🔢 mathematics

Diameter bounds for finite simple Lie algebras

This paper establishes strong, explicit diameter bounds for finite simple Lie algebras over finite fields, proving that their diameters grow polylogarithmically with respect to the algebra size for any generating set and logarithmically for random generators in the classical case, thereby providing an analogue to Babai's conjecture for finite simple groups.

Original authors: Marco Barbieri, Urban Jezernik, Matevž Miščič

Published 2026-06-09

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Marco Barbieri, Urban Jezernik, Matevž Miščič

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 have a giant, complex machine made of many tiny gears and levers. This machine is a Lie algebra, a mathematical structure used to describe symmetry and motion. Now, imagine you are given just a few specific levers (a "generating set") and you want to know: How many times do I need to pull, push, or combine these levers to reach every single part of the machine?

The answer to this question is called the diameter. If the diameter is small, it means the machine is "easy" to explore; you can get anywhere quickly. If the diameter is huge, the machine is a maze where it takes forever to find your way.

This paper is about proving that for a specific, very important class of these machines (called finite simple Lie algebras), the maze is actually very small. No matter how big the machine gets, you can reach every corner in a surprisingly short amount of time.

Here is the breakdown of their discovery, using simple analogies:

1. The Big Goal: "Babai's Conjecture" for Machines

Mathematicians have long wondered about similar machines called "groups." A famous guess (Babai's conjecture) says that for these group machines, the time it takes to explore them grows very slowly—only as fast as the logarithm of the machine's size. Think of it like this: if the machine doubles in size, you only need to take a few extra steps to explore it, not double the steps.

This paper asks: Does this same "fast exploration" rule apply to Lie algebras?
The authors say Yes. They prove that for these specific algebra machines, you can reach every part in a number of steps that is roughly proportional to the logarithm of the size.

2. The Two Main Findings

Finding A: The "Any Lever" Guarantee

The Scenario: You are handed any random set of levers that can eventually build the whole machine. You don't get to pick them; you just have to work with what you're given.
The Result: The authors prove that even with a bad or weird starting set of levers, you can still explore the whole machine very quickly.
The Analogy: Imagine you are in a massive library. Even if you start with a very strange, inefficient map, the authors prove you can still walk to every bookshelf in a number of steps that is surprisingly small (specifically, about $3.11$ times the log of the library's size).

The Catch: The math here is a bit "loose." They prove it works, but the number of steps is a bit higher than the absolute theoretical minimum.

Finding B: The "Random Pick" Superpower

The Scenario: Instead of being handed a random set, you get to pick two levers completely at random from the machine.
The Result: This is where things get amazing. If you pick two levers at random, they will almost certainly allow you to explore the entire machine in a number of steps that is just proportional to the log of the size (a much tighter, "sharper" bound).
The Analogy: If you walk into that same massive library and blindly grab two books off the shelf, there is a near-100% chance that those two books contain the keys to unlock every single aisle in the building, and you can do it in record time.

3. How Did They Do It? (The Secret Sauce)

The authors used a clever trick involving covering maps and growth.

The "Covering" Trick:
Imagine the finite Lie algebra (the machine we care about) is a shadow cast by a much larger, infinite machine built with whole numbers (integers). The authors built a "bridge" (a covering map) from this infinite integer machine down to the finite one.
- Why? In the infinite integer world, it's easier to prove that two random levers create a "snowball effect." Every time you combine them, the number of new things you can make doubles (exponential growth).
- The Jump: They showed that if you take a small "ball" of combinations in the infinite world and project it down to the finite world, it stays distinct and large for a while. This proves that in the finite world, you also get a huge explosion of new elements very quickly.
The "Sum-Product" Engine:
To make sure the "snowball" doesn't get stuck, they used a powerful mathematical tool called the Sum-Product Theorem.
- The Metaphor: Imagine you have a pile of numbers. If you keep adding them together or multiplying them, the pile usually grows fast. The authors proved that in these Lie algebras, the "addition" and "bracketing" (a special multiplication rule for these machines) work together to force the set of reachable elements to grow explosively fast, preventing the process from stalling.

4. The "Split" vs. "Non-Split" Twist

The paper distinguishes between two types of these machines:

Split: The "standard" version, like a regular grid.
Non-Split: A twisted version, like a grid wrapped around a cylinder or a Möbius strip.

For the Split machines, the "random two levers" rule works for all large sizes.
For the Non-Split machines, the rule works for almost all sizes (specifically, for a set of prime numbers that covers 99.9%+ of the possibilities). The authors had to use some advanced number theory (like the Chebotarev Density Theorem) to show that the "bad" cases are so rare they barely matter.

Summary

In plain English:
This paper proves that finite simple Lie algebras are not complex mazes. Whether you are forced to use a random set of tools or you get to pick two at random, you can navigate the entire structure incredibly fast. The time it takes to explore the whole thing grows very slowly as the thing gets bigger.

They achieved this by:

Building a bridge to a simpler, infinite version of the problem.
Showing that random tools create an explosion of new possibilities in that infinite version.
Proving that this explosion translates perfectly to the finite version we care about.

This confirms that these mathematical structures are "rapidly generated," just like their cousins, the finite simple groups.

Technical Summary: Diameter Bounds for Finite Simple Lie Algebras

Problem Statement
The paper addresses the problem of determining the diameter of finite simple Lie algebras over finite fields $\mathbb{F}_p$ . Analogous to Babai's conjecture for finite simple groups, which posits that the diameter of a nonabelian finite simple group $G$ with respect to any generating set $A$ is bounded by $O((\log |G|)^D)$ for absolute constants $C, D$ , this work investigates the "linear version" of the conjecture for Lie algebras. Specifically, it seeks to prove that for a nonabelian finite simple Lie algebra $\mathfrak{g}$ over $\mathbb{F}_p$ and any generating set $A$ , the diameter $\text{diam}(\mathfrak{g}, A)$ is bounded by a polylogarithmic function of the size of the algebra.

A significant challenge in this domain is that, unlike finite simple groups, finite simple Lie algebras are not fully classified. Furthermore, the correspondence between finite simple groups of Lie type and finite simple Lie algebras over $\mathbb{F}_p$ is not global due to the poor behavior of the Baker-Campbell-Hausdorff formula in finite characteristic. The paper aims to establish these bounds independently of a full classification, relying instead on the property of simplicity.

Methodology
The authors build upon and refine methods developed by Dona [Don23], utilizing a two-step strategy to prove rapid generation:

Product Theorem in Lines: The core mechanism involves analyzing the growth of the set $A_k$ (elements generated by $k$ steps) along one-dimensional subspaces (lines). The authors define $\ell(X)$ as the maximum number of elements from a set $X$ lying on any single line in $\mathfrak{g}$ . They prove a "product theorem in lines" (Theorem 1.6), showing that if $\ell(A_k)$ is not already maximal (i.e., equal to $p$ ), then $\ell(A_{2k+\dim \mathfrak{g}})$ grows significantly, specifically satisfying $\ell(A_{2k+\dim \mathfrak{g}}) \geq \ell(A_k)^{5/4-\epsilon}$ . This growth relies on the sum-product theorem for finite fields [MS23].
Transporting Lines to the Whole Algebra: Once a line is fully covered (i.e., $\ell(A_k) = p$ ), the authors demonstrate that the entire algebra can be generated in a number of additional steps proportional to the square of the dimension. This is achieved by constructing a basis using "towers" of Lie brackets relative to a vector spanning the covered line (Proposition 2.3 and Proposition 1.5).

For the specific case of classical Lie algebras with random generators, the authors employ a characteristic 0 lifting strategy:

Covering Objects: They construct characteristic 0 Lie algebras (specifically $\mathbb{Q}$ -coverings of $\mathbb{F}_p$ -forms) that serve as universal covers for families of Lie algebras over $\mathbb{F}_p$ .
Exponential Growth in Characteristic 0: They prove that these covering rings contain pairs of elements that exhibit exponential growth (Lemma 3.6, Proposition 3.7).
Projection and Randomness: By projecting these favorable pairs from the characteristic 0 ring to $\mathbb{F}_p$ , they show that for large $p$ , the images generate a large subset of the algebra. Using the Schwartz-Zippel lemma, they extend this result to generic (random) pairs of elements.
Non-Split Forms: For non-split forms (arising from Dynkin diagram symmetries), the authors use Galois descent and the Chebotarev density theorem to construct a finite collection of number fields. This allows them to cover a set of primes with density arbitrarily close to 1, ensuring the existence of suitable covering rings for almost all primes.

Key Contributions and Results

General Diameter Bound (Theorem 1.3): The authors prove that for any nonabelian finite simple Lie algebra $\mathfrak{g}$ over $\mathbb{F}_p$ and any generating set $A$ , the diameter is bounded by:
$\text{diam}(\mathfrak{g}, A) \leq C(\log |\mathfrak{g}|)^{D+\epsilon}$
where $D = \log 2 / \log(5/4) \approx 3.11$ . This result holds without assuming a classification of the algebras, relying only on simplicity. When $p$ is fixed, the bound improves to $O_p((\log |\mathfrak{g}|)^2)$ , matching the stronger version of Babai's conjecture for groups.
Extremely Rapid Random Generation (Theorem 1.7): For classical Lie algebras (both split and non-split forms), the authors establish a much sharper bound when the generators are chosen uniformly at random.
- Split Forms: For a fixed Lie type, a random pair of elements generates the algebra with diameter $O(\log |\mathfrak{g}|)$ with probability tending to 1 as $p \to \infty$ .
- Non-Split Forms: A similar $O(\log |\mathfrak{g}|)$ bound holds for a set of primes with density at least $1-\epsilon$ .
Refinement of Dona's Work: The paper improves upon Dona's results by providing explicit constants for the exponent $D$ and extending the analysis to non-split forms and random generators, which were not fully addressed in prior work.

Significance and Claims
The paper claims to fully prove the linear version of Babai's conjecture with explicit constants for finite simple Lie algebras. The significance lies in:

Independence from Classification: The general proof bypasses the lack of a complete classification of finite simple Lie algebras, relying solely on the property of simplicity.
Explicit Constants: Unlike asymptotic conjectures, the authors provide an explicit exponent $D \approx 3.11$ and discuss how stronger sum-product theorems could lower this value.
Random Generation: The work demonstrates that classical Lie algebras exhibit "extremely rapid" generation (linear in the logarithm of the size) for random generators, paralleling results for finite simple groups of fixed Lie type.
Methodological Bridge: The use of characteristic 0 coverings and number-theoretic tools (Chebotarev density) to solve problems in finite characteristic Lie algebras represents a novel technical approach in this specific context.

The authors note that while their current method yields an exponent $D \approx 3.11$ (or $D=2$ for fixed $p$ ), the true bound might be $D=1$ . They acknowledge that their strategy, which relies on Proposition 1.5, inherently produces a bound that is at least quadratic in the dimension $\dim \mathfrak{g}$ , suggesting that $D=2$ is the natural limit of this specific approach. They do not claim to have proven $D=1$ but suggest it as a plausible target for future work.