🔢 mathematics

$\mathbb T$ -homogeneous locally nilpotent derivations of trinomial algebras

This paper characterizes the $\mathbb{T}$ -homogeneous locally nilpotent derivations of trinomial algebras, which serve as Cox rings for varieties with a complexity-one torus action.

Original authors: Timofey Krasikov, Kirill Rassolov

Published 2026-05-19

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Timofey Krasikov, Kirill Rassolov

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 an architect trying to understand the hidden "movement rules" of a very specific, complex building made of mathematical blocks. This paper is a blueprint that maps out exactly how these blocks can slide, shift, or rotate without the building collapsing.

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

1. The Building: Trinomial Algebras

Think of a Trinomial Algebra as a special kind of structure built from Lego bricks.

The Bricks: These are variables (like $x, y, z$ ).
The Rules: The building is held together by equations where exactly three terms are connected (e.g., $A + B + C = 0$ ). The authors call these "trinomial" because of the "tri" (three).
The Shape: These structures aren't random; they are built to look like they have a "complexity of one." Imagine a building that is mostly a straight line (simple) but has a little bit of extra wiggle room. This specific shape appears often in advanced geometry, specifically when studying spaces that have a "torus action" (a fancy way of saying the building can be spun or stretched in a very organized, circular way).

2. The Movement: Locally Nilpotent Derivations (LNDs)

The paper studies specific types of "movements" or "flows" that can happen inside this building.

The Derivation: Imagine a machine that takes a block, looks at it, and tells you how to change it. If you apply this machine to a block, you get a new block. If you apply it again to the result, you get another.
Locally Nilpotent: This is the key rule. It means that if you keep pressing the "change" button on any specific block, eventually, the block will turn into zero (nothing). It's like a countdown timer: no matter how big the number is, if you keep subtracting, it will eventually hit zero.
Why it matters: In math, finding these "countdown machines" is like finding a secret door. If a building has no such machines, it's "rigid" (stuck). If it has them, it's flexible.

3. The Compass: Homogeneity

The authors are only interested in movements that respect the building's "spinning" symmetry.

Imagine the building is painted with a gradient of colors that changes as you spin around. A "homogeneous" movement is one where the machine only moves blocks of the same color to blocks of the same color (or a predictable shift in color). It doesn't mix red blocks with blue ones randomly.
The paper focuses on movements that are "T-homogeneous," meaning they follow the rules of the main spinning symmetry of the building.

4. The Two Types of Buildings

The authors split their study into two main types of trinomial buildings, which behave differently:

Type 1 (The Simple, Predictable Ones):
- These buildings are very orderly.
- The Discovery: The authors found that every possible "countdown machine" in these buildings is just a copy (a "replica") of a few basic, standard machines.
- The Result: If you know the basic machines, you know them all. They also figured out exactly when these buildings are "rigid" (have no machines at all) and when they are "semirigid" (all machines are essentially the same type).
Type 2 (The Complex, Surprising Ones):
- These buildings are trickier. They have a hidden layer of complexity.
- The Discovery: Here, things get interesting. Not all machines are just copies of the basic ones. There are new, unique machines that appear only in these complex structures.
- The Analogy: In Type 1, if you have a key, it opens a standard lock. In Type 2, you might find a key that opens a lock in a way that creates a whole new family of keys, or even an infinite number of variations.
- The Method: To solve this, the authors used a "shrink-ray" technique. They showed that no matter how huge and complex the building is, you can shrink it down to a simple 2D surface (a "trinomial surface") to figure out the rules. Once they solved the puzzle for the 2D surface, they could expand the answer back up to the big building.

5. The "Secret Sauce" (The Main Results)

The paper provides a complete "menu" of all possible movements for these buildings:

For Type 1: They gave a formula for every possible movement. It turns out they are all variations of a few simple patterns.
For Type 2: They described the "families" of movements. Sometimes there are just two types, but sometimes (depending on the specific numbers in the equations), there is an infinite family of unique movements.
The "Rigidity" Test: They provided a clear checklist to see if a building is stuck (rigid) or flexible. If the building is flexible, they can tell you exactly how it can move.

Summary

In short, this paper is a master catalog of the "secret dances" that can happen inside a specific class of mathematical structures.

Type 1 buildings dance in a very predictable, repetitive way.
Type 2 buildings can dance in wild, infinite variations.
The authors proved that by understanding the dance of a simple 2D version of the building, you can understand the dance of the entire complex structure.

They didn't just guess; they wrote down the exact mathematical formulas (the "steps") for every possible dance, ensuring that no hidden movement was left undiscovered.

Technical Summary: T-Homogeneous Locally Nilpotent Derivations of Trinomial Algebras

Problem Statement
The paper investigates locally nilpotent derivations (LNDs) on trinomial algebras, which are commutative finitely generated algebras defined by systems of compatible relations, each being a trinomial (a polynomial with three terms). These algebras arise as Cox rings of normal rational semiprojective varieties admitting a torus action of complexity one. The central problem is to describe those LNDs that are homogeneous with respect to the natural torus action $T$ of complexity one.

While homogeneous LNDs under a specific quasitorus $H$ (the "elementary" derivations) were previously studied, this work focuses on $T$ -homogeneous LNDs, where $T$ is the neutral component of $H$ . A key motivation is that for non-factorial varieties, $T$ is a strict subgroup of $H$ , potentially allowing for $T$ -homogeneous LNDs that are not $H$ -homogeneous. Furthermore, understanding homogeneous components of LNDs is crucial for analyzing arbitrary LNDs, as the homogeneous components corresponding to the vertices of the Newton polytope of degrees are guaranteed to be locally nilpotent.

Methodology
The authors employ a combination of algebraic techniques and geometric reduction strategies:

Algebraic Decomposition and Coprimality: The analysis relies heavily on the properties of the grading by the character group of $T$ . The authors utilize the fact that the variables in trinomial algebras are pairwise coprime and that the kernel of an LND is factorially closed. They decompose derivations based on which variables they annihilate, distinguishing between derivations acting on the "trinomial" variables ( $T_{ij}$ ) and the "free" variables ( $S_k$ ).
Reduction to Surface Cases: For trinomial algebras of Type 2 (defined by relations involving vectors in $k^2$ ), the authors reduce the problem of describing LNDs on an arbitrary high-dimensional variety to the case of trinomial surfaces. This is achieved by considering the algebraic closure of the field generated by the variables not involved in the specific derivation's support.
Combinatorial and Toric Geometry: For the surface cases, the paper utilizes the combinatorial description of homogeneous LNDs on varieties with torus actions.
- For certain cases (e.g., $x^2 + y^2 + z^\gamma = 0$ with $\gamma > 2$ ), the variety is related to a toric variety, allowing the use of Demazure roots to classify homogeneous LNDs.
- For the case $x^2 + y^2 + z^2 = 0$ , where the variety is not toric in the standard sense relative to the acting torus, the authors apply the theory of proper polyhedral divisors on semiprojective curves (specifically $\mathbb{P}^1$ ) to classify horizontal type LNDs.
Explicit Construction: The paper constructs explicit formulas for these derivations by lifting solutions from the surface level back to the general trinomial algebra, ensuring the resulting maps are well-defined on the quotient algebra.

Key Contributions and Results

The paper distinguishes between two types of trinomial algebras, defined by the structure of their relations:

1. Type 1 Trinomial Algebras:
Defined by relations of the form $T_i^{l_i} - T_{i+1}^{l_{i+1}} - (a_{i+1} - a_i) = 0$ .

Classification: The authors prove that any $T$ -homogeneous LND on a Type 1 algebra is a "replica" (a scalar multiple by an element in the kernel) of an "elementary" LND or a partial derivative with respect to a free variable $S_k$ .
Explicit Formulas: They provide explicit formulas for these derivations (Theorem 2.2), showing they are determined by a tuple of indices $C = (c_1, \dots, c_r)$ where specific exponents equal 1.
Rigidity and Semirigidity: The paper provides a new proof of the rigidity criterion (Corollary 2.4) and establishes a semirigidity criterion (Corollary 3.2). An algebra is semirigid if all non-zero LNDs are equivalent (share the same kernel). The criteria depend on the existence of free variables and the specific exponents in the relations.

2. Type 2 Trinomial Algebras:
Defined by relations requiring the linear dependence of columns in a matrix involving monomials and vectors in $k^2$ .

Equivalence Classes: Theorem 4.1 describes the equivalence classes of $T$ -homogeneous LNDs. Unlike Type 1, Type 2 algebras admit $T$ -homogeneous LNDs that are not replicas of elementary ones.
Infinite Families: The authors identify conditions under which a single tuple $C$ corresponds to an infinite family of equivalence classes of LNDs (e.g., when certain divisibility conditions on the exponents are met).
Explicit Formulas for Single Relations: For algebras defined by a single trinomial relation $T_0^{l_0} + T_1^{l_1} + T_2^{l_2} = 0$ , Theorem 4.4 and Construction 4.3 provide explicit formulas for all $T$ -homogeneous LNDs. These include cases involving complex coefficients (roots of unity) and specific parity conditions on the exponents.
Semirigidity: A semirigidity criterion is derived (Corollary 4.2), showing that Type 2 algebras are semirigid if and only if they are rigid or reduce to a rigid algebra after removing a single free variable.

Significance
The paper claims significance in several areas:

Extension of Previous Work: It generalizes the study of homogeneous LNDs from the "elementary" $H$ -homogeneous case to the broader $T$ -homogeneous case, revealing new derivations that exist specifically in the non-factorial setting.
Structural Insight: By providing explicit formulas and classification theorems, the work clarifies the structure of the automorphism groups (specifically the $\mathbb{G}_a$ -subgroups) of trinomial varieties.
Methodological Utility: The reduction technique, which maps the problem on high-dimensional trinomial varieties to trinomial surfaces, offers a powerful tool for analyzing LNDs on complex algebras without resorting to the full complexity of multidimensional polyhedral divisors for every case.
Rigidity Criteria: The results offer refined criteria for determining when these algebras are rigid (admit no non-zero LNDs) or semirigid, which are fundamental properties in the study of algebraic varieties and their symmetries.

The authors emphasize that their proofs for Type 2 rely on the combinatorial description of LNDs on complexity-one torus actions, bridging the gap between abstract toric geometry and explicit algebraic computations on trinomial varieties.