🔢 mathematics

From Koszul-Complex Stabilizer Models to Superselection Profiles: Topological Rigidity and Nonsplit Extensions

This paper introduces a topological superselection profile framework based on Koszul-complex stabilizer models to characterize translation-symmetry-enriched topological phases in CSS codes, demonstrating that while single-layer cohomology determines rigidity for regular models, the inter-layer gluing data in nonsplit extensions is essential for distinguishing distinct phases that share identical excitation spectra.

Original authors: Hao Song

Published 2026-08-11

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Hao Song

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

The Quantum Lego Set: Building Unbreakable Codes

Imagine you are trying to build a castle out of Lego bricks, but there's a catch: the bricks are made of a strange, quantum material that can be in two places at once. If you try to look at a single brick to see if it's in the right spot, the whole castle might collapse. This is the world of quantum error correction. Scientists are trying to build "quantum computers" that can solve problems impossible for today's machines, but these computers are incredibly fragile. A tiny bit of noise, like a stray heat wave or a magnetic glitch, can scramble the information. To fix this, researchers use stabilizer codes. Think of these as a set of strict rules or "checks" that tell the computer how the bricks should fit together without ever looking at the bricks directly. If a brick is out of place, the rules detect a "glitch" (an error) without destroying the delicate quantum state.

Now, imagine your Lego castle isn't just sitting on a table; it's on a giant, infinite grid that repeats forever. This is the concept of translation symmetry. In the real world, we build these codes on finite chips, but studying them on an infinite grid helps us understand their deep, unbreakable nature. The paper you are about to read dives into a specific, highly organized way of building these codes using mathematical structures called Koszul complexes. You can think of a Koszul complex as a very specific, elegant recipe for arranging your Lego bricks so that they form a perfect, self-correcting pattern. The big question the author asks is: If we look at the "glitch patterns" (the excitations) created by these codes, can they tell exactly what kind of code they are looking at? And if they change the size of their Lego table, does the number of ways the castle can be built change in a predictable way?

The Paper's Story: Unraveling the Quantum Blueprint

In this paper, the author, Hao Song and colleagues, takes a deep dive into these "Koszul-complex stabilizer models." They treat these models as a unified framework that brings together different types of famous quantum codes, like the toric code (which is like a donut-shaped grid) and newer, more complex bivariate-bicycle codes. Their goal is to create a "superselection profile," which is essentially a detailed fingerprint of the code's behavior.

To understand this fingerprint, imagine the code as a multi-layered cake. The bottom layer represents point-like errors (like a single brick falling out). The next layer up represents loop-like errors (like a string of bricks falling out in a circle). The author introduces a mathematical tool called the topological superselection profile to describe all these layers at once. They find that for a specific, well-behaved family of these models (called "regular Koszul models"), the fingerprint is surprisingly simple: it only has one active layer. It's as if the cake only has one flavor, no matter how many layers they try to stack. This single layer tells them everything they need to know about the code's "topological order"—its fundamental, unchangeable nature.

The author proves a powerful result called topological rigidity. They show that if a code has this "single-layer" fingerprint, they can uniquely identify the code's structure just by looking at that one layer and knowing its "degree" (which layer it is). It's like saying, "If I tell you this Lego castle has exactly one type of unique brick pattern, I can tell you exactly how the whole castle is built." They prove this using a clever mathematical trick involving "Schanuel's lemma," which allows them to compare different code blueprints and show they are essentially the same if their fingerprints match.

However, the story gets more interesting when they look beyond these simple, single-layer codes. The author constructs a set of eight different models based on a 3D toric code. These eight models are tricky: they all have the exact same "fingerprint" layers. If you only looked at the individual layers, you would think they were all the same code. But the author discovers that these models are actually different. They are "nonsplit extensions," meaning the layers are glued together in different, hidden ways.

Here is the twist: even though their fingerprints look identical, these eight models behave differently when you change the size of the system. If you build them on a small table versus a large table, the number of possible ground states (the different ways the castle can sit stably) changes in unique ways for each model. This proves that the "glue" between the layers carries hidden information that the individual layers miss. The author shows that while the single-layer rule works for many codes, it fails to distinguish between these more complex, glued-together structures.

In summary, the paper establishes that for a large class of well-behaved quantum codes, a simple, single-layer mathematical description is enough to perfectly identify the code's structure and predict how it behaves on different-sized grids. But for more complex, "glued" codes, this simple description isn't enough; you need to look at how the layers connect to see the full picture. This work provides a rigorous mathematical dictionary to translate between the abstract algebra of these codes and their physical behavior, helping scientists understand which codes are truly unique and which are just different versions of the same thing.

Technical Summary: From Koszul-Complex Stabilizer Models to Superselection Profiles

Problem Statement
The paper addresses the classification of translation-symmetry-enriched topological (SET) phases within translation-invariant Calderbank–Shor–Steane (CSS) stabilizer codes. While homological methods have successfully extracted topological data (such as anyon types and fusion rules) from stabilizer presentations, a central open question remains: to what extent do these excitation data uniquely determine the stable Clifford class of a code when translation symmetry is preserved? Specifically, the author investigates whether the hierarchy of superselection sectors, encoded via Ext functors, is sufficient to distinguish distinct SET phases or if additional structural data (such as inter-layer gluing) is required. The study focuses on CSS models constructed from Koszul complexes, which unify toric-code hierarchies and bivariate-bicycle (BB) codes, to test the limits of this classification.

Methodology
The author employs a rigorous algebraic framework based on Laurent polynomial rings $R = \mathbb{Z}_N[x_1^{\pm 1}, \dots, x_D^{\pm 1}]$ to model translation-invariant stabilizer codes.

Koszul-Complex Models: They define a class of stabilizer models where the physical qudits and stabilizer maps are derived from the differentials of a Koszul complex $K_\bullet(f)$ associated with a sequence $f = (f_1, \dots, f_s)$ . For regular sequences, these models are strictly topological.
Topological Superselection Profile: The author introduces the topological superselection profile $S_\sigma$ , defined as the positive-degree derived object $\tau_{\ge 1} \text{RHom}_R(\text{coker } \phi_\sigma, R)$ for $\sigma \in \{X, Z\}$ . The cohomology layers $E^\ell_\sigma \cong \text{Ext}^\ell_R(\text{coker } \phi_\sigma, R)$ encode excitation sectors, fusion rules, and translation actions.
Algebraic-Clifford Correspondence: To link algebraic invariants to physical phases, the paper utilizes a matrix-level Schanuel-lemma method. This approach establishes that stable module isomorphism (under free stabilization) corresponds to stable Clifford equivalence (finite-depth translation-invariant Clifford circuits with ancillas).
Finite-Size Analysis: The logical space on a finite periodic lattice is analyzed using the Tor functor, specifically $\text{Tor}^R_j(S, R_L)$ , where $S = R/(f)$ is the Koszul structure module and $R_L$ is the quotient ring imposing periodic boundary conditions.

Key Contributions and Results

Definition of the Superselection Profile: The paper formalizes the topological superselection profile as a derived invariant that organizes all positive-degree superselection layers and their inter-layer gluing. This generalizes previous charge-module frameworks by explicitly retaining the derived structure that may contain information invisible to the graded list of Ext modules alone.
Rigidity Theorem for Single-Layer Profiles: The author proves a topological rigidity theorem (Theorem 2). For translation-invariant topological CSS codes where the X-sector profile is single-layer (at most one positive-degree Ext module is nonzero) and the reduced Pauli-X module admits a finite-length free resolution, the nonzero Ext layer and its degree uniquely determine the stable translation-invariant Clifford class.
- For prime-qudit codes, the finite-resolution hypothesis is automatic.
- In this regime, the translation kernel of the nonzero layer provides the exact minimal coarse graining required to reduce the code to a stack of toric codes.
Realizability: Construction 1 demonstrates that any finitely generated module $E$ satisfying the grade condition ( $\text{grade}_R(E) > j$ ) can be realized as the $j$ -th superselection layer of a valid topological CSS code.
Finite-Size Logical Space: For regular full-length Koszul models, the paper derives an explicit correspondence between the logical space and the Tor functor: $H^Z_j(L) \cong \text{Tor}^R_j(S, R_L)$ . This yields all-degree, all-dimensional logical-superselection correspondences and establishes that the ground-state degeneracy depends on system size only through the greatest common divisors of the system dimensions and the mobility periods of the structure module $S$ .
Nonsplit Extensions and Inter-Layer Gluing: Crucially, the paper demonstrates that the single-layer rigidity does not hold in the multi-layer regime. Using Construction 2, the author realizes eight distinct 3D qubit toric-code extensions that share identical single-layer Ext data ( $E^\ell_\sigma$ $E_{σ}^{ℓ}$ ) and trivial translation actions but possess distinct system-size-dependent ground-state degeneracies.
- These models are inequivalent under finite-depth local unitaries preserving translation symmetry.
- This result proves that inter-layer gluing carries SET data invisible to individual Ext layers, meaning the full derived profile $S_\sigma$ is necessary for a complete classification in multi-layer cases.

Significance and Claims
The paper claims to provide a unified algebraic framework for understanding the relationship between excitation data and topological order in translation-invariant CSS codes. Its primary significance lies in establishing a precise boundary for when excitation data is sufficient for classification:

Rigidity: For "single-layer" profiles (common in regular Koszul models like standard toric codes), the excitation data (Ext modules) is sufficient to uniquely determine the stable Clifford phase.
Limitation of Layer-wise Data: For multi-layer profiles, the paper provides explicit counterexamples showing that individual layers are insufficient. The "inter-layer gluing" within the derived profile is a physical invariant that distinguishes phases with identical excitation spectra.

The work bridges the gap between homological algebra (Ext/Tor functors) and quantum information (Clifford circuits, ground-state degeneracy), offering a "matrix-level" method to classify topological phases that respects translation symmetry. It confirms that while regular Koszul models reduce to toric-code stacks under appropriate coarse graining, the specific manner of this reduction and the resulting SET order are encoded in the translation action on the Koszul structure module.

The Quantum Lego Set: Building Unbreakable Codes

The Paper's Story: Unraveling the Quantum Blueprint

More like this