🔢 mathematics

The Equivalence Problem for Generalized Airy Operators

This paper establishes degree obstructions to the equivalence of generalized Airy operators, thereby resolving a question posed by Nicholas M. Katz, with key results obtained through collaboration between the authors and the MechMath Agent Team.

Original authors: Yichuan Cao, Ruyong Feng, Yunfei Li, Ruichen Qiu

Published 2026-07-02

📖 4 min read🧠 Deep dive

CC BY 4.0

Original authors: Yichuan Cao, Ruyong Feng, Yunfei Li, Ruichen Qiu

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 a master locksmith trying to figure out if two very complex, custom-made keys are actually the same key, just slightly worn or viewed from a different angle. In the world of advanced mathematics, these "keys" are called Generalized Airy Operators. They are special formulas used to describe how things change (differential equations), and they come in specific shapes defined by two numbers, let's call them $n$ and $m$ .

The big question this paper answers is: If two of these "keys" are mathematically equivalent (meaning they can be transformed into one another without losing their essential nature), are they actually identical down to the last detail?

For a long time, a famous mathematician named Nicholas Katz wondered if the answer was "yes." This paper says, "Yes, they are."

Here is how the authors figured it out, using some creative mental models:

1. The "Fingerprint" of the Key

To compare these complex formulas, the authors didn't look at the whole thing at once. Instead, they zoomed in on the "edge" of the problem (mathematically, looking at infinity). They treated the formulas like a recipe that produces a specific flavor.

They discovered that every one of these operators has a unique "fingerprint" made of numbers. If you try to transform one operator into another, this fingerprint has to match perfectly. The authors found that the "shape" of the fingerprint (specifically, the degree or size of the numbers involved) acts like a strict security guard.

2. The "Degree Obstruction" (The Traffic Light)

The core of their discovery is something they call a "degree obstruction."

Think of the numbers in the formula as cars driving on a highway. The authors set up a traffic light system based on the "speed" (or degree) of these cars.

If you try to transform Operator A into Operator B, the "cars" in the transformation must follow specific speed limits.
The authors proved that if the two operators are different (even slightly), the math forces the "cars" to break the speed limit. It's like trying to drive a car through a wall; the physics of the equation simply won't allow it.
Because the "traffic light" turns red for any attempt to make two different operators look the same, the only way the transformation works is if the two operators were already identical to begin with.

3. The "AI Co-Pilot"

A unique part of this story is how the math was done. The authors mention that they worked closely with an artificial intelligence team called MechMath Agent Team (MMAT).

Think of the human authors as the captains of a ship and the AI as a highly advanced navigation system. The captains knew the destination (solving Katz's question), but the journey required navigating through incredibly dense mathematical fog. The AI helped calculate the complex "degree obstructions" and verify the steps, ensuring the ship didn't crash into a reef of errors. It was a true partnership between human intuition and machine precision.

4. The Final Verdict

The paper concludes with a definitive answer to Katz's question from 1987:

The Question: If two Generalized Airy Operators of the same type are equivalent, are they the same?
The Answer: Yes. If you can turn one into the other, they are exactly the same formula. There are no "look-alikes" or "near-misses."

They also applied this logic to a specific property called "self-duality" (whether an operator is its own mirror image). They found that an operator is its own mirror image only if it has a very specific, symmetrical structure (like a butterfly with perfectly matching wings).

Summary

In simple terms, this paper proves that in this specific mathematical universe, you cannot fake it. If two of these special formulas are related, they are twins. The authors used a new method of checking "speed limits" (degree obstructions) to prove that any attempt to make two different formulas look the same is mathematically impossible. They solved a 37-year-old mystery using a mix of human insight and AI assistance.

Technical Summary: The Equivalence Problem for Generalized Airy Operators

Problem Statement

This paper addresses the Equivalence Problem for generalized Airy operators over an algebraically closed field $k$ of characteristic zero. Specifically, it investigates whether two generalized Airy operators of the same type $(n, m)$ that are equivalent (in the sense of differential operator transformation) must necessarily be identical.

A generalized Airy operator of type $(n, m)$ is defined as $L = P(\partial) + Q(x)$ , where $P, Q \in k[x]$ , $P$ is monic of degree $n$ , $Q$ is of degree $m$ , and $Q(0)=0$ . Two operators $L_1$ and $L_2$ are considered equivalent ( $L_1 \sim L_2$ ) if there exists a nonzero differential operator $T$ such that the greatest common right-hand divisor of $T$ and $L_1$ is 1, and the least common left-hand multiple of $T$ and $L_1$ equals $L_2 T$ .

The central question, originally posed by Nicholas M. Katz in 1987, asks:

Question 1.3: Suppose $L_1$ and $L_2$ are two generalized Airy operators of type $(n, m)$ with $\gcd(n, m) = 1$ . If $L_1 \sim L_2$ , does it follow that $L_1 = L_2$ ?

Previous work by Saidane (2000) provided affirmative answers for specific low-order cases (up to order 3), but the general case remained open.

Methodology

The authors employ a combination of formal asymptotic analysis, valuation theory, and degree-theoretic arguments to derive necessary conditions for equivalence. The methodology proceeds through the following stages:

Formal Solutions at Infinity: The authors analyze the formal solutions of generalized Airy operators at the point $x = \infty$ (using the variable $t = 1/x$ ). By applying the Newton–Puiseux theorem, they characterize the roots of the characteristic equation $P(Y) + Q(x) = 0$ in the field of Puiseux series. They establish that the solution space of $L(y)=0$ admits a basis of the form $\exp(\int \eta_i dx) h_i$ , where $\eta_i$ are roots of the characteristic equation and $h_i$ are specific Puiseux series.
Logarithmic Contact Order: A key technical tool introduced is the logarithmic contact order $\mu_{z,T}$ , defined for a solution $z$ and a transformation operator $T$ as:
$\mu_{z,T} = \text{ord}_t \left( \frac{\delta(T(z))}{T(z)} - \frac{\delta z}{z} \right)$
This quantity measures the valuation difference between the logarithmic derivatives of the transformed and original solutions.
Degree Obstructions: The core of the proof involves deriving degree obstructions. By analyzing the valuations of the coefficients of the operators and the solutions, the authors establish necessary conditions that must hold if two distinct operators are equivalent. These conditions relate the degrees of the polynomial parts of the operators ( $n, m$ ) and the degrees of their differences ( $r = \deg(P_1 - P_2)$ , $s = \deg(Q_1 - Q_2)$ ) to the logarithmic contact order.
Newton Polygon Analysis: The authors utilize the Newton polygon of the transformation operator $T$ to constrain its possible slopes. They prove that if the order of $T$ is less than $n$ and the slope conditions are met, the valuation of the transformed solution cannot be zero, leading to contradictions if equivalence is assumed between distinct operators.
Collaboration with AI: The authors explicitly note that the main results in Sections 3 and 4 were obtained through close interactive collaboration with the MechMath Agent Team (MMAT), an artificial intelligence agent system.

Key Contributions and Results

1. Affirmative Answer to Katz's Question

The primary result of the paper is Theorem 4.4, which provides an affirmative answer to Katz's Question 1.3:

Theorem 4.4: Let $L_1$ and $L_2$ be two generalized Airy operators of type $(n, m)$ with $nm \neq 0$ and $\gcd(n, m) = 1$ . If $L_1$ and $L_2$ are equivalent over $k(x)$ , then $L_1 = L_2$ .

The proof relies on Theorem 3.8, which establishes that if $L_1 \neq L_2$ and they are equivalent, at least one of three specific algebraic equalities involving the degrees $n, m, r, s$ and the contact order $\mu$ must hold. Under the condition $\gcd(n, m) = 1$ , the authors demonstrate that all three equalities lead to contradictions (e.g., implying $n=0$ or $s=0$ when $s \neq 0$ ), thereby proving that $L_1$ must equal $L_2$ .

2. Characterization of Self-Duality

As a direct application of the equivalence result, the paper resolves the question of self-duality for generalized Airy operators. An operator $L$ is self-dual if it is equivalent to its adjoint $L^*$ .

Theorem 4.6: Let $L = P(\partial) + Q(x)$ be a generalized Airy operator of type $(n, m)$ with $\gcd(n, m) = 1$ . Then $L$ is self-dual if and only if:

$n$ is even.

$P(Y) = S(Y^2)$ for some polynomial $S \in k[Y]$ .

This result follows immediately from Theorem 4.4, as self-duality implies $L \sim L^*$ , which forces $L = L^*$ . The condition $L = L^*$ imposes strict parity constraints on the coefficients of $P(\partial)$ .

3. Degree Obstructions

The paper establishes Theorem 3.8, a general necessary condition for the equivalence of two generalized Airy operators. It states that if $L_1 \neq L_2$ are equivalent via an operator $T$ , then specific relations between the degrees of the differences of their polynomial components and the logarithmic contact order must hold. These obstructions serve as a powerful tool for ruling out equivalence in the coprime case.

Significance and Claims

The paper claims to settle a long-standing open problem in the theory of differential equations and differential Galois groups. By answering Katz's question, the authors clarify the structure of generalized Airy operators, confirming that within the class of operators where $\gcd(n, m) = 1$ , the equivalence relation is trivial (i.e., it only relates an operator to itself).

The significance of this work lies in:

Completing the classification: It extends previous partial results (order $\le 3$ ) to the general case for coprime $n$ and $m$ .
Differential Galois Theory: The result reinforces the understanding that the differential Galois group of such operators is rigid; the operator is uniquely determined by its equivalence class.
Methodological Innovation: The derivation of degree obstructions provides a new framework for analyzing the equivalence of linear differential operators with polynomial coefficients.
AI Collaboration: The paper highlights the role of the MechMath Agent Team (MMAT) in deriving the main technical proofs, marking a notable instance of human-AI collaboration in advanced mathematical research.

The authors do not propose new applications or future experimental directions, focusing strictly on the theoretical resolution of the equivalence problem and its immediate consequences for self-duality.