Holonomy Asymptotics along Quartic Differential Rays

Weihan Ma

REVIEW 1 major objections 5 minor 17 references

Holonomy Asymptotics along Quartic Differential Rays

T0 review · 1 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read For every nontrivial closed curve, the holonomy of the $tq$-ray of $\mathrm{PSp}(4,\mathbb{R})$-Hitchin representations has all four singular values and eigenvalue moduli growing like $t^{1/4}$ times integrals of the local fourth roots of…

desk verdict A serious, mostly sound extension of the cubic-differential holonomy asymptotics to the quartic PSp(4,R) case; the one real gap is a missing structural lemma about saddle-connection representatives, which looks fixable. read the letter →

arxiv 2608.04729 v1 pith:3IRMOY3P submitted 2026-08-05 math.DG

Weihan Ma This is my paper · ORCID

classification math.DG MSC 53C0753C4357K20

keywords quarticdifferentialHitchincomponentholonomyasymptoticsStokesmatricessingularflatmetricsaddleconnectionsHiggsbundlesWeyl-chamberlengthcompactification

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 aims to prove a complete asymptotic spectral law for the holonomy of a ray of $\mathrm{PSp}(4,\mathbb{R})$-Hitchin representations generated by a holomorphic quartic differential $q$: for each $t>0$ the differential $tq$ defines a Hitchin representation, and the question is what happens to its holonomy along closed curves as $t\to+\infty$. The claimed answer is that the logarithm of every singular value and every eigenvalue modulus grows like $t^{1/4}$ times an explicit number built from the singular flat geometry of $q$: decompose the $|q|^{1/2}$-geodesic representative of the curve into saddle connections, integrate the four local fourth roots of $q$ along each, order the four outcomes, and sum the $k$-th largest over the curve. If correct, the entire spectral asymptotics of holonomy along curves is determined by flat-geometric data, with no restriction on the orders of the zeros of $q$, and the boundary point of the ray in the Weyl-chamber length compactification becomes explicit. The proof proceeds from the cubic-differential analogue [LTW26] but requires a genuinely new Stokes analysis near the zeros, because the relevant products of Stokes matrices have to be computed explicitly and shown to be non-nilpotent.

What carries the argument

The engine of the proof is the local analysis of the flat connection near a zero of $q$. Away from the zero set, known asymptotic decoupling results control parallel transport along straight segments. Near a zero of order $d$, rescaling by the ray turns the local problem into the planar model $q = z^d dz^4$, whose connection exhibits the Stokes phenomenon; the paper computes the relevant Stokes matrices $U_1, U_2, U_3$ explicitly from the radial asymptotics of the $\operatorname{tt}^*$-Toda equations supplied by [GIL26], obtaining constants $s_1 = \sin(4\theta)/\sin\theta$ and $s_2 = \sin(4\theta)\sin(3\theta)/(\sin\theta\sin 2\theta)$ with $\theta = \pi/(d+4)$. The holonomy block passing through a zero is then a product of these matrices with the diagonal asymptotics of the two adjacent saddle connections, encoded through the transition matrix $T = Q^{-1}U_2^{-1}U_1^{-1}$. The decisive algebraic fact is that the assembled products are non-nilpotent, proved from the explicit sequence $b_k = \sin((k-1)\theta)\sin(k\theta)\sin((k+1)\theta)/(\sin\theta\sin 2\theta\sin 3\theta)$ giving the powers $T^k$; the $k=2$ case needs a separate analysis on $\bigwedge^2\mathbb{R}^4$, while symplectic duality reduces $k=3,4$ to $k=1,2$.

What would settle it

On a surface whose flat metric contains a cylinder, take a free homotopy class whose $|q|^{1/2}$-geodesic is the cylinder core, which avoids $Z(q)$: the theorem's hypothesis is violated for this class, so testing whether a limiting form of the formula holds there would settle the scope of the claim. Independently, in the planar model $q = z^d dz^4$, build a closed curve from two radial segments joined by a large circular arc, compute its parallel transport with the explicit Stokes matrices of Proposition 3.13, and compare $t^{-1/4}\log\sigma_k$ with $\sum_i v_{i,k}$ at large $t$; any disagreement beyond numerical error would refute the block asymptotics (Propositions 5.4 and 5.6) on which the global theorem rests.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that the asymptotic spectral data of the holonomy along a curve are governed entirely by the singular flat metric $|q|^{1/2}$. For a nontrivial closed curve $\gamma$, let $c_\gamma = c_\ell \ast \cdots \ast c_1$ be its geodesic representative written as a concatenation of saddle connections, and let $\phi_k = (\sqrt{-1})^{1-k}q^{1/4}$, $k=1,2,3,4$, be the local fourth roots of $q$. For each saddle connection $c_i$, let $v_i = (v_{i,1}, v_{i,2}, v_{i,3}, v_{i,4})$ be the non-increasing rearrangement of $-2\operatorname{Re}\int_{c_i}\phi_k$. The main theorem states that for every $k = 1,2,3,4$, \[ \lim_{t\to+\infty}\frac{1}{$t^{{1/4}}$}\log\sigma_k(\mathrm{Hol}_t(c_\gamma)) = \sum_{i=1}^{\ell}v_{i,k}, \] with the identical limit for $t^{-1/4}\log|\lambda_k(\mathrm{Hol}_t(c_\gamma))|$; equivalently, $t^{-1/4}\log\|\bigwedge^k\mathrm{Hol}_t(c_\gamma)\|$ and the corresponding spectral-radius logarithm both converge to $\sum_i\sum_{j=1}^k v_{i,j}$. A direct corollary is that the ray $\rho_t$ converges in the Weyl-chamber length compactification [Par12] to the boundary point $(\infty,[L_\infty])$ with $L_\infty(\gamma)=\sum_i v_i$, and that the limiting length spectrum is uniformly comparable to the $|q|^{1/2}$-length spectrum, hence has positive systole.

Load-bearing premise

The whole proof rests on the assumption that the closed geodesic representative of every nontrivial free homotopy class, for the singular flat metric $|q|^{1/2}$, is a finite concatenation of saddle connections with endpoints in the zero set; the paper states this rather than proving that such a representative always exists, and in particular classes whose geodesic is a cylinder core avoiding the zeros are not explicitly handled, nor is independence of the chosen representative established.

Editorial extensions

If this is right

The ray $\rho_t$ converges in the Weyl-chamber length compactification to the boundary point $(\infty,[L_\infty])$, with $L_\infty(\gamma)=\sum_i v_i$ uniformly comparable to the $|q|^{1/2}$-length spectrum, so the limit has positive systole.
Since the real-spectrum compactification maps continuously onto the Weyl-chamber length compactification, the result determines the Weyl-chamber image of any real-spectrum limit of the quartic-differential ray.
The formulas pin down all four exterior powers at once: the growth exponents of $\|\bigwedge^k\mathrm{Hol}_t\|$ and of the spectral radius of $\bigwedge^k\mathrm{Hol}_t$ coincide for $k=1,2,3,4$, with the cases $k=3,4$ following from symplectic duality.
The asymptotic law gives a purely combinatorial recipe for the boundary spectrum: the limiting length of any curve is a sum over the saddle connections of its flat geodesic, so the whole limiting marked length spectrum is computable from the flat geometry alone.

Reading between the lines

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

The same Stokes-matrix scheme, fed by the $\operatorname{tt}^*$-Toda asymptotics of [GIL26], is expected to yield analogous holonomy growth laws for cyclic $\mathrm{SL}(n,\mathbb{R})$-Hitchin rays with the top-degree differential scaled; the exterior-power analysis is the main rising cost as $n$ grows.
For free homotopy classes whose $|q|^{1/2}$-geodesic is a cylinder core avoiding the zeros, the theorem as stated is silent; a natural conjecture, not proven here, is that the same limits hold by approximating the core by boundary saddle connections, which would also clarify whether the boundary lengthen spectrum depends only on the flat geometry.
The growth exponents $t^{-1/4}\log|\lambda_k|$ are the effective Lyapunov exponents of the ray, so the flat-geometric formula suggests the same data should appear in the building-valued harmonic map limit along the ray, a limit the paper leaves for future work.

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

This paper does what it says: it gives the first complete spectral asymptotics for holonomy along quartic-differential rays in the PSp(4,R) Hitchin component, covering all four singular values and eigenvalue moduli with no restriction on zero orders. The formulas are natural: the log growth rates are sums over saddle connections of integrals of the local fourth roots. That is a real step beyond the cubic SL(3,R) results of LTW26 and the planar Sp(4,R) model of TW24/GIL26.

What is genuinely new: the explicit Stokes matrices for z^d dz^4, the proof that the composed dominant coefficients are non-nilpotent, and the separate exterior-power analysis needed for the k=2 spectral data. The Stokes constants come from the tt*-Toda expansion, not from fitting, so the circularity burden is near zero. The derivation is coherent and the reliance on CL17, TW24, and GIL26 is clearly cited.

The main soft spot is the one the stress-test flags: Theorem 1.1 is stated for every nontrivial curve class, but the proof assumes the geodesic representative decomposes into saddle connections. For cylinder core classes, the length-minimizing geodesic avoids the zero set; one can use the cylinder boundary, which is a chain of saddle connections in the same free homotopy class, but the paper neither states nor proves this replacement. This is likely fixable with a standard flat-surface lemma or a citation, and I do not think it sinks the theorem, but it is a genuine gap in the written proof. Worth flagging to the author.

Minor things: the paper repeats one sentence in the outline, has a few typos, and some estimates (e.g., Propositions 4.4 and 4.5) are only sketched with references. The proof is long and not machine-checked, so some care is warranted in the Stokes estimates, but I found no internal contradiction.

Bottom line: this deserves a serious referee. The central result is important within the field, the argument is honest and mostly self-contained, and the saddle-connection issue is a patchable assumption rather than a load-bearing flaw. I would send it to review and ask the author to add the missing structural lemma.

Referee Report

1 major / 5 minor

Summary. The paper studies the asymptotic behavior, as t → +∞, of the holonomy of the flat connections associated with the ray t·q in the PSp(4,ℝ)-Hitchin component, where q is a nonzero holomorphic quartic differential on a closed Riemann surface. The main results (Theorems 5.15 and 5.17, Corollaries 5.16 and 5.18) give explicit logarithmic growth rates, at order t^{1/4}, for all singular values and eigenvalue moduli of the holonomy along a closed curve. These rates are expressed as sums over saddle connections of integrals of the four local fourth roots of q. The proof combines asymptotic decoupling away from zeros, a planar Stokes analysis near zeros, explicit computation of the Stokes matrices for the model q = z^d dz^4, and a non-nilpotence argument for the products of the resulting coefficient matrices; the second exterior power is treated separately. No restriction is imposed on the orders of the zeros.

Significance. If the results are correct, this is a substantial contribution to the asymptotic geometry of the rank-two Hitchin component. It extends the cubic-differential results of Loftin–Tamburelli–Wolf to the PSp(4,ℝ) setting, provides the first complete spectral asymptotics for holonomy along quartic-differential rays, and yields a concrete description of the limit point in Parreau's Weyl-chamber length compactification (Corollary 1.3). The paper contains genuinely new explicit computations: the Stokes matrices for the monomial model, including positivity and non-nilpotence properties of their products that were not available in the prior work of Tamburelli–Wolf. The exterior-power analysis for k = 2 is essential and is handled carefully. The proof is largely self-contained after invoking the cited results of Collier–Li, Tamburelli–Wolf, and Guest–Its–Lin, and the overall structure of the derivation is internally consistent.

major comments (1)

[Section 2.4 and beginning of Section 5] The proof of Theorem 1.1 assumes that the geodesic representative cγ of every nontrivial free homotopy class decomposes as a finite concatenation of saddle connections. This is asserted in Section 2.4 but not proved. For a free homotopy class represented by the core of a maximal |q|^{1/2}-cylinder, the length-minimizing closed geodesic lies in the interior of the cylinder and avoids Z(q); it is not a concatenation of saddle connections. One must add a structural lemma that every non-peripheral free homotopy class contains a closed curve that is a finite concatenation of saddle connections (for cylinder classes, a boundary chain of the maximal cylinder is freely homotopic to the core), and that the right-hand side of the formula is independent of the choice of such a representative. Since the arguments in Propositions 5.3–5.6, 5.14, and consequently Theorems 5.15 and 5.17 explicitly require such a decomposition, the universality of Theorem 1.1 for every nontrivial γ is not established as stated without this lemma.

minor comments (5)

[Introduction] The sentence 'We conclude by mentioning several possible directions for future work.' appears twice in succession near the end of the Introduction; the duplication should be removed.
[Section 5.6, proof of Theorem 5.15] The equality ∑_{j=1}^k v_{i,j} = λ_i^{[k]} + μ_{i-1}^{[k]} is not immediate and relies on the collinearity of the two subsegments δ_i and α_{i-1}, which ensures that the four integrals are ordered the same way on each subsegment. This should be stated explicitly in the proof.
[Section 5.3, after equation (59)] The sentence 'By (59), b_1 = b_0 = b_{-1} = 0' refers to a formula that is only established later; it should instead refer to the definition of b_k and the explicit form of T (and its inverse, for b_{-1}).
[Section 3.3.2 and Section 3.4.2] There are several grammar and proof-organization issues: 'We now given the precise formula' (Section 3.3.2), 'We are now evaluate the limits' (Section 3.4.2), and 'This completes the proof of Theorem 3.6 and Lemma 3.12' at the end of the proof of Lemma 3.12. These should be rewritten for clarity.
[Section 2.4] The expression 'the metric is simply |q|^{1/2} = |dw|^2' is notationally imprecise; it should read that the metric tensor is |dw|^2 in the natural coordinate.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the holonomy asymptotics are derived from independent Stokes-data computations, and no fitted parameter or target formula is assumed as input.

full rationale

The paper's derivation is self-contained against independent external results and does not reduce its conclusions to its inputs. The central chain is: (i) the planar Stokes matrices for q = z^d dz^4 are computed in Section 3.4 from the external tt*-Toda asymptotics of Guest–Its–Lin [GIL26, Theorem B.4 and Corollary 8.14], yielding the unipotent matrices U_1, U_2, U_3 in Proposition 3.13; (ii) the Stokes jumps are assembled along circular arcs in Propositions 3.14–3.15; (iii) Section 4 compares the surface connection with the planar model near zeros (Propositions 4.4–4.6); (iv) Section 5 composes holonomy blocks and obtains the leading exponentials (equations (63) and (68)) whose exponents are exactly the saddle-connection integrals. The limiting rates are not inputs: the dominant entries are selected by the geodesic and angular conditions in Propositions 5.4–5.6 and 5.12, and the crucial non-nilpotence of the coefficient matrices is proved independently in Propositions 5.9 and 5.14 using the explicit positivity of the sequence b_k from Proposition 5.8. No parameter is fitted to the holonomy data, and no statement equivalent to Theorem 1.1 is assumed in the proof. The cited works [CL17], [TW24], [GIL26], and [LM25] have no author overlap with the present paper, are published external results, and do not contain the target quartic holonomy asymptotics, so they are independent support rather than self-citation. The only substantive concern raised by the manuscript itself is the assertion in Section 2.4 that every free homotopy class has a closed geodesic representative that decomposes as a finite concatenation of saddle connections, on which the proof in Section 5 relies; this is a possible completeness or correctness gap, not a circular reduction, since no equation or fitted quantity makes the statement true by construction.

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

No free parameters are fitted: the constants s1, s2, b_k, and Δ_k are derived from θ = π/(d+4) using the tt*-Toda expansion and explicit algebraic identities. The listed axioms are the main imported results. The paper introduces no new physical or mathematical entities beyond the computed Stokes matrices, which are derived rather than postulated.

assumptions (6)

domain assumption Existence, uniqueness, and decay estimates for the complete harmonic metric on polynomial quartic models (Theorem 3.1, citing TW24 and LM25).
Section 3.1 uses this to define the planar comparison and to control the error one-form Θ; the Stokes jump theorem depends on these bounds.
domain assumption tt*-Toda asymptotic expansion for A3 solutions at infinity ([GIL26, Theorem B.4 and Corollary 8.14]).
Section 3.4.2 uses this expansion to compute the Stokes constants s1 and s2, hence the matrices U1, U2, U3 and the sequence b_k. If this expansion failed, the dominant coefficients in the holonomy formula would be wrong.
domain assumption Collier-Li asymptotics for parallel transport along straight segments away from zeros ([CL17, Theorem 4.4]).
Proposition 5.1 is a direct translation of this theorem; it gives the diagonal exponential factors for the straight pieces of the modified curve.
domain assumption Every nontrivial free homotopy class admits a geodesic representative for |q|^{1/2} that is a finite concatenation of saddle connections (Section 2.4).
The theorem statement is made in terms of saddle connections; cylinder core curves avoiding zeros require choosing a homotopic boundary representative, which the paper does not explicitly discuss.
standard math A closed saddle-connection geodesic through a zero of order d spans angles at least π on both sides (Proposition 5.4).
This yields the bounds 4 ≤ 2k_i + b_i - a_i ≤ 2N_i - 4 on the number of Stokes directions crossed by the detour arc around the zero.
standard math Sectorial limits and Stokes jumps for the planar connection (Theorems 3.5 and 3.6, following TW24 and DW15 Lemma B.2).
These justify replacing the surface connection by the constant model on each half-plane and give the jump matrices used in Section 3.3.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Holonomy Asymptotics along Quartic Differential Rays." pith.science (2026). https://pith.science/paper/3IRMOY3P

@misc{pith2026260804729,
  author       = {Pith},
  title        = {Pith review of: Holonomy Asymptotics along Quartic Differential Rays},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/3IRMOY3P}},
  note         = {Machine review of arXiv:2608.04729}
}

abstract

Let $X$ be a closed Riemann surface and let $q\in H^0(X,K^4)$ be a nonzero holomorphic quartic differential on $X$. For $t >0$, the ray $tq$ determines a family of Hitchin representations in the $\operatorname{PSp}(4,\mathbb R)$-Hitchin component. We study, as $t\to+\infty$, the asymptotic behavior of their holonomy along closed curves. We obtain explicit asymptotic formulas for all singular values and for the absolute values of all eigenvalues of the holonomy. Their logarithmic growth rates are given by integrating the local fourth roots of $q$ along the saddle connections forming the geodesic representative of the curve with respect to the singular flat metric $\lvert q\rvert^{1/2}$. No restriction is imposed on the orders of the zeros of $q$.

Figures

Figures reproduced from arXiv: 2608.04729 by the authors.

**Figure 4.** The geodesic 𝑐𝛾 and its modification near the zeros, illustrated in the case of three zeros. By multiplicativity of parallel transport, Hol𝑡 ( ̃𝑐𝛾 ) = ℓ ∏ 𝑖=1 [Hol𝑡 ( ̃ 𝛿𝑖 ) Hol𝑡 (𝛽̃ 𝑖 ) Hol𝑡 ( ̃𝛼𝑖 )] . (50) To study the holonomy, we work in a natural coordinate 𝑤 for 𝑞 and use the renormalized frame ℱ𝑡,𝑤 = (𝑡 3 8 (𝑑𝑤) 3 2 , 𝑡 1 8 (𝑑𝑤) 1 2 , 𝑡− 1 8 (𝑑𝑤)− 1 2 , 𝑡− 3 8 (𝑑𝑤)− 3 2 ) . Equivalently, this is the holomorph… view at source ↗

**Figure 5.** Perturbation near a Stokes direction. For the perturbed full segments, define 𝜇 𝜂 𝑖,𝑘 ∶= −2 Re ∫ 𝛼 𝜂 𝑖 𝜙𝑘 , 𝜆𝜂 𝑖,𝑘 ∶= −2 Re ∫ 𝛿 𝜂 𝑖 𝜙𝑘 , and set 𝜇 𝜂 𝑖 ∶= max 1≤𝑘≤4 𝜇 𝜂 𝑖,𝑘, 𝜆𝜂 𝑖 ∶= max 1≤𝑘≤4 𝜆 𝜂 𝑖,𝑘. For 𝜂 > 0 sufficiently small, the paths 𝛼 𝜂 𝑖 and 𝛼𝑖 are also homotopic relative to their endpoints. The same holds for 𝛿 𝜂 𝑖 and 𝛿𝑖 . Since 𝜙𝑘 is exact, we have ∫ 𝛼 𝜂 𝑖 𝜙𝑘 = ∫ 𝛼𝑖 𝜙𝑘 , ∫ 𝛿 𝜂 𝑖 𝜙𝑘 = ∫ 𝛿𝑖 𝜙𝑘 . Consequentl… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

17 extracted references · 11 canonical work pages

[3]

Harmonic metrics of generically regular semisimple Higgs bundles on noncompact Riemann surfaces

[LM23] Qiongling Li and Takuro Mochizuki. “Harmonic metrics of generically regular semisimple Higgs bundles on noncompact Riemann surfaces”. In:Tunisian Journal of Mathematics 5.4 (Nov. 2023), pp. 663–711. issn: 2576-7658. doi: 10.2140/tunis.2023.5.663 ↑9. [LM24] Qiongling Li and Takuro Mochizuki. “Higgs bundles in the Hitchin section over non-compact hy-...

work page doi:10.2140/tunis.2023.5.663 2024
[6]

Asymptotics of Higgs bundles in the Hitchin component

[CL17] Brian Collier and Qiongling Li. “Asymptotics of Higgs bundles in the Hitchin component”. In: Advances in Mathematics 307 (2017), pp. 488–558. doi: 10 . 1016 / j . aim . 2016 . 11 . 031↑5, 12, 13, 17, 39, 40, 49, 50,

work page 2017
[17]

Good wild harmonic bundles and good filtered Higgs bundles

[Moc21] Takuro Mochizuki. “Good wild harmonic bundles and good filtered Higgs bundles”. In:Symmetry, Integrability and Geometry: Methods and Applications (SIGMA) 17 (2021). 66 pages, p. 068.doi: 10.3842/SIGMA.2021.068 ↑18. [Ouy23] Charles Ouyang. “High-Energy Harmonic Maps and Degeneration of Minimal Surfaces”. In:Geom- etry & Topology 27.5 (2023), pp. 16...

work page arXiv 2021
[18]

Limits of cubic differentials and buildings

[LTW26] John Loftin, Andrea Tamburelli, and Michael Wolf. “Limits of cubic differentials and buildings”. In: Proceedings of the London Mathematical Society 132.4 (2026), e70154. doi: 10.1112/plms.70154 ↑2, 3, 5, 6, 47, 49,

work page doi:10.1112/plms.70154 2026
[19]

Local asymptotics for Hitchin’s equations and high energy harmonic maps

[SS26] Nathaniel Sagman and Peter Smillie. “Local asymptotics for Hitchin’s equations and high energy harmonic maps”. In:Mathematische Annalen 394.1 (2026), p. 17.doi: 10.1007/s00208-026-03375- y ↑5. [Sim88] Carlos T Simpson. “Constructing variations of Hodge structure using Yang-Mills theory and applica- tions to uniformization”. In:Journal of the Americ...

work page doi:10.1007/s00208-026-03375- 2026
[26]

The Teichmüller Theory of Harmonic Maps

[Wol89] Michael Wolf. “The Teichmüller Theory of Harmonic Maps”. In:Journal of Differential Geometry 29.2 (1989), pp. 449–479.doi: 10.4310/jdg/1214442885 ↑6. [Wol95] Michael Wolf. “Harmonic Maps from Surfaces to ℝ-Trees”. In: Mathematische Zeitschrift 218.4 (1995), pp. 577–593.doi: 10.1007/BF02571924 ↑6. 75

arXiv 1989
[28]

The Self-Duality Equations on a Riemann Surface

[Hit87] N. J. Hitchin. “The Self-Duality Equations on a Riemann Surface”. In:Proceedings of the London Mathematical Society s3-55.1 (1987), pp. 59–126.doi: 10.1112/plms/s3-55.1.59 ↑2. [Hit92] Nigel J Hitchin. “Lie groups and Teichmüller space”. In:Topology 31.3 (1992), pp. 449–473. doi: 10.1016/0040-9383(92)90044-I ↑2. 73 [KC02] Victor G. Kac and Pokman C...

doi:10.1112/plms/s3-55.1.59 1987
[55]

Aﬀine spheres and convexℝℙ𝑛-manifolds

[Lof01] John C. Loftin. “Aﬀine spheres and convexℝℙ𝑛-manifolds”. In:American Journal of Mathematics 123.2 (2001), pp. 255–274.doi: 10.1353/ajm.2001.0011 ↑2. [MOT24] Giuseppe Martone, Charles Ouyang, and Andrea Tamburelli. “A Closed Ball Compactification of a Maximal Component via Cores of Trees”. In:Algebraic & Geometric Topology 24.7 (2024), pp. 3693–

arXiv 2001

Show all 17 references

[309]

Kobayashi–Hitchin Correspondence for Tame Harmonic Bundles II

Paris: Société Mathématique de France, 2006, pp. viii+117 ↑2. [Moc09] Takuro Mochizuki. “Kobayashi–Hitchin Correspondence for Tame Harmonic Bundles II”. In:Geom- etry & Topology 13.1 (2009), pp. 359–455.doi: 10.2140/gt.2009.13.359 ↑2. [Moc11] Takuro Mochizuki. Wild harmonic bu...

2009 doi
[324]

Anosov flows, surface groups and curves in projective space

Progress in Mathematics. Cham: Birkhäuser, 2017, pp. 203–260.doi: 10.1007/978-3-319-59939- 7_6 ↑6. [Lab06] François Labourie. “Anosov flows, surface groups and curves in projective space”. In:Inventiones Mathematicae 165.1 (2006), pp. 51–114.doi: 10.1007/s00222-005-0487-3 ↑2. ...

2006 doi
[340]

Harmonic bundles and Toda lattices with opposite sign II

Société mathématique de France, 2011.url: https://arxiv.org/abs/0803.1344 ↑18. [Moc14] Takuro Mochizuki. “Harmonic bundles and Toda lattices with opposite sign II”. In:Communications in Mathematical Physics 328.3 (2014), pp. 1159–1198.doi: 10.1007/s00220-014-1994-0 ↑18. [Moc16...

2014 arXiv
[2002]

Harmonic Maps to Build- ings and Singular Perturbation Theory

doi: 10.1007/978-1-4613-0071-7 ↑30. [Kat+15] Ludmil Katzarkov, Alexander Noll, Pranav Pandit, and Carlos Simpson. “Harmonic Maps to Build- ings and Singular Perturbation Theory”. In:Communications in Mathematical Physics 336.2 (2015), pp. 853–903. doi: 10.1007/s00220-014-2276-...

2015 doi
[2009]

Fibrés de Higgs et connexions intégrables : le cas logarithmique (diviseur lisse)

arXiv: 1002.1767 ↑13. [Biq97] Olivier Biquard. “Fibrés de Higgs et connexions intégrables : le cas logarithmique (diviseur lisse)”. In: Annales Scientifiques de l’École Normale Supérieure 30.1 (1997), pp. 41–96.doi: 10.1016/S0012- 9593(97)89915-6 ↑2. [BB04] Olivier Biquard and...

1997 arXiv
[2012]

The tt*-Toda equations of𝐴𝑛 type

doi: 10.48550/arXiv.0909.4487 ↑2. [GIL26] Martin A. Guest, Alexander R. Its, and Chang-Shou Lin. “The tt*-Toda equations of𝐴𝑛 type”. In: Advances in Mathematics 485 (2026), p. 110730.issn: 0001-8708.doi: 10.1016/j.aim.2025.110730 ↑6, 7, 17,

2026 doi arXiv
[2023]

48550 / arXiv.2311.01892 ↑5,

doi: 10 . 48550 / arXiv.2311.01892 ↑5,

doi arXiv
[2024]

[GGM12] Oscar García-Prada, Peter B

aXiv:2208.14409 ↑6. [GGM12] Oscar García-Prada, Peter B. Gothen, and Ignasi Mundet i Riera.The Hitchin–Kobayashi corre- spondence, Higgs pairs and surface group representations

arXiv
[3717]

[Moc06] Takuro Mochizuki

doi: 10.2140/agt.2024.24.3693 ↑6. [Moc06] Takuro Mochizuki. Kobayashi–Hitchin Correspondence for Tame Harmonic Bundles and an Appli- cation. Astérisque

2024 doi

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.

[1] [3]

Harmonic metrics of generically regular semisimple Higgs bundles on noncompact Riemann surfaces

[LM23] Qiongling Li and Takuro Mochizuki. “Harmonic metrics of generically regular semisimple Higgs bundles on noncompact Riemann surfaces”. In:Tunisian Journal of Mathematics 5.4 (Nov. 2023), pp. 663–711. issn: 2576-7658. doi: 10.2140/tunis.2023.5.663 ↑9. [LM24] Qiongling Li and Takuro Mochizuki. “Higgs bundles in the Hitchin section over non-compact hy-...

work page doi:10.2140/tunis.2023.5.663 2024

[2] [6]

Asymptotics of Higgs bundles in the Hitchin component

[CL17] Brian Collier and Qiongling Li. “Asymptotics of Higgs bundles in the Hitchin component”. In: Advances in Mathematics 307 (2017), pp. 488–558. doi: 10 . 1016 / j . aim . 2016 . 11 . 031↑5, 12, 13, 17, 39, 40, 49, 50,

work page 2017

[3] [17]

Good wild harmonic bundles and good filtered Higgs bundles

[Moc21] Takuro Mochizuki. “Good wild harmonic bundles and good filtered Higgs bundles”. In:Symmetry, Integrability and Geometry: Methods and Applications (SIGMA) 17 (2021). 66 pages, p. 068.doi: 10.3842/SIGMA.2021.068 ↑18. [Ouy23] Charles Ouyang. “High-Energy Harmonic Maps and Degeneration of Minimal Surfaces”. In:Geom- etry & Topology 27.5 (2023), pp. 16...

work page arXiv 2021

[4] [18]

Limits of cubic differentials and buildings

[LTW26] John Loftin, Andrea Tamburelli, and Michael Wolf. “Limits of cubic differentials and buildings”. In: Proceedings of the London Mathematical Society 132.4 (2026), e70154. doi: 10.1112/plms.70154 ↑2, 3, 5, 6, 47, 49,

work page doi:10.1112/plms.70154 2026

[5] [19]

Local asymptotics for Hitchin’s equations and high energy harmonic maps

[SS26] Nathaniel Sagman and Peter Smillie. “Local asymptotics for Hitchin’s equations and high energy harmonic maps”. In:Mathematische Annalen 394.1 (2026), p. 17.doi: 10.1007/s00208-026-03375- y ↑5. [Sim88] Carlos T Simpson. “Constructing variations of Hodge structure using Yang-Mills theory and applica- tions to uniformization”. In:Journal of the Americ...

work page doi:10.1007/s00208-026-03375- 2026

[6] [26]

The Teichmüller Theory of Harmonic Maps

[Wol89] Michael Wolf. “The Teichmüller Theory of Harmonic Maps”. In:Journal of Differential Geometry 29.2 (1989), pp. 449–479.doi: 10.4310/jdg/1214442885 ↑6. [Wol95] Michael Wolf. “Harmonic Maps from Surfaces to ℝ-Trees”. In: Mathematische Zeitschrift 218.4 (1995), pp. 577–593.doi: 10.1007/BF02571924 ↑6. 75

arXiv 1989

[7] [28]

The Self-Duality Equations on a Riemann Surface

[Hit87] N. J. Hitchin. “The Self-Duality Equations on a Riemann Surface”. In:Proceedings of the London Mathematical Society s3-55.1 (1987), pp. 59–126.doi: 10.1112/plms/s3-55.1.59 ↑2. [Hit92] Nigel J Hitchin. “Lie groups and Teichmüller space”. In:Topology 31.3 (1992), pp. 449–473. doi: 10.1016/0040-9383(92)90044-I ↑2. 73 [KC02] Victor G. Kac and Pokman C...

doi:10.1112/plms/s3-55.1.59 1987

[8] [55]

Aﬀine spheres and convexℝℙ𝑛-manifolds

[Lof01] John C. Loftin. “Aﬀine spheres and convexℝℙ𝑛-manifolds”. In:American Journal of Mathematics 123.2 (2001), pp. 255–274.doi: 10.1353/ajm.2001.0011 ↑2. [MOT24] Giuseppe Martone, Charles Ouyang, and Andrea Tamburelli. “A Closed Ball Compactification of a Maximal Component via Cores of Trees”. In:Algebraic & Geometric Topology 24.7 (2024), pp. 3693–

arXiv 2001