Pith. sign in

REVIEW 2 major objections 5 minor 5 cited by

The Emperor's New Arc: gigaparsec patterns abound in a $\Lambda$CDM universe

T0 review · 2 major / 5 minor · reviewed 2026-08-09 · deepseek-v4-flash

Pith's one-line read The Giant Arc is a selection artifact: the same friends-of-friends detection applied to a Lambda-CDM simulation produces many equally long, thin, point-overdense structures with matter overdensities of only a few percent.

desk verdict The Giant Arc is a FoF artifact on a sparse sample, and this parameter-matched simulation test shows it convincingly. read the letter →

arxiv 2502.03515 v1 pith:LPW52GA3 submitted 2025-02-05 astro-ph.CO

classification astro-ph.CO PACS 98.80.-k
keywords large-scalestructureoftheUniverseLambda-CDMcosmologycosmologicalsimulationFLAMINGO-10Kfriends-of-friendsclusteringMgIIabsorbersGiantArcisotropyandhomogeneity
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper aims to refute the claim that gigaparsec-scale patterns like the Giant Arc contradict the standard $\Lambda$CDM cosmological model. Reproducing the original detection recipe, the same slice geometry, sparse point density, 95 cMpc linking length, and friends-of-friends clustering, inside FLAMINGO-10K, a 2.83 cGpc $\Lambda$CDM simulation, the authors find Giant-Arc lookalikes in almost every slice. Their key diagnostic separates point overdensity, the quantity observers report, from matter overdensity: the lookalikes have point overdensities near 0.9 but median matter overdensities of only 2 percent. The reported 4.5$\sigma$ significance, the authors show, is an artifact of applying the convex-hull test after the group has already been selected by friends-of-friends. If correct, the Giant Arc is a chance pattern in a sparse tracer sample, not a physical structure, and it poses no challenge to large-scale isotropy.

What carries the argument

The central machinery is the friends-of-friends (FoF) algorithm, single-linkage clustering with a Euclidean metric and a free linking length, applied identically to mock absorber samples and to random Poisson patterns. The diagnostic that carries the argument is the contrast between the point overdensity $\delta_p$, computed from the convex hull around each FoF group, and the matter overdensity $\delta_m$ measured from simulation particles inside the same hull. Because the linking length of 95 cMpc is close to the mean interparticle separation of the sparse slice, roughly 119 cMpc, any FoF group is overdense in points almost by construction, so $\delta_p\sim1$ carries no physical information. The matter overdensity reveals whether the pattern corresponds to real mass, and it is tiny.

What would settle it

Use an independent tracer, such as deep galaxy number counts, weak lensing, or CMB lensing, to measure the matter density inside the Giant Arc's convex hull. The paper predicts a matter overdensity of only a few percent; a measured overdensity comparable to the reported point overdensity of 0.9 would falsify the selection-artifact interpretation.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that Giant Arc-like structures are routine outcomes of $\Lambda$CDM. Applying the identical friends-of-friends algorithm with linking length $l=95\,\mathrm{cMpc}$ to random subsamples of $M_{200,c}=1\text{--}5\times10^{12}\,M_\odot$ subhaloes in FLAMINGO-10K at $z=0.7$, using 1588 points in $2800\times2800\times338\,\mathrm{cMpc}^3$ slices, yields structures as long ($L>750\,\mathrm{cMpc}$), as thin ($b/a<0.5$), and as point-overdense ($\delta_p\gtrsim0.4$) as the Giant Arc in nearly every slice. Roughly 11 to 19 percent of random samples contain at least one structure longer than the Giant Arc's estimated extent. The Giant Arc's quoted overdensity of $\delta_p=0.9$ ranks only at the 87th percentile among simulation groups and the 89th among random point patterns, and the true matter overdensity of these same groups has a median of $\delta_m=0.02$. The paper therefore concludes that the apparent 4.5$\sigma$ overdensity is a consequence of the detection algorithm, not of matter being present, and that the Giant Arc should be called a pattern rather than a structure.

Load-bearing premise

The argument depends on treating FLAMINGO-10K subhaloes in the mass range thought to host Mg II absorbers as faithful stand-ins for those absorbers; if the real absorbers trace matter through different objects or a different mass range, the mock frequency of Giant-Arc lookalikes could change.

Editorial extensions

If this is right

  • The Giant Arc and similarly reported gigaparsec patterns are consistent with $\Lambda$CDM; they do not require new physics or a breakdown of isotropy.
  • A convex-hull overdensity quoted for an FoF-selected group must be calibrated against FoF-selected groups, not against random points; otherwise the significance is inflated by construction.
  • Because different random subsamples of the same simulation volume produce completely different structures, sparse tracer samples cannot be used to claim that a particular pattern is a persistent physical object.
  • Structure can be identified on any scale and at any time in a $\Lambda$CDM realization, so the idea of a fixed homogeneity scale beyond which structure is forbidden is not a valid test of the model.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The same argument should apply to the Big Ring and to other FoF-based 'anomalies' defined by post-hoc slice choices; those claims need the same mock-pipeline treatment before being read as cosmological tensions.
  • A practical rule for future discoveries: run the full detection pipeline on $\Lambda$CDM mock surveys and compare the distribution of the maximum statistic, not just the significance of one group, to account for the look-elsewhere effect.
  • A direct observational test is available: if a dense future survey covers the Giant Arc's volume and the 44-member chain dissolves into many alternative patterns, that would confirm the sampling-artifact interpretation; a coherent overdensity persisting in the dense sample would refute it.
  • The paper's message implies that significance in large-scale-structure searches should be redefined around persistence across independent tracers and random subsamples rather than around the overdensity of an already-selected group.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 5 minor

Summary. The paper tests the claim that the 'Giant Arc' of Mg II absorbers (Lopez et al. 2022) is in tension with the standard ΛCDM cosmological model. Using the FLAMINGO-10K simulation, the authors construct thin slices at z=0.7 with the same volume geometry, point density (N=1588 per 2800×2800×338 cMpc^3), and FoF linking length (95 cMpc) as the observational study. They find numerous structures with length exceeding 700–800 cMpc and axis ratio b/a<0.5, both in subhalo samples over several mass ranges and in random Poisson point sets. They show that the Giant Arc's reported point overdensity δ_p=0.9 is unremarkable (87th/89th percentile in simulated/random samples) and that the same FoF-selected structures have matter overdensities with median δ_m≈0.02. The paper concludes that such patterns are common and expected in ΛCDM and that the apparent overdensity is a selection effect of the FoF algorithm rather than evidence of a physical structure.

Significance. If correct, this paper removes a widely publicized claimed tension between large-scale structure observations and ΛCDM without invoking new physics. The analysis is carefully matched to the observational setup: fixed slice geometry, point density, linking length, and no free parameters tuned to produce the Giant Arc. The inclusion of a Poisson null and the separation of point overdensity from matter overdensity are particularly strong and provide a clean diagnostic for future searches. The paper also makes its analysis reproducible by releasing the script used to produce the figures and numbers. The conclusion carries a falsifiable implication: sparse-tracer FoF searches should routinely yield gigaparsec-scale patterns, so such patterns should not be interpreted as physical structures unless they persist across independent tracer samples and are accompanied by an actual matter overdensity.

major comments (2)
  1. [Section 5.1] The claim that structures do not persist across random subsamples is asserted but not supported by any figure, table, or statistic. The text states that 'different random subsamples of the underlying subhalo population lead to completely different structures' and uses this to conclude that Giant-Arc analogues are 'largely spurious, rather than real underlying structures.' This is a load-bearing step for the 'spurious pattern' conclusion. Please provide the persistence test (e.g., overlap fractions, matched-structure statistics, or a figure) or soften the claim to reflect the evidence actually presented.
  2. [Section 4] The simulation analysis uses real-space subhalo positions, whereas the observed Mg II absorbers are measured in redshift space; the paper does not discuss redshift-space distortions. Because the slice is only 338 cMpc deep and the linking length is 95 cMpc, peculiar-velocity displacements along the line of sight could affect FoF membership and the measured length distribution. This is directly relevant to the quantitative frequency of Giant-Arc analogues (e.g., the 19%/11% fractions quoted in Section 5.1). Please quantify the effect of redshift-space distortions (for example, by applying a simple RSD model to the simulation) or explain in detail why they are negligible.
minor comments (5)
  1. [Sections 2 and 5.1] The paper does not directly address the relation between its random Poisson null (which yields ~11% of slices with a structure as long as the Giant Arc) and Lopez et al.'s random-redshift test (which found no comparable pattern in 1000 randomizations). Please clarify how the two nulls differ and how the present results refute the earlier significance claim.
  2. [Section 4] The sentence 'both the mean interparticle distance and the linking length are similar to the shortest dimension of the slice' is inaccurate: 119 cMpc and 95 cMpc are roughly one-third of 338 cMpc, not similar to it. Please rephrase.
  3. [Figure 1] The Giant Arc panel does not quote a b/a value, while the simulated comparison structures are selected with b/a<0.5. Please state the estimated b/a for the Giant Arc (or note that it is not known) so the comparison population is explicitly justified.
  4. [Section 5.1 / Figure 2] Please specify how many random subsamples per slice were used and confirm whether the quoted 19% and 11% fractions are per slice or per subsample. This would make the statistical interpretation of the CDFs unambiguous.
  5. [Data Availability] The statement that 'access to the underlying simulation data will be provided on reasonable request' may be too weak for full reproducibility; consider depositing the derived subhalo slice samples or the reduced data needed to reproduce Figures 2–5.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the comparison uses an independent ΛCDM simulation and random point baselines, with no parameter fitted to the Giant Arc data.

full rationale

The paper's derivation chain is self-contained and externally benchmarked. It does not fit any parameter to the Giant Arc; instead, it adopts the observational linking length l=95 cMpc and slice geometry from Lopez et al. (2022) and applies the same FoF algorithm to density-matched random subsamples of subhaloes in FLAMINGO-10K. The key comparison is against random Poisson point patterns of equal density, which provides an external, theory-independent baseline. The claim that the reported point overdensity is an algorithmic artefact is supported by a mathematical property of FoF groups with linking length below the mean interparticle distance, and by the direct computation of matter overdensities within the same convex hulls, which are only a few percent. The only self-referential elements are citations to the authors' own FLAMINGO simulation papers, but these are background references to the simulation code and suite, not load-bearing assumptions whose truth is equivalent to the target conclusion. The assumed mapping from MgII absorbers to subhalo mass is acknowledged as uncertain and is tested across mass ranges (Figure 3), so it does not constitute a fitted input renamed as a prediction. No circular step of any enumerated kind is present.

Assumptions & free parameters 4 free parameters · 4 assumptions · 0 invented entities

No parameters are fitted to the Giant Arc data; the analysis compares fixed simulation outputs and random nulls with parameters inherited from the observational paper. The only significant modeling choices are the subhalo mass selection and analysis thresholds, both explicitly tested for sensitivity. No new physical entities are introduced.

free parameters (4)
  • FoF linking length = 95 cMpc
    Adopted from Lopez et al. (2022), not fitted here. The paper shows this value is close to the maximum of group counts in both simulation and random patterns (Figure 3), which is central to the artifact interpretation.
  • Subhalo mass range for MgII analogues = M200,c = 1-5x10^12 solar masses
    Chosen from clustering-based estimates (Section 3). Exact absorber halo masses are unknown; the authors test other ranges and find results unchanged except for the most massive haloes.
  • Minimum structure extent threshold = L greater than 700 cMpc (or 8 times l)
    Used to define giant analogues in Figures 2-4; arbitrary but consistent with the Giant Arc's reported extent.
  • Anisotropy threshold = b/a less than 0.5
    Used to select arc-like structures; arbitrary but representative of the highly elongated Giant Arc.
assumptions (4)
  • domain assumption FLAMINGO-10K provides a faithful Lambda CDM realization of gigaparsec-scale structure at z about 0.7.
    The entire comparison assumes the simulation's gravitational evolution and subhalo population are representative of the real universe on these scales (Section 3).
  • domain assumption Subhaloes with M200,c = 1-5x10^12 solar masses trace MgII absorbers in the same way as the observed absorber population.
    The simulated analogues replace the observed MgII absorbers; because the exact halo masses of absorbers are unknown (Section 3), this mapping is load-bearing. The authors partially validate it by testing other mass ranges.
  • domain assumption The FoF implementation in the analysis is equivalent to the algorithm used by Lopez et al. (2022).
    The paper claims to apply the same methods with the same parameters (Sections 2 and 4); differences in implementation could affect the number of groups found. The random point control reduces this risk.
  • standard math Random Poisson point patterns are a valid null model for isolated FoF groups in sparse samples.
    The Poisson null isolates the algorithmic contribution to group formation (Section 4), following Nadathur (2013).

how reviews work

0 comments
Cite this review

Pith. "Pith review of The Emperor's New Arc: gigaparsec patterns abound in a $\Lambda$CDM universe." pith.science (2026). https://pith.science/paper/LPW52GA3

@misc{pith2026250203515,
  author       = {Pith},
  title        = {Pith review of: The Emperor's New Arc: gigaparsec patterns abound in a $\Lambda$CDM universe},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LPW52GA3}},
  note         = {Machine review of arXiv:2502.03515}
}
abstract

Recent discoveries of apparent large-scale features in the structure of the universe, extending over many hundreds of megaparsecs, have been claimed to contradict the large-scale isotropy and homogeneity foundational to the standard ($\Lambda$CDM) cosmological model. We explicitly test and refute this conjecture using FLAMINGO-10K, a new and very large cosmological simulation of the growth of structure in a $\Lambda$CDM context. Applying the same methods used in the observations, we show that patterns like the "Giant Arc", supposedly in tension with the standard model, are, in fact, common and expected in a $\Lambda$CDM universe. We also show that their reported significant overdensities are an algorithmic artefact and unlikely to reflect any underlying structure.

Figures

Figures reproduced from arXiv: 2502.03515 by the authors.

Figure 1
Figure 1. A selection of prominent structures identified in slices of the FLAMINGO-10K simulation in a single sampling at 𝑧 = 0.7, using the same depth, points density, and linking length as in Lopez et al. (2022), alongside the“Giant Arc” reproduced from Lopez et al. (2022). Each panel shows a region of 1600 × 1600 𝑐Mpc2 with a depth of 338 𝑐Mpc. The background of each simulation panel shows the projected matter surface over… view at source ↗
Figure 2
Figure 2. Probability density (PDF, top) and cumulative distribution (CDF, bottom) of the lengths of the longest structures with 𝑏/𝑎 < 0.5 in random subsamples within 21 non-overlapping slices. The top shows log-normal fits to the individual PDFs for the 21 individual slices. The bottom panel shows the raw CDFs for each slice, and a log-normal fit to the average PDF. The grey band indicates the lower and upper limit of the ex… view at source ↗
Figure 3
Figure 3. Average number of FoF groups per slice more extended than 𝐿 = 700 𝑐Mpc and more anisotropic than 𝑏/𝑎 = 0.5, as a function of linking length, l. Different colours indicate samples of different subhalo mass ranges in the FLAMINGO-10K simulation at 𝑧 = 0.7, as indicated in the legend; the dotted grey line shows the result for random point samples. The number of groups found is a strong function of the linking length an… view at source ↗
Figures from the paper (2 more)
Figure 5
Figure 5. Figure 5: Number of groups more extended than 8× the linking length and more anisotropic than 𝑏/𝑎 = 0.5 in slices of the FLAMINGO-10K simulation at different redshifts, normalised by the corresponding number in random point patterns. Thin lines show raw measurements, thick lines…
Figure 4
Figure 4. Figure 4: Distribution of overdensities associated with giant structures (𝐿 > 700 𝑐Mpc). The top panel shows the point overdensities of groups in samples of FLAMINGO-10K at 𝑧 = 0.7 and in random point patterns. Structures in FLAMINGO-10K are slightly more overdense than those in…

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 5 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Evolution of Cluster Alignments as Evidence of Large-scale Structure Formation in the Universe

    astro-ph.CO 2025-06 conditional novelty 6.0 of 10

    Galaxy cluster orientations are correlated over 200-300 comoving megaparsecs, out to redshift 1, based on 1.58 million clusters, consistent with ΛCDM simulations.

  2. Gigaparsec structures are nowhere to be seen in $\Lambda$CDM: an enhanced analysis of LSS in FLAMINGO-10K simulations

    astro-ph.CO 2025-04 conditional novelty 6.0 of 10

    Using a 65 Mpc linkage scale on the full FLAMINGO-10K box, the authors find no gigaparsec structures and claim the simulation's large-scale structure matches random Poisson data, leaving the Giant Arc as a challenge t...

  3. Optical depth to reionization in a Universe with multiple inhomogeneous domains

    astro-ph.CO 2026-04 unverdicted novelty 5.5 of 10

    Backreaction from multi-domain inhomogeneities, constrained by PantheonPlus+SH0ES, produces τ_reion = 0.0581^{+0.0105}_{-0.0096} and a modest reduction of the Hubble tension.

  4. Constraining the Hubble parameter with the 21 cm brightness temperature signal in a universe with inhomogeneities

    astro-ph.CO 2025-05 reject novelty 5.0 of 10

    In a Buchert-averaged two-domain model, the 21 cm absorption trough is roughly 40 mK deeper for H0=67 than for H0=73, a difference absent in the paper's ΛCDM comparison.

  5. The fallacies of LCDM falsifications

    physics.hist-ph 2025-04 conditional novelty 2.0 of 10

    The paper contends that claimed failures of Lambda-CDM, including the Hubble and S8 tensions, do not meet the strict Popperian falsification standard.

Reference graph

Works this paper leans on

34 extracted references · 5 canonical work pages · cited by 5 Pith papers

  1. [1]

    G., Bagoly Z., Hakkila J

    Bal \'a zs L. G., Bagoly Z., Hakkila J. E., Horv \'a th I., K \'o bori J., R \'a cz I. I., T \'o th L. V., 2015, @doi [ ] 10.1093/mnras/stv1421 , https://ui.adsabs.harvard.edu/abs/2015MNRAS.452.2236B 452, 2236

  2. [2]

    G., Rejt o L., Tusn \'a dy G., 2018, @doi [ ] 10.1093/mnras/stx2550 , https://ui.adsabs.harvard.edu/abs/2018MNRAS.473.3169B 473, 3169

    Bal \'a zs L. G., Rejt o L., Tusn \'a dy G., 2018, @doi [ ] 10.1093/mnras/stx2550 , https://ui.adsabs.harvard.edu/abs/2018MNRAS.473.3169B 473, 3169

  3. [4]

    G., Harris K

    Clowes R. G., Harris K. A., Raghunathan S., Campusano L. E., S \"o chting I. K., Graham M. J., 2013, @doi [ ] 10.1093/mnras/sts497 , https://ui.adsabs.harvard.edu/abs/2013MNRAS.429.2910C 429, 2910

  4. [5]

    R., von Hausegger S., Lukas A., 2023, @doi [Classical and Quantum Gravity] 10.1088/1361-6382/ad0b36 , https://ui.adsabs.harvard.edu/abs/2023CQGra..40x5015C 40, 245015

    Constantin A., Harvey T. R., von Hausegger S., Lukas A., 2023, @doi [Classical and Quantum Gravity] 10.1088/1361-6382/ad0b36 , https://ui.adsabs.harvard.edu/abs/2023CQGra..40x5015C 40, 245015

  5. [6]

    S., White S

    Davis M., Efstathiou G., Frenk C. S., White S. D. M., 1985, @doi [ ] 10.1086/163168 , https://ui.adsabs.harvard.edu/abs/1985ApJ...292..371D 292, 371

  6. [7]

    Fujii H., 2024, @doi [ ] 10.1093/mnras/stad3356 , https://ui.adsabs.harvard.edu/abs/2024MNRAS.527.1982F 527, 1982

  7. [8]

    L., Simcoe R

    Gauthier J.-R., Chen H.-W., Cooksey K. L., Simcoe R. A., Seyffert E. N., O'Meara J. M., 2014, @doi [ ] 10.1093/mnras/stt2443 , https://ui.adsabs.harvard.edu/abs/2014MNRAS.439..342G 439, 342

  8. [9]

    Richard I., et al., 1989, @doi [ ] 10.1086/167425 , https://ui.adsabs.harvard.edu/abs/1989ApJ...340..625G 340, 625

    Gott J. Richard I., et al., 1989, @doi [ ] 10.1086/167425 , https://ui.adsabs.harvard.edu/abs/1989ApJ...340..625G 340, 625

Show all 34 references
  1. [10]

    S., Benitez-Llambay A., Helly J., 2018, @doi [ ] 10.1093/mnras/stx2792 , https://ui.adsabs.harvard.edu/abs/2018MNRAS.474..604H 474, 604

    Han J., Cole S., Frenk C. S., Benitez-Llambay A., Helly J., 2018, @doi [ ] 10.1093/mnras/stx2792 , https://ui.adsabs.harvard.edu/abs/2018MNRAS.474..604H 474, 604

  2. [11]

    R., et al., 2020, @doi [Nature] 10.1038/s41586-020-2649-2 , 585, 357

    Harris C. R., et al., 2020, @doi [Nature] 10.1038/s41586-020-2649-2 , 585, 357

  3. [12]

    P., Geller M

    Huchra J. P., Geller M. J., 1982, @doi [ ] 10.1086/160000 , https://ui.adsabs.harvard.edu/abs/1982ApJ...257..423H 257, 423

  4. [13]

    D., 2007, @doi [Computing in Science & Engineering] 10.1109/MCSE.2007.55 , 9, 90

    Hunter J. D., 2007, @doi [Computing in Science & Engineering] 10.1109/MCSE.2007.55 , 9, 90

  5. [14]

    Kumar Aluri P., et al., 2023, @doi [Classical and Quantum Gravity] 10.1088/1361-6382/acbefc , https://ui.adsabs.harvard.edu/abs/2023CQGra..40i4001K 40, 094001

  6. [15]

    M., Haridasu B

    Lapi A., Boco L., Cueli M. M., Haridasu B. S., Ronconi T., Baccigalupi C., Danese L., 2023, @doi [ ] 10.3847/1538-4357/ad01bb , https://ui.adsabs.harvard.edu/abs/2023ApJ...959...83L 959, 83

  7. [16]

    M., Clowes R

    Lopez A. M., Clowes R. G., Williger G. M., 2022, @doi [ ] 10.1093/mnras/stac2204 , https://ui.adsabs.harvard.edu/abs/2022MNRAS.516.1557L 516, 1557

  8. [17]

    M., Clowes R

    Lopez A. M., Clowes R. G., Williger G. M., 2024a, @doi [arXiv e-prints] 10.48550/arXiv.2409.14894 , https://ui.adsabs.harvard.edu/abs/2024arXiv240914894L p. arXiv:2409.14894

  9. [18]

    M., Clowes R

    Lopez A. M., Clowes R. G., Williger G. M., 2024b, @doi [ ] 10.1088/1475-7516/2024/07/055 , https://ui.adsabs.harvard.edu/abs/2024JCAP...07..055L 2024, 055

  10. [19]

    F., Wake D

    Lundgren B. F., Wake D. A., Padmanabhan N., Coil A., York D. G., 2011, @doi [ ] 10.1111/j.1365-2966.2011.19208.x , https://ui.adsabs.harvard.edu/abs/2011MNRAS.417..304L 417, 304

  11. [20]

    E., Clowes R

    Marinello G. E., Clowes R. G., Campusano L. E., Williger G. M., S \"o chting I. K., Graham M. J., 2016, @doi [ ] 10.1093/mnras/stw1513 , https://ui.adsabs.harvard.edu/abs/2016MNRAS.461.2267M 461, 2267

  12. [21]

    Mazurenko S., Banik I., Kroupa P., 2025, @doi [ ] 10.1093/mnras/stae2758 , https://ui.adsabs.harvard.edu/abs/2025MNRAS.536.3232M 536, 3232

  13. [22]

    Nadathur S., 2013, @doi [ ] 10.1093/mnras/stt1028 , https://ui.adsabs.harvard.edu/abs/2013MNRAS.434..398N 434, 398

  14. [23]

    Richard I., Kim S

    Park C., Choi Y.-Y., Kim J., Gott J. Richard I., Kim S. S., Kim K.-S., 2012, @doi [ ] 10.1088/2041-8205/759/1/L7 , https://ui.adsabs.harvard.edu/abs/2012ApJ...759L...7P 759, L7

  15. [24]

    Park C., Song H., Einasto M., Lietzen H., Heinamaki P., 2015, @doi [Journal of Korean Astronomical Society] 10.5303/JKAS.2015.48.1.75 , https://ui.adsabs.harvard.edu/abs/2015JKAS...48...75P 48, 75

  16. [25]

    Pedregosa F., et al., 2011, Journal of Machine Learning Research, 12, 2825

  17. [26]

    Pizzati E., et al., 2024, @doi [ ] 10.1093/mnras/stae2307 , https://ui.adsabs.harvard.edu/abs/2024MNRAS.534.3155P 534, 3155

  18. [27]

    H., Lavaux G., 2024, @doi [Nature Astronomy] 10.1038/s41550-023-02130-6 , https://ui.adsabs.harvard.edu/abs/2024NatAs...8..247S 8, 247

    Sawala T., Frenk C., Jasche J., Johansson P. H., Lavaux G., 2024, @doi [Nature Astronomy] 10.1038/s41550-023-02130-6 , https://ui.adsabs.harvard.edu/abs/2024NatAs...8..247S 8, 247

  19. [28]

    Schaller M., 2024, @doi [ ] 10.1093/mnrasl/slad199 , https://ui.adsabs.harvard.edu/abs/2024MNRAS.529L..23S 529, L23

  20. [29]

    Schaller M., et al., 2024, @doi [ ] 10.1093/mnras/stae922 , https://ui.adsabs.harvard.edu/abs/2024MNRAS.530.2378S 530, 2378

  21. [30]

    Schaye J., et al., 2023, @doi [ ] 10.1093/mnras/stad2419 , https://ui.adsabs.harvard.edu/abs/2023MNRAS.526.4978S 526, 4978

  22. [31]

    Springel V., et al., 2005, @doi [ ] 10.1038/nature03597 , http://adsabs.harvard.edu/abs/2005Natur.435..629S 435, 629

  23. [32]

    Virtanen P., et al., 2020, @doi [Nature Methods] 10.1038/s41592-019-0686-2 , https://rdcu.be/b08Wh 17, 261

  24. [33]

    White S. D. M., Frenk C. S., Davis M., Efstathiou G., 1987, @doi [ ] 10.1086/164990 , https://ui.adsabs.harvard.edu/abs/1987ApJ...313..505W 313, 505

  25. [34]

    K., Bagla J

    Yadav J. K., Bagla J. S., Khandai N., 2010, @doi [ ] 10.1111/j.1365-2966.2010.16612.x , https://ui.adsabs.harvard.edu/abs/2010MNRAS.405.2009Y 405, 2009

  26. [35]

    Zhu G., M \'e nard B., 2013, @doi [ ] 10.1088/0004-637X/770/2/130 , https://ui.adsabs.harvard.edu/abs/2013ApJ...770..130Z 770, 130

Pith tools

Reviewed August 9, 2026 · model on record in the stance chip above.