Pith. sign in

REVIEW 1 major objections 6 minor 53 references

One formula interface and sparse-precision sampler unifies spatial, temporal, and space-time mixed models for ecological small-area estimation.

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

2026-07-12 06:43 UTC pith:4L3ZY4KG

load-bearing objection Solid package paper: one formula grammar and collapsed sparse-precision engine for SAE-oriented Bayesian LMMs; engineering contribution, not a new model class. the 1 major comments →

arxiv 2607.02836 v1 pith:4L3ZY4KG submitted 2026-07-03 stat.ME stat.CO

stLMM: Bayesian Spatial and Space-Time Linear Mixed Models for Small-Area Ecological Estimation

Andrew O. Finley This is my paper
classification stat.ME stat.CO MSC 62M3062F1562P12
keywords Bayesian modelingsmall-area estimationGaussian processNNGPCARDAGARspace-time modelforest inventory
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.

Ecological monitoring designs often leave user-defined domains with too few samples for reliable direct estimates. This paper presents stLMM, an R package that lets analysts specify Bayesian linear mixed models with a single formula grammar covering iid group effects, AR time effects, point-referenced GP and NNGP terms, areal CAR and DAGAR terms, separable space-time effects, and structured varying coefficients. Fixed effects, residual-variance models for direct estimates, and missing response rows kept as prediction targets sit in the same interface, so users can move between area-level and unit-level small-area workflows without changing software. Under the hood a shared sparse-precision matrix collapses structured latents during MCMC and recovers them afterward, giving one posterior-draw pipeline from fitting through prediction and aggregation. A Washington county biomass example with FIA data and canopy cover shows the workflow producing smoothed county-year means with full uncertainty.

Core claim

stLMM supplies a common formula interface and shared sparse-precision MCMC engine that implements iid, AR, GP/NNGP, CAR/DAGAR, separable areal space-time, and structured varying-coefficient terms; collapses the structured latents during fitting; recovers them for fitted values, diagnostics, and prediction; and thereby supports one posterior-draw workflow for both direct-estimate and unit-level ecological small-area estimation, including missing-response prediction targets.

What carries the argument

The sparse matrix M = Q_w + A^T W A: prior process precision plus likelihood information mapped through the observation-to-process design. Structured latents are integrated out of the MCMC state via sparse Cholesky factorizations of M and recovered post-hoc, so one computational engine serves all supported process terms.

Load-bearing premise

That collapsing the structured latents out of the sampler and recovering them afterward still gives correct joint posterior inference for every supported process type and for the binomial and negative-binomial cases.

What would settle it

Run the package's own parameter-recovery simulations or the documented lme4/NIMBLE comparisons on a known CAR/DAGAR or NNGP model; systematic bias in recovered process parameters or credible-interval coverage would falsify the claim that the collapsed sparse-precision engine is correctly implementing the joint posterior.

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

Share X Bluesky LinkedIn Reddit HN

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

1 major / 6 minor

Summary. The manuscript presents stLMM, an R package for Bayesian linear mixed models with a shared formula interface covering iid grouped effects, AR temporal effects, GP/NNGP point-referenced effects, CAR/DAGAR areal effects, separable areal space-time effects, and structured varying coefficients. The central engineering claim is a common sparse-precision computational engine (Eqs. 1–6) that collapses structured latent processes during MCMC, recovers them for fitted values and prediction, and supports one posterior-draw workflow for ecological small-area estimation, including direct-estimate residual-variance models, unit-level models, and missing-response prediction targets. A Washington county-year biomass example using FIA data and tree canopy cover illustrates the interface for a scaled-variance CAR-time model with a spatially varying coefficient.

Significance. If the package behaves as described, it fills a practical niche: analysts can compare point-referenced, areal, temporal, and space-time SAE models under one formula grammar and one posterior aggregation workflow without rewriting model-specific code. The design strengths are concrete and creditworthy—public source under GPL-3, a documented recover/predict workflow, CHOLMOD-backed sparse precision assembly, support for Pólya-Gamma binomial/negative-binomial cases, and a reproducible FIA application series with diagnostics and related model variants. The contribution is primarily software and interface unification rather than new theory; that is appropriate for a package paper and is useful for ecological monitoring and national forest inventory SAE.

major comments (1)
  1. Section 6 states that simulation-based parameter-recovery checks and validation comparisons with lme4 and NIMBLE for iid and CAR/DAGAR models live on the package website rather than in the manuscript. For the central claim that the collapsed sparse-precision sampler (Section 2, M = Q_w + A^T W A, structured w integrated out then recovered) yields correct joint posterior inference across the advertised term set—including NNGP, CAR-time/DAGAR-time, structured varying coefficients, and Pólya-Gamma cases—the manuscript should include at least a compact table or short appendix summarizing recovery metrics and agreement with the reference implementations. Without that, readers cannot assess the weakest assumption of the paper from the text alone.
minor comments (6)
  1. Section 2, Eq. (1): A is described as mapping observations to latent process nodes and applying covariate scaling for varying-coefficient terms, but the construction of A for interactions such as x:car() is not shown algebraically. A short explicit block for a varying-coefficient term would clarify the design.
  2. Section 5, Eq. (8): the shrinkage weight ω_i = n_i/(n_i + c) with c = 10 is clear in the example, but the manuscript should note whether c is treated as fixed or estimable and how sensitive posterior county-year means are to that choice.
  3. Figure 1 caption and map panels: grey counties are explained as lacking model-ready direct estimates; stating whether those cells still receive full posterior predictive draws from the CAR-time support would help readers interpret the smoothed maps.
  4. Section 4 software context is thorough; a one-sentence contrast of prediction-on-missing-response-rows versus typical newdata-only workflows in spNNGP/sdmTMB would sharpen the niche claim.
  5. Section 7: CRAN status is noted as pending; once accepted, update the permanent archive citation and version pin so the paper remains a stable software reference.
  6. Minor typography: “fixed-sizenegativebinomial” and similar concatenated words in Section 2 should be spaced; “stLMM()” call formatting in Section 5 is otherwise clear.

Circularity Check

0 steps flagged

No significant circularity: software-interface paper implements known model classes under a shared sparse-precision engine without claiming first-principles derivations or forced predictions.

full rationale

stLMM is a package paper whose central claim is engineering and interface unification (common formula grammar; sparse M = Q_w + A^T W A; collapsed MCMC with post-hoc recover/predict), not a derivation of new scientific constants or uniqueness theorems. Section 2 states the standard latent mixed model (Eqs. 1–6) and the computational organization around sparse precision; these are definitions of the implementation, not predictions that reduce to fitted inputs. The Washington biomass example (Section 5, Eqs. 7–8) is a demonstration of the formula syntax on FIA data, not a claim that model-smoothed means are independent discoveries forced by construction. Self-citations (NNGP, spBayes, DAGAR) point to prior published theory and independent implementations; they are not load-bearing uniqueness imports that forbid alternatives or smuggle an ansatz that then becomes the paper’s result. Validation against lme4/NIMBLE and simulation recovery checks are described as external checks, not as re-presentations of fitted parameters. No step equates a claimed prediction to its own fitting target by construction. Score 0 is therefore the correct, proportionate finding.

Axiom & Free-Parameter Ledger

2 free parameters · 5 axioms · 0 invented entities

As a software/methods paper the central claim rests on standard Bayesian LMM theory, published sparse-matrix libraries, and the correctness of the author’s prior process constructions (NNGP, DAGAR). No new physical entities or data-fitted constants are required for the package claim; free parameters appear only as user-chosen priors or example settings.

free parameters (2)
  • residual shrinkage constant c (example) = 10
    In the Washington biomass illustration, shrinkage=10 sets c=10 in the scaled residual-variance weight ω_i = n_i/(n_i+c); this is an analyst choice, not a package-level free parameter, but it does affect the reported maps.
  • user-specified process hyperparameters and priors
    Covariance/precision parameters (range, variance, CAR/DAGAR dependence, AR coefficients, etc.) and their priors are chosen by the user at fit time; the package does not claim universal default values that produce the central result.
axioms (5)
  • domain assumption Gaussian linear mixed model observation equation y = Xβ + Zα + A w + ε with diagonal observation precision W (Eq. 1–3).
    Core modeling assumption stated in Section 2; standard for Bayesian LMMs but restricts the package’s native likelihoods until Pólya-Gamma augmentation is applied.
  • domain assumption Structured latent processes admit sparse precision representations (neighbor sets, adjacency, temporal order, DAG orderings) that can be assembled into block-diagonal Q_w and the sparse matrix M = Q_w + A^T W A.
    Section 2; enables the shared computational engine. Relies on published constructions for NNGP, CAR, DAGAR, AR.
  • standard math Pólya-Gamma data augmentation yields a conditionally Gaussian working likelihood for binomial and fixed-size negative binomial responses.
    Cited to Polson et al. (2013) and related work; used to reuse the same sparse engine for non-Gaussian responses.
  • domain assumption CHOLMOD sparse Cholesky factorizations and BLAS/LAPACK dense kernels correctly implement the required solves and products.
    Section 2 software stack; correctness of numerical linear algebra is assumed from the cited libraries.
  • ad hoc to paper Collapsing structured latents during MCMC and recovering them afterward yields draws from the correct joint posterior for the supported models.
    Design choice of the package (Section 2); standard in some collapsed Gibbs settings but must hold for every process term and for prediction on missing rows.

reviewed 2026-07-12 · how reviews work

0 comments
Cite this review

Pith. "Pith review of stLMM: Bayesian Spatial and Space-Time Linear Mixed Models for Small-Area Ecological Estimation." pith.science (2026). https://pith.science/paper/4L3ZY4KG

@misc{pith2026260702836,
  author       = {Pith},
  title        = {Pith review of: stLMM: Bayesian Spatial and Space-Time Linear Mixed Models for Small-Area Ecological Estimation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/4L3ZY4KG}},
  note         = {Machine review of arXiv:2607.02836}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

stLMM is an R package for Bayesian linear mixed models with spatial, temporal, and space-time latent effects. It provides a common formula interface for independent and identically distributed (iid) grouped effects, autoregressive (AR) temporal effects, Gaussian process (GP) and nearest-neighbor Gaussian process (NNGP) point-referenced effects, conditional autoregressive (CAR) and directed acyclic graph autoregressive (DAGAR) areal effects, separable areal space-time effects, and structured varying coefficients. The package is designed for ecological small-area estimation workflows in which analysts must move between direct-estimate and unit-level models, combine sampling variances or residual-variance models with spatial and temporal borrowing, and retain missing response rows as prediction targets. A shared sparse-precision implementation underlies the model terms. Structured latent effects are collapsed during fitting, then recovered or retained for fitted values, diagnostics, prediction, and posterior summaries. This gives users one posterior-draw workflow for propagating uncertainty from model fitting through prediction and aggregation. The package is demonstrated with a Washington county biomass example from the package article series, using Forest Inventory and Analysis (FIA) data and tree canopy cover to estimate county-year biomass means. The full article series provides reproducible source code, data, diagnostics, and related model variants.

Figures

Figures reproduced from arXiv: 2607.02836 by Andrew O. Finley.

Figure 1
Figure 1. Figure 1: Washington county-year direct estimates and posterior means from the scaled-variance [PITH_FULL_IMAGE:figures/full_fig_p008_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Posterior mean tree canopy cover (TCC) coefficient, in Mg/ha per one standard deviation [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Posterior county-year biomass means and 95% credible intervals for four counties, with [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗

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

53 extracted references · 5 canonical work pages

  1. [1]

    and Gelfand, Alan E

    Datta, Abhirup and Banerjee, Sudipto and Finley, Andrew O. and Gelfand, Alan E. , title =. Journal of the American Statistical Association , year =

  2. [2]

    and Datta, Abhirup and Cook, Bruce D

    Finley, Andrew O. and Datta, Abhirup and Cook, Bruce D. and Morton, Douglas C. and Andersen, Hans-Erik and Banerjee, Sudipto , title =. Journal of Computational and Graphical Statistics , year =

  3. [3]

    Journal of Agricultural, Biological and Environmental Statistics , year =

    Katzfuss, Matthias and Guinness, Joseph and Gong, Wenlong and Zilber, Daniel , title =. Journal of Agricultural, Biological and Environmental Statistics , year =

  4. [4]

    and Gao, Leiwen , title =

    Datta, Abhirup and Banerjee, Sudipto and Hodges, James S. and Gao, Leiwen , title =. Bayesian Analysis , year =

  5. [5]

    and Scott, James G

    Polson, Nicholas G. and Scott, James G. and Windle, Jesse , title =. Journal of the American Statistical Association , year =

  6. [6]

    and Scott, James G

    Windle, Jesse and Carvalho, Carlos M. and Scott, James G. and Sun, Liang , title =. arXiv preprint arXiv:1308.0774 , year =

    Pith/arXiv arXiv

  7. [7]

    and Scott, James G

    Windle, Jesse and Polson, Nicholas G. and Scott, James G. , title =. arXiv preprint arXiv:1405.0506 , year =

    Pith/arXiv arXiv

  8. [8]

    and Smith, Aaron and Pillai, Natesh and Dunson, David B

    Johndrow, James E. and Smith, Aaron and Pillai, Natesh and Dunson, David B. , title =. Journal of the American Statistical Association , year =

  9. [9]

    Ultimate

    Zens, Gregor and Fr. Ultimate. Journal of the American Statistical Association , year =

  10. [10]

    Proceedings of the 29th International Conference on Machine Learning , year =

    Zhou, Mingyuan and Li, Lingbo and Dunson, David and Carin, Lawrence , title =. Proceedings of the 29th International Conference on Machine Learning , year =

  11. [11]

    IEEE Transactions on Pattern Analysis and Machine Intelligence , year =

    Zhou, Mingyuan and Carin, Lawrence , title =. IEEE Transactions on Pattern Analysis and Machine Intelligence , year =

  12. [12]

    and Andersen, Hans-Erik and Babcock, Chad and Cook, Bruce D

    Finley, Andrew O. and Andersen, Hans-Erik and Babcock, Chad and Cook, Bruce D. and Morton, Douglas C. and Banerjee, Sudipto , title =. Journal of Agricultural, Biological and Environmental Statistics , year =

  13. [13]

    and Herriot, Roger A

    Fay, Robert E. and Herriot, Roger A. , title =. Journal of the American Statistical Association , year =

  14. [14]

    and Lei, Xingye and Breslow, Norman , title =

    Leroux, Brian G. and Lei, Xingye and Breslow, Norman , title =. Statistical Models in Epidemiology, the Environment, and Clinical Trials , editor =

  15. [15]

    and Finley, Andrew O

    Shannon, Elliot S. and Finley, Andrew O. and May, Paul B. and Domke, Grant M. and Andersen, Hans-Erik and. Leveraging National Forest Inventory Data to Estimate Forest Carbon Density Status and Trends for Small Areas , journal =. 2025 , volume =

    2025

  16. [16]

    Journal of the American Statistical Association , year =

    Gneiting, Tilmann , title =. Journal of the American Statistical Association , year =

  17. [17]

    A Tutorial on Adaptive

    Andrieu, Christophe and Thoms, Johannes , journal =. A Tutorial on Adaptive. 2008 , publisher =

    2008

  18. [18]

    and Rosenthal, Jeffrey S

    Roberts, Gareth O. and Rosenthal, Jeffrey S. , title =. Journal of Applied Probability , volume =. 2007 , doi =

    2007

  19. [19]

    Roberts and Jeffrey S

    Gareth O. Roberts and Jeffrey S. Rosenthal , title =. Journal of Computational and Graphical Statistics , volume =. 2009 , publisher =. doi:10.1198/jcgs.2009.06134 , url =

    Pith/arXiv arXiv doi:10.1198/jcgs.2009.06134 2009

  20. [20]

    Frontiers in Forests and Global Change , VOLUME=

    Prisley, Steve and Bradley, Jeff and Clutter, Mike and Friedman, Suzy and Kempka, Dick and Rakestraw, Jim and Sonne Hall, Edie , TITLE=. Frontiers in Forests and Global Change , VOLUME=. 2021 , URL=. doi:10.3389/ffgc.2021.746439 , ISSN=

    work page doi:10.3389/ffgc.2021.746439 2021

  21. [21]

    and Liknes, Greg C

    Knott, Jonathan A. and Liknes, Greg C. and Giebink, Courtney L. and Oh, Sungchan and Domke, Grant M. and McRoberts, Ronald E. and Quirino, Valquiria F. and Walters, Brian F. , title =. Methods in Ecology and Evolution , volume =. doi:https://doi.org/10.1111/2041-210X.14084 , url =. https://besjournals.onlinelibrary.wiley.com/doi/pdf/10.1111/2041-210X.1408...

    work page doi:10.1111/2041-210x.14084 2041

  22. [22]

    2015 , publisher=

    Small area estimation , author=. 2015 , publisher=

    2015

  23. [23]

    and Harter, Rachel M

    Battese, George E. and Harter, Rachel M. and Fuller, Wayne A. , title =. Journal of the American Statistical Association , year =

  24. [24]

    and Olsen, Anthony R

    Dumelle, Michael and Higham, Matt and Ver Hoef, Jay M. and Olsen, Anthony R. and Madsen, Lisa , title =. Methods in Ecology and Evolution , volume =. doi:https://doi.org/10.1111/2041-210X.13919 , url =. https://besjournals.onlinelibrary.wiley.com/doi/pdf/10.1111/2041-210X.13919 , year =

    doi:10.1111/2041-210x.13919 2041

  25. [25]

    2014 , publisher=

    Hierarchical Modeling and Analysis for Spatial Data, Second Edition , author=. 2014 , publisher=

    2014

  26. [26]

    and Finley, Andrew O

    White, Grayson W. and Finley, Andrew O. and Yamamoto, Josh K. and Green, Jennifer L. and Frescino, Tracey S. and MacFarlane, David W. and Andersen, Hans-Erik and Domke, Grant M. , title =. Canadian Journal of Forest Research , year =. doi:10.1139/cjfr-2025-0175 , url =

    work page doi:10.1139/cjfr-2025-0175 2025

  27. [27]

    and Finley, Andrew O

    Shannon, Elliot S. and Finley, Andrew O. and May, Paul B. and Domke, Grant M. and Andersen, Hans-Erik and. Leveraging National Forest Inventory Data to Estimate Forest Carbon Density Status and Trends for Small Areas , journal =. 2025 , volume =. doi:10.1016/j.foreco.2025.122999 , url =

    work page doi:10.1016/j.foreco.2025.122999 2025

  28. [28]

    and Finley, Andrew O

    Shannon, Elliot S. and Finley, Andrew O. and Domke, Grant M. and May, Paul B. and Andersen, Hans-Erik and. Toward Spatio-Temporal Models to Support National-Scale Forest Carbon Monitoring and Reporting , journal =. 2025 , volume =. doi:10.1088/1748-9326/ad9e07 , url =

    work page doi:10.1088/1748-9326/ad9e07 2025

  29. [29]

    and Hamm, Nicholas A

    Datta, Abhirup and Banerjee, Sudipto and Finley, Andrew O. and Hamm, Nicholas A. S. and Schaap, Martijn , title =. The Annals of Applied Statistics , year =

  30. [30]

    Annals of the Institute of Statistical Mathematics , year =

    Besag, Julian and York, Jeremy and Mollie, Annie , title =. Annals of the Institute of Statistical Mathematics , year =

  31. [31]

    Journal of the Royal Statistical Society: Series B , year =

    Rue, Havard and Martino, Sara and Chopin, Nicolas , title =. Journal of the Royal Statistical Society: Series B , year =

  32. [32]

    Journal of the Royal Statistical Society: Series B , year =

    Lindgren, Finn and Rue, Havard and Lindstrom, Johan , title =. Journal of the Royal Statistical Society: Series B , year =

  33. [33]

    and Lee, Daniel and Goodrich, Ben and Betancourt, Michael and Brubaker, Marcus A

    Carpenter, Bob and Gelman, Andrew and Hoffman, Matthew D. and Lee, Daniel and Goodrich, Ben and Betancourt, Michael and Brubaker, Marcus A. and Guo, Jiqiang and Li, Peter and Riddell, Allen , title =. Journal of Statistical Software , year =

  34. [34]

    and Anderson-Bergman, Clifford and Temple Lang, Duncan and Bodik, Rastislav , title =

    de Valpine, Perry and Turek, Daniel and Paciorek, Christopher J. and Anderson-Bergman, Clifford and Temple Lang, Duncan and Bodik, Rastislav , title =. Journal of Computational and Graphical Statistics , year =

  35. [35]

    and Banerjee, Sudipto and Gelfand, Alan E

    Finley, Andrew O. and Banerjee, Sudipto and Gelfand, Alan E. , title =. Journal of Statistical Software , year =

  36. [36]

    and Datta, Abhirup and Banerjee, Sudipto , title =

    Finley, Andrew O. and Datta, Abhirup and Banerjee, Sudipto , title =. Journal of Statistical Software , year =

  37. [37]

    and Ward, Eric J

    Anderson, Sean C. and Ward, Eric J. and English, Philina A. and Barnett, Lewis A. K. and Thorson, James T. , title =. Journal of Statistical Software , year =

  38. [38]

    and Finley, Andrew O

    Doser, Jeffrey W. and Finley, Andrew O. and K. Methods in Ecology and Evolution , year =

  39. [39]

    Journal of Open Source Software , year =

    Donegan, Connor , title =. Journal of Open Source Software , year =

  40. [40]

    2023 , note =

    Vieilledent, Ghislain , title =. 2023 , note =

    2023

  41. [41]

    Joint Species Distribution Modelling with the

    Tikhonov, Gleb and Opedal,. Joint Species Distribution Modelling with the. Methods in Ecology and Evolution , year =

  42. [42]

    Journal of Statistical Software , year =

    Zammit-Mangion, Andrew and Cressie, Noel , title =. Journal of Statistical Software , year =

  43. [43]

    WIREs Computational Statistics , year =

    Bakka, Haakon and Rue, Havard and Fuglstad, Geir-Arne and Riebler, Andrea and Bolin, David and Illian, Janine and Krainski, Elias and Simpson, Daniel and Lindgren, Finn , title =. WIREs Computational Statistics , year =

  44. [44]

    Journal of Statistical Software , year =

    Pebesma, Edzer , title =. Journal of Statistical Software , year =

  45. [45]

    Journal of Statistical Software , year =

    Lee, Duncan , title =. Journal of Statistical Software , year =

  46. [46]

    and Skaug, Hans and Bell, Bradley M

    Kristensen, Kasper and Nielsen, Anders and Berg, Casper W. and Skaug, Hans and Bell, Bradley M. , title =. Journal of Statistical Software , year =

  47. [47]

    , title =

    Lindgren, Finn and Seaton, Andy and Ho, Suen Man and Bachl, Fabian E. , title =. 2026 , note =

    2026

  48. [48]

    2025 , note =

    Wickham, Hadley and Hesselberth, Jay and Salmon, Ma. 2025 , note =

    2025

  49. [49]

    Fitting Linear Mixed-Effects Models Using

    Bates, Douglas and M. Fitting Linear Mixed-Effects Models Using. Journal of Statistical Software , year =

  50. [50]

    and Hager, William W

    Chen, Yanqing and Davis, Timothy A. and Hager, William W. and Rajamanickam, Sivasankaran , title =. ACM Transactions on Mathematical Software , year =

  51. [51]

    2025 , note =

    Bates, Douglas and Maechler, Martin and Jagan, Mikael , title =. 2025 , note =

    2025

  52. [52]

    Blackford, L. Susan and Demmel, James and Dongarra, Jack and Duff, Iain and Hammarling, Sven and Henry, Greg and Heroux, Michael and Kaufman, Linda and Lumsdaine, Andrew and Petitet, Antoine and Pozo, Roldan and Remington, Karin and Whaley, R. Clint , title =. ACM Transactions on Mathematical Software , year =

  53. [53]

    Susan and Demmel, James and Dongarra, Jack and Du Croz, Jeremy and Greenbaum, Anne and Hammarling, Sven and McKenney, Alan and Sorensen, Danny , title =

    Anderson, Edward and Bai, Zhaojun and Bischof, Christian and Blackford, L. Susan and Demmel, James and Dongarra, Jack and Du Croz, Jeremy and Greenbaum, Anne and Hammarling, Sven and McKenney, Alan and Sorensen, Danny , title =

This paper was first reviewed by grok-4.5 on July 12, 2026.