In-Situ Simultaneous Magic State Injection on Arbitrary CSS qLDPC Codes
Kun Liu, Shifan Xu, Tomas Jochym-O'Connor, Zhiyang He, Shraddha Singh, Yongshan Ding
Abstract
Quantum low-density parity-check (qLDPC) codes can encode many logical qubits within a single code block at low physical qubit overhead, yet magic state injection into such codes remains largely underexplored. Existing state injection proposals for qLDPC codes predominantly follow an external prepare-and-transfer paradigm, in which raw magic states are prepared outside the target code block and subsequently injected via inter-code operations. We propose the first \emph{in-situ} magic state injection: a scheme in which logical magic states are directly prepared within a qLDPC memory block, only using resources required for syndrome extraction. We show that our scheme is generalizable to any CSS qLDPC code, with examples of circuit-level simulations on the Bivariate Bicycle (BB) code and the Hypergraph Product code. We focus on a regime where correlated injection errors are negligible. In the BB code, this corresponds to a configuration that simultaneously injects four logical states. Under a uniform depolarizing noise model with physical error rate , this achieves an injection error rate of per logical qubit, while the correlated-error contribution is only per logical qubit (about of the injection error rate). Under a hardware-motivated asymmetric noise model where single-qubit gate errors are of two-qubit gate errors, the injection error rate per logical qubit falls to , below the error rate () of the two-qubit gates used to encode the magic states. Its simplicity allows our scheme to be applied to arbitrary CSS qLDPC codes using only the ancilla qubits native to syndrome extraction, and yield a reduction in space overhead relative to both prepare-and-transfer approaches and surface-code-based magic state injection schemes.
AI Impact Assessments
(3 models)Scientific Impact Assessment
Core Contribution
This paper introduces the first in-situ magic state injection (MSI) scheme for arbitrary CSS quantum low-density parity-check (qLDPC) codes. Unlike existing "prepare-and-transfer" approaches that prepare magic states externally and teleport them into the target code block, this scheme prepares logical magic states *directly within* a qLDPC memory block using only the ancilla qubits already required for syndrome extraction. The method can simultaneously inject up to *k* logical magic states into an [[n,k,d]] CSS qLDPC code block. The key technical innovations include: (1) a stabilizer-cleaning procedure to identify "carrier" and "peeled" qubits, (2) a Bell-state-based resolution of support overlaps unique to qLDPC codes (in-logical, cross-logical, and overlap-of-overlaps), formalized through the "compatible pairing" condition, and (3) a noise-aware optimization of initial configurations via mixed-integer linear programming (MILP) to maximize first-round error detection.
Methodological Rigor
The paper is technically thorough. The sufficient conditions for the improved injection scheme (Theorem 1) are cleanly stated and rigorously proven. The framework correctly identifies three types of support overlaps inherent to qLDPC codes and addresses each systematically. The MILP formulation for optimizing initial configurations is well-motivated and practically implementable.
Circuit-level simulations using `stim` under two noise models (uniform depolarizing and hardware-motivated asymmetric) validate the first-order analytical predictions with close agreement. The analytical framework based on detector error models provides interpretable decompositions of both injection error rates and discard rates. The factorization of acceptance proxies (Appendix F) neatly explains why discard rates remain stable across configurations despite varying numbers of fixed stabilizers.
One methodological limitation is the restriction to `stim`-compatible simulations, which requires injecting |Y⟩ states rather than the more practically relevant |T⟩ states. While the authors cite precedent for this approximation and their formalism supports arbitrary rotation angles, the absence of |T⟩-specific simulations leaves a gap. The decoding uses BP+OSD rather than exploring whether tailored decoders could improve performance, though this is a reasonable baseline choice.
Potential Impact
Near-term relevance: The scheme's primary advantage is its dramatically reduced qubit footprint — 288 physical qubits for 4 logical magic states versus 968 for equivalent surface code patches. This makes it particularly relevant for early fault-tolerant demonstrations where qubit count is the binding constraint.
Foundational primitive: More importantly, this work establishes a *qLDPC-native* injection primitive. As the field moves toward qLDPC-based architectures (motivated by recent theoretical and experimental progress), having an injection method that doesn't require auxiliary code blocks or inter-code operations is fundamental infrastructure. The method serves as a natural starting point for future qLDPC cultivation and distillation protocols.
Generality: The applicability to *any* CSS qLDPC code is a significant strength. Demonstrations on both Bivariate Bicycle and Hypergraph Product codes validate this generality, though performance varies substantially between code families.
Limitations on current competitiveness: The authors are transparent that the scheme does not yet match surface code hook injection in per-logical-qubit error rates. The 93% discard rate for the best BB code configuration at p=10⁻³ is concerning for practical throughput, and the discard-amortized spacetime volume is substantially worse than surface code alternatives. The HGP code results (injection error ~8×10⁻³ per logical qubit) similarly suggest considerable room for improvement.
Timeliness & Relevance
This work is highly timely. The quantum computing community is actively transitioning research focus from surface codes toward qLDPC codes, driven by their superior encoding rates. Recent experimental demonstrations on superconducting platforms and proposals for reconfigurable atom array implementations create urgent demand for qLDPC-native non-Clifford resource preparation. The paper directly addresses a recognized gap: while qLDPC memory and Clifford operations have received substantial attention, non-Clifford resource preparation has remained largely tethered to surface-code-inspired paradigms.
Strengths
1. Generality: The scheme applies to any CSS qLDPC code with explicit constructive procedures, not just specific code instances.
2. Resource efficiency: No auxiliary code blocks or additional qubit overhead beyond syndrome extraction resources.
3. Simultaneous injection: The ability to inject multiple logical magic states at once exploits the multi-qubit encoding advantage of qLDPC codes.
4. Analytical tractability: The first-order analytical model closely matches circuit-level simulations, providing a useful design tool for optimizing configurations without expensive Monte Carlo sampling.
5. Sub-gate-error injection: Under asymmetric noise, the per-logical-qubit injection error rate (6.7×10⁻⁴) falls below the two-qubit gate error rate (10⁻³), recovering a desirable property from Li's surface code scheme.
6. Open-source code: Reproducibility is supported by public code availability.
Limitations
1. High discard rates: The 93% discard rate at p=10⁻³ for the best configuration significantly impacts practical throughput and spacetime volume comparisons.
2. Scalability ceiling within a block: Performance degrades sharply when injecting more than 4 out of 12 logical qubits in the BB code, suggesting a fundamental tension between simultaneous injection count and error rate.
3. Code-specific optimization required: While the framework is general, achieving good performance requires code-specific optimization of logical representatives, syndrome extraction circuits, and initial configurations.
4. Correlated errors: The focus on regimes where correlated errors are negligible constrains the operating regime; configurations with more injected logicals exhibit non-negligible correlations.
5. No end-to-end distillation analysis: The paper evaluates raw injection only, without demonstrating integration with distillation or cultivation pipelines.
Overall Assessment
This paper makes a solid foundational contribution by establishing the first in-situ injection primitive for qLDPC codes. While current performance metrics don't yet surpass optimized surface code approaches, the conceptual framework, rigorous analysis, and demonstrated generality position this work as an important building block for the emerging qLDPC computing paradigm. The space overhead advantage is genuine and immediately relevant for near-term experiments.
Generated Apr 8, 2026
Comparison History (62)
Paper 2 demonstrates the first telecom-band nanophotonic quantum memory exceeding 1 microsecond storage time in erbium-doped thin-film lithium niobate, addressing a key missing element for photonic quantum computing and networking. Its experimental breakthrough has broad impact across quantum computing, networking, and sensing, with immediate practical relevance to existing photonic platforms. Paper 1 makes a strong theoretical contribution to magic state injection for qLDPC codes, but its impact is narrower and more incremental within quantum error correction. Paper 2's experimental demonstration on a scalable, practical platform gives it higher near-term and cross-disciplinary impact.
Paper 2 addresses a critical practical bottleneck in quantum error correction—magic state injection for qLDPC codes—which is directly relevant to building fault-tolerant quantum computers. It proposes the first in-situ injection scheme generalizable to any CSS qLDPC code, with concrete circuit-level simulations showing competitive error rates. This has immediate practical impact for the quantum computing community. Paper 1, while intellectually deep in its thermodynamic completeness framework, addresses a more foundational/theoretical question with narrower immediate applicability. The timeliness and engineering relevance of Paper 2 in the rapidly advancing qLDPC field give it higher near-term impact.
Paper 1 addresses a critical practical bottleneck in quantum error correction—magic state injection for qLDPC codes—which is directly relevant to building fault-tolerant quantum computers. It presents the first in-situ injection scheme generalizable to any CSS qLDPC code, with concrete circuit-level simulations showing promising error rates and reduced overhead. This has immediate engineering implications for near-term quantum computing architectures. Paper 2, while theoretically elegant in its thermodynamic completeness framework, addresses a more foundational/conceptual question with narrower immediate practical impact. The timeliness of qLDPC code development and the practical utility of Paper 1 give it higher estimated impact.
Paper 2 addresses a fundamental bottleneck in fault-tolerant quantum computing—magic state injection for qLDPC codes—which is critical for practical quantum computation. Its in-situ approach is the first of its kind, generalizable to any CSS qLDPC code, and reduces space overhead compared to existing methods. This has broad implications for the architecture of future quantum computers. Paper 1, while achieving record-breaking key rates in quantum access networks, represents an incremental advance in quantum communication infrastructure. Paper 2's novelty and centrality to the quantum computing roadmap give it higher potential impact.
Paper 1 proposes a paradigm-shifting 'in-situ' magic state injection scheme for qLDPC codes, addressing a critical bottleneck in fault-tolerant quantum computing. By eliminating external prepare-and-transfer overhead and generalizing to arbitrary CSS codes, it offers profound advancements for scalable quantum architectures. While Paper 2 provides an excellent engineering breakthrough in near-term quantum cryptography networks, Paper 1 tackles a more fundamental challenge in realizing universal quantum computers, giving it higher long-term scientific impact.
Paper 1 addresses a critical bottleneck in fault-tolerant quantum computing—magic state injection for qLDPC codes—with a concrete, generalizable scheme demonstrated through circuit-level simulations. It directly reduces resource overhead for universal quantum computation, which is one of the most pressing practical challenges in the field. Paper 2 presents interesting theoretical insights about quantum coordination advantages using separable states, but its scope is narrower, more abstract, and less immediately applicable. Paper 1's impact spans quantum error correction, fault-tolerant computing, and hardware implementation, making it more broadly impactful and timely.
Paper 1 addresses a critical bottleneck in fault-tolerant quantum computing—magic state injection—using highly relevant qLDPC codes. Its novel in-situ approach reduces space overhead and demonstrates rigorous circuit-level simulations with low error rates. This represents a significant step toward practical quantum advantage. Paper 2, while offering interesting theoretical insights into nanoscale friction, presents a more incremental advance within fundamental tribology and is likely to have a narrower overall impact across disciplines.
Paper 2 has higher likely impact: it addresses a central bottleneck for fault-tolerant quantum computing (magic state injection) in timely, practically relevant qLDPC architectures. It proposes a broadly applicable in-situ method for arbitrary CSS qLDPC codes and backs claims with circuit-level simulations and quantitative error rates under multiple noise models, indicating stronger methodological rigor and nearer-term applicability. Paper 1 is conceptually novel but more speculative (“beyond-quantum” toy model) with narrower adoption prospects and less direct pathway to real-world deployment.
Paper 1 addresses a critical bottleneck in quantum error correction—magic state injection for qLDPC codes—which is central to achieving fault-tolerant quantum computing. It provides the first in-situ injection scheme generalizable to any CSS qLDPC code, with rigorous circuit-level simulations and concrete overhead reductions. This has broad, immediate impact on the quantum computing architecture community. Paper 2 proposes a quantum-accelerated fitness evaluation for bent function search, but the claimed exponential advantage only materializes for n>25 on fault-tolerant hardware that doesn't yet exist, and the validated results (n=6,8) are achievable classically, limiting near-term impact.
Paper 2 addresses a long-standing experimental bottleneck in scalable quantum technologies: generating highly indistinguishable photons from independent, distant sources. By achieving a record 88% indistinguishability matching the intrinsic limit, this experimental breakthrough directly enables tangible progress in both photonic quantum computing and quantum networking. While Paper 1 provides a highly innovative theoretical protocol for quantum error correction (qLDPC codes), experimental hardware milestones that solve physical scaling challenges like those in Paper 2 typically exhibit broader, more immediate real-world impact across multiple quantum information domains.