Existence of infinitely many homotopy classes from $\mathbb S^3$ to $\mathbb S^2$ having a minimimzing $W^{s,\frac 3s}$-harmonic map

Adam Grzela; Katarzyna Mazowiecka

REVIEW 2 major objections 5 minor 13 references

Existence of infinitely many homotopy classes from $\mathbb S^3$ to $\mathbb S^2$ having a minimimzing $W^{s,\frac 3s}$-harmonic map

T0 review · 2 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read The paper proves that for every $s\in(0,1)$ there exist infinitely many homotopy classes of $\pi_3(\mathbb S^2)$ containing a minimizing $W^{s,3/s}$-harmonic map.

desk verdict A sound, well-scoped extension of Riviere's theorem to all s in (0,1); the main estimate is new and checks out, with only minor typos and a qualified optimality claim. read the letter →

arxiv 2506.05060 v1 pith:OVPIDSTX submitted 2025-06-05 math.AP

Adam Grzela , Katarzyna Mazowiecka This is my paper

classification math.AP MSC 58E2035B6535J6035S05

keywords fractionalharmonicmapsHopfinvarianthomotopyclassesminimizingSobolev-SlobodeckijspacesW^{sp}energyco-areaformulaidentity

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 extends Rivière's 1998 theorem to the fractional setting: for every $s\in(0,1)$, there exist infinitely many homotopy classes of maps from $\mathbb S^3$ to $\mathbb S^2$ that contain a minimizer of the fractional energy $W^{s,3/s}$. The proof establishes a sublinear upper bound on the infimum of this energy over maps of Hopf degree $d$, namely $\#_s^d \le C(s)|d|^{3/4}$, and combines it with an energy identity that decomposes any homotopy class into finitely many classes whose energies add and are attained. If correct, the result shows that the fractional Hopf-degree estimate of Schikorra and Van Schaftingen is optimal in the case of maps $\mathbb S^3\to\mathbb S^2$, and it removes the earlier restriction $s>\frac34$ that came from Sobolev embedding.

What carries the argument

The Hopf fibration $h:\mathbb S^3\to\mathbb S^2$ and the co-area formula reduce the fractional energy of a composition $v\circ h$ on $\mathbb S^3$ to a fractional energy of $v$ on $\mathbb S^2$ with the same exponent, effectively lowering the dimension by one. The proof combines this with a construction of degree-$k$ maps $v$ on $\mathbb S^2$ supported on $k$ disjoint geodesic balls of radius $\lambda/\sqrt{k}$, whose localized fractional energy is roughly $k^{sp/2-1}$ per ball and sums to $k^{sp/2}=|d|^{3/4}$ when $sp=3$; an 'opening' lemma that makes a near-minimizer locally constant along a fiber; and a gluing lemma that inserts $d-k^2$ disjoint Hopf-degree-one bubbles of constant energy, turning the degree-$k^2$ bound into a bound for arbitrary $d$.

What would settle it

Compute or bound $\#_s^d$ for a sequence of degrees $d_j\to\pm\infty$ and find, for some $s\in(0,1)$, that $\#_s^{d_j}\ge c|d_j|$ with $c>0$; or exhibit a nonzero homotopy class in $\pi_3(\mathbb S^2)$ for which the energy identity of Theorem 1.3 does not hold. Either would break the contradiction argument that yields Theorem 1.4.

Watch

Extended reading notes

Core claim

The central claim is Theorem 1.4: for every $s\in(0,1)$ there are infinitely many Hopf degrees $d$ for which the infimum $\#_s^d$ of the critical fractional energy $E^{s,3/s}$ among maps $\mathbb S^3\to\mathbb S^2$ of Hopf degree $d$ is attained by some map. The engine is Theorem 3.1, which proves $\#_s^d \le C(s)|d|^{3/4}$ by adapting Rivière's construction: compose degree-$k$ maps on $\mathbb S^2$ with the Hopf fibration and use the co-area formula to show that the fractional energy on $\mathbb S^3$ of the composition is bounded by the corresponding energy on $\mathbb S^2$, then glue in small Hopf-degree-one bubbles to reach arbitrary $d$. With the sublinear bound in hand, the contradiction argument of Rivière carries over: if only finitely many classes had minimizers, the energy identity would force $\#_s^d$ to grow at least linearly in $|d|$, contradicting the sublinear estimate.

Load-bearing premise

The argument relies on the previously established energy identity (Theorem 1.3, cited from [6]), which asserts that every nonzero homotopy class splits into finitely many classes whose energies add and are attained; if that identity has unstated restrictions for small $s$ or for $\pi_3(\mathbb S^2)$, the main theorem would not follow.

Editorial extensions

If this is right

For every $s\in(0,1)$, the energy identity of Mazowiecka and Schikorra plus the sublinear bound yields infinitely many minimizing homotopy classes, removing the earlier restriction $s>\frac34$ obtained by Sobolev embedding.
The exponent $\frac43$ in the fractional Hopf-degree estimate is optimal for maps $\mathbb S^3\to\mathbb S^2$: the upper bound in Theorem 3.1 matches the linear lower bound in $|\deg_H f|^{4/3}$ up to universal constants.
The result sharpens the contrast with the equi-dimensional target $\mathbb S^3$, where only the degrees $-1,0,1$ admit minimizers; the infinitely many minimizers are a phenomenon specific to the Hopf configuration from $\mathbb S^3$ into $\mathbb S^2$.
The contradiction argument uses only the sublinear growth of $\#_s^d$ and the energy identity; the precise value of the exponent is irrelevant as long as it stays below $1$, so the same strategy could apply to other fractional energy exponents.

Reading between the lines

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

The bubble-gluing scheme may extend to the higher Hopf fibrations $\mathbb S^{4m-1}\to\mathbb S^{2m}$ for $m\ge 2$, provided the fiber-integration estimate replacing inequality (3.8) yields the same exponent reduction; that would give sublinear bounds and infinitely many minimizers in those homotopy groups as well.
If the energy identity of Theorem 1.3 were ever shown to fail for some small $s\in(0,1)$, the main theorem could still hold through a different decomposition, since the sublinear bound itself is established independently by construction.
A direct numerical computation of $\#_s^d$ for small degrees and several values of $s$ could reveal growth close to $|d|^{3/4}$, providing independent evidence for the optimality of the exponent and the plausibility of the energy identity.

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.

Desk Editor's Note

read the letter

Dear colleague,

The paper closes the fractional gap: for every s in (0,1) there are infinitely many homotopy classes in pi_3(S^2) with minimizing W^{s,3/s}-harmonic maps. The genuinely new input is Theorem 3.1, #_d^s ≤ C(s)|d|^{3/4} for all s, where previous work only reached s > 3/4 via Sobolev embedding. The proof adapts Riviere's 1998 construction to the nonlocal setting, and the co-area/Cavalieri step that integrates over Hopf fibers is correct.

The main proof is sound. The contradiction is exactly Riviere's, and the energy identity from [6] is used as a black box. That is legitimate: the published theorem in [6] explicitly covers s in (0,1) and l ≥ 2, so pi_3(S^2) falls under its hypotheses. The stress-test worry about a hidden restriction does not land; the statement in this paper quotes the full range.

Two soft spots. First, the displayed '0 < d-k^2 < 2k' in Step 2 is false for d=(k+1)^2. The subsequent bound d-k^2 ≤ 2√d is true, so the estimate (d-k^2) ≲ d^{1/2} survives. A typo, not a gap. Second, the introduction claims Theorem 3.1 proves optimality of the exponent in (1.10) for m=1. That only follows where the lower bound from (1.10) is known, i.e. s ≥ 3/4. For s < 3/4 the lower bound is open, so the claim should be qualified or dropped. The main theorem does not depend on it.

The appendix bubble construction, with degree-one maps supported in balls of radius r and W^{s,3/s}-energy independent of r, is correct and neatly done.

The paper is for specialists in fractional harmonic maps and geometric analysis. It is a well-scoped advance that removes a real obstacle. I would send it to a serious referee; with the typo fixed and the optimality remark qualified, it should be accepted.

Referee Report

2 major / 5 minor

Summary. The paper proves Theorem 1.4: for every s in (0,1), there exist infinitely many homotopy classes in pi_3(S^2) admitting a minimizing W^{s,3/s}(S^3,S^2)-harmonic map. The proof follows Riviere's scheme: Theorem 3.1 establishes the sublinear upper bound #_s^d <= C(s)|d|^{3/4} for the Hopf-degree-d fractional energy, using the Hopf fibration, co-area and Cavalieri estimates to reduce the kernel exponent, and then adding localized Hopf bubbles of controlled energy. This upper bound is combined with the energy decomposition identity stated as Theorem 1.3 (quoted from [6, Theorem 3.2]) to contradict the assumption that only finitely many homotopy classes have attained energy infima.

Significance. If correct, the result extends Riviere's 1998 theorem to the full fractional range s in (0,1) and shows optimality of the exponent in the fractional Hopf-degree estimate (1.10) for m=1. The core new ingredients in Section 3 are explicit and elegant: the co-area/Cavalieri reduction from S^3 to S^2 and the scale-invariant localized bubble construction. The paper is largely self-contained except for its reliance on the cited energy identity [6, Theorem 3.2]; if that identity applies as stated, the contradiction argument is sound.

major comments (2)

[Section 1, Theorem 1.3; Section 3, proof of Theorem 1.4, Eq. (3.22)] The main theorem rests entirely on the cited energy decomposition Theorem 1.3 ([6, Theorem 3.2]). In particular, condition (3), that each #_s^{alpha_i} is attained, is what forces the components d_{a_i} of the decomposition to belong to the assumed finite set of minimizer classes. Please verify explicitly and state in the paper that the hypotheses of [6, Theorem 3.2] hold for all s in (0,1), for (n,l)=(3,2) and p=3/s, and that "attained" means there is a genuine minimizer in the homotopy class. If [6] has any restriction (for example s >= s0 > 0 or a relaxed notion of attainment), Theorem 1.4 would not follow for the affected values of s.
[Section 3, proof of Theorem 1.4, paragraph beginning 'Suppose only finitely many...'] As written, the finite set of homotopy classes containing minimizers includes the degree-zero class, since constant maps are minimizers in their class. Then the smallest energy #_s^{d0} is 0, and the lower bound (3.22) becomes vacuous. The argument should be restricted to nonzero homotopy classes, taking d0 to be a minimizer class of smallest positive energy among the finitely many nonzero minimizer classes. This is fixable, but as written the contradiction does not follow.

minor comments (5)

[Section 3, Step 2, paragraph after choosing k] The displayed inequality 0 < d-k^2 < 2k is false for perfect squares d=(k+1)^2, where d-k^2 = 2k+1. The subsequent estimate only needs d-k^2 <= 3 sqrt(d), so the line should be replaced by d-k^2 <= 2k+1 <= 3 sqrt(d).
[Section 3, Eq. (3.20) and Corollary A.2] The text says E^{s,p}(g_{x_i,r}, S^3) = E, but Corollary A.2 only gives E^{s,p}(g_{x_i,r}, S^3) <= E; the equality should be an inequality.
[Title and Abstract] The title contains the typo 'MINIMIMZING'; this should be corrected to 'MINIMIZING'.
[Section 2, Lemma 2.2] Lemma 2.2 is stated for A subset R^m but is applied on the spheres S^m; the authors should note that the estimate extends to geodesic balls by local coordinates or cite a manifold version.
[Section 3, Eq. (3.5)] The symbol S is reused for the translated fiber S = -y + h^{-1}(z1) in addition to the ambient sphere notation; a different symbol would avoid confusion.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the new upper bound in Theorem 3.1 is proved by explicit independent construction, and the cited energy identity [6] is a load-bearing but non-circular external input.

full rationale

The claimed derivation chain is: Theorem 3.1 supplies a sublinear upper bound #s_d ≤ C|d|^{3/4} for every s∈(0,1), and Theorem 1.3 supplies an energy decomposition d = Σ d_i with #s_d = Σ #s_{d_i} and each #s_{d_i} attained. The proof of Theorem 3.1 is self-contained: it builds test maps from the Hopf fibration, estimates the fractional energy by fiber integration, localizes with Lemmas 2.2–2.5, and uses the explicit degree-one maps of Appendix A. No parameter is fitted to data and no prediction is renamed from an input; the upper bound is derived directly from constructions. Theorem 1.3 is cited from Mazowiecka–Schikorra [6], which is a self-citation by one of the present authors, but it is a general theorem about energy decomposition and attainment for all s∈(0,1) with ℓ≥2; its stated assumptions do not include the target result, it is not a fitted quantity, and it does not by itself assert or imply the existence of infinitely many minimizer classes. The contradiction proof of Theorem 1.4 merely combines the independent sublinear bound with this external identity, exactly as in Rivière's classical argument. Even if the cited identity had hidden restrictions, that would be a correctness risk rather than circularity; the present paper does not define its objects in terms of the conclusion or smuggle the result in through a self-citation chain. Therefore no circular step can be exhibited from the text.

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

No data-fitting parameters appear; constants such as lambda, E, and C(s) are universal and do not vary with the topological degree, so they do not encode the |d|^{3/4} law. The main external load is Theorem 1.3 from [6], plus standard gluing and localization lemmas. No new entities are introduced.

assumptions (3)

standard math Energy identity and attainment for critical fractional harmonic maps in homotopy classes (Theorem 1.3, [6, Theorem 3.2]).
Load-bearing external result used in the final contradiction: every nonzero element of pi_n(S^l) splits into finitely many subclasses with additive energy, each attained. The present paper does not prove it.
standard math Localization lemma, gluing lemma, and opening lemma (Lemmas 2.2, 2.3, 2.4 and Corollary 2.5).
Cited estimates from Monteil-Van Schaftingen and Van Schaftingen used to localize energy, glue maps, and modify maps homotopically while controlling fractional energy.
standard math Hopf fibration properties: deg_H(h)=1 and |h^* omega_{S^2}|=4.
Classical facts needed for the co-area computation in Step 1 of the proof of Theorem 3.1.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Existence of infinitely many homotopy classes from $\mathbb S^3$ to $\mathbb S^2$ having a minimimzing $W^{s,\frac 3s}$-harmonic map." pith.science (2026). https://pith.science/paper/OVPIDSTX

@misc{pith2026250605060,
  author       = {Pith},
  title        = {Pith review of: Existence of infinitely many homotopy classes from $\mathbb S^3$ to $\mathbb S^2$ having a minimimzing $W^s,\frac 3s$-harmonic map},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/OVPIDSTX}},
  note         = {Machine review of arXiv:2506.05060}
}

abstract

In 1998 T. Rivi\`{e}re proved that there exist infinitely many homotopy classes of $\pi_3(\mathbb S^2)$ having a minimizing 3-harmonic map. This result is especially surprising taking into account that in $\pi_3(\mathbb S^3)$ there are only three homotopy classes (corresponding to the degrees $\{-1,0,1\}$) in which a minimizer exists. We extend this theorem in the framework of fractional harmonic maps and prove that for $s\in(0,1)$ there exist infinitely many homotopy classes of $\pi_{3}(\mathbb S^{2})$ in which there is a minimizing $W^{s,\frac{3}{s}}$-harmonic map.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

13 extracted references · 13 canonical work pages

[6]

Mazowiecka and A

K. Mazowiecka and A. Schikorra.s-stability forW s,n/s-harmonic maps in homotopy groups.Annales de l’Institut Henri Poincar´ e C, Analyse non lin´ eaire (2024), published online first. 3

work page 2024
[1]

Bott and L

R. Bott and L. W. Tu.Differential forms in algebraic topology, volume 82 ofGraduate Texts in Mathematics. Springer-Verlag, New York-Berlin, 1982. 2

work page 1982
[2]

Haj lasz

P. Haj lasz. Sobolev mappings, co-area formula and related topics. InProceedings on Analysis and Geometry (Russian) (Novosibirsk Akademgorodok, 1999), pages 227–254. Izdat. Ross. Akad. Nauk Sib. Otd. Inst. Mat., Novosibirsk, 2000. 6

work page 1999
[3]

Hardt and T

R. Hardt and T. Rivi` ere. Connecting rational homotopy type singularities.Acta Math., 200(1):15–83,

work page
[4]

Hatcher.Algebraic topology

A. Hatcher.Algebraic topology. Cambridge University Press, Cambridge, 2002. 4

work page 2002
[5]

Mazowiecka and A

K. Mazowiecka and A. Schikorra. MinimalW s, n s -harmonic maps in homotopy classes.J. Lond. Math. Soc. (2), 108(2):742–836, 2023. 3

work page 2023
[7]

Monteil and J

A. Monteil and J. Van Schaftingen. Uniform boundedness principles for sobolev maps into manifolds. Annales de l’Institut Henri Poincar´ e C, Analyse non lin´ eaire, 36(2):417–449, apr 2019. 5

work page 2019
[8]

Park and A

W. Park and A. Schikorra. Quantitative estimates for fractional Sobolev mappings in rational homo- topy groups.Nonlinear Anal., 235:Paper No. 113349, 16, 2023. 3 12 ADAM GRZELA AND KATARZYNA MAZOWIECKA

work page 2023

Show all 13 references

[9]

Rivi` ere

T. Rivi` ere. Minimizing fibrations andp-harmonic maps in homotopy classes fromS 3 intoS 2.Comm. Anal. Geom., 6(3):427–483, 1998. 2, 6, 7

1998
[10]

Sacks and K

J. Sacks and K. Uhlenbeck. The existence of minimal immersions of 2-spheres.Ann. of Math. (2), 113(1):1–24, 1981. 2

1981
[11]

Schikorra and J

A. Schikorra and J. Van Schaftingen. An estimate of the Hopf degree of fractional Sobolev mappings. Proc. Amer. Math. Soc., 148(7):2877–2891, 2020. 3, 4

2020
[12]

Van Schaftingen

J. Van Schaftingen. Estimates by gap potentials of free homotopy decompositions of critical Sobolev maps.Advances in Nonlinear Analysis, 9(1):1214–1250, 2020. 5

2020
[13]

J. H. C. Whitehead. An Expression of Hopf’s Invariant as an Integral.Proceedings of the National Academy of Sciences of the United States of America, 33(5):117–123, 1947. 1 (Adam Grzela)Institute of Mathematics, University of W arsaw, Banacha 2, 02-097 W arszawa, Poland Email ...

1947

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.

[1] [6]

Mazowiecka and A

K. Mazowiecka and A. Schikorra.s-stability forW s,n/s-harmonic maps in homotopy groups.Annales de l’Institut Henri Poincar´ e C, Analyse non lin´ eaire (2024), published online first. 3

work page 2024

[2] [1]

Bott and L

R. Bott and L. W. Tu.Differential forms in algebraic topology, volume 82 ofGraduate Texts in Mathematics. Springer-Verlag, New York-Berlin, 1982. 2

work page 1982

[3] [2]

Haj lasz

P. Haj lasz. Sobolev mappings, co-area formula and related topics. InProceedings on Analysis and Geometry (Russian) (Novosibirsk Akademgorodok, 1999), pages 227–254. Izdat. Ross. Akad. Nauk Sib. Otd. Inst. Mat., Novosibirsk, 2000. 6

work page 1999

[4] [3]

Hardt and T

R. Hardt and T. Rivi` ere. Connecting rational homotopy type singularities.Acta Math., 200(1):15–83,

work page

[5] [4]

Hatcher.Algebraic topology

A. Hatcher.Algebraic topology. Cambridge University Press, Cambridge, 2002. 4

work page 2002

[6] [5]

Mazowiecka and A

K. Mazowiecka and A. Schikorra. MinimalW s, n s -harmonic maps in homotopy classes.J. Lond. Math. Soc. (2), 108(2):742–836, 2023. 3

work page 2023

[7] [7]

Monteil and J

A. Monteil and J. Van Schaftingen. Uniform boundedness principles for sobolev maps into manifolds. Annales de l’Institut Henri Poincar´ e C, Analyse non lin´ eaire, 36(2):417–449, apr 2019. 5

work page 2019

[8] [8]

Park and A

W. Park and A. Schikorra. Quantitative estimates for fractional Sobolev mappings in rational homo- topy groups.Nonlinear Anal., 235:Paper No. 113349, 16, 2023. 3 12 ADAM GRZELA AND KATARZYNA MAZOWIECKA

work page 2023