Pith. sign in

REVIEW 2 major objections 5 minor 36 references

Optimal smooth approximation of integral cycles

T0 review · 2 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read The paper proves that every integral cycle in an oriented Riemannian manifold can be approximated in flat norm by an integral cycle that is a smooth submanifold off a codimension-5 set, with nearly the same area.

desk verdict A long-awaited complete proof of the Almgren-Browder smooth approximation theorem; the main topological input is sound, with one finite-type detail a referee should have fixed. read the letter →

arxiv 2411.17678 v1 pith:4SJ6G3C4 submitted 2024-11-26 math.DG math.AP

Fredrick Almgren , William Browder , Gianmarco Caldini , Camillo De Lellis This is my paper · ORCID
classification math.DGmath.AP MSC 49Q1557R9555R4057R90
keywords integralcurrentsflatnormsmoothapproximationThomspacehomologyclassesgeometricmeasuretheorycodimension5singularitiescobordism
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

This paper proves that any integral cycle in an oriented Riemannian manifold can be approximated, in the flat norm and with almost no area increase, by an integral cycle that is a smooth submanifold everywhere except on a codimension-5 skeleton of a triangulation. In other words, generalized surfaces arising as area-minimizing currents or arbitrary integer homology cycles are, up to a small flat-norm error, nearly indistinguishable from smooth submanifolds away from a small singular set. When the homology class itself can be represented by a smooth submanifold, the approximation can be chosen completely smooth. The codimension-5 bound is shown to be optimal, using Thom's example of an innately singular homology class.

What carries the argument

The engine is the Thom space T(gamma_n) of the universal oriented n-plane bundle over BSO(n), together with the Thom class u in H^n(T(gamma_n),Z) and Thom's criterion that a homology class is representable by a submanifold exactly when its Poincare dual pulls back u. The paper combines this with two geometric-measure-theory tools: a multi-step approximation of a given integral cycle by polyhedral cycles and then by cycles smooth off the (m-2)-skeleton, and a deformation result that swaps two cycles agreeing off a small neighborhood of the (m-2)-skeleton while controlling mass and flat distance. The lift step is carried by the (n+4)-equivalence h:T(gamma_n) to K(Z,n) representing the Thom class, which lets an n-dimensional cohomology class on the (n+4)-complex Q be realized by a map into T(gamma_n), hence by a smooth preimage of BSO(n).

What would settle it

Compute the relative homotopy group of the pair K(Z,n) and T(gamma_n) in dimension n+5; if it is nonzero and some Poincare dual class on an (n+5)-complex fails to lift, the main construction collapses. On the optimality side, checking whether Thom's 7-dimensional class can be smoothed outside the 1-skeleton of a triangulation directly tests the codimension-5 bound.

Watch

Extended reading notes

Core claim

The central theorem states: given a positive tolerance and an integral m-cycle T representing a nonzero class in the m-dimensional integral homology of a connected smooth closed oriented Riemannian manifold of dimension m+n, there is a smooth triangulation K and an oriented smooth m-dimensional submanifold Sigma of the manifold minus the (m-5)-skeleton such that the volume of Sigma is at most the mass of T plus the tolerance, and the current [[Sigma]] is homologous to T with an integral filling of mass smaller than the tolerance. If the homology class admits a smooth representative, Sigma can be chosen smooth everywhere. The proof realizes the Poincare dual of the homology class as the pullback of the Thom class under a map from a complement of a small neighborhood of the (m-2)-skeleton, then lifts this class along a map into the Thom space of the universal oriented n-plane bundle, using an (n+4)-equivalence with an Eilenberg-MacLane space to extend the smooth preimage across the remaining skeleton.

Load-bearing premise

The whole construction hinges on the fact that, through dimension n+4, the Thom space of the universal oriented n-plane bundle is indistinguishable from the Eilenberg-MacLane space K(Z,n); if that comparison failed one dimension earlier, the smooth preimage needed for the theorem could fail to exist.

Editorial extensions

If this is right

  • The flat norm and mass of every integral cycle can be approximated by smooth submanifolds with singularities confined to codimension 5, so no Lavrentiev gap occurs for the homological Plateau problem: the infimum over integral cycles equals the infimum over these nearly smooth cycles.
  • When the homology class is smoothly representable, which is always true for codimension at most 2 or dimension at most 6, any integral cycle representing it can be approximated by completely smooth submanifolds with almost the same area.
  • There exist sequences of smooth submanifolds with singularities in the (m-5)-skeleta of triangulations that converge to a given integral cycle in the sense of currents while their volumes converge to the mass.
  • The codimension-5 threshold is optimal: Thom's 7-dimensional innately singular class in a 14-dimensional manifold cannot be represented smoothly outside the 1-skeleton, so no smaller singular dimension works in full generality.

Reading between the lines

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

  • The proof tracks all constants through its approximation and deformation propositions, so a quantitative version of the theorem with explicit dependence of the tolerance on the geometry of the manifold and the cycle is likely within reach.
  • The same Thom-space lift strategy may apply to other homology theories or coefficient groups, where the relevant comparison map between the Thom space and an Eilenberg-MacLane space may have a different connectivity.
  • The reliance on a triangulation suggests that a more invariant formulation could remove the skeleton entirely and state the result as: outside a codimension-5 subset, the cycle can be smoothed without changing area or homology class by more than a prescribed amount.
  • If the (n+4)-equivalence in Lemma 5.2 could be improved for special manifolds, the singular dimension might drop below codimension 5 in those cases, though Thom's example shows this cannot happen in general.
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 proves Theorem 1.1: for any integral m-cycle T in a closed oriented Riemannian (m+n)-manifold M and any ε>0, there is a smooth triangulation K and a smooth oriented m-submanifold Σ of M\K^{m-5} with H^m(Σ) ≤ M(T)+ε and flat norm F([[Σ]]-T)<ε, with Σ homologous to T; if the homology class admits a smooth representative, then Σ can be chosen globally smooth. The proof combines Federer–Fleming approximation and deformation (Proposition 4.1), a relative Thom construction (Theorem 2.6), an (n+4)-equivalence between the Thom space T(~γ_n) and K(Z,n) (Lemma 5.2), and a final deformation result (Proposition 4.3). Section 6 proves optimality of the codimension-5 statement using Thom's 7-dimensional class.

Significance. This is a substantial result with a long history: it completes a program announced by Almgren and Browder and establishes the optimal codimension-5 smooth approximation statement for arbitrary integral cycles. It also yields absence of Lavrentiev gaps and clarifies which homology classes admit smooth representatives. The paper is largely self-contained in the geometric measure theory parts and provides useful appendices on triangulations and cohomology operations; no fitted parameters or numerical computations are involved. However, a load-bearing topological input, Lemma 5.2, currently rests on a false and incomplete general theorem, so the proof needs repair before the result can be accepted.

major comments (2)
  1. [§2.2, Theorem 2.3; §5, Lemma 5.2] Theorem 2.3 is false as stated, and its proof contains an invalid implication: from H^i(M_f,X;Z_p)=0 for every prime p it does not follow that H_i(M_f,X;Z)=0 without finite-generation hypotheses and a rational-cohomology check. A concrete counterexample is f:*→K(Q,2), for which f^* is trivially an isomorphism in all degrees with Z_p coefficients but π_2(M_f,*)≅Q. The proof of Lemma 5.2 invokes this theorem to conclude that h:T(~γ_n)→K(Z,n) is an (n+4)-equivalence, and equation (1) in Section 5 needs exactly that conclusion for the lift f:Q→T(~γ_n). The intended application is likely repairable: K(Z,n) and T(~γ_n) have finite-dimensional cohomology in the relevant range, and the missing rational comparison is easy (H^{n+i}(K(Z,n);Q)=0 for i>0, while H^{n+i}(T(~γ_n);Q)≅H^i(BSO(n);Q) vanishes for 1≤i≤3 and is generated by p_1 in degree 4). But as written, the central lift step is not proved.
  2. [§4.1, Proposition 4.6] Proposition 4.6 is asserted with the proof delegated to "an inspection of the argument in [21]" rather than carried out. This proposition is load-bearing because it produces the polyhedral cycle P and the smooth triangulation K in Proposition 4.1, which are the starting objects for the whole construction. Since [21] is an unpublished lecture-notes source and the adaptation to the present one-sided piecewise-smooth setting is not written down, the manuscript should supply a complete proof or a precise statement-and-theorem reference.
minor comments (5)
  1. [§1.3 and §6] There are typographical errors such as "descrived" and "substratifed" (and some OCR-style broken words) that should be corrected throughout.
  2. [§1.2, Theorem 1.5] The notation Z_k(M) is used for cycles of dimension m; please introduce a consistent dimension variable (either use Z_m(M) or define k=m in the statement).
  3. [§5, proof of Theorem 1.1] When applying Proposition 2.1 to Q, the text implicitly treats Q as an (n+4)-dimensional complex. Lemma 5.1 only says Q is homotopy equivalent to such a complex; since Q itself has dimension m+n, the lifting argument should first pass to an (n+4)-dimensional CW model Q' and a homotopy-equivalent map g', then transfer the resulting lift back to Q. This is a routine fix, but the step should be made explicit.
  4. [§5, Lemma 5.2] The opening sentence "The spaces are the same for n∈{1,2}" should read "homotopy equivalent", since T(~γ_1) and K(Z,1) are only homotopy equivalent to S^1, and likewise for n=2 with CP∞.
  5. [§5, after equation (15)] The phrase "T(~γ_n)\∞ is a smooth submanifold" should read "a smooth open manifold" (the complement of the base point in the Thom space), since it is not a submanifold of an ambient manifold in the usual sense.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the central derivation reduces to external algebraic-topology inputs (Serre-Cartan computations, Thom isomorphism, characteristic classes) and to internally proved geometric-measure-theory propositions, while the cited Almgren-Browder announcement is explicitly described as lacking a proof.

full rationale

The paper's derivation chain is self-contained against external benchmarks. Theorem 1.1 is proved by combining two internally proved geometric-measure-theory propositions (Propositions 4.1 and 4.3), the classical relative Thom construction (Theorem 2.6, proved in the paper), and one algebraic-topology input, Lemma 5.2. Lemma 5.2 asserts that the map h : T(γ~n) → K(Z,n) representing the Thom class is an (n+4)-equivalence; its proof invokes Serre's mod-2 computations for Eilenberg-MacLane spaces, Cartan's mod-3 computations, the known cohomology rings of BSO(n), and the Thom isomorphism together with the identities Φ2(wi)=Sq^i(u2) and Φ3(p1)=P^1(u3). These are external, classical results and are not restatements of Theorem 1.1 or of any quantity fitted in the paper. The lift in equation (1) follows from Lemma 5.2 plus Proposition 2.1, so the main conclusion genuinely depends on independent cohomological facts rather than on assuming the conclusion. The cited announcement [6] is not load-bearing: the paper states that 'a proof of the announced result never appeared' and uses [6] only for the general strategy of Thom's criterion in the context of maps to T(γ~n); the proof itself is carried out here. Moreover, Remark 6.4 explicitly corrects a counterexample in [6], further demonstrating that the paper does not treat [6] as authoritative. No fitted parameter is renamed as a prediction, no quantity is defined in terms of the target conclusion, and no uniqueness theorem from the authors' prior work is invoked to force a choice. The skeptic's concern about Lemma 5.2 is a correctness or verification risk: the proof of Theorem 2.3 is terse and the finite-type/rational-cohomology steps are not spelled out, so the (n+4)-equivalence may need further checking. That is an ordinary mathematical-gap concern, not circularity, because the missing justification is independent of the paper's own conclusions. Accordingly, the circularity score is 0.

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

No fitted parameters or invented entities. The proof rests on standard geometric measure theory, algebraic topology, and cobordism facts, plus technical lemmas proved in the paper.

assumptions (4)
  • standard math Federer-Fleming theory of integral currents, including mass, flat norm, deformation theorem, and the flatness criterion for cycles.
    Used throughout Sections 2 and 4; for example, the flat norm definition (3) and the conclusion that a current with boundary supported on a low-dimensional skeleton is a cycle.
  • standard math Thom isomorphism and Thom's criterion for representability of homology classes by smooth submanifolds (Theorems 2.5 and 2.6).
    Bridge between geometric measure theory and algebraic topology; Theorem 2.6 is proved in the paper but relies on the standard Thom construction.
  • standard math Serre's and Cartan's cohomology computations for Eilenberg-MacLane spaces and the polynomial cohomology of BSO(n) with Z_p coefficients.
    Invoked in Lemma 5.2 to establish the n+4-equivalence between T(γ_n) and K(Z,n), the load-bearing lift step in the proof.
  • domain assumption Sullivan's singularity resolution theorem (Theorem D in [29]) and the oriented cobordism groups Ω_5 ≅ Z_2 and Ω_6 = 0.
    Used only in Section 6 to prove optimality (Theorem 6.3), not for the existence part of Theorem 1.1.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Optimal smooth approximation of integral cycles." pith.science (2026). https://pith.science/paper/4SJ6G3C4

@misc{pith2026241117678,
  author       = {Pith},
  title        = {Pith review of: Optimal smooth approximation of integral cycles},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/4SJ6G3C4}},
  note         = {Machine review of arXiv:2411.17678}
}
abstract

In this article we prove that each integral cycle $T$ in an oriented Riemannian manifold $\mathcal{M}$ can be approximated in flat norm by an integral cycle in the same homology class which is a smooth submanifold $\Sigma$ of nearly the same area, up to a singular set of codimension 5. Moreover, if the homology class $\tau$ is representable by a smooth submanifold, then $\Sigma$ can be chosen free of singularities.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

36 extracted references · 36 canonical work pages

  1. [21]

    Lurie, T opics in Geometric T opology (18.937) , Lecture notes available at https://www.math.ias.edu/ lurie/937.html

    J. Lurie, T opics in Geometric T opology (18.937) , Lecture notes available at https://www.math.ias.edu/ lurie/937.html

    work page

  2. [1]

    Annals of Mathematics

    W . Allard, On the first variation of a varifold , in “Annals of Mathematics", 95 (3), 417–491, 1972

    work page 1972

  3. [2]

    Annals of Mathematics

    F. J. Almgren, Some interior regularity theorems for minimal surfaces andan extension of Bernstein’s theorem , in “Annals of Mathematics", 84 (2), 277–292, 1966

    work page 1966

  4. [3]

    Bulletin of the American Mathematical Society (N.S.)

    F. J. Almgren, Q­valued functions minimizing Dirichlet’s integral and th e regularity of area minimizing rectifiable currents up to codimension tw o, in “Bulletin of the American Mathematical Society (N.S.)", 8 (2), 327–328, 1983

    work page 1983

  5. [4]

    Proceeding s of Symposia in Pure Mathematics

    F. J. Almgren, Questions and Answers about Area­Minimizing Surfaces and Geometric Measure Theory, proceedings of Differential Geometry: Partial Different ial Equations on Manifolds (Los Angeles, CA July 8­28 1990) in “Proceeding s of Symposia in Pure Mathematics", 54 (1), 255–259, American Mathematical Society, Providence 1 993

    work page 1990

  6. [5]

    F. J. Almgren, Almgren’s Big Regularity Paper, World Scientific Monograph Series in Mathematics, 1, World Scientific Publishing Co. Inc., River Edge 2000

    work page 2000

  7. [6]

    Pitman monographs and surv eys in pure and applied mathematics

    F. J. Almgren, W . Browder, Homotopy with holes and minimal surfaces , proceedings of Differential Geometry: A Symposium in Honour of Manfredo Do Carmo (Rio de Janeiro, Brazil August 1988) in “Pitman monographs and surv eys in pure and applied mathematics", 52, 15–24, Longman Scientific & Technical, Essex 1991

    work page 1988

  8. [7]

    Benedetti, Lectures on Differential T opology, Graduate studies in mathematics, 218, American Mathematical Society, Providence 2021

    R. Benedetti, Lectures on Differential T opology, Graduate studies in mathematics, 218, American Mathematical Society, Providence 2021

    work page 2021

Show all 36 references
  1. [8]

    manuscripta mathematica

    C. Bohr, B. Hanke, D. Kotschick, Cycles, submanifolds, and structures on normal bundles, in “manuscripta mathematica", 108, 483–494, 2002

    2002

  2. [9]

    Mathematics and Its Applications

    G. Buttazzo, M. Belloni, A Survey on Old and Recent Results about the Gap Phe­ nomenon in the Calculus of V ariations , in “Mathematics and Its Applications", 331, 1–27, 1995

    1995

  3. [10]

    De Giorgi, Frontiere orientate di misura minima, in Seminario di Matematica della Scuola Normale Superiore di Pisa A.A

    E. De Giorgi, Frontiere orientate di misura minima, in Seminario di Matematica della Scuola Normale Superiore di Pisa A.A. 1960­1961 , Editrice Tecnico Scientifica, Pisa 1961, 1–56

    1960

  4. [11]

    Memoirs of the American Mathematical Society

    C. De Lellis and E. Spadaro, Q­valued functions revisited , in “Memoirs of the American Mathematical Society", 211 (991), 2011

    2011

  5. [12]

    Geometric and Functional Analysis

    C. De Lellis and E. Spadaro, Regularity of area minimizing currents I: gradient L p estimates, in “Geometric and Functional Analysis", 24, 1831–1884, 2014

    2014

  6. [13]

    Annali della Scuola Normale Superiore di Pisa ­ Classe di Scienze

    C. De Lellis and E. Spadaro, Multiple valued functions and integral currents, in “Annali della Scuola Normale Superiore di Pisa ­ Classe di Scienze", 14 (5), 1239–1269, 2015

    2015

  7. [14]

    Annals of Mathematics

    C. De Lellis and E. Spadaro, Regularity of area minimizing currents II: center manifold, in “Annals of Mathematics", 183 (2), 499–575, 2016

    2016

  8. [15]

    Annals of Mathematics

    C. De Lellis and E. Spadaro, Regularity of area minimizing currents III: blow­up , in “Annals of Mathematics", 183 (2), 577–617, 2016. 54 Fredrick Almgren† , William Browder , Gianmarco Caldini and Camillo De Lellis

    2016

  9. [16]

    Dubrovin, A.T

    B.A. Dubrovin, A.T. Fomenko, S.P . Novikov, Modern Geometry ­ Methods and Appli­ cations III, Graduate Texts in Mathematics, 124, Springer, Berlin­Heidelberg 1990

    1990

  10. [17]

    Notices of the American Mathematica l Society

    H. Federer, An approximation theorem concerning currents of finite mass , proceedings of Sixty­fourth Summer Meeting and Thirty­eight Colloquiu m (Salt Lake City, Utah September 1­4 1959) in “Notices of the American Mathematica l Society", 6 (4), 326, American Mathematical Soci...

    1959

  11. [18]

    Federer, Geometric measure theory , Classics in Mathematics, Springer, Berlin­ Heidelberg 1996

    H. Federer, Geometric measure theory , Classics in Mathematics, Springer, Berlin­ Heidelberg 1996

    1996

  12. [19]

    Annals of Mathematics

    H. Federer and W . Fleming, Normal and Integral Currents, in “Annals of Mathematics", 72 (3), 458–520, 1960

    1960

  13. [20]

    Goresky , R

    M. Goresky , R. MacPherson, Stratified Morse Theory , Ergebnisse der Mathematik und ihrer Grenzgebiete, 14, Springer, Berlin­Heidelberg 1988

    1988

  14. [22]

    Annali di Matem­ atica Pura ed Applicata

    M. Lavrentiev, Sur quelques probl`emes du calcul des variations , in “Annali di Matem­ atica Pura ed Applicata", 4 (1), 7–28, 1927

    1927

  15. [23]

    Milnor and J

    J. Milnor and J. Stasheff, Characteristic Classes, Annals of Mathematics Studies, 76, Princeton University Press, Princeton 1974

    1974

  16. [24]

    J. R. Munkres, Elements of Algebraic T opology, Addison­Wesley, Menlo Park 1984

    1984

  17. [25]

    Com­ mentarii Mathematici Helvetici

    J. P . Serre, Cohomologie modulo 2 des complexes d’Eilenberg­MacLane , in “Com­ mentarii Mathematici Helvetici", 27, 198–232, 1953

    1953

  18. [26]

    Simon, Lectures on geometric measure theory , Proceedings of the Centre for Mathematical Analysis, 3, Australian National University, Canberra 1984

    L. Simon, Lectures on geometric measure theory , Proceedings of the Centre for Mathematical Analysis, 3, Australian National University, Canberra 1984

    1984

  19. [27]

    E. H. Spanier, Algebraic T opology, McGraw­Hill, New Y ork 1966

    1966

  20. [28]

    N. E. Steenrod and D. B. A. Epstein, Cohomology operations, Annals of Mathematics Studies, 50, Princeton University Press, Princeton 1962

    1962

  21. [29]

    Lecture n otes in mathematics

    D. Sullivan, Singularities in spaces, proceedings of Liverpool Singularities Symposium II (Liverpool, UK September 1969 ­ August 1970) in “Lecture n otes in mathematics", 209, 196–206, Springer, Berlin­Heidelberg 1971

    1969

  22. [30]

    R. M. Switzer, Algebraic T opology ­ Homology and Homotopy, Classics in Mathemat­ ics, Springer, Berlin­Heidelberg 2002

    2002

  23. [31]

    R. E. Mosher and M. C. Tangora, Cohomology operations and applications in homotopy theory, Harper’s Series in Modern Mathematics, Harper & Row, New Y o rk 1968

    1968

  24. [32]

    Annales scientifiques de l’ ´Ecole normale sup ´erieure

    R. Thom, Espaces fibr ´es en sph `eres et carr ´es de Steenrod , in “Annales scientifiques de l’ ´Ecole normale sup ´erieure", 69, 109–182, 1952

    1952

  25. [33]

    Commentarii Mathematici Helvetici

    R. Thom, Quelques propri´et´es globales des vari ´et´es diff ´erentiables, in “Commentarii Mathematici Helvetici", 28, 17–86, 1954. Optimal smooth approximation of integral cycles 55

    1954

  26. [34]

    Bulletin of the American Mathemat­ ical Society

    R. Thom, Ensembles et morphismes stratifi ´es, in “Bulletin of the American Mathemat­ ical Society", 75, 240–284, 1969

    1969

  27. [35]

    C. T. C. Wall, Differential T opology, Cambridge studies in advanced mathematics, 156, Cambridge University Press, Cambridge 2016

    2016

  28. [36]

    Annals of Mathematics

    J. H. C. Whitehead, On C 1 ­complexes, in “Annals of Mathematics", 41 (4), 809–824, 1940. DEPARTMENT OF MA THEMA TICS, PRINCETON UNIVERSITY, PRINCETON NJ 08544, USA. E­mail address: browder@math.princeton.edu DEPARTMENT OF MA THEMA TICS, UNIVERSIT `A DI TRENTO, 38123 TRENTO, I...

    1940

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.