🔢 mathematics

Transversal unfoldings of derived foliations

This paper develops a homotopically invariant, Chevalley-Eilenberg-based formalism for transversal unfoldings of relative derived foliations by introducing a "transverse controller" that classifies integrable transverse deformations and unifies various geometric cases, including classical, Lie-algebroid, and shifted Poisson structures, under a single framework involving derived splitting spaces and global descent.

Original authors: Maurício Corrêa

Published 2026-06-04

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Maurício Corrêa

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are looking at a complex, multi-layered cake. In mathematics, this cake represents a geometric shape with a "foliation"—think of it as a stack of infinitely thin, interwoven sheets or leaves. Sometimes, these sheets are smooth and perfect; other times, they are crumpled, torn, or have holes (singularities).

This paper, written by Maurício Corrêa, introduces a new, high-tech way to study how these "leafy" shapes can change or move when you wiggle a control knob (a parameter).

Here is the breakdown of the paper's ideas using everyday analogies:

1. The Problem: Moving the Leaves

Usually, if you have a stack of paper (a foliation) and you want to see what happens when you slide it slightly, you just look at the movement. But in the "derived" world (a sophisticated branch of math that deals with hidden, higher-dimensional information), things are messy. The leaves might be crumpled, and simply sliding them isn't enough. You need to know how the crumples change and how the whole stack stays connected.

The author asks: If we have a stack of leaves that depends on a parameter (like time or temperature), how do we describe the "unfolding" of that stack?

2. The Solution: The "Transverse Controller"

The paper invents a tool called a Transverse Controller. Think of this as a remote control for the leaf stack.

The Leaves (The Foliations): These are the layers of your cake.
The Parameter (The Knob): This is the thing you turn to change the cake (e.g., time).
The Unfolding: This is the act of turning the knob and having the cake react in a specific, smooth way. It's not just the cake changing shape; it's the cake changing shape because you are moving the knob, and the movement itself is perfectly smooth (flat).

The Transverse Controller is the mathematical object that tells you: "If I turn the knob this way, how do the leaves move sideways?" It acts like a map that connects the movement of the knob to the movement of the leaves.

3. The "Flatness" Requirement

The paper emphasizes that this movement must be "flat" (or integrable).

Analogy: Imagine you are walking on a staircase. If the stairs are "flat" (integrable), you can walk up them smoothly without tripping. If they are "curved" or "twisted" in the wrong way, you might fall.
In this math, "flatness" means the leaves don't get tangled or twisted in a way that breaks the structure. The paper proves that finding a valid "unfolding" is exactly the same as finding a "flat path" on this remote control map.

4. Handling the "Glitchy" Parts (Singularities)

Sometimes the leaf stack has "glitches" or "knots" (singularities). In old math, if you hit a knot, the whole theory would break.

The Paper's Trick: It uses a "homotopically invariant" approach. Think of this as using elastic rubber bands instead of rigid steel rods. If you hit a knot, the rubber band stretches and goes around it, keeping the connection alive.
The author introduces a "Crossed Module" (a fancy name for a two-part controller).
- Part A: The main controller (the remote).
- Part B: The "stabilizers" (the safety nets).
- If the leaves have a knot (central isotropy), the old math would just cut the knot off and ignore it. This paper says, "No, keep the knot!" It keeps the knot as a "stabilizer" in the background, ensuring that even if the leaves are twisted, the math knows exactly how they are twisted.

5. The "Gauss-Manin" Transport (The Conveyor Belt)

One of the coolest results is about Crystals. In this context, a "crystal" is just a piece of data or a pattern sitting on top of the leaves.

The Analogy: Imagine you have a pattern painted on the leaves of a tree. As the tree grows (the parameter changes), you want to carry that pattern along with the leaves without smearing it.
The paper shows that if you have a "flat unfolding" (a smooth remote control), you can build a conveyor belt (called a Gauss-Manin connection). This conveyor belt moves your pattern from one state of the tree to the next perfectly, preserving its shape and color.
The paper proves that this conveyor belt works even for the "glitchy" parts of the tree, as long as you use the new "elastic" math.

6. The "Suwa" Term (The Secret Ingredient)

The paper references an older idea by a mathematician named Suwa.

Analogy: Suwa realized that to move a leaf stack, you need a little extra "glue" (a term like $h d\epsilon$ ) to make sure the movement is smooth.
This paper shows that Suwa's "glue" is actually just a specific way of reading the Maurer-Cartan equation (a complex math formula that ensures things stay flat). It's like realizing that the "glue" is just the instruction manual for how to walk up the stairs without tripping.

Summary of the Big Claim

The paper claims to have built a universal, flexible toolkit for understanding how complex, crumpled leaf-stacks (derived foliations) can be unfolded or moved.

It replaces rigid, breakable math with elastic, rubber-band math (homotopy theory).
It creates a remote control (the Transverse Controller) that tells you exactly how to move the leaves.
It proves that finding a smooth way to move the leaves is the same as finding a flat path on this remote control.
It shows that if you have this smooth path, you can build a conveyor belt to carry any pattern (crystal) along the leaves without breaking it.

In short: The paper gives us a new, robust way to navigate the twists and turns of complex geometric shapes, ensuring that even when things get messy, we can still move them smoothly and carry our data along for the ride.

Technical Summary: Transversal Unfoldings of Derived Foliations

Problem Statement
The paper addresses the problem of controlling the transverse transport of relative derived foliations along a parameter space. While derived foliations, introduced by Toën and Vezzosi, provide a homotopical enhancement of singular foliations via graded mixed de Rham algebras, the behavior of these structures under deformation requires a precise derived framework. Specifically, the author distinguishes between a general deformation (where the leafwise structure changes with a parameter) and an unfolding. An unfolding is defined as an integrable transverse deformation where the parameter directions themselves lift to the total space as flat infinitesimal symmetries. The central challenge is to identify the derived object that classifies such unfoldings, particularly in the presence of singularities and non-effective actions (where stabilizers or central isotropy exist), and to establish the homotopical invariance of this classification independent of specific algebraic models.

Methodology
The paper develops a formalism based on Chevalley–Eilenberg (CE) presentations and homotopical algebra. The methodology proceeds through several stages:

Affine Strict Models: The author first works in the strict affine setting where the relative derived foliation is presented by a cofibrant, strictly perfect relative dg-Lie algebroid $\rho: \mathfrak{g} \to T_{A/B}$ . In this context, the "transverse controller" is constructed as the dg-Lie algebra of basic first-order derivations of $\mathfrak{g}$ modulo inner derivations.
Handling Non-Effectivity: When the inner derivation map $\iota: \mathfrak{g} \to D^1_{\text{bas}}(\mathfrak{g}/B)$ is not injective, the ordinary quotient fails to retain stabilizer data. The paper replaces the quotient with a crossed module (a two-term $L_\infty$ -algebroid) $Uder(\mathfrak{g}/B) = [\mathfrak{g} \to D^1_{\text{bas}}(\mathfrak{g}/B)]$ , which acts as the homotopy quotient.
Intrinsic Formulation: To ensure the results are independent of the chosen CE presentation, the paper constructs an intrinsic "graded-mixed controller" defined as the homotopy quotient of weight-zero graded mixed derivations of the derived de Rham algebra $DR(F/S)$.
Descent and Globalization: Using hypercovers and descent theory for derived stacks, the local affine classification is extended to a global statement. The paper establishes that the classification of global unfoldings is equivalent to the derived space of flat splittings of the descended intrinsic controller.
Deformation and Transport: The paper analyzes the deformation theory of a fixed unfolding via a Maurer–Cartan complex and constructs a Gauss–Manin connection for crystals linearized by the unfolding.

Key Contributions and Results

Classification of Unfoldings: The primary result (Theorem 5.6 for the effective affine case and Theorem 7.5 for the non-effective case) establishes a bijection (or equivalence of derived groupoids) between strict transversal unfoldings and flat splittings of the transverse controller.
- In the effective case, unfoldings are classified by flat splittings of the dg-Lie algebroid $uder(\mathfrak{g}/B) \to T_{B/k}$ .
- In the non-effective case, unfoldings are classified by the derived mapping space of flat higher splittings of the crossed module $Uder(\mathfrak{g}/B)$ .
Homotopy Invariance: Theorem 9.3 and Corollary 9.4 prove that the transverse controller and the resulting classification are independent of the chosen CE presentation. The intrinsic controller is identified as the homotopy quotient $[T_{F/S} \to D^{\text{gm}}_{\text{bas}}(DR(F/S))]$ .
Global Descent Theorem: Under explicit hypotheses regarding CE-presentability and the descent of controllers (Hypothesis 8.6), Theorem 8.8 provides a global equivalence: transversal derived unfoldings of $F/S$ are equivalent to the derived space of flat splittings of the descended controller over the base tangent algebroid.
Obstruction Theory: The paper identifies the Atiyah–Kodaira–Spencer class (Definition 13.1) as the obstruction to the existence of a transverse splitting. The curvature of a chosen splitting serves as the obstruction to flatness (Proposition 13.4). In the abelian case, the existence of a flat splitting is governed by a cohomology class in $H^2(T_{B/k}, \ker(\text{anchor}))$ (Theorem 14.6).
Split Normal Form and Maurer–Cartan: The paper derives a split normal form for the unfolded de Rham algebra (Proposition 15.3). The condition for the total mixed differential to square to zero is shown to be the Maurer–Cartan equation for the transverse Lie-derivative term, generalizing Suwa's classical transverse term $h d\varepsilon$ .
Gauss–Manin Connection and Characteristic Classes: For coherently Cartan-linearized crystals, the paper constructs a flat Gauss–Manin connection (Theorem 10.6). It is shown that this connection preserves the Cartan-linearized weight/Hodge filtration (Proposition 17.2) and that the foliated Chern character of such crystals is horizontal (Proposition 17.3).

Significance and Claims
The paper claims to provide a unified, homotopically invariant framework for transversal unfoldings that subsumes and extends classical theories.

Unification: The construction recovers classical results for smooth foliations, Lie algebroids, and logarithmic structures, while also addressing shifted Poisson geometry and representable leaf spaces.
Derived Necessity: The work emphasizes that for singular derived foliations, the classical truncation is insufficient. The derived structure (specifically the graded mixed differential and the homotopy quotient of derivations) is necessary to correctly capture integrability and transverse symmetries, particularly in the presence of central isotropy where ordinary quotients fail.
Structural Clarity: By identifying the transverse controller as a specific homotopy quotient, the paper clarifies the geometric nature of unfoldings as "integrable transverse deformations" where the parameter directions act as flat symmetries.
Limitations: The global descent theorem is conditional on specific hypotheses regarding CE-presentability and the existence of derived pushforwards of the controller (Hypothesis 8.6). The paper does not claim to construct a general pushforward theory for all Lie-algebroid stacks but isolates the specific descent data required for this classification.

The author notes that the results are derived consequences of the classification theorem rather than additional data, specifically regarding the horizontality of characteristic classes and the existence of the Gauss–Manin connection. The work serves as a foundational formalism for studying the deformation theory of derived foliations and their associated cohomological invariants.