Pith. sign in

REVIEW 3 major objections 6 minor 39 references

Robust Score Matching

T0 review · 3 major / 6 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read Score matching can be made robust to corrupted rows by replacing means with a geometric median of means.

desk verdict Solid robust score matching paper with a real gap: consistency is proved only for sublinear contamination, while the experiments and tuning heuristic target a fixed proportion. read the letter →

arxiv 2501.05105 v1 pith:OODNS4NS submitted 2025-01-09 stat.ML cs.LG

Richard Schwank , Andrew McCormack , Mathias Drton This is my paper · ORCID
classification stat.MLcs.LG MSC 62F3562H1262J07
keywords robustscorematchinggeometricmedianofmeanscontaminationexponentialfamilygraphicalmodelssupportrecoveryconvexityhigh-dimensionalestimation
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 proposes a robust version of score matching that uses the geometric median of means instead of ordinary sample averages. It claims this estimator stays consistent when the observed data has been contaminated, while remaining convex in exponential family models. This matters because score matching is a practical way to fit non-Gaussian graphical models without computing normalizing constants, and contamination would otherwise break the standard estimator. The paper also provides support recovery guarantees for sparse high-dimensional models under corruption.

What carries the argument

The geometric median of means (GMoM) partitions the sample into $K$ blocks, takes the mean of each block, and then computes the geometric median of those block means. The geometric median is a convex combination of its arguments, so applying the GMoM to positive semidefinite matrices keeps the resulting matrix positive semidefinite; this preserves convexity of the score matching objective and guarantees the existence of the estimator. The method inherits a breakdown point near $1/2$, and the block-size parameter $K$ tunes the trade-off between robustness, bias, and variance.

What would settle it

Simulate data from a square root graphical model with a fixed 5% of rows replaced by arbitrary outliers and compute the GMoM score matching estimator at increasing sample sizes $n = 10^3, 10^4, 10^5$. If the squared error to the true parameter does not shrink to zero as $n$ grows, the consistency claim does not hold in the fixed-proportion contamination regime.

Watch

Extended reading notes

Core claim

The central claim is that replacing the score matching statistics $\Gamma(X)$ and $g(X)$ by their geometric-median-of-means versions, $\hat\Gamma_K(X)$ and $\hat g_K(X)$, yields an estimator $\hat\theta(K) = \hat\Gamma_K(X)^{-1}\hat g_K(X)$ that converges in probability to the true parameter $\theta_0$ when at most $n_c = o(n)$ observations are arbitrarily corrupted. With an $\ell^1$ penalty and under an irrepresentability condition, the regularized estimator has support contained in the true support and satisfies $\|\hat\theta(K, \beta, \lambda) - \theta_0\|_\infty \leq c_{\Gamma_0}(2-\alpha)\lambda$ with probability at least $1 - 2\delta$.

Load-bearing premise

The proof only allows the number of corrupted rows to grow like $O(\log(1/\delta))$, so the contamination must be a vanishing fraction of the sample; a fixed positive fraction of corrupted rows is not covered by the consistency theorem.

Editorial extensions

If this is right

  • Score matching can be extended to contaminated data while keeping the convenience of a convex objective in exponential family models.
  • The estimator is consistent when the number of corrupted rows grows slower than the sample size, even if the corruption is adversarial.
  • For sparse graphical models, the regularized robust estimator provably recovers the true conditional independence structure under contamination.
  • In numerical experiments, the robust estimator matches the standard one on clean data and substantially outperforms it when data are contaminated, including on a real precipitation network.

Reading between the lines

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

  • The formal consistency result only covers a vanishing fraction of corrupted rows, while the experiments use a fixed 5% or 10% contamination; the fixed-proportion regime is supported empirically but not by the main theorem.
  • The same GMoM construction could robustify other moment-based estimators that require averaging positive semidefinite matrices, since the convex-combination property preserves definiteness.
  • A natural next test is whether a minimax optimal choice of block size $K$ can be derived for fixed contamination proportions, rather than the heuristic $K = 4\varepsilon n$ used in the paper.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 6 minor

Summary. The paper proposes a robustified score matching estimator for exponential families by replacing the empirical moment quantities Γ(X) and g(X) with geometric median of means (GMoM) aggregates. The resulting objective remains convex in exponential family models. The main theoretical results are a concentration inequality for the GMoM under rowwise corruption (Theorem 3.1), consistency of the robust score matching estimator when the number of corrupted samples is sublinear (Corollary 3.2), and an ℓ1-regularized support recovery guarantee for pairwise interaction models (Theorem 4.3). Numerical experiments on square root graphical models and a precipitation dataset show improved support recovery relative to classical score matching under 5–20% contamination.

Significance. If the advertised claims held for the contamination model in Section 3.1, this would be a valuable extension of score matching to adversarial rowwise corruption: the retention of convexity is a real advantage over naive robustification, and support recovery in non-Gaussian graphical models is of substantial interest. The paper is carefully written, with detailed proofs, a thoughtful discussion of multivariate median properties, and a useful empirical study including a non-synthetic application. The main caveat is that the theoretical guarantees are established only for sublinearly many corrupted samples, whereas the paper's problem statement and experiments use a fixed contamination proportion; the contribution is therefore more modest than the abstract suggests unless that gap is closed.

major comments (3)
  1. [3.1, Theorem 3.1, Corollary 3.2] The contamination model in Section 3.1 allows a fixed proportion ε of rows to be altered arbitrarily, and the simulations in Section 5.1 use ε=5%. However, Theorem 3.1 permits only up to (⌊17 log(1/δ)⌋+1)τ corrupted samples, which for fixed confidence δ is O(1), and Corollary 3.2 obtains consistency only in the regime n_c=o(n). If one sets n_c=ε n with ε>0 fixed, the proof's construction δ=exp(−(⌈n_c/τ⌉−1)/17) yields log(1/δ)=Θ(n), so the error bound in (11) is Θ(1) and no consistency in probability follows; additionally, the required K grows linearly in n, whereas the theorem's framework for fixed δ has K=O(log(1/δ)). The abstract's claim of consistent estimation under contaminated data is therefore not established for the paper's own contamination model. The authors should either restrict the consistency claim to n_c=o(n) and state this limitation prominently, or supply a fixed-ε concentration argument, which would need to address the fact that the GMoM is biased for the mean when K is large.
  2. [3.4, 5.1] The block-size rule K:=4εn used in all experiments is linear in n, while the theoretical results require K=O(log(1/δ)) for a fixed confidence level δ (and K≤n/2). Consequently the empirical configuration K=n/5 for ε=0.05 is outside the scope of Theorem 3.1 and Theorem 4.3. The heuristic is motivated only by breakdown-point considerations, not by the concentration theory; the paper should explicitly state that the experiments are not covered by the theoretical guarantees.
  3. [Theorem 4.3] The support recovery guarantee has the same sublinear-corruption limitation. The statement allows n_c:=τ(⌊17 log(1/δ)⌋+1) corrupted samples, so for a fixed confidence δ this number is O(1) and independent of n. In the fixed-proportion regime n_c=ε n, the theorem can be applied only with δ exponentially small in n; then the lower bound on λ in (16) is of constant order, and the ℓ∞ error bound (17) does not vanish. The discussion following the theorem presents this as a consistency-style guarantee, but for fixed ε the conclusion reduces to a constant-error containment statement rather than convergence to θ0.
minor comments (6)
  1. [3.4] The sentence 'the breakdown point is exceeded when K ≥ 2ε' is dimensionally inconsistent; it should read K ≥ 2ε n.
  2. [Appendix A] The positive-definiteness argument for Γ̂K uses the convex-combination representation (8), which is valid only when the geometric median does not coincide with one of the block means; the K=1 case and the coincident case should be handled separately for completeness.
  3. [2.2] There is a typo: 'subleties' should be 'subtleties'.
  4. [Appendix B] In the simplification of the exponent, the phrase 'First term of ψ negative' is cryptic; the argument would be clearer if the sign of the relevant term were shown explicitly.
  5. [Appendix E] In the statement of Theorem E.2, 'he minimizer' should be 'the minimizer'.
  6. [Figure 1] The caption states that pointwise uncertainty in TPR is at most ±0.75% based on 100–500 simulations; please clarify whether this is a 95% bootstrap band for the average ROC and why the reported uncertainty is so small.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: robust consistency comes from an external median-of-means lemma; self-citations to Yu et al. (2019) are prior published theorems used as building blocks, not as the source of the robustness claim.

full rationale

The paper's main robustness claim is not circular. The estimator in (10) is defined by replacing the empirical matrix Gamma(X) and vector g(X) in the classical score matching loss with their GMoM counterparts in (9); no parameter is fitted to the quantity that is later 'predicted' (the true theta_0). Theorem 3.1 is proved in Appendix B from Minsker's geometric median concentration lemma (Lemma B.1), an external result, and Corollary 3.2 follows by choosing delta(n_c) = exp(-(ceil(n_c/tau_0)-1)/17) and invoking continuity of matrix inversion together with the standard score matching identifiability result (Yu et al. 2019, Prop. 2). The support recovery theorem (Theorem 4.3) is modular: Appendix E shows concentration of the robust Gamma and g estimates (Lemma E.1 plus Theorem 3.1) and then applies Yu et al. (2019, Theorem E.2) as a black box. That citation is a self-citation (M. Drton is a coauthor of Yu et al. 2019), but it is independent support: the cited theorem is a peer-reviewed, parameter-free statement whose assumptions (irrepresentability and closeness of Gamma and g estimates) do not include the GMoM estimator, so the robust claim does not reduce by definition to the citation. The choice K = 4*epsilon*n is explicitly presented as a heuristic validated by simulations (Sections 3.4 and Appendix D), not as a theoretical prediction. The main weakness is a scope gap, not circularity: Section 3.1 and the experiments use a fixed contamination proportion epsilon, while Theorem 3.1 and Corollary 3.2 require n_c = o(n) corrupted rows; this mismatch affects whether the consistency guarantee applies to the paper's own simulation regime, but it is not a circular derivation. For these reasons no step in the derivation is equivalent to its own input.

Assumptions & free parameters 3 free parameters · 6 assumptions · 0 invented entities

The robust estimator depends on three user-chosen hyperparameters (K, beta, lambda). K is the most consequential; the paper's recommended choice requires knowing the contamination proportion. The theoretical guarantees rest on standard exponential family regularity, finite second moments, the irrepresentability condition, and Minsker's robust median concentration lemma. No new physical or conceptual entities are introduced.

free parameters (3)
  • K (number of GMoM blocks) = K = 4*epsilon*n (heuristic; in simulations 200 for n=1000, 5% contamination)
    User-selected; the paper proposes K=4*epsilon*n based on simulation studies in Appendix D. The theoretical results require K to satisfy certain bounds but do not prescribe the value.
  • beta (diagonal multiplier) = 0 for n=1000; about 0.01 for n=80 (upper bound of Theorem 4.3 with Gamma_0 estimated from uncorrupted data)
    Regularization parameter to ensure positive definiteness of the Gram matrix when n < m; set oracle-like in the simulations.
  • lambda (l1 penalty) = Varied over a grid to produce ROC curves
    Standard regularization parameter in Definition 4.1; the theory gives a lower bound but not an explicit value.
assumptions (6)
  • domain assumption Exponential family regularity conditions (A1) and (A2) from Appendix A, plus the conditions of Proposition 2 in Yu et al. (2019) for the score matching loss to identify theta_0.
    Invoked in Corollary 3.2 and Theorem 4.3 to guarantee positive definiteness of Gamma_hat and that the population minimizer is theta_0.
  • domain assumption Finite second moments of Gamma(x) and g(x) under theta_0.
    Assumption 3 in Corollary 3.2 and in Theorem 4.3 to apply Chebyshev and concentration bounds.
  • domain assumption Irrepresentability condition with incoherence parameter alpha in (0,1] for Gamma_0.
    Definition 4.2; required to apply the support recovery theorem E.2 from Yu et al. (2019).
  • domain assumption The contamination model allows at most tau*(floor(17 log(1/delta)) + 1) rows corrupted arbitrarily, with tau < 1/2.
    Theorem 3.1 and Theorem 4.3 statements; this is the key robustness premise and it yields the o(n) limitation.
  • ad hoc to paper Known contamination proportion epsilon (or a bound on the number of corruptions) to choose K.
    Section 3.4 and Section 5.1 set K=4*epsilon*n; the theory does not provide a data-driven selector for K.
  • standard math Minsker's geometric median concentration lemma (Lemma B.1).
    External result from Minsker (2015), stated and used in the proof of Theorem 3.1.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Robust Score Matching." pith.science (2026). https://pith.science/paper/OODNS4NS

@misc{pith2026250105105,
  author       = {Pith},
  title        = {Pith review of: Robust Score Matching},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/OODNS4NS}},
  note         = {Machine review of arXiv:2501.05105}
}
read the original abstract

Proposed in Hyv\"arinen (2005), score matching is a parameter estimation procedure that does not require computation of distributional normalizing constants. In this work we utilize the geometric median of means to develop a robust score matching procedure that yields consistent parameter estimates in settings where the observed data has been contaminated. A special appeal of the proposed method is that it retains convexity in exponential family models. The new method is therefore particularly attractive for non-Gaussian, exponential family graphical models where evaluation of normalizing constants is intractable. Support recovery guarantees for such models when contamination is present are provided. Additionally, support recovery is studied in numerical experiments and on a precipitation dataset. We demonstrate that the proposed robust score matching estimator performs comparably to the standard score matching estimator when no contamination is present but greatly outperforms this estimator in a setting with contamination.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

39 extracted references · 36 canonical work pages

  1. [1]

    , Van Aelst , S

    barticle Alqallaf , F. , Van Aelst , S. , Yohai , V.J. , Zamar , R.H. : Propagation of outliers in multivariate data . Ann. Statist. 37 ( 1 ), 311 -- 331 ( 2009 ) barticle

    work page 2009

  2. [2]

    , Fang , G

    bchapter Bhatt , S. , Fang , G. , Li , P. , Samorodnitsky , G. : Minimax m-estimation under adversarial contamination . In: Proceedings of the 39th International Conference on Machine Learning . Proceedings of Machine Learning Research , vol. 162 , pp. 1906 -- 1924 . PMLR , New York, USA ( 2022 ) bchapter

    work page 1906

  3. [3]

    , Li , J

    botherref Blanchet , J. , Li , J. , Lin , S. , Zhang , X. : Distributionally Robust Optimization and Robust Statistics (2024). https://arxiv.org/abs/2401.14655 botherref

    arXiv 2024

  4. [4]

    : Statistical uses of the spatial median

    barticle Brown , B.M. : Statistical uses of the spatial median . J. Roy. Statist. Soc. Ser. B 45 ( 1 ), 25 -- 30 ( 1983 ) barticle

    work page 1983

  5. [5]

    , Weiss , S.T

    barticle Chu , J.-h. , Weiss , S.T. , Carey , V.J. , Raby , B.A. : A graphical model approach for inferring large-scale networks integrating gene expression and genetic polymorphism . BMC Systems Biology 3 , 1 -- 9 ( 2009 ) barticle

    work page 2009

  6. [6]

    , Gasko , M

    barticle Donoho , D.L. , Gasko , M. : Breakdown properties of location estimates based on halfspace depth and projected outlyingness . Ann. Statist. 20 ( 4 ), 1803 -- 1827 ( 1992 ) barticle

    work page 1992

  7. [7]

    , Kane , D.M

    bbook Diakonikolas , I. , Kane , D.M. : Algorithmic High-Dimensional Robust Statistics . Cambridge University Press , Cambridge ( 2023 ) bbook

    work page 2023

  8. [8]

    , Lerasle , M

    barticle Devroye , L. , Lerasle , M. , Lugosi , G. , Oliveira , R.I. : Sub- G aussian mean estimators . Ann. Statist. 44 ( 6 ), 2695 -- 2725 ( 2016 ) barticle

    work page 2016

Show all 39 references
  1. [9]

    , Perlman , M.D

    barticle Eaton , M.L. , Perlman , M.D. : The non-singularity of generalized sample covariance matrices . Ann. Statist. 1 , 710 -- 717 ( 1973 ) barticle

    1973

  2. [10]

    : An introduction to ROC analysis

    barticle Fawcett , T. : An introduction to ROC analysis . Pattern Recognition Letters 27 ( 8 ), 861 -- 874 ( 2006 ) barticle

    2006

  3. [11]

    , Hastie , T

    barticle Friedman , J. , Hastie , T. , H\" o fling , H. , Tibshirani , R. : Pathwise coordinate optimization . Ann. Appl. Stat. 1 ( 2 ), 302 -- 332 ( 2007 ) barticle

    2007

  4. [12]

    , Hastie , T

    barticle Friedman , J. , Hastie , T. , Tibshirani , R. : Sparse inverse covariance estimation with the graphical lasso . Biostatistics 9 ( 3 ), 432 -- 441 ( 2007 ) barticle

    2007

  5. [13]

    : Estimation of non-normalized statistical models by score matching

    barticle Hyv\" a rinen , A. : Estimation of non-normalized statistical models by score matching . J. Mach. Learn. Res. 6 , 695 -- 709 ( 2005 ) barticle

    2005

  6. [14]

    : Some extensions of score matching

    barticle Hyv\" a rinen , A. : Some extensions of score matching . Comput. Statist. Data Anal. 51 ( 5 ), 2499 -- 2512 ( 2007 ) barticle

    2007

  7. [15]

    , Ravikumar , P

    bchapter Inouye , D. , Ravikumar , P. , Dhillon , I. : Square root graphical models: Multivariate generalizations of univariate exponential families that permit positive dependencies . In: Proceedings of The 33rd International Conference on Machine Learning . Proceedings of Ma...

    2016

  8. [16]

    , Shafiee , S

    botherref Kuhn , D. , Shafiee , S. , Wiesemann , W. : Distributionally Robust Optimization (2024). https://arxiv.org/abs/2411.02549 botherref

    2024 arXiv

  9. [17]

    : Graphical Models

    bbook Lauritzen , S.L. : Graphical Models . Oxford Statistical Science Series , vol. 17 , p. 298 . The Clarendon Press, Oxford University Press , New York, USA ( 1996 ) bbook

    1996

  10. [18]

    , Drton , M

    barticle Lin , L. , Drton , M. , Shojaie , A. : Estimation of high-dimensional graphical models using regularized score matching . Electron. J. Stat. 10 ( 1 ), 806 -- 854 ( 2016 ) barticle

    2016

  11. [19]

    , Mendelson , S

    barticle Lugosi , G. , Mendelson , S. : Mean estimation and regression under heavy-tailed distributions: A survey . Foundations of Computational Mathematics 19 ( 5 ), 1145 -- 1190 ( 2019 ) barticle

    2019

  12. [20]

    , Mosler , K

    barticle Liu , X. , Mosler , K. , Mozharovskyi , P. : Fast computation of T ukey trimmed regions and median in dimension p>2 . J. Comput. Graph. Statist. 28 ( 3 ), 682 -- 697 ( 2019 ) barticle

    2019

  13. [21]

    , Rousseeuw , P.J

    barticle Lopuha\" a , H.P. , Rousseeuw , P.J. : Breakdown points of affine equivariant estimators of multivariate location and covariance matrices . Ann. Statist. 19 ( 1 ), 229 -- 248 ( 1991 ) barticle

    1991

  14. [22]

    , Staerman , G

    bchapter Laforgue , P. , Staerman , G. , Cl \'e men c on , S. : Generalization bounds in the presence of outliers: a median-of-means study . In: Proceedings of the 38th International Conference on Machine Learning . Proceedings of Machine Learning Research , vol. 139 , pp. 593...

    2021

  15. [23]

    , Tan , X.L

    barticle Loh , P.-L. , Tan , X.L. : High-dimensional robust precision matrix estimation: Cellwise corruption under -contamination . Electronic Journal of Statistics 12 ( 1 ), 1429 -- 1467 ( 2018 ) barticle

    2018

  16. [24]

    , Drton , M

    bbook Maathuis , M. , Drton , M. , Lauritzen , S. , Wainwright , M. (eds.): Handbook of Graphical Models . Chapman & Hall/CRC Handbooks of Modern Statistical Methods , p. 536 . CRC Press , Boca Raton, FL ( 2019 ) bbook

    2019

  17. [25]

    : Geometric median and robust estimation in B anach spaces

    barticle Minsker , S. : Geometric median and robust estimation in B anach spaces . Bernoulli 21 ( 4 ), 2308 -- 2335 ( 2015 ) barticle

    2015

  18. [26]

    : Distributed statistical estimation and rates of convergence in normal approximation

    barticle Minsker , S. : Distributed statistical estimation and rates of convergence in normal approximation . Electron. J. Stat. 13 ( 2 ), 5213 -- 5252 ( 2019 ) barticle

    2019

  19. [27]

    , Martin , R.D

    bbook Maronna , R.A. , Martin , R.D. , Yohai , V.J. , Salibi\' a n-Barrera , M. : Robust Statistics , 2 nd edn. Wiley Series in Probability and Statistics , p. 430 . John Wiley & Sons, Inc. , Hoboken, NJ ( 2019 ) bbook

    2019

  20. [28]

    , Oja , H

    barticle Niinimaa , A. , Oja , H. , Tableman , M. : The finite-sample breakdown point of the O ja bivariate median and of the corresponding half-samples version . Statist. Probab. Lett. 10 ( 4 ), 325 -- 328 ( 1990 ) barticle

    1990

  21. [29]

    , Deasy , J.O

    barticle Oh , J.H. , Deasy , J.O. : Inference of radio-responsive gene regulatory networks using the graphical lasso algorithm . BMC Bioinformatics 15 , 1 -- 8 ( 2014 ) barticle

    2014

  22. [30]

    , Dunson , D.B

    barticle Roy , A. , Dunson , D.B. : Nonparametric graphical model for counts . Journal of Machine Learning Research 21 ( 229 ), 1 -- 21 ( 2020 ) barticle

    2020

  23. [31]

    , Oja , H

    bchapter Ronkainen , T. , Oja , H. , Orponen , P. : Computation of the multivariate O ja median . In: Developments in Robust Statistics ( V orau, 2001) , pp. 344 -- 359 . Physica , Heidelberg ( 2003 ) bchapter

    2001

  24. [32]

    : Multivariate symmetry and asymmetry

    barticle Serfling , R.J. : Multivariate symmetry and asymmetry . Encyclopedia of Statistical Sciences 8 , 5338 -- 5345 ( 2006 ) barticle

    2006

  25. [33]

    , Kolar , M

    bchapter Sun , S. , Kolar , M. , Xu , J. : Learning structured densities via infinite dimensional exponential families . In: Advances in Neural Information Processing Systems , vol. 28 . Curran Associates, Inc. , New York, USA ( 2015 ) bchapter

    2015

  26. [34]

    : A survey of multidimensional medians

    barticle Small , C.G. : A survey of multidimensional medians . International Statistical Review / Revue Internationale de Statistique 58 ( 3 ), 263 -- 277 ( 1990 ) barticle

    1990

  27. [35]

    barticle Tank , K. , A.M.G. , Coauthors : Daily dataset of 20th-century surface air temperature and precipitation series for the european climate assessment . International Journal of Climatology 22 , 1441 -- 1453 ( 2002 ) barticle

    2002

  28. [36]

    , Zhang , C.-H

    barticle Vardi , Y. , Zhang , C.-H. : A modified W eiszfeld algorithm for the F ermat- W eber location problem . Math. Program. 90 ( 3 ), 559 -- 566 ( 2001 ) barticle

    2001

  29. [37]

    , Drton , M

    barticle Yu , S. , Drton , M. , Shojaie , A. : Generalized score matching for non-negative data . J. Mach. Learn. Res. 20 ( 1 ), 2779 -- 2848 ( 2019 ) barticle

    2019

  30. [38]

    , Kolar , M

    bchapter Yu , M. , Kolar , M. , Gupta , V. : Statistical inference for pairwise graphical models using score matching . In: Advances in Neural Information Processing Systems , vol. 29 . Curran Associates, Inc. , New York, USA ( 2016 ) bchapter

    2016

  31. [39]

    write newline

    " write newline "" before.all 'output.state := FUNCTION string.to.integer 't := t text.length 'k := #1 'char.num := t char.num #1 substring 's := s is.num s "." = or char.num k = not and char.num #1 + 'char.num := while char.num #1 - 'char.num := t #1 char.num substring FUNCTI...

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.