are correct, massive early-type galaxies were systematically less centrally concentrated at higher redshift, supporting inside-out assembly.","Denser galaxies having steeper profiles implies a tight coupling between stellar surface density and total mass distribution, useful for galaxy formation models.","Lensed quasars become a validated, independent population for joint lensing–dynamics studies, complementing galaxy–galaxy lenses.","The measured γ(z, Σ) relation can serve as a prior for time-delay cosmography, reducing a key systematic in Hubble constant measurements.","The spectroscopic technique for handling quasar contamination extends directly to the much larger lensed-quasar samples expected from upcoming wide-field surveys."],"fun_headline_variants":["Lensed quasars: mass slope steepens with density, flattens with redshift","Quasar lenses reveal mass profile slope depends on density and redshift","24 lensed quasars: denser galaxies have steeper mass profiles","New lensing-dynamics results: mass slope up with density, down with redshift"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The 24 lens systems with measurable velocity dispersions are treated as representative of the 106-system parent population, even though selection depends on the availability and quality of archival spectra; if those systems are biased toward brighter or more massive lenses, the inferred redshift and density trends could be wrong.","fun_headline_variants_meta":{"raw":{"variants":["Lensed quasars: mass slope steepens with density, flattens with redshift","Quasar lenses reveal mass profile slope depends on density and redshift","24 lensed quasars: denser galaxies have steeper mass profiles","New lensing-dynamics results: mass slope up with density, down with redshift"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00063,"raw_usage":{"total_tokens":2860,"prompt_tokens":970,"completion_tokens":1890,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":714,"completion_tokens_details":{"reasoning_tokens":1808}},"tokens_in":714,"tokens_out":1890,"duration_ms":13363,"temperature":1.0,"reasoning_tokens":1808,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-02T23:29:32.194830+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure stellar velocity dispersions for all 106 parent-sample lenses from new deep spectroscopy and rerun the joint analysis; if the recovered γz and γs shift outside the quoted error bars, the reported trends are an artifact of which systems happened to have archival spectra.","supporting_citations":[],"review_version":1}