Pith. sign in

REVIEW 4 major objections 6 minor 27 references

HiKO: A Hierarchical Framework for Beyond-Second-Order KO Codes

T0 review · 4 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read HiKO trains neural Kronecker-operation codes to beat Reed-Muller at orders 3 and 4.

desk verdict Plausible first extension of KO codes to r=3/4 with a transferable training recipe, but the RM baseline and power normalization need to be nailed down before the gains are trusted. read the letter →

arxiv 2506.10121 v1 pith:CJWMLISX submitted 2025-06-11 cs.IT math.IT

Shubham Srivastava , Adrish Banerjee This is my paper · ORCID
classification cs.ITmath.IT MSC 94B3594B0568T07
keywords neuralchannelcodingKroneckeroperationcodesReed-MullerhierarchicaltrainingprogressiveunfreezingPlotkinconstructionAWGNBERperformance
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper claims that neural error-correcting codes built on the Kronecker-operation (KO) architecture can be trained beyond second order, where standard KO codes degrade, by decomposing a high-rate code into simpler constituent codes and transferring learned parameters from those parts. The authors introduce HiKO, a three-phase training protocol: train small KO components, map their parameters into the larger Plotkin tree, then progressively unfreeze the components while fine-tuning with adaptive learning rates. On the AWGN channel, HiKO codes are reported to outperform classical Reed-Muller codes of the same length and dimension for orders $r=3$ and $r=4$, with gains of 0.1 to 0.3 dB at a bit error rate of $10^{-4}$. If the comparison holds, this would be the first demonstration that learned KO-style codes work at high rates, a regime relevant to practical high-throughput links.

What carries the argument

The central object is the hierarchical Plotkin tree shared by RM and KO codes. The paper's mechanism is parameter transfer within that tree: a target $\mathrm{HiKO}(m,r)$ code reuses the trained encoder and decoder networks of constituent $\mathrm{KO}(m-1,r)$ and $\mathrm{KO}(m-1,r-1)$ codes, maps those weights onto the matching subtrees, freezes them, and then unfreezes level by level. The only architectural change to the per-node KO cell is a learnable scalar $\alpha_i$ multiplying the classical Plotkin term $u \oplus v$ in the encoder plus dropout and SELU regularized networks in the decoder; this keeps the parameter count essentially unchanged while making the optimization tractable at rates $r>2$.

What would settle it

Run the identical HiKO training pipeline against an independently implemented RM decoder with the same $10^7$ bits per SNR point and report binomial confidence intervals; if the intervals at the claimed operating points overlap, or if a stronger RM decoder closes the 0.1--0.3 dB gap, the reported consistent gains would not survive. A minimal check: at $-1$ dB, compare the standard error of the reported $9.98\times10^{-5}$ and $2.48\times10^{-4}$ BER values for HiKO(9,3) and RM(9,3).

Watch

Extended reading notes

Core claim

HiKO shows that the obstacle to high-rate KO codes is training, not architecture: the Plotkin tree of $\mathrm{RM}(m,r)$ can be initialized from pre-trained $\mathrm{KO}(m-1,r)$ and $\mathrm{KO}(m-1,r-1)$ constituents, and then jointly optimized with dropout, learnable skip connections (realized as one extra scalar per node), and a progressive unfreezing schedule. With this recipe, HiKO(8,3), HiKO(8,4), HiKO(9,3), and HiKO(9,4) all report lower bit error rates than their Reed-Muller counterparts across the tested SNR range, with the largest margin at long block length: at $-1$ dB, HiKO(9,3) reports BER $9.98\times10^{-5}$ versus $2.48\times10^{-4}$ for RM(9,3). The paper further reports that HiKO codeword distance distributions resemble random Gaussian codebooks more than the concentrated profile of algebraic RM codes, which it reads as evidence that the learned codes approach Shannon-optimal behavior while retaining recursive decoding.

Load-bearing premise

The comparison assumes the Reed-Muller baselines are produced by a fair and correctly tuned implementation of Dumer's recursive decoder under the same test conditions, and that the reported BER values are stable enough to support margins of 0.1 to 0.3 dB.

Editorial extensions

If this is right

  • If the reported BER curves are correct, HiKO codes give the first learned KO-style codes that beat RM codes at orders 3 and 4, extending neural channel coding to rates previously out of reach.
  • The 0.1--0.3 dB improvement at BER $10^{-4}$ for block lengths 256 and 512 would be directly usable in short-to-medium block-length regimes where classical codes still compete.
  • Because the parameter count per node is nearly the same as standard KO codes, the gain costs almost no extra decode complexity at inference time.
  • The Gaussian-like codeword distance distribution suggests that hierarchical initialization steers learned codes toward random-codebook-like behavior, the classical route to capacity, which the paper argues explains the gains.
  • The recursive mapping principle stated for HiKO(5,3) extends to the larger codes studied and in principle to even deeper hierarchies, so the framework is not limited to the four reported configurations.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Inference: The hierarchical recipe is not tied to RM or KO codes; any code family whose generator has a recursive Plotkin-like decomposition could receive the same pretraining-then-unfreezing treatment, so the method may transfer to polar or algebraic-geometry constructions.
  • Inference: A natural stress test is to compare HiKO against RM codes decoded by a stronger soft-decision decoder rather than Dumer's recursive decoder, since the claimed margin is small and could shrink against a stronger baseline.
  • Inference: The reported distance analysis is geometric, not information-theoretic; proving that the learned encoder-decoder pair actually approaches Gaussian-codebook capacity would require computing achievable rates or mutual information, which the paper does not do.
  • Inference: A testable extension is to measure the same hierarchical protocol on channels with higher-order modulation or fading, where the Gaussian-codebook approximation is less obviously optimal.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The paper introduces HiKO, a hierarchical training framework for Kronecker Operation (KO) codes that targets high-order, high-rate configurations (r=3 and r=4) with block lengths 256 and 512. The framework trains smaller constituent KO codes, transfers their parameters into a larger Plotkin-structured encoder/decoder, and then fine-tunes with progressive unfreezing, dropout, and learnable skip connections. The authors report that HiKO codes outperform classical Reed-Muller codes at BER levels around 1e-4, with claimed gains of 0.1 to 0.3 dB, and they present a distance-distribution analysis suggesting that HiKO codes resemble random Gaussian codebooks. The paper also provides a public code repository.

Significance. If the reported comparisons are reliable, this is a meaningful step: it is, to my knowledge, the first claim of KO codes working beyond second order, and it proposes a concrete training strategy (hierarchical parameter transfer plus progressive unfreezing) that could be useful for other neural code families. The strengths are that the HiKO training protocol is described in some detail (Section IV-A), the code is released, and the central claim is an empirical comparison against an external family, not a circular derivation from fitted constants. However, the significance is conditional: the Reed-Muller baseline decoder is not specified, the power/SNR normalization is not stated, and no conventional KO baseline at the target configurations is shown. These gaps directly affect whether the headline 'consistently outperform RM codes' is established, so the contribution, while potentially valuable, is not yet fully supported.

major comments (4)
  1. [Section IV-A and Figure 3] The Reed-Muller baseline is not specified. Section IV-A details the HiKO training protocol (batch size, training SNRs, epochs), but says nothing about how the RM comparison curves in Figure 3 are generated beyond the label 'classical Reed-Muller codes.' Section II-B gives the equations of Dumer's recursive decoder, but a mathematical description is not an implementation specification: the algorithm has variants (soft vs. hard decisions, list size, exact vs. approximate recursion), and Dumer decoding is known to have order-dependent behavior for higher-order RM codes. Without stating the exact decoder, its hyperparameters, and whether it was run under comparable computational conditions, the reported 0.1-0.3 dB gains and the specific point 9.98e-5 vs. 2.48e-4 for (9,3) at -1 dB may reflect a weak baseline rather than a code improvement. This is load-bearing for the central claim and must be fixed by reporting the RM decoder implementation and, ideally, comparing against more than one RM decoding variant.
  2. [Section IV-A and Figure 3] The power constraint and SNR definition are not stated. The HiKO encoder produces real-valued outputs (Section IV-C discusses a quantized version only in the distance analysis), while the RM baseline is BPSK. The paper never specifies the average power constraint (e.g., E[||x||^2] = n) or whether the horizontal axis is Eb/N0 or Es/N0. Without a common normalization, the BER comparison in Figure 3 is not a clean code comparison; a real-valued code can appear better simply because it uses more energy per transmitted symbol. The relation between the training noise variances sigma^2_enc and sigma^2_dec and the test SNR axis should also be clarified.
  3. [Section I-C and Section III] No conventional KO baseline at the target configurations is shown. The introduction says standard KO codes degrade at higher rates, and Figure 1 illustrates two initial KO variants for KO(3,8), but the paper does not compare HiKO(8,3), HiKO(8,4), HiKO(9,3), or HiKO(9,4) against conventionally trained KO codes of the same size and order. Without such a comparison, the reported improvements cannot be attributed to the hierarchical training framework rather than to any end-to-end neural training with the enhanced architecture. This also weakens the 'first successful extension beyond second order' claim, since the reader cannot see whether a plain KO code at r=3 or r=4 indeed fails or how much the hierarchical part contributes.
  4. [Section IV-A and IV-B] No statistical uncertainty or multiple-run results are reported. The paper states that 10^7 bits are used per SNR point, which for the cited BER values around 1e-4 gives small binomial standard errors (roughly 1e-5 or less), so the pointwise differences are not obviously noise. However, 'consistently outperform' across configurations is a claim about training robustness as well as decoding performance, and no confidence intervals, multiple random seeds, or training-run variance are given. This is secondary to the baseline issue, but it should be addressed with seed variation or a statement that the displayed curves are representative.
minor comments (6)
  1. [Figure 1] The notation is inconsistent: Figure 1 labels the codes RM(3,8) and KO(3,8), while the rest of the paper uses RM(m,r) and KO(m,r) with m as the length exponent, so these should be RM(8,3) and KO(8,3).
  2. [Equation (3)] The LLR definition writes P[yi|xi=0] and P[yi|xi=1], but yi is real-valued for the AWGN channel; the notation should use probability densities (e.g., p(yi|xi)) to be precise.
  3. [Section IV-C] The quantized version of HiKO(8,3) in Figure 4 is not described: no quantization resolution, mapping to binary symbols, or resulting rate/power constraint is given, so the claim of 'binary transmission compatibility' is not supported by the presented details.
  4. [Section IV-D] The parameter-count formula uses 'l' as the input codeword length, but at internal Plotkin nodes the dimension is half of the current block length; the notation should be clarified to avoid confusion between the full code length and the node input size.
  5. [Equation (17)] The cyclical learning-rate scheduler S(eta_max, t) is described verbally, but the cycle period is not stated; without it, the schedule is not fully reproducible.
  6. [Reference [2]] The bibliography entry for the BCH code reference is incomplete: it lacks author names and a citation key, making it difficult to verify.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: HiKO's central BER claim is an external empirical comparison, not a derivation from its own equations or fitted parameters.

full rationale

The paper's central claim is that trained HiKO codes outperform classical Reed-Muller codes in bit error rate. That claim is established by Monte Carlo simulation against an external code family, and none of the reported BER numbers is produced by plugging a fitted parameter back into the definition of the HiKO code. The HiKO encoder and decoder equations (12)-(16) define the architecture; they do not mathematically entail the performance curves in Figure 3. The only author self-citation, [22] (Dense KO codes), appears in the related-work enumeration and is not load-bearing. The distance-distribution analysis in Section IV-C is post hoc and interpretive; it does not enter the BER comparison. The main weakness identified in the manuscript, namely the underspecified RM decoder implementation and unclear SNR normalization for the comparison, is a reproducibility and fairness concern rather than a circularity, because the RM baseline is an external benchmark rather than an input to the HiKO construction. No reduction of any claimed result to the paper's own assumptions or fitted values can be exhibited, so there is no circular step to report.

Assumptions & free parameters 5 free parameters · 5 assumptions · 0 invented entities

The central claim rests on empirical training experiments. The trainable weights, schedule constants, and architectural choices are fitted or hand-chosen; the only structurally novel assumption is that hierarchical parameter transfer from small pretrained codes lands the optimizer in a good basin, which is not proven. Standard RM recursion and Dumer decoding are inherited from prior work.

free parameters (5)
  • Learnable skip coefficients alpha_i = not reported per node
    Eq. 12 introduces alpha_i as a trainable scale on the residual u XOR v; the paper counts this as the only added parameter per node but reports no learned values.
  • Hidden layer width H = 120 hidden units
    Section III.B.2 fixes H=120 for decoder networks; no sensitivity analysis is given, but capacity and results depend on it.
  • Training SNR protocol = 0 dB encoder, -1 dB decoder
    Section IV.A fixes these hand-chosen operating conditions; performance at other training SNRs is not analyzed.
  • Progressive unfreezing and learning-rate schedule constants = T=300, K=2, eta_initial 2e-4 and 1e-4, scaling 1/25 and 1/(25*10000)
    Algorithm 1 and Eq. 17 define schedule constants chosen by hand; no ablation shows they are optimal or necessary.
  • Dropout rate = not specified
    Dropout regularization is claimed as an architectural innovation, but the dropout probability is never reported.
assumptions (5)
  • standard math Reed-Muller codes satisfy the Plotkin recursion RM(m,r) = {(u, u XOR v)} with u in RM(m-1,r) and v in RM(m-1,r-1).
    Eqs. 1-2 in Section II.A; this is the classical construction inherited by the paper.
  • standard math Dumer's recursive decoder correctly and efficiently decodes RM codes by alternating soft and hard decisions on the Plotkin tree.
    Section II.B, Eqs. 4-5; used as the baseline decoder and as the skeleton for the KO decoder.
  • domain assumption The AWGN channel with binary transmission and soft-MAP decoding at leaves is the correct evaluation model.
    Section IV.A uses AWGN; the Gaussian-codebook optimality discussion also assumes this channel.
  • ad hoc to paper Weight transfer from pretrained KO(4,3) and KO(4,2) constituents to target KO(8,3), KO(8,4), KO(9,3), and KO(9,4) codes initializes the optimizer in a favorable basin.
    No theoretical analysis supports this transfer; the paper relies on empirical results alone. This is the core new assumption of the hierarchical method.
  • domain assumption Coordinate-wise neural networks can represent the nonlinear combining functions needed at each Plotkin node.
    Inherited from KO codes [21] and used throughout Eqs. 6-8 and 12-16.

how reviews work

0 comments
Cite this review

Pith. "Pith review of HiKO: A Hierarchical Framework for Beyond-Second-Order KO Codes." pith.science (2026). https://pith.science/paper/CJWMLISX

@misc{pith2026250610121,
  author       = {Pith},
  title        = {Pith review of: HiKO: A Hierarchical Framework for Beyond-Second-Order KO Codes},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CJWMLISX}},
  note         = {Machine review of arXiv:2506.10121}
}
abstract

This paper introduces HiKO (Hierarchical Kronecker Operation), a novel framework for training high-rate neural error-correcting codes that enables KO codes to outperform Reed-Muller codes beyond second order. To our knowledge, this is the first attempt to extend KO codes beyond second order. While conventional KO codes show promising results for low-rate regimes ($r < 2$), they degrade at higher rates -- a critical limitation for practical deployment. Our framework incorporates three key innovations: (1) a hierarchical training methodology that decomposes complex high-rate codes into simpler constituent codes for efficient knowledge transfer, (2) enhanced neural architectures with dropout regularization and learnable skip connections tailored for the Plotkin structure, and (3) a progressive unfreezing strategy that systematically transitions from pre-trained components to fully optimized integrated codes. Our experiments show that HiKO codes consistently outperform traditional Reed-Muller codes across various configurations, achieving notable performance improvements for third-order ($r = 3$) and fourth-order ($r = 4$) codes. Analysis reveals that HiKO codes successfully approximate Shannon-optimal Gaussian codebooks while preserving efficient decoding properties. This represents the first successful extension of KO codes beyond second order, opening new possibilities for neural code deployment in high-throughput communication systems.

Figures

Figures reproduced from arXiv: 2506.10121 by the authors.

Figure 1
Figure 1. Performance comparison of classical Reed-Muller [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. The HiKO hierarchical training framework: (a) The three-phase training process for HiKO, showing the progression from [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. BER performance comparison between classical Reed [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Pairwise distance distribution for RM(8,3), HiKO(8,3), [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

27 extracted references · 23 canonical work pages

  1. [22]

    Dense KO codes: Faster convergence and reduced complexity through dense connectivity,

    S. Srivastava and A. Banerjee, “Dense KO codes: Faster convergence and reduced complexity through dense connectivity,” in 2024 IEEE International Symposium on Information Theory (ISIT) , 2024, pp. 1415– 1420

    work page 2024

  2. [1]

    A class of multiple-error-correcting codes and the decoding scheme,

    I. Reed, “A class of multiple-error-correcting codes and the decoding scheme,” Transactions of the IRE Professional Group on Information Theory, vol. 4, no. 4, pp. 38–49, 1954

    1954

  3. [2]

    On a class of error correcting binary group codes,

    “On a class of error correcting binary group codes,” Information and control, vol. 3, no. 1, pp. 68–79, 1960

    work page 1960

  4. [3]

    Near Shannon limit error-correcting coding and decoding: Turbo-codes,

    C. Berrou, A. Glavieux, and P. Thitimajshima, “Near Shannon limit error-correcting coding and decoding: Turbo-codes,” in IEEE Interna- tional Conference on Communications , vol. 2, 1993, pp. 1064–1070 vol.2

    work page 1993

  5. [4]

    Gallager, Low Density Parity Check Codes

    R. Gallager, Low Density Parity Check Codes . Cambridge, MA: MIT Press, 1963

    work page 1963

  6. [5]

    Channel polarization: A method for constructing capacity- achieving codes,

    E. Arikan, “Channel polarization: A method for constructing capacity- achieving codes,” in IEEE International Symposium on Information Theory (ISIT) , 2008, pp. 1173–1177

    work page 2008

  7. [6]

    Tinyturbo: Efficient turbo decoders on edge,

    S. A. Hebbar, R. K. Mishra, S. K. Ankireddy, A. V . Makkuva, H. Kim, and P. Viswanath, “Tinyturbo: Efficient turbo decoders on edge,” in IEEE International Symposium on Information Theory (ISIT) . IEEE, 2022, pp. 2797–2802

    work page 2022

  8. [7]

    Multi-label and concatenated neural block decoders,

    C. T. Leung, M. Motani, and R. V . Bhat, “Multi-label and concatenated neural block decoders,” in IEEE International Symposium on Informa- tion Theory (ISIT) , 2020, pp. 274–279

    2020

Show all 27 references
  1. [8]

    Belief propagation decoding of short graph-based channel codes via reinforcement learning,

    S. Habib, A. Beemer, and J. Kliewer, “Belief propagation decoding of short graph-based channel codes via reinforcement learning,” IEEE Journal on Selected Areas in Information Theory , vol. 2, no. 2, pp. 627–640, 2021

    2021

  2. [9]

    Pruning and quantizing neural belief propagation decoders,

    A. Buchberger, C. H ¨ager, H. D. Pfister, L. Schmalen, and A. Graell i Amat, “Pruning and quantizing neural belief propagation decoders,” IEEE Journal on Selected Areas in Communications , vol. 39, no. 7, pp. 1957–1966, 2021

    1957

  3. [10]

    Learning to decode linear codes using deep learning,

    E. Nachmani, Y . Be’ery, and D. Burshtein, “Learning to decode linear codes using deep learning,” in 54th Annual Allerton Conference on Communication, Control, and Computing (Allerton) , 2016, pp. 341– 346

    2016

  4. [11]

    Deep learning methods for improved decoding of linear codes,

    E. Nachmani, E. Marciano, L. Lugosch, W. J. Gross, D. Burshtein, and Y . Be’ery, “Deep learning methods for improved decoding of linear codes,” IEEE Journal of Selected Topics in Signal Processing , vol. 12, no. 1, pp. 119–131, 2018

    2018

  5. [12]

    DEEPTURBO: Deep turbo decoder,

    Y . Jiang, S. Kannan, H. Kim, S. Oh, H. Asnani, and P. Viswanath, “DEEPTURBO: Deep turbo decoder,” in IEEE 20th International Workshop on Signal Processing Advances in Wireless Communications (SPA WC), 2019, pp. 1–5

    2019

  6. [13]

    Neural successive cancel- lation decoding of polar codes,

    N. Doan, S. Ali Hashemi, and W. J. Gross, “Neural successive cancel- lation decoding of polar codes,” in IEEE 19th International Workshop on Signal Processing Advances in Wireless Communications (SPA WC) , 2018, pp. 1–5

    2018

  7. [14]

    Learning to communicate: Channel auto-encoders, domain specific regularizers, and attention,

    T. J. O’Shea, K. Karra, and T. C. Clancy, “Learning to communicate: Channel auto-encoders, domain specific regularizers, and attention,” in IEEE International Symposium on Signal Processing and Information Technology (ISSPIT), 2016, pp. 223–228

    2016

  8. [15]

    An introduction to deep learning for the physical layer,

    T. O’Shea and J. Hoydis, “An introduction to deep learning for the physical layer,” IEEE Transactions on Cognitive Communications and Networking, vol. 3, no. 4, pp. 563–575, 2017

    2017

  9. [16]

    AI coding: Learning to construct error correction codes,

    L. Huang, H. Zhang, R. Li, Y . Ge, and J. Wang, “AI coding: Learning to construct error correction codes,” IEEE Transactions on Communi- cations, vol. 68, no. 1, pp. 26–39, 2020

    2020

  10. [17]

    DRF codes: Deep SNR-robust feedback codes,

    M. B. Mashhadi, D. Gunduz, A. Perotti, and B. Popovic, “DRF codes: Deep SNR-robust feedback codes,” arXiv preprint arXiv:2112.11789 , 2021

    2021 arXiv

  11. [18]

    MIND: Model indepen- dent neural decoder,

    Y . Jiang, H. Kim, H. Asnani, and S. Kannan, “MIND: Model indepen- dent neural decoder,” in IEEE 20th International Workshop on Signal Processing Advances in Wireless Communications (SPA WC) , 2019, pp. 1–5

    2019

  12. [19]

    Model-driven DNN decoder for turbo codes: Design, simulation, and experimental results,

    Y . He, J. Zhang, S. Jin, C.-K. Wen, and G. Y . Li, “Model-driven DNN decoder for turbo codes: Design, simulation, and experimental results,” IEEE Transactions on Communications , vol. 68, no. 10, pp. 6127–6140, 2020

    2020

  13. [20]

    Learning to denoise and decode: A novel residual neural network decoder for polar codes,

    H. Zhu, Z. Cao, Y . Zhao, and D. Li, “Learning to denoise and decode: A novel residual neural network decoder for polar codes,” IEEE Transactions on V ehicular Technology, vol. 69, no. 8, pp. 8725–8738, 2020

    2020

  14. [21]

    KO codes: inventing nonlinear encoding and decoding for reliable wireless communication via deep-learning,

    A. V . Makkuva, X. Liu, M. V . Jamali, H. Mahdavifar, S. Oh, and P. Viswanath, “KO codes: inventing nonlinear encoding and decoding for reliable wireless communication via deep-learning,” in 38th Interna- tional Conference on Machine Learning , ser. Proceedings of Machine Lear...

    2021

  15. [23]

    ProductAE: Towards training larger channel codes based on neural product codes,

    M. Vahid Jamali, H. Saber, H. Hatami, and J. H. Bae, “ProductAE: Towards training larger channel codes based on neural product codes,” arXiv e-prints , pp. arXiv–2110, 2021

    2021

  16. [24]

    Reed-Muller subcodes: Machine learning-aided design of efficient soft recursive decoding,

    M. V . Jamali, X. Liu, A. V . Makkuva, H. Mahdavifar, S. Oh, and P. Viswanath, “Reed-Muller subcodes: Machine learning-aided design of efficient soft recursive decoding,” in IEEE International Symposium on Information Theory (ISIT) . IEEE, 2021, pp. 1088–1093

    2021

  17. [25]

    Deeppolar: Inventing nonlinear large-kernel polar codes via deep learning,

    S. A. Hebbar, S. K. Ankireddy, H. Kim, S. Oh, and P. Viswanath, “Deeppolar: Inventing nonlinear large-kernel polar codes via deep learning,” arXiv preprint arXiv:2402.08864 , 2024

    2024 arXiv

  18. [26]

    Binary codes with specified minimum distance,

    M. Plotkin, “Binary codes with specified minimum distance,” IRE Transactions on Information Theory , vol. 6, no. 4, pp. 445–450, 1960

    1960

  19. [27]

    Recursive decoding and its performance for low-rate reed- Muller codes,

    I. Dumer, “Recursive decoding and its performance for low-rate reed- Muller codes,” IEEE Transactions on Information Theory, vol. 50, no. 5, pp. 811–823, 2004

    2004

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.