🔢 mathematics

The finite basis problem for matrix semirings $\mathbf{M}_n(S_7)$

This paper establishes an embedding theorem for matrix semirings over additively idempotent semirings and proves that the varieties generated by matrix semirings over the nonfinitely based semiring $S_7$ are themselves nonfinitely based, forming a strictly ascending chain of distinct varieties.

Original authors: Jun Jiao, Miaomiao Ren

Published 2026-07-14

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Jun Jiao, Miaomiao Ren

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 a universe built from tiny, magical building blocks called matrices. These aren't just grids of numbers; they are special "ai-semirings," a fancy name for a system where adding things together has a weird rule: if you add a thing to itself, it doesn't get bigger, it just stays the same. Think of it like a bucket that's already full; pouring more water in doesn't make it overflow, it just stays full.

In this paper, two math detectives, Jun Jiao and Miaomiao Ren, investigate a specific, mysterious set of these blocks called $S_7$ . This set has only three elements, but it's the "boss" of a very strange problem: the Finite Basis Problem.

The Mystery: Can We Write the Rules?

Every set of these blocks follows a set of secret rules (called "identities") that tell you how they behave. The big question is: Can we write down all these rules using a short, finite list?

For most sets, the answer is "Yes." You can write a cheat sheet with a few lines, and you're done. But for some sets, the rules are so complex and endless that no matter how long your cheat sheet is, there's always a new rule you missed. These sets are called nonfinitely based. It's like trying to write a dictionary for a language that keeps inventing new words every time you turn the page.

The Big Discovery: The $S_7$ Virus

The authors prove something incredible about the matrix versions of $S_7$ . They show that if you take $S_7$ and arrange it into a square grid of size $n \times n$ (where $n$ is 2 or bigger), the resulting matrix system is nonfinitely based.

Think of $S_7$ as a tiny, invisible virus. The authors prove that this virus is so contagious that if you put it inside a matrix, the whole matrix system becomes infected with the "infinite rules" disease. No matter how big the matrix gets (2x2, 3x3, 100x100), the rules remain endless and unwriteable.

They also found that this infection spreads even further. If you look at any group of systems that sits "between" the tiny $S_7$ and the big matrix $M_n(S_7)$ , they are all infected too. In fact, the space between them is so crowded with different types of these systems that there are at least countably infinitely many distinct varieties hiding there. It's like finding an infinite number of different species of bugs living in a single drop of water.

The Ladder of Varieties

The paper also builds a ladder. They proved that you can always fit the $n \times n$ matrix system inside the $(n+1) \times (n+1)$ system. This creates a chain:
$V(M_2) \le V(M_3) \le V(M_4) \dots$

For some other types of blocks (like the two-element distributive lattice), this ladder keeps climbing forever without ever reaching the same rung twice. But for our $S_7$ virus, the authors aren't 100% sure if the ladder stops or keeps going forever. They suspect it might stop at the second rung, but they haven't proven it yet.

The "Five-Matrix" Trick

Here is the most playful part of their investigation. The authors looked at what happens when you multiply these matrices together. They discovered a strange "nilpotent" property.

Imagine you have a stack of these matrices. If you multiply five of them together in a row, the result is always a "dead" matrix (filled with a special symbol $\infty$ that acts like zero). It's as if the system has a memory limit: after five steps, everything collapses into nothingness.

However, if you only multiply four of them, they don't always collapse. Sometimes they still have life. This means the system is 5-nilpotent but not 4-nilpotent.

What This Means for the Mystery

Because the system collapses so quickly after five multiplications, the authors strongly suggest (but do not prove) that the ladder of varieties might actually stop. They think that the rules for a 2x2 matrix might be exactly the same as the rules for a 3x3 matrix, and so on. If this is true, the "infinite rules" disease is real, but the ladder of sizes might not be as tall as we thought.

The Verdict

Proven: The matrix systems $M_n(S_7)$ are nonfinitely based. They have no finite list of rules.
Proven: The interval between $S_7$ and $M_n(S_7)$ contains infinitely many distinct varieties.
Proven: The multiplicative part of these matrices (without the identity matrix) is 5-nilpotent (five matrices in a row always equal zero).
Suggested: The ladder of varieties $V(M_n(S_7))$ might stabilize (stop changing) at $n=2$ , meaning $V(M_2) = V(M_3) = \dots$ . The authors believe this is likely because of the "five-step collapse," but they admit they don't have a proof yet.

In short, the paper solves the mystery of why these matrices are so complex (they are nonfinitely based) and gives us a strong hint that their complexity might not get any worse as the matrices get bigger, even though we haven't officially closed that door yet.

Title: The Finite Basis Problem for Matrix Semirings $M_n(S_7)$

Problem Statement
The paper addresses the finite basis problem for the class of additively idempotent semirings (ai-semirings). Specifically, it investigates the variety generated by the matrix semiring $M_n(S)$ over an ai-semiring $S$ . While the finite basis problem for small ai-semirings has been extensively studied—most notably the classification of three-element ai-semirings, where $S_7$ is identified as the unique nonfinitely based example—the behavior of matrix semirings $M_n(S_7)$ for $n \ge 2$ remained unresolved. The authors aim to determine whether $M_n(S_7)$ is finitely based and to characterize the lattice of varieties contained within the interval $[V(S_7), V(M_n(S_7))]$ .

Methodology
The authors employ a combination of embedding theorems, hypergraph theory, and structural analysis of matrix operations over $S_7$ .

Embedding Theorem: The paper first establishes a general embedding theorem proving that for any ai-semiring $S$ and $n \ge 2$ , $M_n(S)$ embeds into $M_{n+1}(S)$ . This is achieved by constructing an injective homomorphism $\phi: M_n(S) \to M_{n+1}(S)$ that appends a copy of the last row and column to a matrix, utilizing the idempotence of addition in $S$ . This result implies an ascending chain of varieties $V(M_2(S)) \le V(M_3(S)) \le \dots$ .
Hypergraph Criteria for Non-Finite Basis: To prove non-finite basis properties, the authors utilize a criterion by Gao et al. [10] involving 3-uniform hypergraphs. They construct a specific non-2-colorable hypergraph $H_n$ with high girth and associate it with an ai-semiring term $t_{H_n}$ . The core strategy involves showing that $M_n(S_7)$ satisfies the identity $t_{H_n} \approx t_{H_n} + x_{v_n}$ for a specific vertex $v_n$ , which, combined with the presence of the flat semiring $Sc(abc)$, implies the variety is nonfinitely based.
Structural Analysis of $M_n(S_7)$ : The authors perform a detailed case analysis of matrix multiplication in $M_n(S_7)$ , leveraging the specific Cayley tables of $S_7$ (where elements are $\{1, a, \infty\}$ ). They establish properties regarding when products result in the additive maximum $\infty$ or the minimal elements $1$ and $a$ .
Nilpotency Analysis: In the final section, the authors analyze the multiplicative reduct of $M'_n(S_7)$ (the set $M_n(S_7)$ excluding the constant matrix $[1]_n$ ). They investigate the nilpotency index of this semigroup to infer potential stabilization of the variety chain.

Key Contributions and Results

Non-Finite Basis of $M_n(S_7)$ : The primary result is that for all $n \ge 2$ , the matrix semiring $M_n(S_7)$ is nonfinitely based. Furthermore, the authors prove a stronger result: every variety in the interval $[V(Sc(abc)), V(M_n(S_7))]$ is nonfinitely based. Since $V(S_7)$ contains $Sc(abc)$, this implies that every variety in $[V(S_7), V(M_n(S_7))]$ is nonfinitely based.
Strict Ascending Chain: The paper demonstrates that the chain of varieties generated by matrix semirings over the two-element distributive lattice $D_2$ is strictly ascending ( $V(D_2) < V(M_2(D_2)) < \dots$ ). While the strictness of the chain for $S_7$ ( $V(M_n(S_7)) < V(M_{n+1}(S_7))$ ) is not fully proven, the authors show that $V(S_7) < V(M_2(S_7))$ .
Infinite Variety Interval: The authors prove that the interval $[V(S_7), V(M_2(S_7))]$ contains at least countably infinitely many distinct varieties. This is achieved by constructing a sequence of varieties $V_k$ generated by flat semirings $S(w_k)$ associated with linear words $w_k$ , showing that $V_k$ is a proper subvariety of $V_{k+1}$ .
Application to Five-Element Semirings: Applying their main theorem, the authors identify 73 five-element ai-semirings as nonfinitely based. They provide a specific example, $S(5, 5158)$ , and demonstrate that its variety lies within the interval $[V(S_7), V(M_2(S_7))]$ .
Nilpotency of the Multiplicative Reduct: The paper shows that the multiplicative reduct of $M'_n(S_7)$ is 5-nilpotent (the product of any five matrices is the zero matrix $[\infty]_n$ ) but not 4-nilpotent.

Significance and Claims
The paper claims to make a significant contribution to the finite basis problem for ai-semirings by resolving the status of $M_n(S_7)$ , a structure of central importance due to the unique non-finite basis property of $S_7$ . The results confirm that the "nonfinitely based" property of $S_7$ propagates to all matrix semirings $M_n(S_7)$ and to all varieties lying between $S_7$ and these matrix semirings.

Regarding the open question of whether the chain $V(M_n(S_7))$ stabilizes (i.e., whether $V(M_n(S_7)) = V(M_{n+1}(S_7))$ ), the authors remain modest. They state that while they cannot prove the equality, the fact that the multiplicative reduct of $M'_n(S_7)$ is 5-nilpotent "strongly suggests" that the equality may indeed hold for all $n \ge 2$ . The paper positions itself as a "modest step" toward a complete solution to the broader problem of classifying finite ai-semirings with non-finite basis properties, particularly those containing $S_7$ . The authors explicitly note that general guiding results for $M_n(S)$ for arbitrary $S$ are still lacking and that their work, alongside previous studies by Dolinka, Gusev, and Volkov, serves as an initial direction in this area.