# Techniques: High-threshold, low-overhead and single-shot decodable fault-tolerant quantum memory

## What the paper contributes (one paragraph)
The paper introduces quantum radial codes, a new family of quantum low-density parity-check (qLDPC) codes built from the lifted product of classical quasi-cyclic radial codes. Each code is set by a pair of integers (r,s) with parameters [[2r^2s, 2(r-1)^2, <=2s]], giving independently tunable numbers of logical qubits and distance, an intuitive visual representation, a canonical basis of logical operators, and optimal-length stabiliser measurement circuits. Under circuit-level noise simulated with stim and decoded with BP+OSD in an overlapping-window scheme, small instances match the error suppression of surface codes of comparable distance while using roughly five times fewer physical qubits, and they show evidence of single-shot (constant-depth) decodability. The authors argue the codes' small size, flexibility and confinement properties make them strong candidates for near-term hardware such as neutral-atom platforms with long-range connectivity.

## Techniques offered
- **quantum radial codes (qLDPC code family via lifted product)** - a new family of low-density parity-check quantum memory codes with independently tunable k and d, a canonical logical-operator basis, and roughly five times lower qubit overhead than surface codes at comparable distance.
  - guarantee: exact value of k = 2(r-1)^2 and a proven upper bound d <= 2s on distance for every code in the family, with numerical evidence of average-case distance linear in s.
  - quote: "We present a new family of quantum low-density parity-check codes, which we call radial codes, obtained from the lifted product of a specific subset of classical quasi-cyclic codes."
- **single-shot (constant-depth) overlapping-window decoding under circuit-level noise** - a decoding strategy that slices the detector error model into time windows and commits local corrections so logical error rate per cycle saturates, enabling faster logical clock speeds and reduced decoding complexity.
  - guarantee: statistical evidence of a sustainable single-shot threshold, with constant error rate per decoding cycle and exponential suppression of word error rate as physical error rate falls.
  - quote: "In our work, we use an overlapping window~\cite{dennis_topological_2002, berent_analog_2024, gong_toward_2024} (sometimes referred to as a sliding window~\cite{skoric_parallel_2023, huang_improved_2023}) decoding approach that fulfils these requirements."
- **optimal-length stabiliser measurement (syndrome extraction) circuits** - a parallelised stabiliser-measurement schedule for these codes in which ancilla qubits are never idle and each syndrome cycle has minimal possible length.
  - guarantee: static construction guaranteeing collision-free, valid syndrome extraction with minimal-length cycles; the authors note it is not certified optimal from a fault-tolerance (effective-distance) perspective.
  - quote: "We can use the previously described structure of QRCs to define an optimal-length (in the sense that ancilla qubits are never idle) schedule for parallelised stabiliser measurements."

## Where it could apply
- Target primitive(s): fault-tolerant quantum memory (logical qubit storage); not a specific algorithmic primitive such as VQE or QPE.
- Target application group(s): error_correction; platforms_tooling.
- Code family / hardware assumptions: qLDPC (lifted-product radial codes); compared against surface codes and bivariate bicycle codes; suited to long-range-connectivity hardware such as neutral atom platforms.

## Caveats
The contributions are suppliable to error-correction and platform results. Distance is supported by numerical estimates rather than an exact value (only an upper bound is proven), the chosen decoding parameters are non-optimal, the measurement circuits are not certified optimal for effective distance, and single-shot performance is demonstrated only for small instances; large radial codes are outperformed by large bivariate bicycle codes.