Quantum Error Correction by Purification
Jonathan Raghoonanan, Tim Byrnes · 2026
Read Quantum Error Correction by Purification by Jonathan Raghoonanan, Tim Byrnes (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-11568.txt.
Introduction
Raghoonanan and Byrnes blend quantum error correction with error mitigation in a hybrid scheme, PQES. Correction removes errors from a state; mitigation cleans noisy measurement data instead. PQES runs SWAP tests on many noisy copies to purify the state. The full outcome record combines all branches, with no postselection. Purification can also sit between unitary blocks, shrinking errors mid-computation. For local depolarizing noise, the error threshold sits at 3/4 for any register size. Local dephasing gives a 1/2 threshold, and Clifford twirling can lift it back up.
Catalog record
| CREATOR | Jonathan Raghoonanan, Tim Byrnes |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 83465 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.11568
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Quantum Error Correction by Purification Jonathan Raghoonanan, Tim Byrnes Abstract We present a general-purpose quantum error correction primitive based on state purification via the SWAP test, which we refer to as purification quantum error correction (PQEC). This method operates on N N noisy copies, requires minimally O ( M log 2 N ) O(M\log_{2}N) data qubits to process the M M -qubit inputs. In a similar way to standard QEC, the purification steps may be interleaved within a quantum algorithm to suppress the logical error rate. No postselection is performed and no knowledge of the state is required. We analyze its performance under a variety of error channels and find that PQEC is highly effective at boosting fidelity and reducing logical error rates, particularly for the depolarizing channel. Error thresholds for the local depolarizing channel are found to be 75 % 75\% for any register size. For local dephasing, the error threshold is reduced to 50 % 50\% but may be boosted using twirling. I Introduction Quantum error correction (QEC) is essential for reliable quantum computation: without active suppression of noise, circuit depth is limited by decoherence, control errors, and leakage 28 ; 10 ; 34 ; 36 . Stabilizer codes—notably the surface code—form the dominant pathway toward scalable architectures due to locality and high thresholds 16 ; 15 . Yet, the resource overheads for full fault tolerance (magic state factories, repeated rounds of syndrome extraction, and real-time decoding) remain a central challenge for current devices, motivating alternative techniques 25 ; 33 . Alongside fully fledged fault-tolerant QEC, a wide range of lighter-weight strategies has been an alternative direction of interest. An early example of this is the use of decoherence-free subspaces and noiseless subsystems, where information is encoded in symmetry-protected sectors that are immune to specific collective noise models 23 ; 8 ; 33 . Dynamical decoupling uses sequences of fast control pulses to average out system–bath couplings and extend coherence times without encoding into a code space 32 . More recently, quantum error-mitigation protocols infer zero-noise expectation values from noisy data via extrapolation in noise strength or quasi-probability cancellation, and have become a leading approach for near-term devices where full fault tolerance is out of reach 31 ; 5 . A closely related line of work, traditionally framed in the context of long-distance communication and resource preparation rather than computation, is entanglement purification 11 ; 37 . In a typical purification protocol, one starts with several noisy copies of a quantum state, and using only local operations and classical communication (LOCC), a single higher fidelity copy of the target state is obtained after post-selection. In most purification protocols, the target state is fixed and known , such as a specific Bell pair, GHZ state, or graph state. Furthermore, they typically employ postselection where particular measurement outcomes are desirable and otherwise the resultant state is discarded. For example, the original protocols of Bennett, Wootters, and co-workers distilled high-fidelity Bell pairs from many noisy copies, together with twirling operations to an isotropic Bell-diagonal form 2 . Deutsch, Sanpera, and co-workers incorporated such purification steps into quantum privacy amplification for quantum key distribution over noisy channels 9 . Subsequent work generalized these ideas to multipartite entangled states 27 ; 1 . Purification is considered essential for quantum repeaters and networking architectures 30 ; 3 ; 22 . Modern variants optimize gate counts, success probabilities, and robustness for concrete network architectures 26 ; 20 , but the basic setting remains one in which a known resource state is distilled offline from many noisy copies. Historically, purification has been mainly considered in the context of quantum state initialization due to the fact that the target state is fixed and known. However, a more general unknown-state purification can be performed, where an arbitrary unknown state is purified. This can be performed when a typical requirement of purification is relaxed, such as only using LOCC operations. Here, the task is to recover a nearly pure copy of an unknown state from many noisy copies, without prior knowledge of which pure state was prepared. Early attempts include optimal collective processing and universal purification procedures, but are generally limited to single qubits, asymptotic regimes, and may require having limited knowledge of the initial state 7 ; 21 . More recently, Childs and co-workers showed that a recursive, SWAP-test-based protocol can efficiently boost the fidelity of an unknown pure state given multiple noisy copies, for qudits of arbitrary dimension 6 . Here, the SWAP test consists of a quantum primitive that determines the fidelity between two states 4 . In the purification context, by post-selecting on the symmetric outcome of the test, one can iteratively amplify overlap with the dominant eigenvector of the input density operator. Because the protocol acts on whatever pure state occupies the dominant eigenvalue, it naturally aligns with the idea of protecting unknown encoded information rather than a fixed resource state. In parallel, several recent purification-based quantum error mitigation (QEM) approaches exploit the expectation that the ideal output of a computation is approximately pure. A prominent example is given by Huggins, McClean and co-workers 19 , which uses collective measurements on multiple copies to estimate expectation values with respect to an effectively purified state. The approach is called virtual distillation because the state itself does not become purified, but the use of a modified observable makes the effective (or virtual) state \rho^{N} , while physically only the state \rho^{\otimes N} is present. The state \rho^{N} approaches the dominant eigenvector of \rho as N increases, and thus corresponds to a purified state. Related multi-copy purification/verification techniques—including postselected variants—have been experimentally benchmarked in quantum simulation settings and shown to reduce error by orders of magnitude, with sampling overhead becoming the dominant practical limitation at scale 29 . These works primarily target the mitigation of measured observables without necessarily outputting a purified state for further coherent processing. Figure 1: Purification quantum error correction implementation in a quantum circuit. (a) Depth-efficient configuration of SWAP-based purifiers being interleaved within a M -qubit quantum algorithm consisting of a sequence of unitary gates U_{n} . The whole quantum circuit is given by the unitary U=\prod_{n}U_{n} . The state \ket{\psi_n}^{(\ell)} refers to the state at the n th stage of the quantum algorithm, with \ell rounds of purification. (b) The SWAP gadget and its definition. The \Pi_{+} measurement corresponds to the ancilla outcome |0\rangle and \Pi_{-} corresponds to |1\rangle . The second copy of the state is discarded (traced out). Thick horizontal lines denote qubit registers consisting of M qubits, thin horizontal lines are single qubits. In this paper, we merge several of the concepts above — namely unknown-state purification, virtual distillation, and quantum error correction — to show that the SWAP-based purification can implement unknown-state purification as an error-correction primitive. We refer to this scheme as purification quantum error correction (PQEC). Figure 1 illustrates the basic logical flow: noisy copies arrive sequentially and are pairwise merged via SWAP tests to form progressively purified states. Layers can be interwoven between unitary steps in a larger algorithm such that errors may be corrected during the evolution of a quantum circuit. Rather than relying upon codewords as in traditional QEC, the