Universal Formulas for de Jonqui`eres Loci and Equivariant Contact Problems

arXiv:2609.30892 · math.AG · Submitted 2026-09-25 · 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: I'm Ada.

Emmy: And I'm Emmy, here with Kolya.

Ada: We're looking at the paper: "Universal Formulas for de Jonqui`eres Loci and Equivariant Contact Problems".

Emmy: Here's the gist of the paper. Universal formulas for de Jonquières loci and equivariant contact problems provide universal formulas for relative de Jonquières loci that recover those above under specialization The paper develops universal formulas for relative de Jonquières loci that allow both the curve and the line bundle to vary in moduli Definition of Relative Loci and Stratification The paper studies the generalization of a classical problem to varying families, defining locally closed subsets called relative de Jonquières loci For a family of genus g curves f: Ψ → S and a relative degree d line bundle L on Ψ such that f∗L is a vector bundle, the projectivization P(f∗L) is stratified by relative de Jonquières loci, which parameterize sections of L with prescribed vanishing orders.

Ada: The closures Xµ are called the relative de Jonquières loci of type µ.

Emmy: Construction via Weighted Bundles The key ingredient in the construction is a vector bundle Fν(L) on C n g → Picd g which we call the weighted bundle of principal parts.

Ada: This bundle generalizes relative bundles of principal parts

EH16: and secant bundles on symmetric products

Mat65: to the weighted setting.

Title and authors: Ada: We’re now looking at the paper's title, "Universal Formulas for de Jonquières Loci and Equivariant Contact Problems," and who wrote it. This title immediately tells us this is about creating universal tools for these specific geometric objects.

Emmy: It suggests they are building a set of formulas that don't depend on a single curve or line bundle, but rather on the underlying moduli space itself, allowing both to vary simultaneously.

Kolya: For me, the title points toward solving problems in enumerative geometry where you have constraints imposed by contact orders between curves and divisors.

Ada: That’s right, it’s about those contact problems—how divisors can intersect curves with specific multiplicities or vanishing orders defined by a line bundle.

Emmy: It sounds like they are taking classical results, maybe from earlier work on lines in projective space, and generalizing them to this much broader setting.

Kolya: I hope they manage to keep the complexity manageable enough so that these universal formulas are actually usable for computation rather than just theoretical exercises.

Ada: The authors seem to be tackling a very deep problem here, connecting algebraic geometry with some of those more combinatorial aspects of intersection theory.

Emmy: It’s interesting that they are using tools like the weighted bundle of principal parts; it shows a clear structural idea for organizing the data in this generalized setting.

The paper's summary: Ada: So, what does the paper actually summarize? It boils down to them defining relative de Jonquières loci and then deriving universal formulas for their fundamental classes.

Emmy: They define these loci as locally closed subsets that parameterize sections of a line bundle subject to specific vanishing orders, which is the starting point for all their computation.

Kolya: So they are essentially studying how divisors behave when you have constraints on how they touch curves in a certain way, and then calculating the size or degree of those loci.

Ada: Exactly. And the key part is Theorem one point two, which states that these universal formulas for the classes of these loci exist in terms of twisted kappa classes and those weighted bundles <ref:2609.30892#pg1,universal formulas for the classes of these loci>.

Emmy: The construction involves using a vector bundle F nu L to pick up higher-order zeros, and then pushing forward its Chern classes via the map to the universal Picard stack.

Kolya: So they are essentially saying that you can calculate the size of these loci by summing up certain combinations of these kappa classes based on those vanishing orders.

Ada: That’s right, and they show this works even when you specialize to simpler cases, recovering known results from things like the family of lines in projective space.

Emmy: It’s a powerful statement because it shows that this general machinery doesn't just work for the simplest examples; it has real power across the board.

The paper's improvements: Ada: Now let’s talk about what they improved or added. They are really improving the scope of their tools by making them more general than previously possible.

Emmy: Specifically, they introduced this vector bundle F nu L to handle the structure better than just using a direct sum of pullbacks of relative bundles of principal parts.

Kolya: That’s a big structural improvement because when points collide, their multiplicity in this new bundle is added up rather than just taking the maximum, which changes things fundamentally.

Ada: It simplifies how we handle those collisions, and that structure leads directly to the way they calculate the Chern classes of F nu L as sums of products of pullbacks.

Emmy: And they use those Chern classes to define the universal polynomials in Theorem one point two, which are then computed using a combinatorial pushforward lemma <ref:2609.30892#pg1>.

Kolya: I like that they explicitly show how these universal polynomials look for specific examples, like the one with d equals six and nu equal to two three two.

Ada: And they extend this concept to complete intersections in projective space, which is a significant leap because it moves us out of just curves into higher dimensions.

Emmy: Theorem five point six gives us the class of those universal strata for these complete intersections, which involves a sum over terms involving the tangent bundle and these weighted bundles.

Conclusion: Ada: So to wrap up this whole discussion, what are the main implications of this paper? It really moves us toward having universal tools for de Jonquières loci that depend on varying families.

Emmy: It means we have a unified way to approach these contact problems across different geometric variations, which is a big step in making these calculations systematic.

Kolya: For me, the implication is that we can now use these formulas to compute classes for things like coincident root loci of binary forms in a GL2 setting using established methods.

Ada: It’s about connecting the abstract structure of the generalized loci to concrete, known results, which validates this entire construction.

Emmy: I think the real value is in how they handle the structural aspects—the weighted bundle—which provides a clear way to manage complexity when things get messy.

Kolya: And one thing they flag as a limitation is that their method for complete intersections involves summing over the degrees of those components, which means you have to deal with that summation structure.

Ada: Exactly, so while they’ve given us universal formulas, the computational cost might still be high due to that summation.

Emmy: It’s a trade-off between having a general formula and having an easily computable one for every single instance.

Kolya: But overall, this paper provides a solid framework, and it gives us concrete results for several important areas in the field of enumerative geometry.

Ada: So we’ve heard about "Universal Formulas for de Jonquières Loci and Equivariant Contact Problems," which gives us those universal formulas that let both the curve and line bundle vary in moduli.

SMITA RAJAN

math.AG

Submitted: 2026-09-25

Updated: 2026-11-03

Comments: v2: 25 pages. Added additional assumption to Theorem 1.2, and additional Lemma 3.3

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

Importance score: 88/100

The gist: Universal formulas for de Jonquières loci and equivariant contact problems provide universal formulas for relative de Jonquières loci that recover those above under specialization The paper…

Key concepts

Relative de Jonquières Loci
These are locally closed subsets that parameterize sections of a line bundle with prescribed vanishing orders. They are studied in families where both the curve and the line bundle can change, providing a way to study geometric constraints across varying moduli.
Weighted Bundle of Principal Parts Fν(L)
This is a key vector bundle constructed on the space of curves to generalize classical bundles. Its fibers represent cohomology groups related to sections of L twisted by specific divisors. This bundle is essential for constructing the universal formulas and computing characteristic classes.
Twisted Kappa Classes
These are specific characteristic classes defined using the pullbacks of Chern classes from the universal Picard stack. They form a basis for expressing the class of a relative de Jonquières locus in terms of combinatorial data related to partitions and vanishing orders.
Complete Intersections in Projective Space
This refers to geometric objects formed by the intersection of several hypersurfaces within projective space. The paper extends its universal formulas to these settings, allowing for the computation of classes like [Σeµ] in the moduli space MGLr+1(d).

Terminology

Summary

Universal formulas for de Jonquières loci and equivariant contact problems provide universal formulas for relative de Jonquières loci that recover those above under specialization The paper develops universal formulas for relative de Jonquières loci that allow both the curve and the line bundle to vary in moduli

Definition of Relative Loci and Stratification

The paper studies the generalization of a classical problem to varying families, defining locally closed subsets called relative de Jonquières loci For a family of genus g curves f: Ψ → S and a relative degree d line bundle L on Ψ such that f∗L is a vector bundle, the projectivization P(f∗L) is stratified by relative de Jonquières loci, which parameterize sections of L with prescribed vanishing orders. The closures Xµ are called the relative de Jonquières loci of type µ. p. 0

Where the paper says it

, µm) a partition of d, define the locally closed subsets X◦ µ = ( (C, s) : div(s) = Xm i=1 µipi ) ⊆ P(f∗L), which give rise to a natural stratification of P(f∗L). We call their closures Xµ = X◦ µ the relative de Jonqui`eres loci of type µ.

Construction via Weighted Bundles

The key ingredient in the construction is a vector bundle Fν(L) on C n g → Picd g which we call the weighted bundle of principal parts. This bundle generalizes relative bundles of principal parts [EH16] and secant bundles on symmetric products [Mat65] to the weighted setting. p. 1

Where the paper says it

The key ingredient in our construction is a vector bundle Fν(L ) on C n g → Picd g which we call the weighted bundle of principal parts. The vector bundle Fν(L ) generalizes relative bundles of principal parts [EH16] and secant bundles on symmetric products [Mat65] to the weighted setting.

The fiber of Fν(L) over a geometric point (C,L, p1,. p. 1
Where the paper says it

The vector bundle Fν(L ) generalizes relative bundles of principal parts [EH16] and secant bundles on symmetric products [Mat65] to the weighted setting. The fiber of Fν(L ) over a geometric point (C,L, p1, .

.., pn) is H0(C,L ⊗ OC/I ν1 p1. p. 1
Where the paper says it

We write µ = (µ1, . , µm) and ν = (ν1, .

. I νn pn). The construction commutes with base change on S. p. 2
Where the paper says it

UNIVERSAL FORMULAS FOR DE JONQUIERES LOCI AND EQUIVARIANT CONTACT PROBLEMS 3 ` The vector bundle Fν(L ) has rank c(ν) and its construction commutes with base change. We use Fν(L ) to pick up higher-order zeros of sections of L.

Universal Formulas in Twisted Kappa Classes

Theorem 1.2 provides universal formulas for relative de Jonquières loci, stating that the class of the relative de Jonquières locus Xµ is [Xµ] = 1/Q a ma! p. 1

Where the paper says it

In the case that S = Spec k, when the relative de Jonqui`eres loci Xµ have codimension r in P(H0 (C,L)), the degree of Xµ is equal to the number of de Jonqui`eres divisors of length m for a general r-dimensional linear subsystem (L, V ). In this paper, we provide universal formulas for relative de Jonqui`eres loci that, under specialization, recover those above.

Xc(ν) i=n γ c(ν)−i g i ν, where g i ν:= α∗ f i ν. These formulas are expressed in terms of the twisted kappa classes κi,j = f∗(c1(ωf)i+1c1(L)j). p. 1
Where the paper says it

From c1(ωf ) and c1(L ), we can define tautological classes in A∗ (Picd g ) called the twisted kappa classes: κi,j = f∗(c1(ωf ) i+1c1(L ) j ). In the following theorem, we give universal formulas for relative de Jonqui`eres loci.

The construction involves combinatorial pushforward lemma to compute f i ν as an explicit polynomial in the twisted kappa classes. p. 1
Where the paper says it

Suppose that Xµ has codimension c(ν)−n. Then, there exist explicit polynomials f i ν ∈ Ai−n (Picd g ) in the twisted kappa classes such that the class of the relative de Jonqui`eres locus Xµ is [Xµ] = 1 Q a ma!

Application to Specific Cases

The paper provides examples showing that these universal formulas recover known results, such as for the family of lines in P r and L = OPS(d). For the case f: PW → BGL2 and L = OPW(d), the GL2-equivariant class of the coincident root locus is computed to agree with known formulas in [FNR03] and [ST18]. p. 1

Where the paper says it

Additionally, when f : PW → BGL2 is the projectivization of the universal rank 2 vector bundle and L = OPW(d), the relative de Jonqui`eres loci are the GL2-equivariant coincident root loci of binary forms. Their fundamental classes were computed in [FNR03] and PGL2-equivariantly in [ST18].

Furthermore, for a degree d hypersurface X, the class of µ-incident lines is given by [Sµ(X)] = 1/Q a ma! p. 4
Where the paper says it

UNIVERSAL FORMULAS FOR DE JONQUIERES LOCI AND EQUIVARIANT CONTACT PROBLEMS 5 ` where β : Gr(2, W) → BGLr+1. (2) When d(µ) ⩾ 0, the class of µ-incident lines to a general degree d hypersurface X is given by evaluating g c(ν) ν at ci = 0: [Sµ(X)] = 1 Q a ma!

g c(ν) ν ci=0.

Generalization to Complete Intersections

The formulas are extended to complete intersections in projective space, where the moduli space of degree d hypersurfaces is replaced by the moduli space of type d complete intersections MGLr+1 (d). Theorem 5.6 gives the class of Σeµ as [Σeµ] = Yl i=1 ctop(T ∨ i ⊠ Fµ(Li)). p. 2

Where the paper says it

Taking f : C → Mg, and L = ωf gives us α ∗κi,j = κi+j . The fundamental class of the stratum of differentials for this partition when k = 1 and g = 4 in A3 (PH4) is [Xµ] = 1 6 h γ 3 27κ 3 0 − 198κ 2 0 + 336κ0 + γ 2 54κ 2 0κ1 − 474κ0κ1 + 1024κ1 + γ 36κ0κ 2 1 − 228κ 2 1 − 180κ0κ2 + 1320κ2 + 8κ 3 1 − 120κ1κ2 + 592κ3 i .

This shows that the expected codimension of Σeµ is c(µ) X l j=1 k j. p. 4
Where the paper says it

In the case that the hypersurface X is a plane curve, our formulas for [Vµ] are equivariant enrichments of the formulas in [CMR26]. The following table shows the GL3-equivariant classes of these loci up to codimension 2.

The class [Sµ(X)] can be computed from [Σeµ] in the same way as in the hypersurface case. p. 4
Where the paper says it

Equivalently, it is the degree of the scroll of tritangent lines to a general sextic surface, which was computed in [CV24, Proposition 4. We note that the equivariant class [Vµ] determines [Sµ(X)] in the following way.

Invariance and Equivariant Classes

The paper notes that invariance under a change of variables implies that the coefficient of c1 in Vµ is-d/3 times the coefficient of γ. This implies that in codimension 1, the non-equivariant fundamental class fully determines the equivariant fundamental class. p. 5

Where the paper says it

Invariance under this change of variables implies that the coefficient of c1 in Vµ is −d/3 times the coefficient of γ. Hence, in codimension 1, the non-equivariant fundamental class of Vµ fully determines the equivariant fundamental class.

In codimension 2 and higher, any class can be written as Aγ2 + Bc1γ + Cc2 1 + Dc2, and invariance under the above change of variables implies that B = -d/3 A C = d/9 A - 1/3 D. p. 5
Where the paper says it

Hence, in codimension 1, the non-equivariant fundamental class of Vµ fully determines the equivariant fundamental class. In codimension 2, any class can be written as Aγ2 + Bc1γ + Cc2 1 + Dc2, and invariance under the above change of variables implies that B = − 2d 3 A C = d 2 9 A − 1 3 D.

The paper also shows that the class [Sµ] is determined by g c(ν) ν ci=0. p. 4
Where the paper says it

The following table shows the GL3-equivariant classes of these loci up to codimension 2. ν Codimension Equivariant Fundamental Class of Vµ (4) 1 − 2d(3d − 2)(d − 3)c1 + 6(3d − 2)(d − 3)γ (3, 2) 1 − d(d − 4)(d − 3)(d 2 + 6d − 4)c1 + 3(d − 4)(d − 3)(d 2 + 6d − 4)γ

Summary of Key Findings

The main contribution is providing universal formulas for relative de Jonquières loci that allow both the curve and the line bundle to vary in moduli. These formulas are expressed in terms of twisted kappa classes and weighted bundles of principal parts The results successfully compute fundamental classes for various geometric objects, including strata of differentials and coincident root loci of binary forms. p. 1

Where the paper says it

The fundamental classes of strata of k-differentials were computed in [Che13], [Sau19], and also over the moduli space of curves with rational tails in [GT24]. Additionally, when f : PW → BGL2 is the projectivization of the universal rank 2 vector bundle and L = OPW(d), the relative de Jonqui`eres loci are the GL2-equivariant coincident root loci of binary forms.

The formulas are applicable to complete intersections in projective space as well. p. 6
Where the paper says it

In Section 5. 3, we also generalize these contact problems to the case of complete intersections in projective space, where we replace P(SymdW∨) with a partial compactification of Di Lorenzo’s moduli stack of complete intersections [Lor22].

References

[Arb+85] E. Arbarello et al. Geometry of Algebraic Curves: Volume I. p. 0

Where the paper says it

UNIVERSAL FORMULAS FOR DE JONQUIERES LOCI ` AND EQUIVARIANT CONTACT PROBLEMS SMITA RAJAN Abstract. Given a family of genus g curves f : Ψ → S and a relative degree d line bundle L on Ψ such that f∗L is a vector bundle, the projectivization P(f∗L) is stratified by relative de Jonqui`eres loci, which parameterize sections of L with prescribed vanishing orders.

Vol. 267. A Series of Comprehensive Studies in Mathematics. p. 0
Where the paper says it

(1 + µ 2 1 t1 + · · · + µ 2 mtm) g (1 + µ1t1 + · · · + µmtm) d−r−g . This paper studies the generalization of this problem to the case where (C,L) varies in families.

New York: Springer, 1985

[Che13] Dawei Chen “Strata of abelian differentials and the Teichmüller dynamics” In Journal of Modern Dynamics 7.1 (2013), pp. 135–152 issn: 1930-5311, 1930-532X doi: 10.3934/jmd.2013.7.135 url: https://doi.org/ [Che]

[CMR26] Ciro Ciliberto Rick Miranda and Joaquim Roe Contact Invariants for Plane Curves in a Pencil 2026 arXiv: 2605.27623 math/AG url: https://arxiv. p. 0

Where the paper says it

1 arXiv:2609. 30892v1 [math.

p. 0
Where the paper says it

1 arXiv:2609. 30892v1 [math.

p. 0
Where the paper says it

1 arXiv:2609. 30892v1 [math.

org/abs/math/0311312 [CMR]

[CV24] Ciro Ciliberto and Alessandro Verra On the variety of tritangential planes to a general K3 surface of degree 6 and genus 4 in P 4 2024 arXiv: 2406.00822 math/AG url: https://arxiv. p. 0

Where the paper says it

1 arXiv:2609. 30892v1 [math.

p. 0
Where the paper says it

1 arXiv:2609. 30892v1 [math.

p. 0
Where the paper says it

1 arXiv:2609. 30892v1 [math.

org/abs/math/AG [CV]

[EH16] David Eisenbud and Joe Harris All That A Second Course in Algebraic Geometry Vol. 159 Cambridge Studies in Advanced Mathematics Cambridge University Press 2016 doi: 10. p. 0

Where the paper says it

(1 + µ 2 1 t1 + · · · + µ 2 mtm) g (1 + µ1t1 + · · · + µmtm) d−r−g . This paper studies the generalization of this problem to the case where (C,L) varies in families.

p. 0
Where the paper says it

(1 + µ 2 1 t1 + · · · + µ 2 mtm) g (1 + µ1t1 + · · · + µmtm) d−r−g . This paper studies the generalization of this problem to the case where (C,L) varies in families.

p. 0
Where the paper says it

(1 + µ 2 1 t1 + · · · + µ 2 mtm) g (1 + µ1t1 + · · · + µmtm) d−r−g . This paper studies the generalization of this problem to the case where (C,L) varies in families.

1017/CBO9781139062046 [EH]

[FNR03] L M Feher A Nemethi and R Rimanyi Coincident root loci of binary forms 2003 arXiv: math/0311312 math/AG url: https://arxiv.org/abs/math/0311312 [FNR]

[FNR05] L M Fehér A Némethi and R Rimanyi Degeneracy of Two and Three Forms In Canadian Mathematical Bulletin 48.4 (2005) pp. 547–560 url: https://rimanyi.web.unc.edu/wp-content/uploads/sites/9870/2018/07/formak6. p. 6

Where the paper says it

1, for assistance generating SageMath code for computing examples with the weighted bundles of principal parts, and for proofreading. The manuscript was written entirely by the author, who assumes responsibility for all content.

p. 6
Where the paper says it

1, for assistance generating SageMath code for computing examples with the weighted bundles of principal parts, and for proofreading. The manuscript was written entirely by the author, who assumes responsibility for all content.

p. 6
Where the paper says it

1, for assistance generating SageMath code for computing examples with the weighted bundles of principal parts, and for proofreading. The manuscript was written entirely by the author, who assumes responsibility for all content.

pdf [FNR]

[GT24] Iulia Gheorghita and Nicola Tarasca Incidence varieties in the projectivized kth Hodge bundle over curves with rational tails In Communications in Contemporary Mathematics 26.5 (2024) Paper No. p. 0

Where the paper says it

(1 + µ 2 1 t1 + · · · + µ 2 mtm) g (1 + µ1t1 + · · · + µmtm) d−r−g . This paper studies the generalization of this problem to the case where (C,L) varies in families.

p. 0
Where the paper says it

(1 + µ 2 1 t1 + · · · + µ 2 mtm) g (1 + µ1t1 + · · · + µmtm) d−r−g . This paper studies the generalization of this problem to the case where (C,L) varies in families.

p. 0
Where the paper says it

(1 + µ 2 1 t1 + · · · + µ 2 mtm) g (1 + µ1t1 + · · · + µmtm) d−r−g . This paper studies the generalization of this problem to the case where (C,L) varies in families.

2350015, 47 issn: 0219-1997,1793-6683 doi: 10.1142/S0219199723500153 url: https://doi.org/ [GT]

[Lar24] Hannah Larson The Chow ring of the universal Picard stack over the hyperelliptic locus 2024 arXiv: 2404.12607 math/AG url: https://arxiv. p. 0

Where the paper says it

1 arXiv:2609. 30892v1 [math.

p. 0
Where the paper says it

1 arXiv:2609. 30892v1 [math.

p. 0
Where the paper says it

1 arXiv:2609. 30892v1 [math.

org/abs/math/AG [Lar]

[Lor22] Andrea Di Lorenzo Intersection theory on moduli of smooth complete intersections 2022 arXiv: 2301.06024 math/AG url: https://arxiv. p. 0

Where the paper says it

1 arXiv:2609. 30892v1 [math.

p. 0
Where the paper says it

1 arXiv:2609. 30892v1 [math.

p. 0
Where the paper says it

1 arXiv:2609. 30892v1 [math.

org/abs/math/AG [Lor]

[Mat65] Arthur Mattuck Secant Bundles on Symmetric Products In American Journal of Mathematics 87.4 (1965) pp. 779–797 doi: 10.2307 / 2373208 url: https://ams.org/ [Mat]

[Sau19] Adrien Sauvaget Cohomology Classes of Strata of Differentials In Geometry & Topology 23.3 (2019) pp. 1085–1171 doi: 10.2140 / gt. 2019. 23. 1085 url: https://doi.org/ [Sau]

[ST18] Hunter Spink and Dennis Tseng P GL2-equivariant strata of point configurations in P 1. 2018 arXiv: math/AG url: https://arxiv. p. 0

Where the paper says it

1 arXiv:2609. 30892v1 [math.

p. 0
Where the paper says it

1 arXiv:2609. 30892v1 [math.

p. 0
Where the paper says it

1 arXiv:2609. 30892v1 [math.

org/abs/math/AG [ST]

[Stacks] The Stacks Project Authors The Stacks Project 2026 url: https://stacks.math.columbia.edu [Stacks]

[Arb+85] E Arbarello et al Geometry of Algebraic Curves Volume I Vol 267 A Series of Comprehensive Studies in Mathematics New York Springer 1985 isbn: 978-0-387-90997-4 [Arb]

[Che13] Dawei Chen Strata of abelian differentials and the Teichmüller dynamics In Journal of Modern Dynamics 7.1 (2013) pp. 135–152 issn: 1930-5311, 1930-532X doi: 10.3934/jmd.2013.7.135 url: https://doi.org/ [Che]

[CMR] Ciro Ciliberto Rick Miranda and Joaquim Roe Contact Invariants for Plane Curves in a Pencil 2026 arXiv: 2605.27623 math/AG url: https://arxiv.org/abs/math/0311312 [CMR]

[CV] Ciro Ciliberto and Alessandro Verra On the variety of tritangential planes to a general K3 surface of degree 6 and genus 4 in P 4 2024 arXiv: 2406.00822 math/AG url: https://arxiv.org/abs/math/AG [CV]

[EH] David Eisenbud and Joe Harris All That A Second Course in Algebraic Geometry Vol. 159 Cambridge Studies in Advanced Mathematics Cambridge University Press 2016 doi: 10.1017/CBO9781139062046 [EH]

[FNR] L M Feher A Nemethi and R Rimanyi Coincident root loci of binary forms 2003 arXiv: math/AG math/AG url: https://arxiv.org/abs/math/0311312 [FNR]

[FNR5] L M Fehér A Némethi and R Rimanyi Degeneracy of Two and Three Forms In Canadian Mathematical Bulletin 48.4 (2005) pp. 547–560 url: https://rimanyi.web.unc.edu/wp-content/uploads/sites/9870/2018/07/formak6.pdf [FNR]

[GT] Iulia Gheorghita and Nicola Tarasca Incidence varieties in the projectivized kth Hodge bundle over curves with rational tails In Communications in Contemporary Mathematics 26.5 (2024) Paper No. 2350015, 47 issn: 0219-1997,1793-6683 doi: 10.1142/S0219199723500153 url: https://doi.org/ [GT]

[Lar] Hannah Larson The Chow ring of the universal Picard stack over the hyperelliptic locus 2024 arXiv: 2404.12607 math/AG url: https://arxiv.org/abs/math/AG [Lar]

[Lor] Andrea Di Lorenzo Intersection theory on moduli of smooth complete intersections 2022 arXiv: 2301.06024 math/AG url: https://arxiv.org/abs/math/AG [Lor]

[Mat] Arthur Mattuck Secant Bundles on Symmetric Products In American Journal of Mathematics 87.4 (1965) pp. 779–797 doi: 10.2307 / 2373208 url: https://ams.org/ [Mat]

[Sau] Adrien Sauvaget Cohomology Classes of Strata of Differentials In Geometry & Topology 23.3 (2019) pp. 1085–1171 doi: 10.2140 / gt. 2019. 23. 1085 url: https://doi.org/ [Sau]

[ST] Hunter Spink and Dennis Tseng P GL2-equivariant strata of point configurations in P 1. 2018 arXiv: math/AG url: https://arxiv.org/abs/math/AG [ST]

[Stacks] The Stacks Project Authors The Stacks Project 2026 url: https://stacks.math.columbia.edu [Stacks]

[Arb+85] E Arbarello et al Geometry of Algebraic Curves Volume I Vol 267 A Series of Comprehensive Studies in Mathematics New York Springer 1985 isbn: 978-0-387-90997-4 [Arb]

[Che13] Dawei Chen Strata of abelian differentials and the Teichmüller dynamics In Journal of Modern Dynamics 7.1 (2013) pp. 135–152 issn: 1930-5311, 1930-532X doi: 10.3934/jmd.2013.7.135 url: https://doi.org/ [Che]

[CMR] Ciro Ciliberto Rick Miranda and Joaquim Roe Contact Invariants for Plane Curves in a Pencil 2026 arXiv: 2605.27623 math/AG url: https://arxiv.org/abs/math/0311312 [CMR]

[CV] Ciro Ciliberto and Alessandro Verra On the variety of tritangential planes to a general K3 surface of degree 6 and genus 4 in P 4 2024 arXiv: 2406.00822 math/AG url: https://arxiv.org/abs/math/AG [CV]

[EH] David Eisenbud and Joe Harris All That A Second Course in Algebraic Geometry Vol. 159 Cambridge Studies in Advanced Mathematics Cambridge University Press 2016 doi: 10.1017/CBO9781139062046 [EH]

[FNR] L M Feher A Nemethi and R Rimanyi Coincident root loci of binary forms 2003 arXiv: math/AG math/AG url: https://arxiv.org/abs/math/0311312 [FNR]

[FNR5] L M Fehér A Némethi and R Rimanyi Degeneracy of Two and Three Forms In Canadian Mathematical Bulletin 48.4 (2005) pp. 547–560 url: https://rimanyi.web.unc.edu/wp-content/uploads/sites/9870/2018/07/formak6.pdf [FNR]

[GT] Iulia Gheorghita and Nicola Tarasca Incidence varieties in the projectivized kth Hodge bundle over curves with rational tails In Communications in Contemporary Mathematics 26.5 (2024) Paper No. 2350015, 47 issn: 0219-1997,1793-6683 doi: 10.1142/S0219199723500153 url: https://doi.org/ [GT]

[Lar] Hannah Larson The Chow ring of the universal Picard stack over the hyperelliptic locus 2024 arXiv: 2404.12607 math/AG url: https://arxiv.org/abs/math/AG [Lar]

[Lor] Andrea Di Lorenzo Intersection theory on moduli of smooth complete intersections 2022 arXiv: 2301.06024 math/AG url: https://arxiv.org/abs/math/AG [Lor]

[Mat] Arthur Mattuck Secant Bundles on Symmetric Products In American Journal of Mathematics 87.4 (1965) pp. 779–797 doi: 10.2307 / 2373208 url: https://ams.org/ [Mat]

[Sau] Adrien Sauvaget Cohomology Classes of Strata of Differentials In Geometry & Topology 23.3 (2019) pp. 1085–1171 doi: 10.2140 / gt. 2019. 23. 1085 url: https://doi.org/ [Sau]

[ST] Hunter Spink and Dennis Tseng P GL2-equivariant strata of point configurations in P 1. 2018 arXiv: math/AG url: https://arxiv.org/abs/math/AG [ST]

[Stacks] The Stacks Project Authors The Stacks Project 2026 url: https://stacks.math.columbia.

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. Enhanced enumerative geometry computation for curves by providing universal formulas for relative de Jonquières loci. The improved system can compute classes of these loci in terms of tautological classes pulled back from the universal Picard stack Picd g, allowing for calculations even when both the curve and line bundle vary in moduli.

  2. Computation of GL2-equivariant classes of coincident root loci for binary forms is possible using universal polynomials derived from Theorem 1.5. This allows AI systems to determine fundamental classes like [Xmu] for these loci, matching known results from literature such as [FNR03].

  3. Calculation of the Chern classes of weighted bundles of principal parts Fnu(L) is enabled through Proposition 2.5 and Lemma 3.1. The system can compute c(Fnu(L)) explicitly by summing monomials in twisted kappa classes, which are then pushed forward via g to yield the polynomials f i nu in A∗ (Picd g).

  4. Computation of the fundamental class of universal mu-incident lines is possible using Corollary 5.3. The system can determine [Smu] by evaluating g c(nu) nu at ci = 0, which yields a formula that depends on the partition nu and the universal polynomials f i nu.

  5. Extension of enumerative formulas to complete intersections in projective space is achieved via Theorem 5.6 and Proposition 5.2. The improved system can compute the class of universal mu-incident lines [Sigmae mu] in the moduli space MGLr+1 (d) by using a summation over the degrees of the complete intersection components.

  6. Determination of characteristic classes for complete intersection loci is possible by utilizing Theorem 5.6 and Proposition 5.2 to compute [Sigmae mu] as a sum involving ctop(T v i). This allows for the computation of classes in moduli spaces constructed through towers of Grassmann bundles like MGLr+1 (d).

Abstract

Given a family of genus g curves f:Ψ to S and a relative degree d line bundle L on Ψ such that f*L is a vector bundle, the projectivization P(f*L) is stratified by relative de Jonquières loci, which parameterize sections of L with prescribed vanishing orders. We give universal formulas for the classes of these loci in terms of the relative O(1) on P(f*L) and tautological classes pulled back from the universal Picard stack Pic g d. When f is the family of lines in P r, we use the universal polynomials to compute GL r+1-equivariant classes of loci of hypersurfaces and complete intersections admitting lines with prescribed contact orders.

Sources

Related papers