betaFeedback

WorksPhysics and Astronomy

Logical-Class Diagnostic Layer for Stabilizer Codes and Decoders via Supercharge Structures

Abstract

We develop a decoder-agnostic logical-class diagnostic layer for abelian Pauli stabilizer codes, motivated by fault-tolerant quantum computing. An ancilla fermionic register defines a nilpotent supercharge whose clean degree-zero harmonic sector is the code space. Projected-return blocks derived from graded commutators organize the recovery test of whether a decoder-selected correction preserves every encoded input for every error in a declared same-syndrome family; the passing syndromes form the certified faithful window. Demoting the supercharge to an exact binary diagnostic formulation classifies residuals modulo stabilizers, separating a check violation, a harmless stabilizer residual, and a syndrome-silent nontrivial logical action, with a two-sector form for CSS codes. Because the unknown fault cancels, the relative logical class of two record-compatible candidates is computable online. We verify the diagnostic formalisms on the Steane code and specify a static–circuit bridge audit protocol for the Gross bivariate-bicycle quantum low-density parity-check (qLDPC) code. The method supplements existing decoders without additional physical qubits for software use.

Introduction

Consider a quantum processor storing one or more encoded qubits in a low-density parity-check memory during repeated syndrome extraction. Every reported check can be quiet while the encoded information has moved to a different logical state. This is not a defect of the check measurement. A syndrome reports local stabilizer violations; it does not reveal a logical Pauli that commutes with all the checks. Let SS be an abelian Pauli stabilizer group and N(S)N(S) its normalizer. A post-correction residual in SS is harmless, whereas a residual in N(S)\SN(S)\backslash S changes the encoded state without producing a syndrome. The local–global distinction is familiar from the stabilizer formalism and the Knill–Laflamme criterion [1, 2, 3, 4]. Our question is how to preserve that distinction when auditing a decoder’s proposed recovery.

In this paper, we propose and develop, to supply quantum decoders with a logical-address diagnostic and audit layer: an exact, decoder-agnostic apparatus that certifies not merely that a correction has returned the state to the code space, but that it has returned the state to the correct logical fibre inside it. It does not replace a base decoder, and it does not attempt to infer an unobserved fault from a silent record. It answers two sharper questions: what does a proposed correction do on a declared error family, and do two record-compatible candidates differ logically.

The need is practical, as well as conceptual. High-rate quantum low-density parity-check codes have matured from asymptotic constructions into a developing field with end-to-end memory protocols and real-time decoder stacks [5, 6, 7, 8]. These codes are strongly degenerate [9]: one detector record is compatible with a structurally large family of fault histories, and repeated noisy syndrome extraction enlarges that family further [10], in a way shaped by the chosen extraction circuit [11]. Belief propagation with ordered-statistics post-processing [9] and localized-statistics decoding [12] are built to decode such degenerate codes, and post-processing strategies already attack decoder degeneracy directly [13]. Degeneracy is thus the normal operating condition of a real decoder, and it is the mechanism by which two detector-compatible accounts of one observation can differ by a protected logical action. What is still missing is an exact, per-record logical verdict.

We organize this audit with a nilpotent supercharge. Section 2 associates one fermionic ancilla-register mode with each stabilizer generator SaS_a. The violated-check projector Πa1=(I−Sa)/2\Pi_a^1=(I-S_a)/2 controls creation in that mode: Q=∑a=1rΠa1⊗ca†Q=\sum_{a=1}^{r}\Pi_a^1\otimes c_a^\dagger, Q2=0Q^2=0. The commuting projectors and anticommuting register operators make the square vanish. Acting on an empty register, QQ creates a coherent one-excitation superposition supported on the violated checks; it does not itself perform a stabilizer measurement. The auxiliary operator HSUSY={Q,Q†}H_{\mathrm{SUSY}}=\{Q,Q^\dagger\} counts those violations; Its “energy” is syndrome weight. Bright, syndrome-bearing directions are paired by QQ. A code state is dark, but a code state carrying an old register excitation is dark too. The clean protected manifold therefore needs the ancilla-degree restriction K=ker⁡Q∩ker⁡Q†∩ker⁡Nanc=CQEC⊗∣0⟩anc\mathcal{K}=\ker Q\cap\ker Q^\dagger\cap\ker N_{\mathrm{anc}}=\mathcal{C}_{\mathrm{QEC}}\otimes|0\rangle_{\mathrm{anc}}. This cut removes stale flags and identifies the landing space that a recovery must actually reach. It also gives the code space a degree-zero harmonic interpretation familiar from supersymmetric complexes [14]. The value of this language is not a new physical protection mechanism: it makes the syndrome-active directions and the clean encoded sector visible in one algebra.

Three perspectives organize the paper: we see the departure, because QQ writes the syndrome record into the ancilla register exactly once; we certify the return, because a syndrome-resolved block tests the landing of a whole sector; we scale the audit, because the same test survives demotion to finite-field arithmetic and then circuit-level detector histories.

Four cooperating objects carry these promises, and naming their roles once makes the formal development easier to follow. The destination is K\mathcal{K}: the arena every certified recovery must hit. The route is the projected-return block, displayed here in its decoder-facing form Bs=PKCsP~sB_s=P_{\mathcal{K}}C_s\widetilde{P}_s, in which P~s\widetilde{P}_s selects a clean bright syndrome sector, the decoder-selected factor CsC_s proposes the return, and PKP_{\mathcal{K}} tests the landing; § 3 derives this block from a graded commutator internal to the QQ-complex, so the apparatus is intrinsic rather than appended. The warrant is the certified faithful window SfaithS_{\mathrm{faith}}, the syndrome addresses on which the selected correction is proved, not merely scored, to carry one declared family to one logical fibre [2, 3]. The diagnosis is the software residual class H1(Q~full)H^1(\widetilde{Q}_{\mathrm{full}}), obtained by demoting QQ to an exact finite-field differential in § 4 using the binary chain complex associated with the CSS code construction [15, 16]. Route and diagnosis answer different questions: one is a statement about transport of an operator block, the other a quotient-theoretic statement about a committed residual. Keeping both visible is what dinstiguishes a scalar cost, a soft score, or a symmetry label, from a certificate of logical faithfulness.

The construction is checked at two scales, and the two audits test different interfaces. The Steane [[7,1,3]][[7, 1, 3]] analysis [16] of Subsec. 4.5 exhausts the complete weight-one catalogue by hand, an executability proof of the conventions rather than a benchmark of large-code performance. The Steane code also supports experimentally demonstrated fault-tolerant logical gate operations, including magic-state injection [17]. The gross [[144,12,12]][[144, 12, 12]] bivariate-bicycle analysis of § 5 moves to a fixed extraction circuit in the setting of Ref. [5]. There a spacetime fault history ff produces a detector record d=Dfd = Df, and its logical action is recorded by the linear map AA. Two candidates faf_a and fbf_b sharing a detector record share a logical address if and only if A(fa+fb)=0A(f_a + f_b) = 0, the exact criterion established in Subsec. 5.2. The unknown true fault cancels from this comparison, so the test needs no online oracle. The audit is specified for the static–circuit bridge of Subsec. 5.3 on a declared low-weight domain.

Two downstream refinements preserve the same logical order. The noise-calibrated weighted lift of § 6 may reorder candidates without changing the exact binary logical quotient. The character-resolved compression of Subsec. 3.7 may reorganize an already-certified ledger whenever a finite Abelian action on syndrome labels is verified. Finally, the static Witten index of § 7 counts the clean protected states [14]. The order is strict: first certify, and only then weight, rerank, resolve symmetry, and compress.

The main new results are four. First, an ancilla-register supercharge for an arbitrary Abelian Pauli stabilizer code, whose clean degree-zero harmonic sector is exactly CQECC_{\mathrm{QEC}}. Second, the graded-commutator return block, its sectorwise collapse, and the certified faithful window built on it. Third, the demotion Q↦Q~Q \mapsto\widetilde{Q} used as an accept–reject certificate on ρfs\rho_{f_s}, together with an online ambiguity flag computed from relative classes of candidates, in Subsec. 4.4. Fourth, the static–circuit bridge with its mismatch ledger, and a matched ablation that separates the value of logical information from the value of its implementation, in Subsec. 5.6. Three supporting results keep the architecture consistent: the invariance of the logical address under noise-calibrated weighting, a character-resolved symmetry reorganization of the certified ledger, and a projected Witten index that fixes the size of K\mathcal{K}. All these statements concern logical classification.

The paper follows the order of dependence. Section 2 builds QQ, K\mathcal{K}, and the return blocks, and records the compatibility of an external non-Clifford resource with K\mathcal{K} in Subsec. 2.8. Section 3 introduces the decoder map, the faithfulness test, and SfaithS_{\mathrm{faith}}, and ends with the certified symmetry reorganization. Section 4 develops the finite-field certificate and checks it by hand on the Steane [[7,1,3]][[7, 1, 3]] code in Subsec. 4.5. Section 5 carries the construction to circuit level on the gross code. Section 6 adds bias-calibrated weighting, and § 7 fixes the size of the protected arena. Section 8 concludes and discusses related perspectives.

Ancilla-Register Supercharge and Clean Protected Manifold for Stabilizer Codes

This section develops, in a model-independent and intrinsic way, a construction that embeds an arbitrary stabilizer code, of Calderbank–Shor–Steane type or of general abelian Pauli type, into a graded supersymmetric complex, by adjoining a fermionic ancilla-register sector and by defining a single nilpotent supercharge QQ. The output is the static operator-theoretic structure used throughout the paper: a nilpotent QQ, a non-negative operator HSUSY={Q,Q†}H_{\mathrm{SUSY}} = \{Q,Q^\dagger\}, an ancilla-degree grading of the enlarged Hilbert space, a Hodge decomposition adapted to that grading, and a clean protected manifold K\mathcal{K} whose degree-zero harmonic representatives are precisely the encoded code states.

The construction in this section is kinematic. Here, the identification of K\mathcal{K} follows from the algebra of QQ alone together with the ancilla-degree grading that QQ already respects, in the standard sense in which a supersymmetric complex is read degree by degree [14]. The scope is deliberately narrow, and stating it once prevents later over-reading. What transfers from the general supersymmetric formalism to a stabilizer code is the static algebraic structure alone: nilpotency, the fermion-number grading, the Hodge decomposition, and the cohomological identification of the code subspace. The operator HSUSYH_{\mathrm{SUSY}} is used throughout as a non-negative auxiliary grading operator that resolves syndrome sectors. The objects fixed here are the intrinsic input used later in § 3 for the syndrome-resolved projected-return blocks and the certified faithful window, in § 4 for the exact finite-field residual classifier, in § 7 for the static index count of protected logical fibres, in § 2.8 for the logical non-Clifford interface, and in § 5 for the circuit-level detector-fibre audit on the gross code. The separation between local check information and global logical action that motivates the whole architecture is the essential insight from the stabilizer formalism and the Knill–Laflamme criterion [1, 2, 3], and the present section is its operator-level realization.

Setup: Stabilizer code and ancilla-register Hilbert space

Let S=⟨S1,…,Sr⟩⊂Pn\mathcal{S} = \langle S_1,\ldots,S_r\rangle\subset\mathcal{P}_n be an abelian Pauli stabilizer group on nn data qubits, with Hdata=(C2)⊗n\mathcal{H}_{\mathrm{data}} = (\mathbb{C}^2)^{\otimes n}. Each generator SaS_a, with a=1,…,ra = 1,\ldots,r, is Hermitian, unitary, and squares to the identity,

Sa†=Sa,Sa2=I,[Sa,Sb]=0.(1)S_a^\dagger= S_a,\qquad S_a^2 = I,\qquad[S_a,S_b] = 0. \tag*{(1)}

The single-stabilizer syndrome projectors are

Πa0=12(I+Sa),Πa1=12(I−Sa),(2)\Pi_a^0 = \frac{1}{2}(I+S_a),\qquad\Pi_a^1 = \frac{1}{2}(I-S_a), \tag*{(2)}

which are mutually orthogonal projectors obeying Πa1=I−Πa0\Pi_a^1 = I-\Pi_a^0, Πa0Πa1=0\Pi_a^0\Pi_a^1 = 0, and [Πas,Πbs′]=0[\Pi_a^s,\Pi_b^{s'}] = 0 for all a,b,s,s′a,b,s,s'. The code subspace is the simultaneous +1+1 eigenspace of all generators,

CQEC={∣ψ⟩∈Hdata∣Sa∣ψ⟩=∣ψ⟩ for all a}=∏a=1rΠa0Hdata,(3)\mathcal{C}_{\mathrm{QEC}} = \{|\psi\rangle\in\mathcal{H}_{\mathrm{data}} \mid S_a|\psi\rangle= |\psi\rangle\ \text{for all }a\} = \prod_{a=1}^{r}\Pi_a^0\mathcal{H}_{\mathrm{data}}, \tag*{(3)}

with dim⁡CQEC=2k\dim\mathcal{C}_{\mathrm{QEC}} = 2^k and k=n−rk = n-r the number of logical qubits for an independent generating set. This is the standard stabilizer-code starting point [1, 3]. Its complement is resolved by the violation projectors Πa1\Pi_a^1, and that is the property the supercharge structure of Subsec. 2.2 is closely related to.

To each stabilizer generator SaS_a we adjoin one fermionic ancilla-register mode, with creation and annihilation operators ca†c_a^\dagger and cac_a satisfying the canonical anticommutation relations

{ca,cb†}=δab,{ca,cb}=0,{ca†,cb†}=0.(4)\left\{c_a,c_b^\dagger\right\}=\delta_{ab},\qquad\left\{c_a,c_b\right\}=0,\qquad\left\{c_a^\dagger,c_b^\dagger\right\}=0. \tag*{(4)}

The ancilla-register Fock space is Hanc=⋀∙Cr\mathcal{H}_{\mathrm{anc}}=\bigwedge^\bullet\mathbb{C}^r, with fermion-number operator and parity

Nanc=∑a=1rca†ca,(−1)F=(−1)Nanc.(5)N_{\mathrm{anc}}=\sum_{a=1}^{r}c_a^\dagger c_a,\qquad(-1)^F=(-1)^{N_{\mathrm{anc}}}. \tag*{(5)}

The full Hilbert space of the construction is the graded tensor product

H=Hdata⊗Hanc,dim⁡H=2n2r.(6)\mathcal{H}=\mathcal{H}_{\mathrm{data}}\otimes\mathcal{H}_{\mathrm{anc}},\qquad\dim\mathcal{H}=2^n2^r. \tag*{(6)}

The grading (−1)F(-1)^F defined by fermion number plays the role of the fermion parity throughout the paper.

It is convenient, and it will turn out to be indispensable in Subsec. 2.3, to resolve H\mathcal{H} into its ancilla-degree slices from the very beginning,

H=⨁p=0rH(p),H(p)=ker⁡(Nanc−p)=Hdata⊗⋀pCr,(7)\mathcal{H}=\bigoplus_{p=0}^{r}\mathcal{H}^{(p)},\qquad\mathcal{H}^{(p)}=\ker(N_{\mathrm{anc}}-p)=\mathcal{H}_{\mathrm{data}}\otimes\bigwedge^p\mathbb{C}^r, \tag*{(7)}
H(0)=Hdata⊗∣0⟩anc,dim⁡H(0)=2n.(8)\mathcal{H}^{(0)}=\mathcal{H}_{\mathrm{data}}\otimes|0\rangle_{\mathrm{anc}},\qquad\dim\mathcal{H}^{(0)}=2^n. \tag*{(8)}

The integer pp is the number of ancilla-register excitations, that is, the number of check flags currently written into the register. It is the degree of the syndrome record. The slice p=0p=0 is the clean register, in which no check has been flagged, while p>0p>0 records an occupied subset of violated checks. The operator NancN_{\mathrm{anc}} is the fermion-number, equivalently form-degree, operator of a standard supersymmetric complex, in which the graded structure and its degreewise cohomology are part of the original formulation rather than an addition to it [14]. Recording the grading explicitly at this point is what allows every cohomological statement in Subsecs. 2.3 and 2.5 to be degree-resolved, and it is what prevents the full kernel of the supercharge from being confused with the degree-zero protected sector required by the projected-return blocks of Subsec. 3.3 and by the fibre tests of Sec. 3.

The Supercharge QQ

The supercharge that promotes the stabilizer code to the ancilla-register construction is the nilpotent differential

Q=∑a=1rΠa1⊗ca†=∑a=1r12(I−Sa)⊗ca†,(9)Q=\sum_{a=1}^{r}\Pi_a^1\otimes c_a^\dagger=\sum_{a=1}^{r}\frac{1}{2}(I-S_a)\otimes c_a^\dagger, \tag*{(9)}

with adjoint

Q†=∑a=1rΠa1⊗ca.(10)Q^\dagger=\sum_{a=1}^{r}\Pi_a^1\otimes c_a. \tag*{(10)}

By construction QQ raises the ancilla-register number by one and Q†Q^\dagger lowers it by one,

References

  1. [1]D. Gottesman, “Stabilizer codes and quantum error correction,” arXiv:quant-ph/9705052 (1997).DOI
  2. [2]E. Knill and R. Laflamme, Theory of quantum error-correcting codes, Physical Review A 55, 900 (1997).DOI
  3. [3]M. A. Nielsen and I. L. Chuang, Quantum Computation and Quantum Information, Cambridge University Press, Cambridge, UK (2010).
  4. [4]B. M. Terhal, Quantum error correction for quantum memories, Rev. Mod. Phys. 87, 307–346 (2015).DOI
  5. [5]S. Bravyi, A. W. Cross, J. M. Gambetta, D. Maslov, P. Rall, and T. J. Yoder, High-threshold and low-overhead fault-tolerant quantum memory, Nature 627, 778–782 (2024); arXiv:2308.07915 [quant-ph].DOI
  6. [6]N. P. Breuckmann and J. N. Eberhardt, Quantum low-density parity-check codes, PRX Quantum 2, 040101 (2021), arXiv:2103.06309.DOI
  7. [7]T. Müller, T. Alexander, M. E. Beverland, M. Bühler, B. R. Johnson, T. Maurer and D. Van­deth, “Improved belief propagation is sufficient for real-time decoding of quantum memory,” [arXiv:2506.01779 [quant-ph]].arxiv.org/abs/2506.01779
  8. [8]T. Maurer, M. Bühler, M. Kröner, F. Haverkamp, T. Müller, D. Vandeth and B. R. Johnson, “Real-time decoding of the gross code memory with FPGAs,” [arXiv:2510.21600 [quant-ph]].arxiv.org/abs/2510.21600
  9. [9]P. Panteleev and G. Kalachev, “Degenerate Quantum LDPC Codes With Good Finite Length Performance,” Quantum 5 (2021), 585 doi:10.22331/q-2021-11-22-585 [arXiv:1904.02703 [quant-ph]].DOI
  10. [10]E. Dennis, A. Kitaev, A. Landahl, and J. Preskill, ”Topological quantum memory”, J. Math. Phys. 43, 4452–4505 (2002), arXiv:quant-ph/0110143.DOI
  11. [11]M. H. Shaw and B. M. Terhal, “Lowering connectivity requirements for bivariate bicycle codes using morphing circuits,” Phys. Rev. Lett. 134, 090602 (2025). doi:10.1103/PhysRevLett.134.090602. arXiv:2407.16336.DOI
  12. [12]T. Hillmann, L. Berent, A. O. Quintavalle, J. Eisert, R. Wille and J. Roffe, “Localized statistics decoding for quantum low-density parity-check codes,” Nature Commun. 16 (2025) no.1, 8214 doi:10.1038/s41467-025-63214-7 [arXiv:2406.18655 [quant-ph]].DOI
  13. [13]K. Tsubouchi, H. Yamasaki and S. Tamiya, “Degeneracy Cutting: A Local and Efficient Post-Processing for Belief Propagation Decoding of Quantum Low-Density Parity-Check Codes,” [arXiv:2510.08695 [quant-ph]].arxiv.org/abs/2510.08695
  14. [14]E. Witten, “Constraints on supersymmetry breaking,” Nucl. Phys. B 202, 253–316 (1982).DOI
  15. [15]A. R. Calderbank and P. W. Shor, Good quantum error-correcting codes exist, Phys. Rev. A 54, 1098–1105 (1996).pubmed.ncbi.nlm.nih.gov/9913578
  16. [16]A. M. Steane, “Error correcting codes in quantum theory,” Phys. Rev. Lett. 77, 793–797 (1996).pubmed.ncbi.nlm.nih.gov/10062908
  17. [17]L. Postler, et al. “Demonstration of fault-tolerant universal quantum gate operations,” Nature 605 (2022) no.7911, 675-680 [arXiv:2111.12654 [quant-ph]].DOI
  18. [18]S. Bravyi and A. Kitaev, “Universal quantum computation with ideal Clifford gates and noisy ancillas,” Phys. Rev. A 71, 022316 (2005).DOI
  19. [19]G. Kemp and S. Ramgoolam, “BPS states, conserved charges and centres of symmetric group algebras,” JHEP 01 (2020), 146 [arXiv:1911.11649 [hep-th]].DOI
  20. [20]D. Berenstein and S. Wang, “BPS coherent states and localization,” JHEP 08 (2022), 164 [arXiv:2203.15820 [hep-th]].DOI
  21. [21]J. P. B. Ataides, D. K. Tuckett, S. D. Bartlett, S. T. Flammia and B. J. Brown, “The XZZX surface code,” Nature Commun. 12 (2021) no.1, 2172 [arXiv:2009.07851 [quant-ph]].DOI
  22. [22]C. Gidney, “Stim: a fast stabilizer circuit simulator,” Quantum 5, 497 (2021), arXiv:2103.02202.DOI
  23. [23]K. Chen, Y. Liu, W. Fang, J. Paykin, X. C. Wu, A. Schmitz, S. Zdancewic and G. Li, “Verifying Fault-Tolerance of Quantum Error Correction Codes,” doi:10.1007/978-3-031-98685-7_1 [arXiv:2501.14380 [quant-ph]].
  24. [24]W. Fang and M. Ying, “Symbolic Execution for Quantum Error Correction Programs,” Proc. ACM Program. Lang. 8 (2024) 189 [arXiv:2311.11313[quant-ph]].DOI
  25. [25]C.-F. Kung, K.-Y. Kuo and C.-Y. Lai, “Efficient approximate degenerate ordered statistics decoding for quantum codes via reliable subset reduction,” arXiv:2412.21118 [quant-ph].
  26. [26]J. Kinney, J. M. Maldacena, S. Minwalla and S. Raju, “An Index for 4 dimensional super conformal theories,” Commun. Math. Phys. 275 (2007), 209-254 [arXiv:hep-th/0510251 [hep-th]].DOI
  27. [27]S. Koutsioumpas, H. Sayginel, M. Webster, and D. E. Browne, “Automorphism ensemble decoding of quantum LDPC codes,” arXiv:2503.01738 (2025).
  28. [28]N. Raveendran and B. Vasic, “Trapping Sets of Quantum LDPC Codes,” Quantum 5 (2021), 562 [arXiv:2012.15297 [cs.IT]].DOI
  29. [29]Z. P. Bradshaw, M. L. LaBorde and D. Montero, “Symmetry-based quantum codes beyond the Pauli group,” Phys. Rev. A 113 (2026) no.4, 042431 [arXiv:2512.07908 [quant-ph]].
  30. [30]P. Fendley, K. Schoutens and J. de Boer, “Lattice models with N=2 supersymmetry,” Phys. Rev. Lett. 90 (2003), 120402 [arXiv:hep-th/0210161 [hep-th]].DOI
  31. [31]F. Alam et al. “Programmable digital quantum simulation of 2D Fermi-Hubbard dynamics using 72 superconducting qubits,” [arXiv:2510.26845 [quant-ph]].
  32. [32]H. Bernien et al., “Probing many-body dynamics on a 51-atom quantum simulator,” Nature 551, 579 (2017).
  33. [33]A. P. Burgers et al., “Controlling Rydberg excitations using ion–core transitions in alkaline–earth atom–tweezer arrays,” PRX Quantum 3, 020326 (2022).
  34. [34]A. Browaeys and T. Lahaye, “Many-body physics with individually controlled Rydberg atoms,” Nat. Phys. 16, 132 (2020).DOI

Paper details

vlnc:W_zme5r4qv
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 2

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents