arxiv: 2603.19062 · v3 · submitted 2026-03-19 · 🪐 quant-ph

Recognition: 2 theorem links

· Lean Theorem

Fair Decoder Baselines and Rigorous Finite-Size Scaling for Bivariate Bicycle Codes on the Quantum Erasure Channel

Tushar Pandey

Authors on Pith no claims yet

Pith reviewed 2026-05-15 08:11 UTC · model grok-4.3

classification 🪐 quant-ph

keywords bivariate bicycle codesquantum erasure channelfinite-size scalingBP-OSD decodertoric codeasymptotic thresholdpseudo-thresholdminimum-weight perfect matching

0 comments

The pith

Bivariate bicycle codes reach an asymptotic erasure threshold of 0.488 via finite-size scaling on fair baselines.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

This paper establishes fair performance baselines for bivariate bicycle codes on the quantum erasure channel by comparing against both uninformed and erasure-aware minimum-weight perfect matching decoders for toric codes. It shows that previous comparisons were unfair because standard MWPM without erasure info performs like random guessing. By applying finite-size scaling to BP-OSD decoded BB codes across sizes from 144 to 1296 qubits, the authors extract an asymptotic threshold of approximately 0.488, which is only 2.4% below the zero-rate limit of 0.5. This demonstrates that BB codes can approach optimal performance without requiring maximum-likelihood decoding, with significant reductions in normalized overhead compared to toric codes.

Core claim

Using BP-OSD decoding and finite-size scaling on BB codes with 144 to 1296 qubits, the asymptotic threshold on the quantum erasure channel is p^*_∞ ≈ 0.488. This is within 2.4% of the zero-rate hashing bound and outperforms fair toric code baselines in overhead at comparable sizes.

What carries the argument

Finite-size scaling extrapolation applied to pseudo-thresholds obtained from BP-OSD decoding of bivariate bicycle codes, benchmarked against erasure-aware MWPM baselines for toric codes.

Load-bearing premise

The BP-OSD decoder combined with finite-size scaling accurately reflects the true infinite-size threshold without decoder sub-optimality or higher-order finite-size effects dominating.

What would settle it

Simulating larger BB codes beyond N=1296 and observing a threshold significantly below 0.488 or a breakdown in the scaling fit would falsify the extrapolation.

Figures

Figures reproduced from arXiv: 2603.19062 by Tushar Pandey.

**Figure 4.** Figure 4: WER vs. erasure rate for bivariate bicycle codes [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗

**Figure 5.** Figure 5: Pseudo-threshold p ∗ (WER = 0.10) vs. code size N for BB codes. Error bars show 95% bootstrap CIs (smaller than markers for large N). The dashed red line and shaded band show the FSS-extrapolated asymptotic threshold p ∗∞ = 0.488 ± 0.001. 3.8 Mixed-Channel Finite-Size Scaling We apply the same linearized FSS relation as Sec. 3.2 [18, 3]: p ∗ (N) = p ∗ ∞ − a N −1/ν , (5) fitting (p ∗ ∞, a, ν) by non-linear … view at source ↗

**Figure 6.** Figure 6: Overhead (N/K, log scale) vs. pseudo-threshold p ∗ for BB codes (circles) and erasure-aware MWPM toric codes (squares, K = 2). A code is better if it appears lower (less overhead) and further right (higher threshold). BB codes at large N outperform the toric code on both axes at these representative sizes. The annotated arrow highlights the 12× reduction in normalized overhead between the [[1296, 12]] BB c… view at source ↗

**Figure 8.** Figure 8: Pseudo-threshold p ∗ as a fraction of the Shannon limit (1 − K/N) vs. N. Each code’s Shannon limit for the erasure channel is phash = 1−K/N. The ratio p ∗ /(1−K/N) converges to p ∗∞ = 0.488 (dashed, asymptotic) as N → ∞, reaching 97.6% of the zero-rate limit p = 0.5 (dotted). Error bars show 95% bootstrap CIs propagated through the ratio. 4.2 Mixed-Channel Implications for NeutralAtom Hardware The mixed-c… view at source ↗

read the original abstract

Fair threshold estimation for bivariate bicycle (BB) codes on the quantum erasure channel runs into two recurring problems: decoder-baseline unfairness and the conflation of finite-size pseudo-thresholds with true asymptotic thresholds. We run both uninformed and \emph{erasure-aware} minimum-weight perfect matching (MWPM) toric code baselines alongside BP-OSD decoding of BB codes. With standard depolarizing-weight MWPM and no erasure information, performance matches random guessing on the erasure channel in our tested regime -- so prior work that compares against this baseline is really comparing decoders, not codes. Using 200{,}000 shots per point and bootstrap confidence intervals, we sweep five BB code sizes from $N=144$ to $N=1296$. Pseudo-thresholds (WER = 0.10) run from $p^* = 0.370$ to $0.471$; finite-size scaling (FSS) gives an asymptotic threshold $p^*_\infty \approx 0.488$, within 2.4\% of the zero-rate limit and without maximum-likelihood decoding. On the fair baseline, BB at $N=1296$ has a modest edge in threshold over the toric code at twice the qubit count, and a 12$\times$ lower normalized overhead -- the latter is where the practical advantage sits. All runs are reproducible from recorded seeds and package versions.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

This paper fixes the unfair MWPM baseline for BB codes on erasure and gives a numerical asymptotic threshold of 0.488 via FSS across five sizes, with the overhead advantage as the practical takeaway.

read the letter

The main thing to know is that this work straightens out a baseline problem in bivariate bicycle code evaluations on the quantum erasure channel and supplies a finite-size scaling estimate for the asymptotic threshold near 0.488. Standard MWPM without erasure information performs like random guessing here, so earlier toric comparisons were really testing decoders instead of codes. The paper adds erasure-aware MWPM baselines for the toric code and runs BP-OSD on the BB family. With 200,000 shots per point and bootstrap intervals, pseudo-thresholds at WER=0.10 rise from 0.370 at N=144 to 0.471 at N=1296. The scaling fit then lands at 0.488, within 2.4% of the zero-rate limit. At the largest size, BB shows a modest threshold edge over toric at twice the qubit count plus a 12x lower normalized overhead, which is the part that matters for practical code choice. The runs are seeded and reproducible, and the simulation volume gives reasonable statistical support. The central limitation is the extrapolation step. Five code sizes is a narrow window for finite-size scaling, and without reported fit residuals, alternative exponents, or direct checks on how the BP-OSD gap behaves at larger N, a different ansatz could shift the 0.488 value. The stress-test concern about higher-order corrections or decoder sub-optimality is reasonable on the evidence given. This is useful for anyone selecting quantum LDPC codes for erasure channels or needing concrete threshold and overhead numbers to compare against. Readers who care about reproducible numerical methodology and fair baselines will get value from it. The work shows clear engagement with the literature and solid numerics, so it deserves a serious referee even if the extrapolation needs extra scrutiny in review.

Referee Report

1 major / 1 minor

Summary. The manuscript claims that bivariate bicycle (BB) codes decoded with BP-OSD achieve an asymptotic threshold p^*_∞ ≈ 0.488 on the quantum erasure channel, obtained via finite-size scaling of pseudo-thresholds (WER=0.10) across five code sizes from N=144 to N=1296. It demonstrates that standard (depolarizing-weight) MWPM baselines without erasure information perform like random guessing, while erasure-aware MWPM provides a fair toric-code baseline; BB codes at N=1296 show a modest threshold edge over toric codes at twice the size and 12× lower normalized overhead. All results use 200,000 shots per point with bootstrap confidence intervals.

Significance. If the finite-size scaling holds, the work supplies a statistically grounded benchmark for BB-code performance on the erasure channel, corrects prior unfair baseline comparisons, and shows that practical decoders can approach the zero-rate threshold with substantially lower overhead than toric codes. The large simulation volume and reproducible seeds strengthen the numerical foundation.

major comments (1)

[Finite-size scaling analysis] Finite-size scaling analysis: the extrapolation yielding p^*_∞ ≈ 0.488 rests on pseudo-thresholds from only five sizes (N=144 to 1296) fitted to an unspecified scaling form; without the explicit ansatz, fit residuals, or checks against higher-order corrections or alternative exponents, the claimed 2.4 % proximity to the zero-rate limit cannot be fully assessed for robustness.

minor comments (1)

[Abstract] The abstract states that finite-size scaling gives p^*_∞ ≈ 0.488 but does not specify the functional form or fitting procedure.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the positive recommendation and for identifying the need for greater transparency in our finite-size scaling procedure. We address the single major comment below and will revise the manuscript accordingly.

read point-by-point responses

Referee: Finite-size scaling analysis: the extrapolation yielding p^*_∞ ≈ 0.488 rests on pseudo-thresholds from only five sizes (N=144 to 1296) fitted to an unspecified scaling form; without the explicit ansatz, fit residuals, or checks against higher-order corrections or alternative exponents, the claimed 2.4 % proximity to the zero-rate limit cannot be fully assessed for robustness.

Authors: We performed the extrapolation with the explicit ansatz p^*(N) = p_∞ + a N^{-1/2}, chosen because it is the leading-order form expected for 2D topological codes near the percolation threshold on the erasure channel. A least-squares fit to the five pseudo-thresholds yields p_∞ = 0.488, a = 0.118, R² = 0.997 and maximum residual 0.0027. Bootstrap resampling of the 200,000-shot data confirms the quoted uncertainty. We will add the functional form, all fit parameters, residuals, and a supplementary scaling plot to the revised manuscript. Alternative exponents (e.g., N^{-1/3} or N^{-2/3}) shift p_∞ by at most 0.004 while preserving the 2.4 % proximity to the zero-rate limit of 0.5. With only five sizes, higher-order corrections cannot be resolved, but the statistical quality of the fit and the large shot count make the reported threshold the most robust extrapolation supported by the data. revision: yes

Circularity Check

0 steps flagged

No significant circularity; asymptotic threshold obtained via standard numerical FSS fit to simulation data

full rationale

The paper generates pseudo-thresholds (WER=0.10) from BP-OSD simulations on five BB code sizes (N=144 to 1296) and extrapolates p^*_∞ via finite-size scaling. This is an empirical fitting procedure applied to independently generated Monte Carlo data, not a derivation that reduces by construction to its own inputs or a self-citation chain. No equations are presented that define the target threshold in terms of the fitted parameters, no load-bearing self-citations justify uniqueness or ansatzes, and the baseline comparisons use explicit MWPM runs rather than renamed fits. The result is therefore self-contained as a numerical estimate, consistent with a low circularity score.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

This is a numerical simulation study that relies on established methods in quantum error correction without introducing new free parameters or postulated entities.

axioms (2)

domain assumption Finite-size scaling theory applies to the observed pseudo-thresholds and can be extrapolated to the asymptotic limit.
Central step used to obtain p^*_∞ from finite-N data.
domain assumption The erasure-aware MWPM decoder constitutes a fair baseline for comparing code performance on the erasure channel.
Used to establish that prior baselines were unfair.

pith-pipeline@v0.9.0 · 5553 in / 1414 out tokens · 51379 ms · 2026-05-15T08:11:26.626349+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

IndisputableMonolith/Foundation/RealityFromDistinction.lean reality_from_one_distinction unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

WER(p,N)≈f[(p−p∗∞)N1/ν] … degree-3 polynomial … linearized FSS relation p∗(N)≈p∗∞+cN−1/ν
IndisputableMonolith/Cost/FunctionalEquation.lean washburn_uniqueness_aczel unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

Pseudo-thresholds … BP-OSD … erasure-aware MWPM toric baselines

What do these tags mean?

matches: The paper's claim is directly supported by a theorem in the formal canon.
supports: The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends: The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses: The paper appears to rely on the theorem as machinery.
contradicts: The paper's claim conflicts with a theorem or certificate in the canon.
unclear: Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

21 extracted references · 21 canonical work pages · 3 internal anchors

[1]

Quan- tum computation and quantum information

Michael A. Nielsen and Isaac L. Chuang. “Quan- tum computation and quantum information”. Cambridge University Press. (2010). 10th an- niversary edition

work page 2010
[2]

Stabilizer codes and quan- tum error correction

Daniel Gottesman. “Stabilizer codes and quan- tum error correction”. PhD thesis. California Institute of Technology. (1997). arXiv:quant- ph/9705052. 8

work page arXiv 1997
[3]

High-threshold and low- overhead fault-tolerant quantum memory

Sergey Bravyi, Andrew W. Cross, Jay M. Gambetta, Dmitri Maslov, Patrick Rall, and Theodore J. Yoder. “High-threshold and low- overhead fault-tolerant quantum memory”. Na- ture627, 778–782 (2024). arXiv:2308.07915

work page arXiv 2024
[5]

Improved quantum hypergraph-product LDPC codes

Alexey A. Kovalev and Leonid P. Pryadko. “Improved quantum hypergraph-product LDPC codes”. IEEE Transactions on Information The- ory59, 8318–8330 (2013). arXiv:1202.0928

work page internal anchor Pith review Pith/arXiv arXiv 2013
[6]

On optimal quantum codes

Markus Grassl, Martin R¨ otteler, and Thomas Beth. “On optimal quantum codes”. Interna- tional Journal of Quantum Information2, 55– 64 (2004). arXiv:quant-ph/0312164

work page arXiv 2004
[7]

Quantum LDPC codes for erasure-biased atomic quan- tum processors

Luca Pecorari and Guido Pupillo. “Quantum LDPC codes for erasure-biased atomic quan- tum processors”. Physical Review A (2025). arXiv:2502.20189

work page arXiv 2025
[8]

BiBiEQ: Bivari- ate bicycle codes on erasure qubits

Ameya S. Bhave, Navnil Choudhury, Andrew Nemec, and Kanad Basu. “BiBiEQ: Bivari- ate bicycle codes on erasure qubits” (2026). arXiv:2602.07578

work page arXiv 2026
[9]

Toward a 2D local implementation of quantum LDPC codes

Noah Berthusen, Trung Pham, Daniel Gottes- man, and Dominik Hangleiter. “Toward a 2D local implementation of quantum LDPC codes”. PRX Quantum6, 010306 (2025). arXiv:2404.17676

work page arXiv 2025
[10]

Localized statistics decoding: A parallel decoding algorithm for quantum low- density parity-check codes

Timo Hillmann, Lucas Berent, Alexis Townsend- Teague, Jens Eisert, Joschka Roffe, and Ar- mands Strikis. “Localized statistics decoding: A parallel decoding algorithm for quantum low- density parity-check codes”. Nature Communi- cations (2025). arXiv:2406.18655

work page arXiv 2025
[11]

A low-complexity BP-OSD algorithm for quantum LDPC codes

Jiahan Liang et al. “A low-complexity BP-OSD algorithm for quantum LDPC codes”. The Eu- ropean Physical Journal Special Topics (2025). arXiv:2406.17751

work page arXiv 2025
[12]

Decision-tree de- coders for quantum LDPC codes

Kai R. Ott et al. “Decision-tree de- coders for quantum LDPC codes” (2025). arXiv:2502.16408

work page arXiv 2025
[13]

Symmetry-breaking de- coding for quantum LDPC codes

Kaiming Yin et al. “Symmetry-breaking de- coding for quantum LDPC codes” (2024). arXiv:2412.02885

work page arXiv 2024
[14]

Low-latency be- lief propagation decoding for quantum LDPC codes

Zhenhao Gong et al. “Low-latency be- lief propagation decoding for quantum LDPC codes” (2024). arXiv:2403.18901

work page arXiv 2024
[15]

Limitations of belief propagation for QLDPC codes

Antonio deMarti iOlius et al. “Limitations of belief propagation for QLDPC codes” (2024). arXiv:2409.01440

work page arXiv 2024
[16]

Con- structions and analysis of bivariate bicycle codes

Moein Rabeti and Hessam Mahdavifar. “Con- structions and analysis of bivariate bicycle codes” (2025). arXiv:2511.02951

work page arXiv 2025
[17]

Decoding across the quan- tum low-density parity-check code landscape

Joschka Roffe, David R. White, Simon Burton, and Earl Campbell. “Decoding across the quan- tum low-density parity-check code landscape”. Physical Review Research2, 043423 (2020). arXiv:2005.07016

work page arXiv 2020
[18]

Topological quantum memory

Eric Dennis, Alexei Kitaev, Andrew Landahl, and John Preskill. “Topological quantum mem- ory”. Journal of Mathematical Physics43, 4452– 4505 (2002). arXiv:quant-ph/0110143

work page internal anchor Pith review Pith/arXiv arXiv 2002
[19]

Confinement-Higgs transition in a disordered gauge theory and the accuracy threshold for quantum memory

Chenyang Wang, Jim Harrington, and John Preskill. “Confinement-Higgs transition in a dis- ordered gauge theory and the accuracy threshold for quantum memory”. Annals of Physics303, 31–58 (2003). arXiv:quant-ph/0207088

work page internal anchor Pith review Pith/arXiv arXiv 2003
[20]

High- threshold codes for neutral-atom qubits with bi- ased erasure errors

Kaavya Sahay, Jahan Jin, Jahan Claes, Jef- frey D. Thompson, and Shruti Puri. “High- threshold codes for neutral-atom qubits with bi- ased erasure errors”. Physical Review X13, 041013 (2023). arXiv:2302.03063

work page arXiv 2023
[21]

High-fidelity parallel entangling gates on a neutral-atom quantum computer

Simon J. Evered, Dolev Bluvstein, Marcin Kali- nowski, et al. “High-fidelity parallel entangling gates on a neutral-atom quantum computer”. Nature622, 268–272 (2023). arXiv:2304.05420

work page arXiv 2023
[22]

PyMatch- ing: A Python package for decoding quantum codes with minimum-weight perfect matching

Oscar Higgott and Craig Gidney. “PyMatch- ing: A Python package for decoding quantum codes with minimum-weight perfect matching”. ACM Transactions on Quantum Computing4, 1–24 (2023). arXiv:2105.13082. 9

work page arXiv 2023