Pith. sign in

REVIEW 4 major objections 3 minor 28 references

Equitable Federated Learning with NCA

T0 review · 4 major / 3 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read A 284 KB neural cellular automaton model matches UNet segmentation in federated learning while shrinking communication and encryption overhead by orders of magnitude.

desk verdict Useful measurement study of a tiny NCA backbone in FL, but the segmentation-parity claim needs much stronger evaluation before it can be trusted. read the letter →

arxiv 2506.21735 v1 pith:IE4A6O6M submitted 2025-06-26 cs.CV

Nick Lemke , Mirko Konstantin , Henry John Krumb , John Kalkhof , Jonathan Stieber , Anirban Mukhopadhyay This is my paper · ORCID
classification cs.CV
keywords FederatedLearningEquityResourceLimitedNeuralCellularAutomataMedicalImageSegmentationHomomorphicEncryptionCommunicationEfficiencyEdgeComputing
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

Federated learning lets hospitals train a shared medical AI without sharing patient images, but the approach has been impractical where internet is slow and clinics lack GPUs. This paper claims that FedNCA removes those barriers by swapping the usual heavy segmentation network for Med-NCA, a neural cellular automaton that repeatedly applies a tiny learned update rule. The entire model is 284 KB, roughly 5,000 times smaller than a UNet, so each federated round transmits far less data and homomorphic encryption of the weight updates takes milliseconds instead of minutes. On fetal ultrasound and chest X-ray segmentation, the authors report Dice scores of 74% and 78%, matching the UNet and TransUNet baselines while cutting communication costs by 500–2000× and encryption runtime by up to 1800×. If the result holds, privacy-preserving collaborative segmentation becomes feasible on sub-€300 smartphones, giving low-resource clinics a route into medical AI.

What carries the argument

The mechanism is Med-NCA, a neural cellular automaton: a segmentation network built from a small learned rule that is applied iteratively over time steps rather than through a deep stack of layers. FedNCA couples two such automata, a coarse stage ($T_0=20$ steps) and a fine stage ($T_1=40$ steps), trained end-to-end via backpropagation through time. The same weights are reused at every time step, which is what keeps the parameter count, and therefore the update size, minuscule; that small size is what lets the CKKS homomorphic scheme encrypt the updates quickly and lets the server average ciphertexts directly.

What would settle it

Re-run the ultrasound experiment with a conventional split, for example 70% of patients in the federated training set across 5 clients and 30% held out, and compare Dice to the UNet baseline with confidence intervals; if FedNCA no longer matches UNet, the equality claim is refuted. A complementary check is to measure actual wall-clock time for one federated round over a simulated 1 Mbps link to confirm the claimed 500–2000× communication saving in practice.

Watch

Extended reading notes

Core claim

The paper's central discovery is that the Med-NCA backbone is not merely a lightweight alternative to UNet but is naturally suited to federated learning: because each client's model is so small, the server can aggregate encrypted updates without decryption, and the whole training loop fits on a low-cost phone. The authors demonstrate this with a two-stage NCA in which a coarse stage runs 20 update steps on a downsampled image and a fine stage runs 40 steps on the upsampled features, trained end-to-end by backpropagation through time. Under federated averaging with the CKKS homomorphic scheme, this model achieves 74% Dice on fetal abdominal ultrasound and 78% Dice on chest X-ray lung segmentation, values the authors report as equal to UNet and TransUNet, while transmitting 500–2000× fewer bytes per round and encrypting weights in under 20 milliseconds versus about 27 seconds for a UNet.

Load-bearing premise

The central claim rests on the ultrasound experiment's inverted data split: 118 of 169 patients (70%) are held out for testing while only 51 patients are split across the 5 federated clients, with no rationale given, so the reported 74% Dice may reflect an unusually small training set and an easy test set rather than a true match to baseline quality.

Editorial extensions

If this is right

  • Federated training becomes possible on smartphones and tablets costing under €300, removing the need for GPU clusters in low-resource clinics.
  • Per-round communication cost drops to roughly a five-hundredth to a two-thousandth of a UNet's, so clients on weak or metered connections can join frequent aggregation rounds.
  • Homomorphic encryption becomes practical for federated segmentation: encrypting and decrypting updates takes under 20 milliseconds, enabling secure aggregation even against untrusted servers.
  • Compression techniques such as 4-bit quantization and top-k sparsification, which the authors find degrade the Dice of UNet and TransUNet, become unnecessary for FedNCA, preserving accuracy while saving bandwidth.

Reading between the lines

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

  • The paper does not report the byte size of the ciphertexts per round; measuring the actual encrypted payload would tell whether the 500–2000× saving survives CKKS packing overhead.
  • If the NCA design generalizes, the same two-stage backbone could be applied to federated training for other dense prediction tasks, such as lesion detection or multi-organ segmentation, where communication constraints are the bottleneck.
  • The authors' results on a single ultrasound split suggest a broader empirical program: benchmarking FedNCA across multiple client counts, non-i.i.d. data distributions, and modalities would test how the efficiency gains trade off against segmentation quality.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 3 minor

Summary. The paper proposes FedNCA, a federated learning system for medical image segmentation built on the Med-NCA backbone. The central claims are that FedNCA matches the segmentation quality of UNet and TransUNet (Dice 74% on fetal ultrasound, 78% on chest X-ray) while reducing communication cost by 500–2000x, enabling training on smartphones, and reducing homomorphic-encryption runtime by about 1800x compared with TransUNet. The authors argue this makes privacy-preserving federated learning practical in low- and middle-income countries with limited bandwidth and hardware.

Significance. If the claims were fully supported, FedNCA would be a useful contribution to resource-constrained federated learning: the model is genuinely tiny, the paper reports measurements on real low-cost smartphones, and the code is public. The combination of a cellular-automata backbone with homomorphic encryption is a sensible direction. However, the significance of the paper rests on the 'equal segmentation quality at drastically reduced cost' claim, and that claim is currently not well supported by the experimental design. The inconsistent communication-reduction factors also make the quantitative headline unreliable. With a stronger evaluation and consistent accounting, the work could be a solid systems contribution, but in its present form the evidence is too thin for the paper's central claims.

major comments (4)
  1. [§4 and Fig. 3] The segmentation-parity claim is not statistically supported. The ultrasound experiment reserves 118 of 169 patients for testing and trains on 51 patients across 5 clients; the X-ray experiment uses only 50 images total, with 25 for testing and 25 for training split across 5 clients. All results are from a single run, with no standard deviations, no multiple seeds, and no significance test. With a 25-image test set, the reported Dice values of 74% and 78% carry wide confidence intervals, so the claim that FedNCA 'achieves equal segmentation quality' to UNet/TransUNet cannot be distinguished from noise. The paper should report error bars over repeated runs and, ideally, a proper patient-wise split rationale, especially given the inverted 70%-test/30%-train split.
  2. [§3 and §5] The communication-reduction factors are mutually inconsistent and unsubstantiated. Section 3 claims a 'nearly 500×' reduction versus U-Net, Section 5 claims '2000× lower communication overhead,' and the same section later states that compressed U-Net/TransUNet updates remain 'at least 300×' larger. The paper also states in Section 1 that Med-NCA has '5000× fewer parameters than a UNet' without deriving this number. These discrepancies matter because communication cost is the paper's headline efficiency measure. The authors should provide a single, clearly defined accounting of model sizes and per-round transmission bytes (including metadata and compression overhead) and derive each claimed ratio explicitly.
  3. [§4 XRay paragraph] The X-ray dataset is misidentified. The text says 'The MIMIC-III [27] dataset consists of chest XRay images,' but MIMIC-III is a clinical database, not a chest X-ray image collection; the relevant imaging resource is MIMIC-CXR. This is not a trivial citation slip, because it affects reproducibility and the exact composition of the 50-image subset. The dataset source and patient-level split should be corrected and described precisely.
  4. [§5, Homomorphic encryption paragraph] The 1400x/1800x homomorphic-encryption speedup is reported without the parameters that determine its validity: CKKS security level, polynomial modulus degree, scale, and whether the same plaintext encoding and ciphertext parameters are used for all three models. Without these details the comparison is hard to reproduce and the claimed speedup may partly reflect an unfavorable choice of encryption parameters for the larger models. Please specify the full HE configuration and, if possible, report ciphertext sizes and end-to-end round times.
minor comments (3)
  1. [Throughout] There are numerous typographical errors and spacing issues, e.g., 'FederatedLearning' in the abstract, 'fine and course NCA' in Algorithm 1, 'Intrinisic gradient compression' in reference [6], and 'Lenovo T ab P11' in Figure 4. The manuscript should be carefully proofread.
  2. [Fig. 3 caption] The caption states that FedNCA 'achieves the best Dice,' but the text claims only equality with baselines; these statements should be reconciled, especially since no error bars are shown.
  3. [Fig. 3 and Baselines paragraph] The notation 'top-25' and 'top-01' is ambiguous: does 'top-01' mean 1% or 0.1%? Define the sparsification percentage clearly and report the resulting transmission cost per round for each setting.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: FedNCA's efficiency and quality claims are backed by direct measurements, not by construction or self-citation chains.

full rationale

The paper's central chain is architectural: a lightweight Med-NCA backbone produces small model updates, which yields lower communication costs and faster homomorphic encryption. Each link in this chain is assessed by direct measurement in Section 5 (Dice scores, transmission cost in MiB, training time on smartphones, and encryption/decryption runtime), not by a fitted parameter or by a result that is equivalent to its input. The self-citations (Med-NCA [10], NCA edge-device training [11]) are used as prior architectural building blocks and as motivation; they are not invoked as a uniqueness theorem, do not forbid alternative architectures, and do not supply the quantitative reduction numbers, which are empirically measured. No equation in the paper reduces another equation to an input by definition. The inverted 70% test split and single-run Dice comparisons are serious evaluation-validity concerns, but they are evidentiary weaknesses rather than circular reasoning. Therefore, the circularity score is 0.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The central claim depends on the parameter count of Med-NCA and the measurement of transmission cost and encryption runtime. These are not derived in this paper but taken from prior work and experimental setup. The hand-chosen values (T0, T1, client count, split) affect the results and are not justified. No invented entities are introduced.

free parameters (3)
  • NCA time steps T0, T1 = T0=20, T1=40
    Hand-set in the Med-NCA architecture from prior work; not fitted here, but they determine the computational cost and hence the communication claim.
  • Number of FL clients = 5
    Chosen for both datasets; no sensitivity analysis provided.
  • Ultrasound train/test split = 51 train, 118 test (30%/70%)
    Authors chose an inverted split; affects the reliability of the Dice results.
assumptions (4)
  • domain assumption Med-NCA architecture as described in [10] has the stated parameter count and behavior.
    The communication and encryption claims rely on the parameter count of Med-NCA being as reported in prior work.
  • domain assumption CKKS homomorphic encryption correctly supports averaging of encrypted model weights without accuracy loss in this setting.
    The paper assumes CKKS can aggregate the encrypted weights; no experiments on accuracy after encryption are reported.
  • domain assumption The transmission cost includes all relevant metadata for each method, making the 300-2000x comparisons valid.
    The paper does not detail how metadata is counted; comparisons may not be apples-to-apples.
  • domain assumption Training on the listed smartphones is representative of low-cost devices in LMICs.
    Only 5 devices are measured; no model of device diversity is provided.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Equitable Federated Learning with NCA." pith.science (2026). https://pith.science/paper/IE4A6O6M

@misc{pith2026250621735,
  author       = {Pith},
  title        = {Pith review of: Equitable Federated Learning with NCA},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IE4A6O6M}},
  note         = {Machine review of arXiv:2506.21735}
}
read the original abstract

Federated Learning (FL) is enabling collaborative model training across institutions without sharing sensitive patient data. This approach is particularly valuable in low- and middle-income countries (LMICs), where access to trained medical professionals is limited. However, FL adoption in LMICs faces significant barriers, including limited high-performance computing resources and unreliable internet connectivity. To address these challenges, we introduce FedNCA, a novel FL system tailored for medical image segmentation tasks. FedNCA leverages the lightweight Med-NCA architecture, enabling training on low-cost edge devices, such as widely available smartphones, while minimizing communication costs. Additionally, our encryption-ready FedNCA proves to be suitable for compromised network communication. By overcoming infrastructural and security challenges, FedNCA paves the way for inclusive, efficient, lightweight, and encryption-ready medical imaging solutions, fostering equitable healthcare advancements in resource-constrained regions.

Figures

Figures reproduced from arXiv: 2506.21735 by the authors.

Figure 1
Figure 1. FedNCA a) circumvents problems with low bandwidth internet connec [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Our FedNCA setup including a) the two-stage Med-NCA backbone, b) [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Dice score and transmission cost in MiB. FedNCA achieves the best Dice [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Training time per epoch of FedNCA on heterogeneous hardware. [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]
Figure 5
Figure 5. Figure 5: Time (in s) taken for homomorphic encryption and decryption of FedNCA, [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

28 extracted references · 24 canonical work pages

  1. [27]

    Mimic-iii, a freely accessible critical care database.Scientific data, 3(1):1–9, 2016

    Alistair EW Johnson, Tom J Pollard, Lu Shen, Li-wei H Lehman, Mengling Feng, Mohammad Ghassemi, Benjamin Moody, Peter Szolovits, Leo Anthony Celi, and Roger G Mark. Mimic-iii, a freely accessible critical care database.Scientific data, 3(1):1–9, 2016

    2016

  2. [1]

    Federated learning for medical image analysis: A survey.Pattern Recognition, page 110424, 2024

    Hao Guan, Pew-Thian Yap, Andrea Bozoki, and Mingxia Liu. Federated learning for medical image analysis: A survey.Pattern Recognition, page 110424, 2024

    work page 2024

  3. [2]

    Privacy-enhancing technologies in federated learning for the internet of healthcare things: a survey.Electronics, 12(12):2703, 2023

    Fatemeh Mosaiyebzadeh, Seyedamin Pouriyeh, Reza M Parizi, Quan Z Sheng, Meng Han, Liang Zhao, Giovanna Sannino, Caetano Mazzoni Ranieri, Jó Ueyama, and Daniel Macêdo Batista. Privacy-enhancing technologies in federated learning for the internet of healthcare things: a survey.Electronics, 12(12):2703, 2023

    work page 2023

  4. [3]

    Federated learning for smart healthcare: A survey.ACM Computing Surveys (Csur), 55(3):1–37, 2022

    DinhCNguyen,Quoc-VietPham,PubuduNPathirana,MingDing,ArunaSenevi- ratne, Zihuai Lin, Octavia Dobre, and Won-Joo Hwang. Federated learning for smart healthcare: A survey.ACM Computing Surveys (Csur), 55(3):1–37, 2022

    work page 2022

  5. [4]

    Artificial intelligence for strengthening healthcare systems in low- and middle-income countries: a systematic scoping review.npj Digital Medicine, 5(1):162, 2022

    Tadeusz Ciecierski-Holmes, Ritvij Singh, Miriam Axt, Stephan Brenner, and San- dra Barteit. Artificial intelligence for strengthening healthcare systems in low- and middle-income countries: a systematic scoping review.npj Digital Medicine, 5(1):162, 2022

    work page 2022

  6. [5]

    PePR: Performance Per Resource Unit as a Metric to Promote Small-Scale Deep Learning in Medical Image Analysis

    Raghavendra Selvan, Bob Pepin, Christian Igel, Gabrielle Samuel, and Erik B Dam. Equity through access: A case for small-scale deep learning.arXiv preprint arXiv:2403.12562, 2024

    work page Pith review arXiv 2024

  7. [6]

    Intrinisic Gradient Compression for Federated Learning

    Luke Melas-Kyriazi and Franklyn Wang. Intrinisic gradient compression for fed- erated learning. arXiv preprint arXiv:2112.02656, 2021

    work page Pith review arXiv 2021

  8. [7]

    FedFMS: Exploring Federated Foun- dation Models for Medical Image Segmentation

    Yuxi Liu, Guibo Luo, and Yuesheng Zhu. FedFMS: Exploring Federated Foun- dation Models for Medical Image Segmentation . Inproceedings of Medical Image Computing and Computer Assisted Intervention – MICCAI 2024, volume LNCS 15008. Springer Nature Switzerland, October 2024

    work page 2024

Show all 28 references
  1. [8]

    Fedcv: a federated learning framework for diverse computer vision tasks.arXiv preprint arXiv:2111.11066, 2021

    Chaoyang He, Alay Dilipbhai Shah, Zhenheng Tang, Di Fan1Adarshan Naiynar Sivashunmugam, Keerti Bhogaraju, Mita Shimpi, Li Shen, Xiaowen Chu, Mahdi Soltanolkotabi, and Salman Avestimehr. Fedcv: a federated learning framework for diverse computer vision tasks.arXiv preprint arXi...

    2021 arXiv

  2. [9]

    No one left behind: Inclusive federated learning over heterogeneous devices

    Ruixuan Liu, Fangzhao Wu, Chuhan Wu, Yanlin Wang, Lingjuan Lyu, Hong Chen, and Xing Xie. No one left behind: Inclusive federated learning over heterogeneous devices. InProceedings of the 28th ACM SIGKDD Conference on Knowledge Dis- covery and Data Mining, pages 3398–3406, 2022

    2022

  3. [10]

    Med-nca: Robust and lightweight segmentation with neural cellular automata

    John Kalkhof, Camila González, and Anirban Mukhopadhyay. Med-nca: Robust and lightweight segmentation with neural cellular automata. InInternational Con- ference on Information Processing in Medical Imaging, pages 705–716. Springer, 2023. 10 N. Lemke et al

    2023

  4. [11]

    Unsupervised training ofneuralcellularautomataonedgedevices

    John Kalkhof, Amin Ranem, and Anirban Mukhopadhyay. Unsupervised training ofneuralcellularautomataonedgedevices. In International Conference on Medical Image Computing and Computer-Assisted Intervention, pages 498–507. Springer, 2024

    2024

  5. [12]

    The evolution of mobile phone surveys in low- and middle-income countries: A study of coverage structure.International Journal of Public Opinion Research, 35(4):edad031, 10 2023

    Mahmoud Elkasabi and Azam Khan. The evolution of mobile phone surveys in low- and middle-income countries: A study of coverage structure.International Journal of Public Opinion Research, 35(4):edad031, 10 2023

    2023

  6. [13]

    Building trusted federated learning: Key technologies and challenges.Journal of Sensor and Actuator Net- works, 12(1), 2023

    Depeng Chen, Xiao Jiang, Hong Zhong, and Jie Cui. Building trusted federated learning: Key technologies and challenges.Journal of Sensor and Actuator Net- works, 12(1), 2023

    2023

  7. [14]

    Preservingprivacyandsecurityinfederatedlearning

    TrucNguyenandMyTThai. Preservingprivacyandsecurityinfederatedlearning. IEEE/ACM Transactions on Networking, 32(1):833–843, 2023

    2023

  8. [15]

    Fedml-he: An efficient homomorphic- encryption-based privacy-preserving federated learning system

    Weizhao Jin, Yuhang Yao, Shanshan Han, Jiajun Gu, Carlee Joe-Wong, Srivatsan Ravi, Salman Avestimehr, and Chaoyang He. Fedml-he: An efficient homomorphic- encryption-based privacy-preserving federated learning system. arXiv preprint arXiv:2303.10837, 2023

    2023 arXiv

  9. [16]

    Transunet: Transformers make strong encoders for medical image segmentation.arXiv preprint arXiv:2102.04306, 2021

    JienengChen,YongyiLu,QihangYu,XiangdeLuo,EhsanAdeli,YanWang,LeLu, Alan L Yuille, and Yuyin Zhou. Transunet: Transformers make strong encoders for medical image segmentation.arXiv preprint arXiv:2102.04306, 2021

    2021 arXiv

  10. [17]

    U-net: Convolutional networks for biomedical image segmentation

    Olaf Ronneberger, Philipp Fischer, and Thomas Brox. U-net: Convolutional networks for biomedical image segmentation. In Medical image computing and computer-assisted intervention–MICCAI 2015: 18th international conference, Mu- nich, Germany, October 5-9, 2015, proceedings, par...

    2015

  11. [18]

    Growing neural cellular automata.Distill, 5(2):e23, 2020

    Alexander Mordvintsev, Ettore Randazzo, Eyvind Niklasson, and Michael Levin. Growing neural cellular automata.Distill, 5(2):e23, 2020

    2020

  12. [19]

    Topology-aware federated learning in edge computing: A comprehensive survey

    Jiajun Wu, Fan Dong, Henry Leung, Zhuangdi Zhu, Jiayu Zhou, and Steve Drew. Topology-aware federated learning in edge computing: A comprehensive survey. ACM Computing Surveys, 56(10):1–41, 2024

    2024

  13. [20]

    Communication- efficient federated learning over capacity-limited wireless networks.IEEE Trans- actions on Cognitive Communications and Networking, 2024

    Jaewon Yun, Yongjeong Oh, Yo-Seb Jeon, and H Vincent Poor. Communication- efficient federated learning over capacity-limited wireless networks.IEEE Trans- actions on Cognitive Communications and Networking, 2024

    2024

  14. [21]

    Federated dropout—a simple approach for enabling federated learning on resource constrained devices.IEEE wireless communications letters, 11(5):923–927, 2022

    Dingzhu Wen, Ki-Jun Jeon, and Kaibin Huang. Federated dropout—a simple approach for enabling federated learning on resource constrained devices.IEEE wireless communications letters, 11(5):923–927, 2022

    2022

  15. [22]

    Fedsz: Leveraging error-bounded lossy compression for federated learning communications

    Grant Wilkins, Sheng Di, Jon C Calhoun, Zilinghan Li, Kibaek Kim, Robert Un- derwood, Richard Mortier, and Franck Cappello. Fedsz: Leveraging error-bounded lossy compression for federated learning communications. In2024 IEEE 44th Inter- national Conference on Distributed Compu...

    2024

  16. [23]

    In- verting gradients-how easy is it to break privacy in federated learning?Advances in neural information processing systems, 33:16937–16947, 2020

    Jonas Geiping, Hartmut Bauermeister, Hannah Dröge, and Michael Moeller. In- verting gradients-how easy is it to break privacy in federated learning?Advances in neural information processing systems, 33:16937–16947, 2020

    2020

  17. [24]

    Source inference attacks in federated learning

    Hongsheng Hu, Zoran Salcic, Lichao Sun, Gillian Dobbie, and Xuyun Zhang. Source inference attacks in federated learning. In2021 IEEE International Con- ference on Data Mining (ICDM), pages 1102–1107. IEEE, 2021

    2021

  18. [25]

    Homomorphic encryption for arithmetic of approximate numbers

    Jung Hee Cheon, Andrey Kim, Miran Kim, and Yongsoo Song. Homomorphic encryption for arithmetic of approximate numbers. In Advances in Cryptology– ASIACRYPT 2017: 23rd International Conference on the Theory and Applications of Cryptology and Information Security, Hong Kong, Chi...

    2017

  19. [26]

    Fetal ab- dominal structures segmentation dataset using ultrasonic images.Mendeley Data, 2023

    Roberto; Santos Luís Otávio; Soares Muylaert Barroso Felipe; Zimmermann Loureiro Chaves Thiago Da Correggio, Karine Souza; Noya Galluzzo. Fetal ab- dominal structures segmentation dataset using ultrasonic images.Mendeley Data, 2023

    2023

  20. [28]

    Adaptive gradient sparsification for efficient federated learning: An online learning approach

    Pengchao Han, Shiqiang Wang, and Kin K Leung. Adaptive gradient sparsification for efficient federated learning: An online learning approach. In2020 IEEE 40th international conference on distributed computing systems (ICDCS), pages 300–

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.