This paper develops a general way to make non‑Abelian quantum low‑density parity‑check (qLDPC) codes that keep a fixed fraction of logical qubits while the code grows and that can correct errors up to a distance proportional to system size. In plain terms, the authors show how to get sparse, high‑quality quantum error‑correcting codes outside the usual stabilizer (Pauli) setting by combining algebraic topology tools with a procedure called gauging.

The core idea is to start from sheaf codes. A sheaf code is built from many small classical codes attached to pieces of a complex and then combined using standard algebraic structures. The authors “gauge” these sheaf codes using a mathematical operation called a cup product. Gauging here means adding non‑Pauli constraints that couple logical degrees of freedom. That produces non‑Abelian stabilizer generators: the new generators include conventional Pauli operators dressed with multi‑qubit controlled phase gates. The paper gives an explicit description of the full coded subspace and an orthonormal basis for it.

A major technical step is showing these new codes actually protect against low‑weight errors. The authors do this by using the Knill–Laflamme condition, which is the general requirement for quantum error correction, rather than relying on Pauli‑specific arguments. They combine expansion properties of the parent sheaf codes with a method called cleaning of logical representatives to bound how heavy any logical error must be. This is how they prove linear distance: any nontrivial logical error must act on a number of physical components that grows linearly with system size.

The paper also uses these constructions to address quantum magic, a measure of how far states or codes are from being classically simulable. The authors build an “almost‑good” family of codes whose entire code spaces show long‑range magic. They further describe protocols, based on gauging and ungauging measurements, that implement logical Clifford measurements and prepare encoded magic states. These elements connect the coding results to questions in quantum many‑body physics and complexity, in particular to conjectures about low‑energy states that are not prepareable by shallow stabilizer circuits.