Reconstruction of F-cohomological field theories on moduli of compact type

This paper establishes an analogue of the Givental-Teleman reconstruction theorem for F-cohomological field theories on the moduli space of compact type, which is then applied to reconstruct extended $r$-spin classes and derive new relations between $\kappa$-classes.

The Gist

Imagine you are an architect trying to understand the blueprints of a massive, infinitely complex city. This city isn't made of bricks and mortar, but of mathematical shapes called "moduli spaces." These spaces represent every possible way you can draw a curve (like a circle, a figure-eight, or a pretzel) with specific points marked on it.

In the world of mathematics, there is a powerful tool called a Cohomological Field Theory (CohFT). Think of a CohFT as a universal "stamp" or "rulebook" that tells you how to assign a specific value (a number or a shape) to every possible configuration of these curves. It's like a master key that unlocks the secrets of the city's geometry.

For a long time, mathematicians had a perfect way to rebuild this master key if they knew the "ground floor" of the city (the simplest, most basic shapes). This was the Givental–Teleman reconstruction. It was like saying, "If I know the rules for the flat, empty parks, I can mathematically predict the rules for the entire skyscraper district."

The Problem: A Broken Compass

However, there is a newer, slightly different version of this rulebook called an F-CohFT (F-cohomological field theory). It's a bit more flexible and handles "open" boundaries (like a curve with a loose end) differently.

The problem was that the old "master key" (Givental–Teleman) didn't work for this new F-version. When mathematicians tried to use the ground floor rules to predict the complex skyscrapers for F-CohFTs, the math broke down. The "compass" they used to navigate the city was spinning wildly. It seemed impossible to reconstruct the full theory just from the simple parts.

The Solution: The "Compact Type" Shortcut

The authors of this paper, Gaëtan Borot, Silvia Raggi, and Paolo Rossi, found a clever workaround. They realized that while the whole city is chaotic, there is a specific, well-behaved neighborhood called the "Moduli of Compact Type."

The Analogy:
Imagine the city has a chaotic, tangled downtown (where curves can loop back on themselves in complex ways) and a very orderly, tree-like suburb (where the curves are simple and don't have loops).

• The old method failed because it tried to map the whole tangled downtown.
• The authors realized that if you only look at the orderly, tree-like suburb (the "compact type"), the chaos disappears.

They proved that for this specific neighborhood, the "broken compass" actually works perfectly again. They showed that if you have a "flat" starting point (a simple algebraic structure), you can uniquely and perfectly reconstruct the entire rulebook for this neighborhood using a specific set of mathematical tools (the F-Givental group).

How They Did It (The Metaphor)

To prove this, they used a technique similar to patching a quilt.

1. The Strata: They broke the complex city down into smaller, manageable patches (strata).
2. The Boundary: They looked at how these patches touch each other.
3. The Stitching: They showed that if you know the rules for the "smooth" parts of the city (the open curves), you can mathematically "stitch" them together to figure out the rules for the "compact" parts. They proved that there is only one way to stitch them together without creating ripples or tears.

The Big Payoff: The "Extended r-Spin" Mystery

Why does this matter? The authors applied their new method to solve a specific, long-standing puzzle involving something called the "extended r-spin class."

The Analogy:
Imagine you have a special type of rubber band (the r-spin class) that can be stretched in $r$ different ways. Mathematicians had been trying to figure out exactly how these bands behave when you twist them in complex ways.

• Using their new reconstruction method, the authors were able to take the simple, flat rules of the rubber band and predict exactly how it behaves in the complex, compact neighborhood.
• The Result: They discovered a hidden pattern. They found that certain complicated mathematical formulas (involving $\kappa$-classes, which are like "twist counters" for the curves) actually vanish (become zero) under specific conditions.

It's as if they predicted that if you twist a rubber band exactly $r$ times in a specific way, it will suddenly snap back to zero length, and they proved exactly when and why this happens.

Summary

• The Goal: Rebuild a complex mathematical rulebook (F-CohFT) from its simplest parts.
• The Obstacle: The standard method failed for the new, flexible version of the rulebook.
• The Breakthrough: The authors realized that if you restrict your view to a specific, orderly part of the mathematical city (the "compact type"), the reconstruction works perfectly.
• The Result: They successfully rebuilt the rulebook and used it to solve a mystery about how certain mathematical "rubber bands" behave, finding new relationships and proving that some complex formulas are actually zero.

In short, they found a way to navigate a chaotic mathematical maze by focusing on the one part of the maze that follows a straight line, proving that the whole maze is actually much more orderly than anyone thought.

Technical Summary

1. Problem Statement and Context

Background:
Cohomological Field Theories (CohFTs) are algebraic structures on the cohomology of moduli spaces of stable curves ($\overline{\mathcal{M}}_{g,n}$) that formalize Gromov-Witten theory. A landmark result by Givental and Teleman established that in the semi-simple case, a CohFT is uniquely determined by its genus 0 data (a Frobenius manifold) via the action of the Givental group. This allows for the reconstruction of higher-genus invariants from genus 0 data.

The Challenge:
F-Cohomological Field Theories (F-CohFTs) are a generalization introduced in [BR21] where the compatibility condition is relaxed to the moduli space of curves with compact type (dual graphs are trees), and permutation invariance is reduced to a single marked point.

• The genus 0 part of an F-CohFT defines a flat F-manifold (a weaker structure than a Frobenius manifold, lacking a flat metric).
• While an F-Givental group action exists on F-CohFTs, it is not transitive in general. Specifically, the reconstruction of an F-CohFT from its underlying flat F-manifold fails for the full moduli space $\overline{\mathcal{M}}_{g,n}$ because higher-genus potentials are sensitive to classes on the non-compact-type locus (loops in the dual graph).

The Goal:
The authors aim to prove an analogue of the Givental–Teleman reconstruction theorem for F-CohFTs by restricting the domain to the moduli space of compact type ($\mathcal{M}^{ct}_{g,n}$). They seek to determine under what conditions an F-CohFT is uniquely reconstructible from its genus 0 data (flat F-manifold) within this restricted setting.

2. Methodology

The authors employ a strategy inspired by Teleman's original proof for CohFTs but adapted to the specific geometric and algebraic constraints of F-CohFTs.

Key Technical Steps:

1. Restriction to Compact Type:
   The authors focus on the restriction of F-CohFTs to $\mathcal{M}^{ct}_{g,n}$. They define invertible F-CohFTs (where the degree 0 part $\alpha$ is invertible) and semi-simple flat F-manifolds.

2. Geometric Patching and Stability:

   • They introduce free-boundary and pinned-boundary F-CohFTs defined on moduli spaces of hyperbolic surfaces with boundaries.
   • Using hyperbolic geometry, they construct tubular neighborhoods of strata in $\mathcal{M}^{ct}_{g,n}$ (strata corresponding to stable trees).
   • They utilize stability theorems (Harer, Looijenga, Madsen-Weiss) to analyze the cohomology of these moduli spaces in the "stable range" (large genus). This allows them to treat $\kappa$ and $\psi$ classes as free generators in the limit.

3. The F-Givental Group Action:
   They define the action of the F-Givental group (generated by $R(z) \in \text{End}(V)[[z]]$ and $T(z) \in z^2 V[[z]]$) on F-CohFTs.

   • Theorem A: They prove that the F-Givental group acts freely and transitively on the set of invertible compact-type F-CohFTs with a fixed underlying F-TFT (degree 0 part). This establishes that any such F-CohFT can be written as $R^T \omega$ restricted to compact type.

4. Differential Equations from Flat F-Manifolds:

   • They analyze the formal shift of the F-CohFT to relate the group elements $R(z)$ and $T(z)$ to the geometry of the underlying flat F-manifold.
   • Assuming semi-simplicity, they derive a system of differential equations for $R(z)$ and the vacuum vector $\Upsilon(z)$.
   • These equations characterize $R(z)$ as a matrix of flat sections for a deformed connection $\nabla - z^{-1} \cdot$ on the flat F-manifold.
   • Theorem B: This shows that $R(z)$ is determined by the flat F-manifold up to a diagonal ambiguity (related to Hodge classes and $\kappa$-relations).

5. Conformal Fixing of Ambiguities:

   • For conformal F-CohFTs (those satisfying a scaling property), they introduce an Euler vector field.
   • Theorem C: They prove that for conformal, invertible, semi-simple F-CohFTs, the diagonal ambiguity vanishes. The F-CohFT is uniquely determined by the 1-jet of the conformal flat F-manifold at the origin.

3. Key Contributions and Results

• Theorem A (Transitivity on Compact Type):
  The F-Givental group acts freely and transitively on invertible compact-type F-CohFTs with a given F-TFT. This is the first major step, proving that the "missing" information in the full reconstruction is precisely the data on the non-compact-type locus.
  $$(R^T \omega)|_{ct} = \Omega|_{ct}$$

• Theorem B (Reconstruction from Flat F-Manifolds):
  For semi-simple invertible F-CohFTs, the element $R(z)$ is determined by the flat F-manifold structure. Specifically, the columns of $R(z)H^{-1}e^{U/z}$ form a basis of flat sections for the deformed connection. This links the algebraic reconstruction to the differential geometry of the flat F-manifold.

• Theorem C (Unique Reconstruction for Conformal Cases):
  If the F-CohFT is conformal, the reconstruction is unique. The diagonal ambiguities (which correspond to multiplication by exponentials of Hodge classes or specific $\kappa$-relations) are fixed by the conformal dimension.

• Theorem D (Application to Extended r-Spin Classes):
  The authors apply their theory to the extended $r$-spin class.

  • They construct a 1-dimensional compact-type F-CohFT from the extended $r$-spin theory restricted to the "extended direction" (the subspace $V'$).
  • They derive explicit vanishing relations for polynomials in $\kappa$-classes on $\mathcal{M}^{ct}_{g,n}$.
  • Specifically, they define polynomials $P^{(r)}_m(\kappa)$ and prove:
    $$P^{(r)}_m(\kappa) = 0 \quad \text{if} \quad (r-1)(2g-2+n) < rm$$
    This provides a new family of relations in the tautological ring of the moduli space of compact type, distinct from but related to Pixton's relations.

4. Significance

1. Resolution of the Reconstruction Problem: The paper successfully adapts the powerful Givental–Teleman machinery to the F-CohFT setting, a crucial step for understanding non-hamiltonian integrable hierarchies associated with F-CohFTs (like the Double Ramification hierarchies).
2. Geometric Insight: It clarifies the role of the "compact type" restriction. The failure of reconstruction on the full moduli space is due to the sensitivity to non-tree dual graphs; restricting to trees (compact type) restores the transitivity of the group action.
3. New Relations in Tautological Rings: The application to $r$-spin theory yields explicit, non-trivial relations among $\kappa$-classes. These relations are significant for the study of the cohomology of moduli spaces of curves.
4. Bridge between Geometry and Integrability: The results provide a rigorous foundation for constructing integrable hierarchies from F-CohFTs, as the flows of these hierarchies often depend only on the restriction to the compact type.

5. Conclusion

Borot, Ragni, and Rossi establish that while F-CohFTs are generally not reconstructible from their genus 0 data on the full moduli space, they are uniquely reconstructible on the moduli space of compact type provided the theory is invertible and semi-simple. In the conformal case, this reconstruction is entirely determined by the flat F-manifold structure. This work not only generalizes the Givental–Teleman theory but also produces concrete new relations in algebraic geometry, specifically for the extended $r$-spin theory.