{"id":"17d94f51-759e-4e3d-9846-6df555717a86","arxiv_id":"2509.09798","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"A new compilation of over 30,000 strong-lensing candidates with multi-survey images yields an independent ground-based constraint γ_PPN = 1.023 ± 0.028 and a combined constraint γ_PPN = 1.034 ± 0.025.","lead":"The authors assembled a catalog of more than 30,000 known or candidate strong gravitational lenses and generated image cutouts for them in seven ground-based surveys, preparing a 'watch list' for the Rubin Observatory. As a demonstration, they used 206 of these systems to test General Relativity through the PPN parameter γ, finding γ = 1.023 ± 0.028, consistent with Einstein's theory.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The γPPN headline precision inherits prior-dominated β and δ; the quoted error omits a systematic term for population scatter in slopes and anisotropy, and the aperture-correction shift (~0.04) is comparable to the statistical error.","rationale":"The reader's weakest_assumption identifies exactly the load-bearing point: the γPPN inference depends on prior-dominated β and δ, and the statistical error does not include a population-scatter or prior-location systematic. My reading of the full text confirms this: Section 7.1.2 reports the priors, states that β and δ are prior-dominated, and presents the aperture-correction sensitivity as a consistency check rather than as a systematic term. The database construction itself is a strong, independently useful deliverable: provenance is preserved, cross-matches are documented, cutout counts are reported per survey, and code is public, so the database claim does not need to be rejected. The γPPN claim should therefore remain conditional: the resource is acceptable, but the headline constraint should carry a qualifier and a robustness test. I found no internal inconsistency in the compilation steps that would change the reader's verdict. The concrete test above would settle whether the prior-dominance concern actually shifts γPPN beyond the quoted error.","tokens_in":34392,"tokens_out":6778,"duration_ms":59584,"concrete_test":"Re-run the §7.1.2 MCMC on LaStBeRu_cosmo_ground with β and δ priors changed to wide uniform priors over physically allowed ranges (e.g., β∈[−0.5,0.5], δ∈[1.5,3.5]), and separately with a hierarchical population model that draws per-system α, δ, and β from distributions with free scatter. Compare the resulting γPPN central value and uncertainty with 1.023±0.028. Also recompute the aperture-corrected versus uncorrected difference on the identical 280-system combined sample. If the central value shifts by more than about 0.02 or the uncertainty grows by more than about 50%, the quoted precision is not robust to the homology assumption.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central scientific deliverable, the γPPN constraint in §7.1.2, rests on Equation (8) evaluated with global power-law slopes α, δ and constant anisotropy β, with Gaussian priors ⟨α⟩=2.00±0.08, ⟨β⟩=0.18±0.13, ⟨δ⟩=2.40±0.11. The paper explicitly states that the posteriors for β and δ were prior-dominated. In this regime, γPPN is marginalized over parameters whose values are set by the priors, not by the lensing and dynamics data. If the true ETG population has different mean slopes or anisotropy, or appreciable intrinsic scatter not represented by those priors, the central value of γPPN will shift. The quoted 1.023±0.028 (ground) and 1.034±0.025 (combined) therefore propagate the widths of the priors but not the possibility that the prior locations are wrong. The aperture-corrected versus uncorrected results (0.979±0.029 vs 1.023±0.028) differ by about 1.5σ, showing a systematic of order 0.04 that is not included in the error budget. Because a single homologous population is assumed for 206 or 280 lenses spanning heterogeneous θ_E and σ_v measurements, this is the least secure condition for the headline claim.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper presents LaStBeRu, a consolidated database of 31,569 unique strong-lensing candidates compiled from the literature, cross-matched with photometric and spectroscopic catalogs from seven wide-field surveys, and accompanied by 366,905 image cutouts in two angular sizes. The catalog preserves provenance for every value, merges duplicates by JNAME with explicit 1 and 10 arcsec rules, and applies objective quality filters (seeing, exposure time, empty-cutout removal) for image selection. As a demonstration application, the authors construct a 206-system ground-based sample (LaStBeRu_cosmo_ground) of isolated early-type galaxy lenses with SDSS velocity dispersions and use the Schwab et al. (2010) method to constrain the PPN parameter γ_PPN, reporting 1.023 ± 0.028 for the ground sample and 1.034 ± 0.025 for the combined sample of 280 unique systems, which they describe as the tightest constraint from this type of analysis.","tokens_in":34676,"tokens_out":8723,"duration_ms":71522,"significance":"The catalog construction is a substantial and careful contribution: provenance is tracked for each measurement, the JNAME merging rules are explicit, cross-matching uses a 1 arcsec nearest-neighbour criterion, cutout selection uses header-based seeing and exposure criteria, empty cutouts are removed with SEP, and the data and code are publicly available. This resource should be useful for LSST preparation, machine-learning training, and follow-up target selection. The γ_PPN application is a reasonable demonstration of the database, but the headline precision claim is not yet supported by the error budget because the posterior for γ_PPN is conditioned on prior-dominated population parameters (β, δ) and the quoted uncertainty omits the corresponding systematic. The catalog itself is not affected by this issue, but the application's conclusions need revision before publication.","major_comments":[{"comment":"The paper states that 'Posteriors for β and δ were prior-dominated.' Because γ_PPN is inferred jointly with the global power-law slopes α, δ and anisotropy β, a prior-dominated β and δ means the γ_PPN posterior is effectively conditional on the prior means ⟨β⟩=0.18 and ⟨δ⟩=2.40. The quoted errors (1.023±0.028 and 1.034±0.025) propagate the prior widths but not the uncertainty in the prior locations, nor any intrinsic population scatter. The comparison between aperture-corrected (1.023±0.028) and uncorrected (0.979±0.029) results shows a shift of ~0.04, comparable to the statistical error, which indicates a systematic not captured by the 3% line-of-sight term in Eq. (9). The authors should either marginalize over population hyper-parameters, add a systematic term, or explicitly frame the result as conditional on the adopted priors and avoid calling it the tightest constraint.","section":"7.1.2, Eqs. (8)-(9)"},{"comment":"The abstract claims 'the most stringent constraint on γ_PPN', but Section 7.1.2 correctly qualifies this as 'the tightest constraint on γ_PPN from this type of analysis to date'. Solar-system tests constrain γ_PPN to ~10^-5, orders of magnitude tighter than the values reported here. The abstract should carry the same qualifier as the body text, or the claim will be misleading.","section":"Abstract"},{"comment":"The concluding remarks state 'For the first time, we used priors on the brightness and density slopes from the same data, providing a more self-consistent analysis compared to previous works.' However, Section 7.1.2 lists Gaussian priors ⟨α⟩=2.00±0.08, ⟨β⟩=0.18±0.13, ⟨δ⟩=2.40±0.11 without any derivation from the LaStBeRu data. If these priors are taken from the literature, the claim in Section 8 is unsupported and should be removed or corrected; if they were estimated from the sample, the estimation procedure must be described. This inconsistency is central to the claimed novelty of the γ_PPN analysis.","section":"8 and 7.1.2"}],"minor_comments":[{"comment":"The text writes 'SWEELS' catalog, but the standard acronym for the Sloan WFC Edge-on Late-type Lens Survey is SWELLS (Treu et al. 2011).","section":"7.1.1"},{"comment":"The sentence 'which is relevant for the modeling and has not typically been used in previous statistical analyses of this type' is repeated twice in the same paragraph.","section":"7.1.1"},{"comment":"The MCMC setup (64 walkers, 15,000 steps, 500 burn-in) is reported, but no convergence diagnostics (e.g., autocorrelation time or Gelman-Rubin statistic) are given; please report them for the quoted γ_PPN values.","section":"7.1.2"},{"comment":"The numbers in the Venn-diagram labels are not explained in the caption; please define what each number represents (e.g., number of systems in each intersection).","section":"Figure 13"},{"comment":"The empty-cutout fractions (11% for CFHTLenS, 18% for RCSLenS) should be reported as fractions of attempted cutouts, and the authors should comment on whether the removal of empty cutouts biases the final catalog against faint lenses or poor-seeing images.","section":"5.1"},{"comment":"The data are hosted on a GitHub repository; for long-term reproducibility, a versioned DOI or Zenodo record would be preferable.","section":"6"}],"recommendation":"major_revision","confidential_remarks":"The catalog itself is a solid data product and likely well within the journal's scope. My main concern is the γ_PPN application's headline claims, which overstate the precision and independence of the result. These are fixable by revision; I would not reject the paper. The authors should also be asked to clarify the provenance of the priors in Section 7.1.2, as the conclusion's claim of using 'priors from the same data' appears to contradict the methods description."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know. First, the LaStBeRu database is a genuinely useful community resource: 31,569 lensing candidates with aggregated tabular data, provenance, and 366,905 uniform ground-based cutouts from seven surveys. If you work on strong lensing or Rubin preparation, this is a real product, and the catalog construction is careful. Second, the gamma_PPN application is a legitimate but not novel analysis, and its headline precision is partly borrowed from priors the data do not constrain.\n\nThe new stuff: the database itself, and the 206-system ground-based sample with 103 systems not used in earlier gamma_PPN analyses. The duplicate merging via JNAME, the 1-arcsecond nearest-neighbor cross-matches, the cutout selection using seeing and exposure headers, and the empty-cutout rejection are all described transparently, with code and data public. That is exactly what a resource paper should look like.\n\nThe soft spots are in the gamma_PPN section, and they are real but not fatal. The posteriors for the luminosity slope delta and anisotropy beta are explicitly prior-dominated, so the quoted gamma_PPN error propagates the widths of those priors but not the risk that the prior locations are wrong. The aperture-corrected versus uncorrected results (1.023 ± 0.028 vs 0.979 ± 0.029) differ by about 0.04, comparable to the statistical error, and that systematic is not in the error budget. The abstract's 'most stringent constraint' should carry the qualifier 'from this type of analysis' that the text already has. None of this undermines the catalog; it means the gamma_PPN result is less solid than the '±0.025' suggests.\n\nWho this is for: anyone needing a consolidated watchlist, an ML training set, or a follow-up target selector for ground-based lensing before LSST. That audience gets real value. I would send it to a referee: the database deserves careful review, and the gamma_PPN analysis needs the systematic treatment above. I would recommend conditional acceptance, with the gamma_PPN claims toned down and either a systematic term or a sensitivity analysis added.","headline":"LaStBeRu is a genuinely valuable strong-lensing database with careful construction; the gamma_PPN result is a legitimate but prior-limited application whose quoted error needs a systematic term.","tokens_in":35261,"tokens_out":2575,"would_cite":true,"duration_ms":380869,"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":"This paper builds the LaStBeRu database of 31,569 unique strong-lensing systems with 366,905 image cutouts from seven wide-field surveys, and uses a 206-system ground-based subsample to measure $\\gamma_{\\mathrm{PPN}} = 1.023 \\pm 0.028$…","keywords":["strong gravitational lensing","galaxy-galaxy lensing","catalogues","wide-field surveys","post-Newtonian parameter","tests of general relativity","stellar velocity dispersions","ground-based imaging"],"falsifier":"Re-run the MCMC analysis with the same 206 systems but replace the global Gaussian priors on $\\alpha$, $\\beta$, $\\delta$ with per-system values measured from independent data, such as spatially resolved IFU kinematic profiles for a subset of lenses. If the resulting $\\gamma_{\\mathrm{PPN}}$ moves outside the quoted $1.034 \\pm 0.025$ by more than the statistical uncertainty, the prior-dominated assumption of a homologous population is doing the work. A simpler check: the authors' own aperture-corrected and uncorrected results differ by 0.044 (1.023 vs 0.979), which is larger than either statistical error bar, so an analysis that fully propagates uncertainties in effective radii and seeing would reveal whether that missing systematic explains the difference.","tokens_in":34173,"feed_emoji":"🔭","tokens_out":11047,"duration_ms":93455,"temperature":0.7,"pith_summary":"This paper assembles a single consolidated database, LaStBeRu, of 31,569 unique strong gravitational lensing systems drawn from the literature, enriched by cross-matching with photometric and spectroscopic surveys and accompanied by 366,905 image cutouts from seven wide-field ground-based surveys. The authors argue this is the most complete ground-based watch list of known lenses available before the next generation of wide-field surveys begins, useful for training automated finders, planning follow-up observations, and archival cosmological analyses. As a demonstration, they select 206 isolated early-type galaxy lenses that are clearly identifiable in ground-based images and use them, together with published samples, to constrain the post-Newtonian parameter that measures how much space curvature light feels compared with matter. Their MCMC analysis yields $\\gamma_{\\mathrm{PPN}} = 1.023 \\pm 0.028$ for the new ground-based sample alone and $\\gamma_{\\mathrm{PPN}} = 1.034 \\pm 0.025$ when combined with two earlier catalogs, which they describe as the tightest constraint of this kind to date; both results agree with general relativity at the quoted uncertainty level.","feed_headline":"31,569-lens archive yields tightest lens-based gravity check","feed_subtitle":"The catalog preps wide-field surveys and keeps general relativity intact at galactic scales.","key_machinery":"The framework has two load-bearing parts. The aggregation pipeline ('slcomp') gives every system a unique coordinate-based JNAME, merges duplicates within a 10-arcsecond radius, cross-matches against photometric and spectroscopic catalogs at a 1-arcsecond radius, and applies hierarchical rules to assign one adopted value per physical quantity while keeping full provenance. The gravity test then rests on an analytic model that equates the luminosity-weighted, seeing-convolved line-of-sight velocity dispersion of an early-type galaxy to the value predicted from its Einstein radius and the lens-source distance ratio, assuming power-law mass and light profiles $\\rho(r) \\propto r^{-\\alpha}$, $\\nu(r) \\propto r^{-\\delta}$, and a constant velocity-anisotropy parameter $\\beta$. In this model lensing responds to the sum of the two metric potentials $\\Phi + \\Psi$ while stellar dynamics respond to $\\Phi$ alone, and the factor $2/(1+\\gamma_{\\mathrm{PPN}})$ in the predicted dispersion is what the data constrain.","core_discovery":"The central claim has two parts. First, the LaStBeRu database consolidates 31,569 unique strong-lensing systems from many separate literature catalogs, resolves duplicate entries by averaging positions into a single coordinate-based identifier, cross-matches them with photometric and spectroscopic surveys to add magnitudes, redshifts, and velocity dispersions, and delivers 20-arcsecond and 4-arcminute image cutouts in every available band of seven ground-based surveys. Second, using a visually inspected subsample of 206 isolated early-type galaxy lenses that show clear lensing features in ground-based images and whose velocity dispersions come from a single spectroscopic survey, the paper obtains an independent measurement of the post-Newtonian parameter $\\gamma_{\\mathrm{PPN}} = 1.023 \\pm 0.028$ by combining strong-lensing geometry with stellar dynamics. Combining this sample with two earlier published catalogs, for a total of 280 unique systems, yields $\\gamma_{\\mathrm{PPN}} = 1.034 \\pm 0.025$, which the authors report as the tightest constraint from this kind of analysis to date and consistent with general relativity.","pith_inferences":["If the database is actively maintained as new lens discoveries appear, it could become a standard calibration set for automated lens finders; that wider adoption is not something the paper can guarantee.","The 0.044 gap between the aperture-corrected and uncorrected $\\gamma_{\\mathrm{PPN}}$ estimates is larger than the quoted statistical errors, hinting that the true systematic floor of this measurement may be set by assumptions about galaxy profiles and aperture corrections rather than by sample size.","A natural extension, not pursued here, would be to measure velocity dispersions for the hundreds of LaStBeRu lenses that lack them, which could roughly halve the uncertainty on $\\gamma_{\\mathrm{PPN}}$ and test whether the gravitational slip parameter is constant across redshift or environment."],"forward_implications":["The LaStBeRu database can serve as a watch list for lensed transient events and as a benchmark for training machine-learning lens finders on ground-based data, directly addressing the expected flood of candidates from the next generation of wide-field surveys.","The 206-system ground-based sample shows that competitive $\\gamma_{\\mathrm{PPN}}$ constraints can be obtained without space-based imaging, using homogeneous ground-based images and a single spectroscopic survey's velocity dispersions.","Combining LaStBeRu with the two earlier catalogs produces the tightest $\\gamma_{\\mathrm{PPN}}$ constraint of this method so far, so future measurements of additional velocity dispersions can push the constraint further.","The database's cross-matching adds thousands of new redshifts and velocity-dispersion measurements (an 851% increase in velocity dispersions over the original literature compilation), enabling archival-only cosmological analyses of hundreds of lenses."],"supporting_citations":[{"why":"Supplies the method and analytic formula that connect observed velocity dispersions to $\\gamma_{\\mathrm{PPN}}$ through lensing geometry and power-law galaxy profiles.","marker":"Schwab et al. (2010)"},{"why":"Reports the first $\\gamma_{\\mathrm{PPN}}$ measurement from the SLACS sample, the baseline this paper's result extends and compares against.","marker":"Bolton et al. (2006a)"},{"why":"The Master Lens Database is the main meta-catalog from which thousands of duplicate lens entries are consolidated into LaStBeRu.","marker":"Moustakas (2012)"},{"why":"One of the two earlier compilations whose systems are combined with LaStBeRu to reach the 280-system $\\gamma_{\\mathrm{PPN}}$ constraint.","marker":"Cao et al. (2015)"},{"why":"The other earlier compilation used in the combined 280-system sample and a source of comparison literature values.","marker":"Chen et al. (2019a)"},{"why":"Provides the likelihood and uncertainty model, including the 3% systematic term, adopted for the $\\gamma_{\\mathrm{PPN}}$ MCMC analysis.","marker":"Liu et al. (2022)"},{"why":"Gives the aperture-correction slope $\\zeta = -0.066 \\pm 0.035$ used to homogenize velocity dispersions measured in different apertures.","marker":"Cappellari et al. (2006)"},{"why":"Justifies identifying the gravitational slip parameter $\\eta$ with the PPN parameter $\\gamma_{\\mathrm{PPN}}$ under the assumptions of this analysis.","marker":"Toniato & Rodrigues (2021)"}],"fun_headline_variants":["31,569-lens archive tightens gravity's toughest test","Mega-lens catalog preps Rubin, nails GR with 31k+ systems","Largest lens sample yet delivers tightest gamma_PPN test","31,569 lenses in one catalog, plus a fierce gravity test"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that all 206 lenses form one homologous population of early-type galaxies, so a single global power-law mass slope, light slope, and velocity-anisotropy parameter can be enforced through Gaussian priors centered at $\\langle\\alpha\\rangle = 2.00$, $\\langle\\beta\\rangle = 0.18$, and $\\langle\\delta\\rangle = 2.40$. If the real lenses spread more widely in slope or anisotropy, the inferred $\\gamma_{\\mathrm{PPN}}$ shifts; the fact that the posteriors for $\\beta$ and $\\delta$ are prior-dominated means the priors, not the data, anchor those parameters.","fun_headline_variants_meta":{"raw":{"variants":["31,569-lens archive tightens gravity's toughest test","Mega-lens catalog preps Rubin, nails GR with 31k+ systems","Largest lens sample yet delivers tightest gamma_PPN test","31,569 lenses in one catalog, plus a fierce gravity test"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000782,"raw_usage":{"total_tokens":3542,"prompt_tokens":1119,"completion_tokens":2423,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":735,"completion_tokens_details":{"reasoning_tokens":2346}},"tokens_in":735,"tokens_out":2423,"duration_ms":16894,"temperature":1.0,"reasoning_tokens":2346,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T15:59:32.519153+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Re-run the MCMC analysis with the same 206 systems but replace the global Gaussian priors on $\\alpha$, $\\beta$, $\\delta$ with per-system values measured from independent data, such as spatially resolved IFU kinematic profiles for a subset of lenses. If the resulting $\\gamma_{\\mathrm{PPN}}$ moves outside the quoted $1.034 \\pm 0.025$ by more than the statistical uncertainty, the prior-dominated assumption of a homologous population is doing the work. A simpler check: the authors' own aperture-corrected and uncorrected results differ by 0.044 (1.023 vs 0.979), which is larger than either statistical error bar, so an analysis that fully propagates uncertainties in effective radii and seeing would reveal whether that missing systematic explains the difference.","supporting_citations":[{"cited_title":"S., Rappaport S","cited_arxiv_id":null,"evidence_quote":"Supplies the method and analytic formula that connect observed velocity dispersions to $\\gamma_{\\mathrm{PPN}}$ through lensing geometry and power-law galaxy profiles."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the likelihood and uncertainty model, including the 3% systematic term, adopted for the $\\gamma_{\\mathrm{PPN}}$ MCMC analysis."},{"cited_title":"D., Rodrigues D","cited_arxiv_id":null,"evidence_quote":"Justifies identifying the gravitational slip parameter $\\eta$ with the PPN parameter $\\gamma_{\\mathrm{PPN}}$ under the assumptions of this analysis."}],"review_version":2}