Flexibility of the isometric immersion system in arbitrary dimension and codimension and the energy scaling of prestrained thin films

arXiv:2607.15838 · math.AP · Submitted 2026-07-17 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Mathematics. The day in mathematics: new theorems, new proofs, and what they open.

Ada: We're looking at the paper: "Flexibility of the isometric immersion system in arbitrary dimension and codimension and the energy scaling of prestrained thin films".

Paper summary: Ada: To recap, we're discussing this paper, "Flexibility of the isometric immersion system in arbitrary dimension and codimension and the energy scaling of prestrained thin films." We established that for a given C r, beta regular metric on a d-dimensional domain, every short immersion into the Euclidean space R d plus k can be uniformly approximated by exact isometric immersions of regularity C one alpha for any alpha less than some threshold.

Emmy: The main point is the flexibility statement itself, but what’s particularly new is how they unify the threshold exponent alpha zero across all dimensions and codimensions k. They show that this exponent agrees with results from other systems, like the Monge-Ampere system.

Kolya: I want to focus on that range of codimensions k between one and d times d plus one minus d plus one where they previously lacked a general result for the flexibility exponent alpha zero. This is where they claim their novelty lies.

Ada: That’s right, and they connect this to a concrete physical problem by proving a new quantitative immersability estimate for thin non-Euclidean films. They look at an energy functional E h g that measures the averaged pointwise deficit of a deformation from being an isometric immersion of the metric g on a domain Omega h, as the thickness h goes to zero.

Emmy: When they consider the case where d equals two and k equals one, this estimate models the elastic energy of deformations of a prestrained film, which is what people study in that area. They show that for d equals two and k equals one, this infimum of non-Euclidean energies scales like h theta where theta is between one and two.

Kolya: I'm interested in the quantitative estimate itself. It proves that the energy falls with a rate of four alpha zero alpha plus one when you take the limit as the film’s thickness goes to zero, which gives us a concrete scaling law for that physical energy.

Ada: So, essentially, they give us a general rule for how regular these approximations can be and how that relates to the geometry of the underlying space and the codimension we are working in. This is important because it links analysis directly to physical modeling of thin materials.

Conclusion: Emmy: So looking at this paper, "Flexibility of the isometric immersion system in arbitrary dimension and codimension and the energy scaling of prestrained thin films," it feels like they’ve successfully bridged a gap between pure geometric analysis and applied physics. The authors are Marta Lewicka and Hui Li.

Ada: They achieved a unified description for how flexible these isometric immersion systems are by giving a general threshold exponent alpha zero that works for any dimension d and codimension k, which is what this paper is all about.

Kolya: What does it mean in simpler terms? It means we have a consistent way to predict the regularity of the best possible approximation you can get for an isometric immersion, no matter how complicated the space or how high the codimension.

Emmy: And beyond that, they use this to give a quantitative estimate on energy scaling for prestrained thin films in the limit of vanishing thickness. This helps physical models understand how elastic energy behaves when things become incredibly thin.

Ada: The real implication is that we now have a tool to predict the behavior of these systems in a much wider range of geometric settings than before, specifically covering those codimensions k between one and d times d plus one minus d plus one where general results were missing.

Kolya: It’s about moving from specific cases to a unified framework, which is a big step in understanding the limits of what we can approximate geometrically.

MARTA LEWICKA, HUI LI

math.AP

Submitted: 2026-07-17

Updated: 2026-11-16

Comments: 32 pages. The paper is now self-contained, whereas previously it relied on an external result concerning bounds for the normal frame which was partially erroneous, as it claimed that the constant in Corollary 2.7 was independent of the immersion. New version: we now added Example 2.8 providing an explicit construction showing that this constant cannot, in general, be independent of the immersion

License: http://creativecommons.org/licenses/by/4.0/

Importance score: 89/100

The gist: Flexibility of the isometric immersion system in arbitrary dimension and codimension and the energy scaling of prestrained thin films proves that every short immersion into Euclidean space can be…

Key concepts

Isometric Immersion
This is a mathematical concept where you map a surface or manifold from one space (like a curved domain) into another space (like Euclidean space) while preserving all distances and angles. In this context, it means finding a way to embed the geometry of the original object perfectly into the target space without stretching or distorting it.
Regularity Exponent ($\alpha_0$)
This exponent defines how smooth or 'regular' an approximating immersion must be. The paper finds a specific threshold, $\alpha_0$, below which the approximation is possible. This exponent depends on the dimension ($d$) and codimension ($k$), setting a limit on how well you can approximate the complex immersion using simpler, smoother ones.
Flexibility Statement
This refers to the mathematical principle that states whether a given geometric problem (like finding an isometric immersion) has many solutions or only one. The paper provides a 'unified' statement, meaning it proves flexibility—that approximations exist—for a wide range of dimensions and codimensions, which was previously unknown.

Terminology

Summary

Flexibility of the isometric immersion system in arbitrary dimension and codimension and the energy scaling of prestrained thin films proves that every short immersion into Euclidean space can be uniformly approximated by exact isometric immersions with a specific regularity exponent, which provides a unified flexibility statement for arbitrary dimensions and codimensions

Main Theorem Statement

The central finding is the proof of Theorem 1.1, which states that for a given C r,β-regular Riemann metric posed on a d-dimensional domain, every short immersion into the Euclidean space R d+k can be uniformly approximated by exact isometric immersions of regularity C 1,α for any α < α0 = min r+β 2,1 1+d(d+1)/k The novelty lies in providing a unified flexibility statement for arbitrary dimensions d and codimensions k, treating the uncharted range k ∈ (1,d(d+1) 2 − d + 1) with the exponent α0 agreeing with that obtained in [27] for the Monge-Amp’ere system

Flexibility Exponent Analysis

The paper establishes a new threshold flexibility exponent α0 = 1/(1+2d∗/k) for the system (1.1) This exponent is shown to recover several known results as special cases, such as the result obtained in [35, 36] when k = 1. p. 1

Where the paper says it

Our theorem recovers several previously known results as special cases. When k = 1, it yields the exponent obtained in [12], although it does not reach the improved threshold in [20].

The theorem recovers the exponent obtained in [12] when k = 1 but does not reach the improved threshold in [20] When d = k, it includes the result of [8]. p. 1
Where the paper says it

Our theorem recovers several previously known results as special cases. When k = 1, it yields the exponent obtained in [12], although it does not reach the improved threshold in [20].

The novelty lies also in treating the so far uncharted range k ∈ (1,d(d+1) 2 − d + 1) with no corresponding general theorem previously available

Quantitative Immersability Estimate

The paper derives a new estimate in the quantitative immersability of thin non-Euclidean films, proving that inf E h g ≤ Ch 4α α+1 for all α in the range (1.2) For d = 2, k = 1, this implies inf E h g ∼ h θ where 1 ≤ θ < 2. p. 2

Where the paper says it

Since the current state of the art sets the flexibility exponent at α0 = 1 3 in virtue of [5], resulting in 4α0 α0+1 = 1, Theorem 1. 2 consequently implies the following uncharted scaling regime in which questions on the asymptotics and the arising types of singularities should be studied: inf E h g ∼ h θ where 1 ≤ θ < 2.

Conversely, for d = 2, k = 2, flexibility holds up to α0 = 1 in virtue of [30], resulting in inf E h g ≤ Chθ for all θ < 2

Technical Construction and Proof Strategy

The proof of Theorem 1.3 relies on a stage construction, whose iteration via the Nash-Kuiper algorithm yields Theorem 1. p. 2

Where the paper says it

Our result relies on the stage construction, whose iteration via the Nash-Kuiper algorithm in Theorem 1. 5 yields Theorem 1.

1. This construction involves N = lcm(d∗, k) steps where each step applies the Kuiper corrugation to replace one rank-one primitive defect of the defect D by higher order error terms The fundamental blow-up/decay ratio obtained is J/S = d∗/k

Energy Scaling Proof Details

The proof of Theorem 1.2 involves using the Kirchhoff-Love extension adapted to the metric g, showing that E h g(u h) ≤ Ch 4α α+1 for all h ≪ 1 This is achieved by relating the energy density W to the ambient metric and using a choice of t = 1/(α+1) such that 2tα = 1 + t(α − 1)

Normal Frame Propagation

The existence and bounds for the normal frame are established through Lemma 2.6, which shows that for any C 1 immersion u satisfying (2. p. 1

Where the paper says it

Let g ∈ Cr,β(¯ω, R d×d sym,>) be defined on the closure of an open set ω ⊂ R d , diffeomorphic to B1, for some r+β > 0. Then, for every immersion u ∈ C1 (¯ω, R d+k ) satisfying: (∇u) T ∇u < g in ω, ¯ for every ϵ > 0 and for every regularity exponent α in: 0 < α < min nr + β 2 , 1 1 + 2d∗/k o where d∗ = d(d + 1) 2 , (1.

3), there exists a normal frame with gradient bounds whose constants depend only on γ and the modulus of continuity of ∇u Corollary 2. p. 6
Where the paper says it

We also present several auxiliary lemmas concerning the propagation of the normal frame to the immersions generated in the inductive course of the stage. As a byproduct of our estimates, we deduce existence of the normal frame to any C 1 -regular immersion, with its derivatives obeying a natural bound with multiplicative constants that depend only on the immersability parameter and the modulus of continuity of the tangent space.

7 concludes that this leads to the final result by showing that ∇(m)E i u0 ≤ N X−1 j=0 ∇(m) (E i u j+1 − E i u j)0 ≤ CN∇um for all m = 1. p. 5
Where the paper says it

The same statement as in Theorem 1. 3 also holds and one concludes the final result by a version of Theorem 1.

.. n

Literature Comparison

The work contributes to the growing body of work on the flexibility of C 1,α isometric immersions and solutions to related PDEs The results build upon classical Nash-Kuiper theorem [35, 26] for codimension k = 1 and the study of H¨older-regular isometric immersions by Borisov The paper addresses the gap between results at k = 1 and k = d(d+1) 2 − d + 1

Appendix Details

Appendix A proves Lemma 2.3 by calculating the gradient of the modified immersion, showing that (∇u˜)T∇u˜ − (∇u)T∇u is equal to 2Γ¯′ + (Γ′)2 a squared η ⊗ η - 2Γλa⟨∂ ij u, Eu⟩ i=1. p. 6

Where the paper says it

4, respectively, yields our main Theorem 1. The modified immersion ˜u produced in a single stage is obtained through N = lcm(d∗, k) steps (i.

.. d + 4Γ + 2Γ Γ¯' + 2Γ¯' λa sym∇a ⊗ η + 2Γ + Γ¯2 λ2∇a ⊗∇a Appendix B provides the proofs of propagation and existence of normal frame Lemmas The proof of Corollary 2.7 involves an inductive extension on i = 1. p. 6
Where the paper says it

Finally, in section 4 we provide the proof of Theorem 1. 2 by applying a suitable version of the Kirchhoff-Love extension, formulated with respect to the ambient metric g, to the H¨older regular isometric immersions constructed in the preceding sections.

.. k, leading to the final bound ∇(m)E i u0 ≤ N X−1 j=0 ∇(m) (E i u j+1 − E i u j)0 ≤ CN∇um The proof of Theorem 1. p. 5
Where the paper says it

Third, using geometric arguments, the same rigidity has been proved by Borisov in a series of papers [1, 2], for C 1,α isometric immersions at α > 2 3 , of C 2 metrics, with a simpler analytic proof of this result obtained in [12]. In particular, we see that 2 3 is an upper bound on the range of H¨older exponents that can be reached using convex integration for isometric immersions of 2-dimensional metrics into R 2+1.

2 uses the polar decomposition theorem to show that the energy density is bounded by Ch 4α α+1

References

[1] Borisov, Y., The parallel translation on a smooth surface. p. 5

Where the paper says it

a composition of a rotation and a translation. Second, any C 2 surface with positive Gauss’s curvature must be locally convex; for generalizations of this statement we refer to [18], and to [37, Chapter II], [38, Chapter IX] where the requirement of the C 2 regularity is replaced by the requirement that the immersion is C 1 and that the measure on S 2 induced by the Gauss map has bounded variation.

I-IV., Vestnik Leningrad, 13(4):160-171, and 13(19):45–54, and 14(1):34–50, and 14(13):20–26

[2] Borisov, Y., C 1,α-isometric immersions of Riemannian spaces, Doklady Akad. p. 0

Where the paper says it

Introduction In this paper, we prove a new result in the theory of convex integration for the system of isometric immersions and, as an application, derive a new estimate in the quantitative immersability of thin non-Euclidean films. The central departing question is whether, for a given Riemannian metric g posed on a domain ω ⊂ R d , every short immersion into R d+k , where k ≥ 1 denotes the codimension, can be approximated by exact isometric immersions u: (∇u) T ∇u = g in ω ⊂ R d .

Nauk SSSR 163:869–871, (1965)

[3] Borisov, Y., Irregular C 1,β-surfaces with analytic metric, Sib. p. 5

Where the paper says it

The H¨older-regular isometric immersions were studied by Borisov [1, 2], who conjectured that the Nash-Kuiper theorem extended to C 1,α-regularity for α < 1 1+d 2+d and α < 1 5 for d = 2. He provided a proof for the exponent α < 1 7 , dimension d = 2 and the analytic metric g in [3].

Mat. Zh. 45(1):25-–61, (2004)

[4] Cao, W., Hirsch, J. p. 5

Where the paper says it

These results were generalized to compact manifolds by Cao and Sz´ekelyhidi in [10]. Locally, it has been recently improved to α < 1 d 2−d+1 by Cao, Hirsch and Inauen in [5] and to α < 1 2d−1 by Inauen in [20].

p. 5
Where the paper says it

These results were generalized to compact manifolds by Cao and Sz´ekelyhidi in [10]. Locally, it has been recently improved to α < 1 d 2−d+1 by Cao, Hirsch and Inauen in [5] and to α < 1 2d−1 by Inauen in [20].

p. 5
Where the paper says it

These results were generalized to compact manifolds by Cao and Sz´ekelyhidi in [10]. Locally, it has been recently improved to α < 1 d 2−d+1 by Cao, Hirsch and Inauen in [5] and to α < 1 2d−1 by Inauen in [20].

and Inauen, D. p. 0
Where the paper says it

A number of partial results support this intuition. In codimension k = 1, the best result currently available is due to Inauen [20] and it is set at α0 = 1 2d−1 .

p. 0
Where the paper says it

A number of partial results support this intuition. In codimension k = 1, the best result currently available is due to Inauen [20] and it is set at α0 = 1 2d−1 .

, C 1,1/3− very weak solutions to the two dimensional Monge-Amp’ere equation, Calc. p. 5
Where the paper says it

We point out that no such rigidity statement is possible for the isometric immersions already into R 4 , in view of [30]. The parallel version of the Nash-Kuiper scheme as in the papers cited above, was first developed to handle flexibility of C 1,α weak solutions to the Monge-Amp`ere equation by Lewicka and Pakzad in [34], where the result was proved for α < 1 7 .

Var., 64:160, (2025)

[5] Cao, W., Hirsch, J. and Inauen, D., A Nash-Kuiper theorem for isometric immersions beyond Borisov’s exponent, arXiv:2503. p. 5

Where the paper says it

The H¨older-regular isometric immersions were studied by Borisov [1, 2], who conjectured that the Nash-Kuiper theorem extended to C 1,α-regularity for α < 1 1+d 2+d and α < 1 5 for d = 2. He provided a proof for the exponent α < 1 7 , dimension d = 2 and the analytic metric g in [3].

13867

[6] Cao, W., Hirsch, J., Inauen, D. p. 0

Where the paper says it

A number of partial results support this intuition. In codimension k = 1, the best result currently available is due to Inauen [20] and it is set at α0 = 1 2d−1 .

p. 0
Where the paper says it

A number of partial results support this intuition. In codimension k = 1, the best result currently available is due to Inauen [20] and it is set at α0 = 1 2d−1 .

and Lewicka, M. p. 0
Where the paper says it

FLEXIBILITY OF THE ISOMETRIC IMMERSION SYSTEM IN ARBITRARY DIMENSION AND CODIMENSION AND THE ENERGY SCALING OF PRESTRAINED THIN FILMS MARTA LEWICKA AND HUI LI Abstract. We prove that, for a given C r,β-regular Riemann metric posed on a d-dimensional domain, every short immersion into the Euclidean space R d+k , can be uniformly approximated by exact isometric immersions of regularity C 1,α for any α < α0 = min{ r+β 2 , 1 1+d(d+1)/k }.

p. 0
Where the paper says it

FLEXIBILITY OF THE ISOMETRIC IMMERSION SYSTEM IN ARBITRARY DIMENSION AND CODIMENSION AND THE ENERGY SCALING OF PRESTRAINED THIN FILMS MARTA LEWICKA AND HUI LI Abstract. We prove that, for a given C r,β-regular Riemann metric posed on a d-dimensional domain, every short immersion into the Euclidean space R d+k , can be uniformly approximated by exact isometric immersions of regularity C 1,α for any α < α0 = min{ r+β 2 , 1 1+d(d+1)/k }.

p. 0
Where the paper says it

FLEXIBILITY OF THE ISOMETRIC IMMERSION SYSTEM IN ARBITRARY DIMENSION AND CODIMENSION AND THE ENERGY SCALING OF PRESTRAINED THIN FILMS MARTA LEWICKA AND HUI LI Abstract. We prove that, for a given C r,β-regular Riemann metric posed on a d-dimensional domain, every short immersion into the Euclidean space R d+k , can be uniformly approximated by exact isometric immersions of regularity C 1,α for any α < α0 = min{ r+β 2 , 1 1+d(d+1)/k }.

, Full flexibility of the Monge-Amp’ere system in codimension d∗ − d + 1, arXiv:2603. p. 0
Where the paper says it

up to α0 = 1. This full flexibility phenomenon was subsequently extended in the context of the closely related Monge-Amp`ere system in [6], for the codimension k = d 2−d+2 2 .

28909 (2026)

[7] Cao W. and Inauen D. p. 0

Where the paper says it

A number of partial results support this intuition. In codimension k = 1, the best result currently available is due to Inauen [20] and it is set at α0 = 1 2d−1 .

p. 0
Where the paper says it

A number of partial results support this intuition. In codimension k = 1, the best result currently available is due to Inauen [20] and it is set at α0 = 1 2d−1 .

, Rigidity and flexibility of isometric extension. p. 5
Where the paper says it

We point out that no such rigidity statement is possible for the isometric immersions already into R 4 , in view of [30]. The parallel version of the Nash-Kuiper scheme as in the papers cited above, was first developed to handle flexibility of C 1,α weak solutions to the Monge-Amp`ere equation by Lewicka and Pakzad in [34], where the result was proved for α < 1 7 .

Comment. Math. Helv., 99(1):39–80, (2024)

[8] Cao W. and Szekelyhidi Jr. p. 0

Where the paper says it

1) of regularity C 1 . The analogous flexibility result in the H¨older classes C 1,α, for every α < 1 1+d(d+1) , was later provided by Conti, De Lellis and Szekelyhidi [12].

p. 0
Where the paper says it

1) of regularity C 1 . The analogous flexibility result in the H¨older classes C 1,α, for every α < 1 1+d(d+1) , was later provided by Conti, De Lellis and Szekelyhidi [12].

p. 0
Where the paper says it

1) of regularity C 1 . The analogous flexibility result in the H¨older classes C 1,α, for every α < 1 1+d(d+1) , was later provided by Conti, De Lellis and Szekelyhidi [12].

, L. ´, Very weak solutions to the two-dimensional Monge-Amp’ere equation, Sci. p. 5
Where the paper says it

We point out that no such rigidity statement is possible for the isometric immersions already into R 4 , in view of [30]. The parallel version of the Nash-Kuiper scheme as in the papers cited above, was first developed to handle flexibility of C 1,α weak solutions to the Monge-Amp`ere equation by Lewicka and Pakzad in [34], where the result was proved for α < 1 7 .

China Math. p. 0
Where the paper says it

1 arXiv:2607. 15838v2 [math.

, 62(6): 1041–1056, (2019)

[9] Cao W. and Szekelyhidi, Jr. p. 0

Where the paper says it

1) of regularity C 1 . The analogous flexibility result in the H¨older classes C 1,α, for every α < 1 1+d(d+1) , was later provided by Conti, De Lellis and Szekelyhidi [12].

p. 0
Where the paper says it

1) of regularity C 1 . The analogous flexibility result in the H¨older classes C 1,α, for every α < 1 1+d(d+1) , was later provided by Conti, De Lellis and Szekelyhidi [12].

, L. ´, On the isometric version of Whitney’s strong embedding theorem, Adv. p. 6
Where the paper says it

Finally, in section 4 we provide the proof of Theorem 1. 2 by applying a suitable version of the Kirchhoff-Love extension, formulated with respect to the ambient metric g, to the H¨older regular isometric immersions constructed in the preceding sections.

Math., 460:110040, (2025)

[10] Cao W. and Szekelyhidi, Jr., L. ´, Global Nash-Kuiper theorem for compact manifolds, J. p. 5

Where the paper says it

Borisov’s conjecture was confirmed by Conti, De Lellis and Sz´ekelyhidi in [12], whereas the case α < 1 5 , d = 2 was resolved by De Lellis, Inauen and Sz´ekelyhidi in [15]. These results were generalized to compact manifolds by Cao and Sz´ekelyhidi in [10].

Differential Geom. 122(1):35–68, (2022)

[11] Cohn-Vossen, S., Zwei S¨atze ¨uber die Starrheit der Eifl¨achen, Nachrichten G¨ottingen, pp. 125–137, (1927)

[12] Conti, S., De Lellis, C. p. 0

Where the paper says it

1) of regularity C 1 . The analogous flexibility result in the H¨older classes C 1,α, for every α < 1 1+d(d+1) , was later provided by Conti, De Lellis and Szekelyhidi [12].

and Szekelyhidi Jr., L. ´, h-principle and rigidity for C 1,α isometric embeddings, Proceedings of the Abel Symposium, in: Nonlinear Partial Differential Equations, pp. p. 5
Where the paper says it

1) with the metric g ∈ Cr,β admits an approximating sequence of C 1,α isometric immersions, for all α < min{ r+β 2 , 1}. When d = 2, k = 1, we quote the following rigidity results.

83–116, (2010)

[13] De Lellis, C., The masterpieces of John Forbes Nash Jr, in: H. p. 0

Where the paper says it

for u : ω → R d+k , g : ω → R d×d sym,>. 1) The seminal works by Nash [35, 36] and Kuiper [26] established that, already in codimension k = 1, any short immersion in the sense that (∇u) T ∇u < g on the chart, is the uniform limit of solutions to (1.

Holden and R. Piene (editors), The Abel Prize 2013–2017, Springer Verlag (2017)

[14] De Lellis C. and Inauen D., C 1,α isometric embeddings of polar caps. p. 0

Where the paper says it

FLEXIBILITY OF THE ISOMETRIC IMMERSION SYSTEM IN ARBITRARY DIMENSION AND CODIMENSION AND THE ENERGY SCALING OF PRESTRAINED THIN FILMS MARTA LEWICKA AND HUI LI Abstract. We prove that, for a given C r,β-regular Riemann metric posed on a d-dimensional domain, every short immersion into the Euclidean space R d+k , can be uniformly approximated by exact isometric immersions of regularity C 1,α for any α < α0 = min{ r+β 2 , 1 1+d(d+1)/k }.

Adv. Math., 363:106996, (2020)

[15] De Lellis, C., Inauen, D. and Szekelyhidi Jr., L. ´, A Nash-Kuiper theorem for C 1, 1/5 −δ immersions of surfaces in 3 dimensions, Revista Matematica Iberoamericana, 34(3):1119–1152, (2018)

[16] Gromov M., Partial differential relations, Volume 9 of “Ergebnisse der Mathematik und ihrer Grenzgebiete” (3) [Results in Mathematics and Related Areas (3)]. p. 0

Where the paper says it

Since then, considerable effort has been devoted to determining the maximal H¨older exponent α0 for which flexibility persists and it is natural to expect that α0 should increase with k. A number of partial results support this intuition.

Springer-Verlag, Berlin, (1986)

[17] Gunther M. ¨, Isometric embeddings of Riemannian manifolds. p. 0

Where the paper says it

Introduction In this paper, we prove a new result in the theory of convex integration for the system of isometric immersions and, as an application, derive a new estimate in the quantitative immersability of thin non-Euclidean films. The central departing question is whether, for a given Riemannian metric g posed on a domain ω ⊂ R d , every short immersion into R d+k , where k ≥ 1 denotes the codimension, can be approximated by exact isometric immersions u: (∇u) T ∇u = g in ω ⊂ R d .

In “Proceedings of the International Congress of Mathematicians”, Vol. I, II (Kyoto, 1990), pp. 1137—1143, Math. p. 0
Where the paper says it

1 arXiv:2607. 15838v2 [math.

Soc. Japan, Tokyo, (1991)

[18] Hartmann, P. and Nirenberg, L., On spherical image maps whose Jacobians do not change sign, Amer. p. 6

Where the paper says it

If d = k, the symmetric part of ∇f is: sym∇f = (∇f + (∇f) T )/2. By C we denote a universal constant that may change from line to line.

J. Math. 81, pp. 901–920, (1959)

[19] Herglotz, G., Uber die Starrheit der Eifl¨achen ¨, Abh. Math. Semin. Hansische Univ., 15, pp. 127–129, (1943)

[20] Inauen D., Flexibility of codimension one C 1,θ isometric immersions arXiv:2603. p. 5

Where the paper says it

Locally, it has been recently improved to α < 1 d 2−d+1 by Cao, Hirsch and Inauen in [5] and to α < 1 2d−1 by Inauen in [20]. If the codimension k is larger than 1, one can construct more regular isometric immersions.

08382 (2026)

[21] Inauen, D. and Lewicka, M., The Monge-Amp’ere system in dimension two and codimension three, to appear in Revista Matematica Iberoamericana, arXiv:2501. p. 6

Where the paper says it

ISOMETRIC IMMERSION SYSTEM IN ARBITRARY DIMENSION AND CODIMENSION 7 on the value of k, were obtained in the sequence of papers [28, 29, 21]. In [22], Inauen and Lewicka established the full flexibility of the Monge-Amp´ere system for d = k = 2, and in [6] their result was extended to arbitrary d and the corresponding k = d∗ − d + 1.

12474 (2025)

[22] Inauen, D. and Lewicka, M., The Monge-Amp´ere system in dimension two is fully flexible in codimension two, 2504. p. 6

Where the paper says it

ISOMETRIC IMMERSION SYSTEM IN ARBITRARY DIMENSION AND CODIMENSION 7 on the value of k, were obtained in the sequence of papers [28, 29, 21]. In [22], Inauen and Lewicka established the full flexibility of the Monge-Amp´ere system for d = k = 2, and in [6] their result was extended to arbitrary d and the corresponding k = d∗ − d + 1.

03582 (2025)

[30] Lewicka, M., Full flexibility of isometric immersions of metrics with low H¨older regularity in Poznyak theorem’s dimension, arXiv:2511. p. 5

Where the paper says it

The H¨older-regular isometric immersions were studied by Borisov [1, 2], who conjectured that the Nash-Kuiper theorem extended to C 1,α-regularity for α < 1 1+d 2+d and α < 1 5 for d = 2. He provided a proof for the exponent α < 1 7 , dimension d = 2 and the analytic metric g in [3].

16305

[31] Lewicka, M., Calculus of Variations on Thin Prestressed Films: Asymptotic Methods in Elasticity, Progress in Nonlinear Differential Equations and Their Applications, 101, Birkh¨auser, (2022)

[34] Lewicka, M. and Pakzad, M. p. 5

Where the paper says it

We point out that no such rigidity statement is possible for the isometric immersions already into R 4 , in view of [30]. The parallel version of the Nash-Kuiper scheme as in the papers cited above, was first developed to handle flexibility of C 1,α weak solutions to the Monge-Amp`ere equation by Lewicka and Pakzad in [34], where the result was proved for α < 1 7 .

, Convex integration for the Monge-Amp’ere equation in two dimensions, Analysis and PDE, 10(3):695–727, (2017)

[35] Nash, J., The imbedding problem for Riemannian manifolds, Ann. p. 2

Where the paper says it

4) parametrized by their “thickness” h ≪ 1. Consider the Riemannian metric g on Ω1 and pose the problem of minimizing the following energy functionals, as h → 0: E h g (u) = 1 h k ˆ Ωh W (∇u)g −1/2 det g 1/2 d(x, z) for all u ∈ H1 (Ωh , R d+k ).

Math., 63:20–63, (1956)

[36] Nash, J., C 1 isometric imbeddings, Ann. p. 0

Where the paper says it

FLEXIBILITY OF THE ISOMETRIC IMMERSION SYSTEM IN ARBITRARY DIMENSION AND CODIMENSION AND THE ENERGY SCALING OF PRESTRAINED THIN FILMS MARTA LEWICKA AND HUI LI Abstract. We prove that, for a given C r,β-regular Riemann metric posed on a d-dimensional domain, every short immersion into the Euclidean space R d+k , can be uniformly approximated by exact isometric immersions of regularity C 1,α for any α < α0 = min{ r+β 2 , 1 1+d(d+1)/k }.

Math., 60:383–396, (1954)

[40] Weyl, W., Uber die Bestimmung einer geschlossenen konvexen Fl¨ache durch ihr Linienelement ¨, Vierteljahrsschrift der naturforschenden Gesellschaft, Zurich, 61, pp. 40-72, (1916)

[41] Isett P., A proof of Onsager’s conjecture. p. 5

Where the paper says it

He provided a proof for the exponent α < 1 7 , dimension d = 2 and the analytic metric g in [3]. Borisov’s conjecture was confirmed by Conti, De Lellis and Sz´ekelyhidi in [12], whereas the case α < 1 5 , d = 2 was resolved by De Lellis, Inauen and Sz´ekelyhidi in [15].

p. 5
Where the paper says it

He provided a proof for the exponent α < 1 7 , dimension d = 2 and the analytic metric g in [3]. Borisov’s conjecture was confirmed by Conti, De Lellis and Sz´ekelyhidi in [12], whereas the case α < 1 5 , d = 2 was resolved by De Lellis, Inauen and Sz´ekelyhidi in [15].

p. 5
Where the paper says it

He provided a proof for the exponent α < 1 7 , dimension d = 2 and the analytic metric g in [3]. Borisov’s conjecture was confirmed by Conti, De Lellis and Sz´ekelyhidi in [12], whereas the case α < 1 5 , d = 2 was resolved by De Lellis, Inauen and Sz´ekelyhidi in [15].

p. 5
Where the paper says it

He provided a proof for the exponent α < 1 7 , dimension d = 2 and the analytic metric g in [3]. Borisov’s conjecture was confirmed by Conti, De Lellis and Sz´ekelyhidi in [12], whereas the case α < 1 5 , d = 2 was resolved by De Lellis, Inauen and Sz´ekelyhidi in [15].

p. 5
Where the paper says it

He provided a proof for the exponent α < 1 7 , dimension d = 2 and the analytic metric g in [3]. Borisov’s conjecture was confirmed by Conti, De Lellis and Sz´ekelyhidi in [12], whereas the case α < 1 5 , d = 2 was resolved by De Lellis, Inauen and Sz´ekelyhidi in [15].

Ann. of Math. (2), 188(3):871–963, (2018)

[42] Jacobowitz, H., Implicit function theorems and isometric embeddings, Annals of Mathematics, 95(2):191–225, (1972)

[43] Isett P., A proof of Onsager’s conjecture. Ann. of Math. (2), 188(3):871–963, (2018)

[44] Jacobowitz, H., Implicit function theorems and isometric embeddings, Annals of Mathematics, 95(2):191–225, (1972)

[45] Isett P., A proof of Onsager’s conjecture. Ann. of Math. (2), 188(3):871–963, (2018)

[46] Jacobowitz, H., Implicit function theorems and isometric embeddings, Annals of Mathematics, 95(2):191–225, (1972)

[47] Isett P., A proof of Onsager’s conjecture. Ann. of Math. (2), 188(3):871–963, (2018)

[40] Weyl, W., Uber die Bestimmung einer geschlossenen konvexen Fl¨ache durch ihr Linienelement ¨, Vierteljahrsschrift der naturforschenden Gesellschaft, Zurich, 61, pp. 40-72, (1916)

[48] Isett P., A proof of Onsager’s conjecture. Ann. of Math. (2), 188(3):871–963, (2607. p. 0

Where the paper says it

was partially supported by NSF grant DMS-2006439. 1 arXiv:2607.

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Improved systems can uniformly approximate short immersions into Euclidean space R d+k by exact isometric immersions of regularity C 1, alpha for any alpha less than a specific threshold α0 = min(r+beta/2, 1/(1+d(d+1)/k)).

  2. The system can provide a unified flexibility statement for arbitrary dimensions d and codimensions k, including the uncharted range k in (1, d(d+1)/2 - d + 1) excluding d.

  3. Quantitative estimates for thin non-Euclidean films can be derived, setting the scaling exponent of the infimum of non-Euclidean energies at 4alpha0 alpha0+1 in the limit of vanishing film thickness.

Abstract

We prove that, for a given C r,β-regular Riemann metric posed on a d-dimensional domain, every short immersion into the Euclidean space R d+k, can be uniformly approximated by exact isometric immersions of regularity C 1,α for any α<α 0= r+β over 2, 1 over 1+d(d+1)/k. Our theorem recovers several previously known results as special cases. The novelty thereof lies in providing a unified flexibility statement for arbitrary dimensions d and codimensions k, while also treating the so far uncharted range k in (1, d(d+1) over 2-d+1) d, where no corresponding general result was previously available. Our threshold flexibility exponent α 0 agrees with that previously obtained for the closely related Monge-Ampère system. As an application, we prove a new estimate in the quantitative immersability of thin prestrained films, setting the scaling exponent of the infimum of non-Euclidean energies in presence of an arbitrary prestrain metric, and in the limit of the film's vanishing thickness, at 4α 0 over α 0+1.

Sources

Related papers