Pith. sign in

REVIEW 2 major objections 56 references

A test-time method turns single-view material predictors into multi-view consistent decompositions by aligning them to geometric consensus targets, without retraining.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · grok-4.5

2026-07-14 21:35 UTC pith:XFB2L3RF

load-bearing objection Document mismatch: title/abstract are Geo-ID (cs.CV), but the supplied full text is a Magellanic Clouds reddening paper; Geo-ID cannot be evaluated from this body. the 2 major comments →

arxiv 2603.13859 v2 pith:XFB2L3RF submitted 2026-03-14 cs.CV

Geo-ID: Test-Time Geometric Consensus for Cross-View Consistent Intrinsics

Alara Dirik , Stefanos Zafeiriou This is my paper
classification cs.CV
keywords intrinsic image decompositioncross-view consistencytest-time optimizationgeometric consensusPBR materialsalbedomulti-view
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

Intrinsic image decomposition estimates material properties such as albedo, roughness, and metallicity from photographs. Strong single-view models exist, but applying them independently to different views of the same scene produces inconsistent materials, which limits editable neural scenes and 3D reconstruction. This paper claims that inconsistency can be fixed at test time: take any off-the-shelf single-view predictor, couple its per-view outputs through sparse geometric correspondences between views, and treat those links as uncertainty-aware consensus targets that pull the predictions into agreement. The approach needs no retraining and no inverse rendering, works on sparse unordered image sets rather than dense video, and improves cross-view consistency as more views are added while keeping single-view quality comparable. The resulting consistent materials then support coherent appearance editing and relighting in neural scene representations.

Core claim

Geo-ID shows that pretrained single-view intrinsic predictors can be repurposed, without retraining or inverse rendering, to produce cross-view consistent PBR decompositions by coupling independent per-view predictions through sparse geometric correspondences that form uncertainty-aware consensus targets. Cross-view consistency improves as the number of views grows, single-view performance stays comparable, and the consistent maps enable coherent downstream editing and relighting.

What carries the argument

Uncertainty-aware geometric consensus targets: sparse correspondences between views define shared targets that couple independent single-view intrinsic predictions at test time, forcing material estimates to agree where geometry links the views.

Load-bearing premise

The method assumes sparse geometric matches between views are accurate enough, and that consensus targets built from them stand for true physical material consistency rather than averaging errors or spreading bad matches.

What would settle it

On a multi-view synthetic scene with known ground-truth materials, check whether Geo-ID raises cross-view material agreement and absolute material accuracy over independent single-view runs as views are added; if agreement rises while absolute error does not improve (or worsens) versus ground truth, the consensus is only smoothing predictions, not recovering true materials.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • Sparse, unordered multi-view photo sets can receive consistent material maps without video models or dense ordered sequences.
  • Off-the-shelf single-view intrinsic networks become usable inputs to editable neural scenes and multi-view 3D pipelines.
  • Adding more views further improves material agreement under the method while single-view quality holds.
  • Downstream relighting and appearance editing become coherent across viewpoints when materials come from Geo-ID rather than independent per-view runs.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same geometric-consensus idea could be applied to other single-view estimators that should be multi-view consistent, such as surface normals or lighting, not only PBR materials.
  • Results will hinge on correspondence quality in textureless or highly specular regions, so pairing with stronger matchers or explicit uncertainty masks is a natural next stress test.
  • If consensus targets can be built incrementally, the method could support progressive capture sessions rather than only batch multi-view sets.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

2 major / 0 minor

Summary. The manuscript claims to present Geo-ID, a model-agnostic test-time framework that couples independent single-view intrinsic (PBR) predictors via sparse geometric correspondences to form uncertainty-aware consensus targets, thereby improving cross-view consistency of albedo/roughness/metallicity without retraining or inverse rendering, while preserving single-view quality and enabling coherent neural-scene editing. The supplied full text, however, is an entirely different paper: three-dimensional reddening maps of the Magellanic Clouds constructed from OGLE-IV RRab stars, adaptive quadtree partitioning, and iterative E(V-I)–distance fits (arXiv:2603.13860). No Geo-ID method, equations, experiments, or results appear in the body.

Significance. If the Geo-ID claims in the abstract were supported by a matching manuscript, a training-free, correspondence-driven consensus procedure that improves multi-view intrinsic consistency for sparse unordered views would be of clear practical value for editable neural scenes and 3D reconstruction. The actual body instead delivers the first 3D reddening maps of the LMC/SMC from RR Lyrae, with publicly released partitions and a query tool—an independently useful astronomical product. Because the abstract and body describe different works, neither contribution can be properly assessed or credited under the submitted identity.

major comments (2)
  1. Document identity failure: the title/abstract describe Geo-ID (cs.CV, test-time geometric consensus for cross-view intrinsics), while the full manuscript text is the Magellanic Clouds 3D reddening paper (astro-ph.GA, arXiv:2603.13860). No Geo-ID sections, correspondence construction, uncertainty model, consensus objective, baselines, or tables exist in the body. The central claims of the abstract therefore cannot be verified, and the submission is not reviewable as Geo-ID.
  2. Because the body is the wrong paper, load-bearing elements required by the abstract—accuracy/density of sparse geometric correspondences, validity of uncertainty-aware consensus targets as surrogates for multi-view material consistency, quantitative consistency gains versus single-view baselines, and preservation of single-view decomposition quality—are entirely absent. No evaluation of those claims is possible from the provided text.

Circularity Check

0 steps flagged

No assessable circularity: supplied manuscript body is a different paper (Magellanic Clouds reddening maps), so Geo-ID derivation chain cannot be walked; abstract alone shows no definitional reduction.

full rationale

The CACHEABLE PAPER SOURCE CONTEXT contains the full text of an unrelated astronomy paper (arXiv 2603.13860 on 3D reddening maps of the LMC/SMC from OGLE-IV RR Lyrae stars, adaptive quadtree partitions, and iterative E(V-I)–distance fits), not the Geo-ID computer-vision manuscript. Consequently no Geo-ID equations, consensus objective, correspondence construction, uncertainty model, or experimental tables exist to inspect. The abstract-level description of Geo-ID (coupling independent single-view PBR predictions to uncertainty-aware consensus targets formed from sparse geometric correspondences) is an empirical test-time procedure; nothing in the available text reduces a claimed prediction to a fitted input by construction, imports a uniqueness theorem via self-citation, or renames a known result. Default non-finding therefore applies: score 0, empty steps. (The mismatched Magellanic text itself uses ordinary iterative self-consistency for map construction and likewise exhibits none of the six circularity patterns.)

Axiom & Free-Parameter Ledger

2 free parameters · 3 axioms · 1 invented entities

Abstract-only review of Geo-ID. Load-bearing premises are domain assumptions about geometry, uncertainty, and the sufficiency of consensus targets; free parameters (weights, thresholds, correspondence settings) are expected in any test-time optimization but not specified. No new physical entities are introduced.

free parameters (2)
  • consensus / alignment loss weights and uncertainty thresholds
    Any test-time coupling of per-view predictions to consensus targets requires trade-off weights and uncertainty handling; values are not given in the abstract but the central consistency claim depends on them.
  • geometric correspondence sparsity and matching criteria
    Sparse correspondences define the consensus graph; how many matches, which matcher, and outlier rejection are free design choices that control the result.
axioms (3)
  • domain assumption Sparse multi-view geometric correspondences identify the same surface points across views with sufficient accuracy for material consensus.
    Core mechanism in the abstract: coupling predictions through sparse geometric correspondences.
  • ad hoc to paper Uncertainty-aware consensus targets derived from independent single-view predictions are a valid proxy for true multi-view intrinsic consistency.
    The paper’s distinctive claim is that these targets improve consistency without inverse rendering; this is a methodological assumption, not a standard theorem.
  • domain assumption Off-the-shelf single-view intrinsic predictors produce estimates that are useful enough that geometric consensus can improve them without retraining.
    Model-agnostic reuse of pretrained predictors is stated as a design premise.
invented entities (1)
  • Geo-ID uncertainty-aware geometric consensus targets no independent evidence
    purpose: Provide shared multi-view targets that couple independent per-view intrinsic predictions at test time.
    Named methodological construct of the paper; not a physical entity. Independent evidence would be external benchmarks showing improved ground-truth PBR accuracy and downstream editing, which are claimed but not inspectable here.

pith-pipeline@v1.1.0-grok45 · 11186 in / 2692 out tokens · 25046 ms · 2026-07-14T21:35:35.747027+00:00 · methodology

0 comments
read the original abstract

Intrinsic image decomposition aims to estimate physically based rendering (PBR) parameters such as albedo, roughness, and metallicity from images. While recent methods achieve strong single-view predictions, applying them independently to multiple views of the same scene often yields inconsistent estimates, limiting their use in downstream applications such as editable neural scenes and 3D reconstruction. Video-based models can improve cross-frame consistency but require dense, ordered sequences and substantial compute, limiting their applicability to sparse, unordered image collections. We propose Geo-ID, a novel test-time framework that repurposes pretrained single-view intrinsic predictors to produce cross-view consistent decompositions by coupling independent per-view predictions through sparse geometric correspondences that form uncertainty-aware consensus targets. Geo-ID is model-agnostic, requires no retraining or inverse rendering, and applies directly to off-the-shelf intrinsic predictors. Experiments on synthetic benchmarks and real-world scenes demonstrate substantial improvements in cross-view intrinsic consistency as the number of views increases, while maintaining comparable single-view decomposition performance. We further show that the resulting consistent intrinsics enable coherent appearance editing and relighting in downstream neural scene representations.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

56 extracted references · 16 canonical work pages · 3 internal anchors

  1. [1]

    Bell, C. P. M., Cioni, M.-R. L., Wright, A. H., et al. 2020, MNRAS, 499, 993, doi: 10.1093/mnras/staa2786

    doi:10.1093/mnras/staa2786 2020

  2. [2]

    Bell, C. P. M., Cioni, M.-R. L., Wright, A. H., et al. 2022, MNRAS, 516, 824, doi: 10.1093/mnras/stac1545

    doi:10.1093/mnras/stac1545 2022

  3. [3]

    2022, Universe, 8, 122, doi: 10.3390/universe8020122

    Bhardwaj, A. 2022, Universe, 8, 122, doi: 10.3390/universe8020122

    work page doi:10.3390/universe8020122 2022

  4. [4]

    Q., Guo, H

    Chen, B. Q., Guo, H. L., Gao, J., et al. 2022, MNRAS, 511, 1317, doi: 10.1093/mnras/stac072

    doi:10.1093/mnras/stac072 2022

  5. [5]

    Q., Huang, Y., Yuan, H

    Chen, B. Q., Huang, Y., Yuan, H. B., et al. 2019, MNRAS, 483, 4277, doi: 10.1093/mnras/sty3341

    doi:10.1093/mnras/sty3341 2019

  6. [6]

    C., Lebouteiller, V., et al

    Chevance, M., Madden, S. C., Lebouteiller, V., et al. 2016, A&A, 590, A36, doi: 10.1051/0004-6361/201527735

    doi:10.1051/0004-6361/201527735 2016

  7. [7]

    L., Olsen, K., et al

    Choi, Y., Nidever, D. L., Olsen, K., et al. 2018, ApJ, 866, 90, doi: 10.3847/1538-4357/aae083

    doi:10.3847/1538-4357/aae083 2018

  8. [8]

    Clark, C. J. R., Roman-Duval, J. C., Gordon, K. D., et al. 2023, ApJ, 946, 42, doi: 10.3847/1538-4357/acbb66

    doi:10.3847/1538-4357/acbb66 2023

  9. [9]

    2019, A&A, 622, A60, doi: 10.1051/0004-6361/201833374

    Clementini, G., Ripepi, V., Molinaro, R., et al. 2019, A&A, 622, A60, doi: 10.1051/0004-6361/201833374

    doi:10.1051/0004-6361/201833374 2019

  10. [10]

    2023, A&A, 674, A18, doi: 10.1051/0004-6361/202243964

    Clementini, G., Ripepi, V., Garofalo, A., et al. 2023, A&A, 674, A18, doi: 10.1051/0004-6361/202243964

    doi:10.1051/0004-6361/202243964 2023

  11. [11]

    I., Clementini, G., et al

    Cusano, F., Moretti, M. I., Clementini, G., et al. 2021, MNRAS, 504, 1, doi: 10.1093/mnras/stab901 de Grijs, R., Wicker, J. E., & Bono, G. 2014, AJ, 147, 122, doi: 10.1088/0004-6256/147/5/122

    doi:10.1093/mnras/stab901 2021

  12. [12]

    Distance, reddening and three dimensional structure of the SMC - I: Using RRab stars

    Deb, S. 2017, arXiv e-prints, arXiv:1707.03130, doi: 10.48550/arXiv.1707.03130 D´ ek´ any, I., Grebel, E. K., & Pojma´ nski, G. 2021, ApJ, 920, 33, doi: 10.3847/1538-4357/ac106f

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.1707.03130 2017

  13. [13]

    Draine, B. T. 2003, ARA&A, 41, 241, doi: 10.1146/annurev.astro.41.011802.094840

    Pith/arXiv arXiv doi:10.1146/annurev.astro.41.011802.094840 2003

  14. [14]

    L., Madore, B

    Freedman, W. L., Madore, B. F., Hoyt, T., et al. 2020, ApJ, 891, 57, doi: 10.3847/1538-4357/ab7339

    doi:10.3847/1538-4357/ab7339 2020

  15. [15]

    2022, Habilitation Thesis, 1, doi: 10.48550/arXiv.2202.01868

    Galliano, F. 2022, Habilitation Thesis, 1, doi: 10.48550/arXiv.2202.01868

    Pith/arXiv arXiv doi:10.48550/arxiv.2202.01868 2022

  16. [16]

    Galliano, F., Galametz, M., & Jones, A. P. 2018, ARA&A, 56, 673, doi: 10.1146/annurev-astro-081817-051900

    doi:10.1146/annurev-astro-081817-051900 2018

  17. [17]

    D., Roman-Duval, J., Bot, C., et al

    Gordon, K. D., Roman-Duval, J., Bot, C., et al. 2014, ApJ, 797, 85, doi: 10.1088/0004-637X/797/2/85 G´ orski, M., Zgirski, B., Pietrzy´ nski, G., et al. 2020, ApJ, 889, 179, doi: 10.3847/1538-4357/ab65ed

    doi:10.1088/0004-637x/797/2/85 2014

  18. [18]

    2019, ApJ, 887, 93, doi: 10.3847/1538-4357/ab5362

    Finkbeiner, D. 2019, ApJ, 887, 93, doi: 10.3847/1538-4357/ab5362

    doi:10.3847/1538-4357/ab5362 2019

  19. [19]

    Groenewegen, M. A. T. 2024, in IAU Symposium, Vol. 376, IAU Symposium, ed. R. de Grijs, P. A. Whitelock, & M. Catelan, 128–149, doi: 10.1017/S1743921323002995

    work page doi:10.1017/s1743921323002995 2024

  20. [20]

    L., Chen, B

    Guo, H. L., Chen, B. Q., Yuan, H. B., et al. 2021, ApJ, 906, 47, doi: 10.3847/1538-4357/abc68a

    doi:10.3847/1538-4357/abc68a 2021

  21. [21]

    K., & Duffau, S

    Haschke, R., Grebel, E. K., & Duffau, S. 2011, AJ, 141, 158, doi: 10.1088/0004-6256/141/5/158

    doi:10.1088/0004-6256/141/5/158 2011

  22. [22]

    2025, ApJS, 278, 2, doi: 10.3847/1538-4365/adbcad

    He, S.-X., Huang, Y., Li, X.-Y., et al. 2025, ApJS, 278, 2, doi: 10.3847/1538-4365/adbcad

    work page doi:10.3847/1538-4365/adbcad 2025

  23. [23]

    Henize, K. G. 1956, ApJS, 2, 315, doi: 10.1086/190025

    doi:10.1086/190025 1956

  24. [24]

    2025, The Innovation, 6, 100907, doi: 10.1016/j.xinn.2025.100907

    Huang, Y., & Li, Z. 2025, The Innovation, 6, 100907, doi: 10.1016/j.xinn.2025.100907

    work page doi:10.1016/j.xinn.2025.100907 2025

  25. [25]

    2016, ApJ, 832, 176, doi: 10.3847/0004-637X/832/2/176

    Inno, L., Bono, G., Matsunaga, N., et al. 2016, ApJ, 832, 176, doi: 10.3847/0004-637X/832/2/176

    doi:10.3847/0004-637x/832/2/176 2016

  26. [26]

    M., Skowron, D

    Jacyszyn-Dobrzeniecka, A. M., Skowron, D. M., Mr´ oz, P., et al. 2017, AcA, 67, 1, doi: 10.32023/0001-5237/67.1.1

    work page doi:10.32023/0001-5237/67.1.1 2017

  27. [27]

    C., & Panchal, A

    Joshi, Y. C., & Panchal, A. 2019, A&A, 628, A51, doi: 10.1051/0004-6361/201834574

    doi:10.1051/0004-6361/201834574 2019

  28. [28]

    C., et al

    Kim, S., Kwon, E., Madden, S. C., et al. 2010, A&A, 518, L75, doi: 10.1051/0004-6361/201014645

    work page doi:10.1051/0004-6361/201014645 2010

  29. [29]

    2018, A&A, 616, A132, doi: 10.1051/0004-6361/201832832

    Lallement, R., Capitanio, L., Ruiz-Dern, L., et al. 2018, A&A, 616, A132, doi: 10.1051/0004-6361/201832832

    doi:10.1051/0004-6361/201832832 2018

  30. [30]

    C., & Zhang, H.-W

    Li, X.-Y., Huang, Y., Liu, G.-C., Beers, T. C., & Zhang, H.-W. 2023, ApJ, 944, 88, doi: 10.3847/1538-4357/acadd5 13

    doi:10.3847/1538-4357/acadd5 2023

  31. [31]

    2024, Research Notes of the American Astronomical Society, 8, 85, doi: 10.3847/2515-5172/ad3540

    Mateu, C. 2024, Research Notes of the American Astronomical Society, 8, 85, doi: 10.3847/2515-5172/ad3540

    work page doi:10.3847/2515-5172/ad3540 2024

  32. [32]

    2020, MNRAS, 496, 3291, doi: 10.1093/mnras/staa1676

    Mateu, C., Holl, B., De Ridder, J., & Rimoldini, L. 2020, MNRAS, 496, 3291, doi: 10.1093/mnras/staa1676

    work page doi:10.1093/mnras/staa1676 2020

  33. [33]

    2021, MNRAS, 508, 245, doi: 10.1093/mnras/stab2399

    Mazzi, A., Girardi, L., Zaggia, S., et al. 2021, MNRAS, 508, 245, doi: 10.1093/mnras/stab2399

    doi:10.1093/mnras/stab2399 2021

  34. [34]

    M., & Finkbeiner, D

    Meisner, A. M., & Finkbeiner, D. P. 2015, ApJ, 798, 88, doi: 10.1088/0004-637X/798/2/88

    work page doi:10.1088/0004-637x/798/2/88 2015

  35. [35]

    2014, in American Astronomical Society Meeting Abstracts, Vol

    Meixner, M. 2014, in American Astronomical Society Meeting Abstracts, Vol. 224, American Astronomical Society Meeting Abstracts #224, 217.01

    2014

  36. [36]

    E., Clementini, G., Sarro, L

    Muraveva, T., Delgado, H. E., Clementini, G., Sarro, L. M., & Garofalo, A. 2018a, MNRAS, 481, 1195, doi: 10.1093/mnras/sty2241

    doi:10.1093/mnras/sty2241

  37. [37]

    2018b, MNRAS, 473, 3131, doi: 10.1093/mnras/stx2514

    Muraveva, T., Subramanian, S., Clementini, G., et al. 2018b, MNRAS, 473, 3131, doi: 10.1093/mnras/stx2514

    doi:10.1093/mnras/stx2514

  38. [38]

    Riess, A. G. 2021, ApJ, 910, 121, doi: 10.3847/1538-4357/abe530

    doi:10.3847/1538-4357/abe530 2021

  39. [39]

    Pejcha, O., & Stanek, K. Z. 2009, ApJ, 704, 1730, doi: 10.1088/0004-637X/704/2/1730

    work page doi:10.1088/0004-637x/704/2/1730 2009

  40. [40]

    M., Bono, G., & Ripepi, V

    Piersimoni, A. M., Bono, G., & Ripepi, V. 2002, AJ, 124, 1528, doi: 10.1086/341821 Planck Collaboration, Abergel, A., Ade, P. A. R., et al. 2014, A&A, 571, A11, doi: 10.1051/0004-6361/201323195 Planck Collaboration, Aghanim, N., Ashdown, M., et al. 2016, A&A, 596, A109, doi: 10.1051/0004-6361/201629022

    work page doi:10.1086/341821 2002

  41. [41]

    Prudil, Z., Kunder, A., D´ ek´ any, I., & Koch-Hansen, A. J. 2024, A&A, 684, A176, doi: 10.1051/0004-6361/202347338

    work page doi:10.1051/0004-6361/202347338 2024

  42. [42]

    Reyes, R. E. C., Navarro, F. A. R., Mel´ endez, J., Steiner, J., & Elizalde, F. 2015, Revista Mexicana de Astronom´ ıa y Astrof´ ısica, 51, 253, doi: 10.48550/arXiv.1503.02066

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.1503.02066 2015

  43. [43]

    2019, ApJ, 876, 85, doi: 10.3847/1538-4357/ab1422

    Scolnic, D. 2019, ApJ, 876, 85, doi: 10.3847/1538-4357/ab1422

    doi:10.3847/1538-4357/ab1422 2019

  44. [44]

    D., Meixner, M., et al

    Roman-Duval, J., Gordon, K. D., Meixner, M., et al. 2014, ApJ, 797, 86, doi: 10.1088/0004-637X/797/2/86

    doi:10.1088/0004-637x/797/2/86 2014

  45. [45]

    2023, in Physics and Chemistry of Star Formation: The Dynamical ISM Across Time and Spatial Scales, ed

    Rubio, M. 2023, in Physics and Chemistry of Star Formation: The Dynamical ISM Across Time and Spatial Scales, ed. V. Ossenkopf-Okada, R. Schaaf, I. Breloy, & J. Stutzki, 44

    2023

  46. [46]

    2020, ARA&A, 58, 529, doi: 10.1146/annurev-astro-032620-021933

    Salim, S., & Narayanan, D. 2020, ARA&A, 58, 529, doi: 10.1146/annurev-astro-032620-021933

    doi:10.1146/annurev-astro-032620-021933 2020

  47. [47]

    J., Finkbeiner, D

    Schlegel, D. J., Finkbeiner, D. P., & Davis, M. 1998, ApJ, 500, 525, doi: 10.1086/305772

    doi:10.1086/305772 1998

  48. [48]

    The OGLE Collection of Variable Stars. Over 45 000 RR Lyrae Stars in the Magellanic System

    Skowron, D. M., Skowron, J., Udalski, A., et al. 2021, ApJS, 252, 23, doi: 10.3847/1538-4365/abcb81 Soszy´ nski, I., Udalski, A., Szyma´ nski, M. K., et al. 2016, AcA, 66, 131, doi: 10.48550/arXiv.1606.02727

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.3847/1538-4365/abcb81 2021

  49. [49]

    2005, A&A, 430, 421, doi: 10.1051/0004-6361:20041279

    Subramaniam, A. 2005, A&A, 430, 421, doi: 10.1051/0004-6361:20041279

    work page doi:10.1051/0004-6361:20041279 2005

  50. [50]

    2012, ApJ, 744, 128, doi: 10.1088/0004-637X/744/2/128

    Subramanian, S., & Subramaniam, A. 2012, ApJ, 744, 128, doi: 10.1088/0004-637X/744/2/128

    doi:10.1088/0004-637x/744/2/128 2012

  51. [51]

    L., van Loon, J

    Tatton, B. L., van Loon, J. T., Cioni, M. R., et al. 2013, A&A, 554, A33, doi: 10.1051/0004-6361/201321209

    doi:10.1051/0004-6361/201321209 2013

  52. [52]

    Trumpler, R. J. 1930, PASP, 42, 214, doi: 10.1086/124039

    doi:10.1086/124039 1930

  53. [53]

    K., & Szyma´ nski, G

    Udalski, A., Szyma´ nski, M. K., & Szyma´ nski, G. 2015, AcA, 65, 1, doi: 10.48550/arXiv.1504.05966

    Pith/arXiv arXiv doi:10.48550/arxiv.1504.05966 2015

  54. [54]

    2023, ApJ, 946, 43, doi: 10.3847/1538-4357/acb647

    Wang, S., & Chen, X. 2023, ApJ, 946, 43, doi: 10.3847/1538-4357/acb647

    doi:10.3847/1538-4357/acb647 2023

  55. [55]

    2025, ApJS, 280, 15, doi: 10.3847/1538-4365/adea39

    Wang, T., Yuan, H., Chen, B., et al. 2025, ApJS, 280, 15, doi: 10.3847/1538-4365/adea39

    doi:10.3847/1538-4365/adea39 2025

  56. [56]

    Zhang, X., & Green, G. M. 2025, Science, 387, 1209, doi: 10.1126/science.ado9787

    doi:10.1126/science.ado9787 2025