Pith. sign in

REVIEW 3 major objections 4 minor 33 references

AutoDFL: A Scalable and Automated Reputation-Aware Decentralized Federated Learning

T0 review · 3 major / 4 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read AutoDFL claims that layering zk-Rollups under a permissioned blockchain lets a reputation-aware federated learning system reach over 3000 transactions per second and cut gas by up to 20 times.

desk verdict AutoDFL has a plausible architecture and a real prototype, but the 3000 TPS headline is arithmetic, not evidence — their own latency numbers contradict it. read the letter →

arxiv 2501.04331 v1 pith:AKY6ZSVV submitted 2025-01-08 cs.DC cs.CRcs.ETcs.LG

Meryem Malak Dif , Mouhamed Amine Bouchiha , Mourad Rabah , Yacine Ghamri-Doudane This is my paper · ORCID
classification cs.DCcs.CRcs.ETcs.LG
keywords federatedlearningblockchainzk-RollupsLayer-2scalingreputationmanagementdecentralizedoraclenetworksmartcontractsscalability
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

This paper argues that reputation-aware federated learning, which normally clogs a blockchain with model submissions, evaluation calls, and reputation updates, can be moved onto a Layer-2 zk-Rollup to regain scalability without giving up the security guarantees of the Layer-1 chain. The proposed framework, AutoDFL, batches federated learning transactions off-chain and posts only validity proofs, while a decentralized oracle network performs the model evaluation and aggregation automatically. The paper reports a proof-of-concept evaluation reaching an average throughput of over 3000 TPS and a gas reduction of up to 20 times compared to a single-layer deployment. If accurate, this makes blockchain-based federated learning practical for large cross-device scenarios while keeping the incentive and reputation system on-chain.

What carries the argument

The load-bearing mechanism is the zk-Rollup batching pipeline, where a batch of up to 20 transactions is committed, proven, and executed on Layer-1 in three phases, so that the gas cost of the prove and execute phases stays roughly constant as the number of transactions grows. The second mechanism is the Decentralized Oracle Network, which retrieves local model weights from IPFS using on-chain hashes, evaluates each model against the task's validation set, computes the score-weighted FedAvg aggregation, and triggers the on-chain reputation update, thereby moving heavy computation off-chain while keeping the result verifiable on-chain.

What would settle it

Run a controlled experiment with AutoDFL where a batch of 20 transactions is submitted to the L2 and measure the time from the first submission to the L1 finalization of the batch. If the measured end-to-end throughput (20 divided by that time) is well below 3000 TPS, or if the total latency per batch grows faster than linearly with batch size, then the claimed throughput scaling fails; the reported L1 throughput under the same workload would also need to be reproduced to confirm the 150 TPS baseline.

Watch

Extended reading notes

Core claim

The central claim is that a dual-layer blockchain architecture, with zk-Rollups as a Layer-2 and a permissioned Byzantine-fault-tolerant chain as Layer-1, can carry a full reputation-aware federated learning workflow—task publishing, local model submission, automated evaluation, score-weighted aggregation, and reputation update—at over 3000 transactions per second and with up to 20 times lower gas cost than a traditional single-layer BFL system, while preserving the same security as the underlying Layer-1. The paper also claims that the proposed reputation model, which derives an objective score from model accuracy, participation rate, and distance from the global model and combines it with a subjective trust rating based on task-publisher interaction history, is automated and fair because task publishers are never allowed to score their own trainers.

Load-bearing premise

The 3000 TPS figure is computed by multiplying the Layer-1 throughput (150 TPS) by the zk-Rollup batch size (20 transactions), which assumes that Layer-1 consensus is the only bottleneck and that the Layer-2 sequencer and proof generation can keep pace without adding latency.

Editorial extensions

If this is right

  • If the claimed throughput is sustained, a BFL system can handle hundreds of trainers across multiple tasks without congesting the base chain.
  • Gas costs become roughly flat per batch beyond the commit phase, so per-task overhead declines as more transactions are batched.
  • The automated oracle-based evaluation removes task publishers from the scoring loop, closing the false-reporting and collusion vectors they would otherwise have.
  • The framework can be repurposed to any decentralized collaborative computation that needs reputation scoring, such as crowdsourcing, data marketplaces, or model exchange.

Reading between the lines

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

  • The 3000 TPS claim is an extrapolation (L1 throughput times batch size) that assumes the off-chain sequencer and proof generator are never the bottleneck; a fair test would measure end-to-end latency from L2 submission to L1 finalization for a full batch.
  • The collusion resistance of the reputation model rests on the assumption that at least two-thirds of the oracle nodes are honest; the paper does not analyze how an adversary controlling a majority of the oracle network could manipulate scores.
  • The gas reduction factor compares on-chain gas only; the cost of running the oracle nodes and generating zero-knowledge proofs is externalized, so a full-system cost comparison would need to include those off-chain expenses.
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

3 major / 4 minor

Summary. The paper proposes AutoDFL, a decentralized federated learning framework that uses a permissioned Layer-1 blockchain with Quorum-based BFT consensus, zk-rollups as a Layer-2 scaling solution, IPFS for off-chain storage, and a Decentralized Oracle Network for automated evaluation and aggregation. A reputation model combines objective and subjective scores to incentivize good behavior. The authors present a proof of concept and report an average throughput of over 3000 TPS and gas reductions of up to 20 times.

Significance. If the performance claims were supported, AutoDFL would be a useful step toward scalable blockchain-based federated learning, and the combination of zk-rollups with an automated reputation mechanism is a relevant design direction. The authors provide a proof of concept with publicly available code, which is commendable. However, the central throughput claim is not substantiated by the reported measurements, and the reputation model is validated only against its own behavior, so the significance of the experimental contributions is currently limited.

major comments (3)
  1. [Section VI-D-2] The headline result of over 3000 TPS is not measured but computed as batch size times L1 throughput (20 × 150 = 3000 TPS). This calculation assumes that L1 consensus is the only bottleneck and that every L1 transaction can carry a full zk-rollup batch, ignoring sequencer capacity, proof-generation time, and calldata/gas limits. The paper's own Table II shows end-to-end latencies of about 3.2 seconds for 20 publishTask calls, implying roughly 6 TPS, not 3000 TPS. Because this is the paper's strongest quantitative claim, the evaluation does not support the abstract's statement.
  2. [Section VI-C, Fig. 3] The validation of the reputation model is circular: the simulation applies the proposed update equations and then confirms that 'good' behavior leads to high reputation and 'malicious' behavior leads to low reputation, which is true by construction. There is no comparison against an alternative reputation scheme, no sensitivity analysis for the numerous free parameters (τ, σ, γ, λ, θ, Rmin, Rinit), and no quantitative metric of how well the model distinguishes honest from malicious participants in a realistic setting. Thus the claim of a 'fair' and 'automated' reputation model is not demonstrated.
  3. [Section III-C and Abstract] The abstract and Section III-C assert that zk-rollups maintain 'the same level of security as the underlying Layer-1 blockchain,' but the paper provides no security analysis for the L2 layer, the sequencer, the proof system, or the interaction with the DON. The threat model assumes that at least 2/3 of blockchain and DON nodes behave correctly, yet the paper does not analyze how this assumption interacts with the zk-rollup trust assumptions. This is a load-bearing claim that should be either formally analyzed or substantially weakened.
minor comments (4)
  1. [Table I] Table I is difficult to parse because the column layout is ambiguous; the labels for dual-layer versus single-layer gas values and the relationship between '#Batches' and 'Consumed Gas' should be clarified, ideally with a revised header and explanatory notes.
  2. [Fig. 5] Figure 5 is presented as a performance comparison, but the AutoDFL throughput values are derived from the same batch-size multiplication described in Section VI-D-2, not from a direct L2 measurement; the caption and text should state this explicitly.
  3. [Section IV] Several symbols in the reputation model are introduced without precise definitions, such as the recency term C_j in Equation (6) and the exact semantics of the interaction counts X_{TA→TP} and X_{TP} in Equation (5). Adding a notation table or more detailed explanations would improve reproducibility.
  4. [Section VI-D-2] The phrase 'gas reduction of up to 20X' is attributed to the batch of 20 function calls; the text should state the batch size and the amortization assumption clearly, since the reduction factor is not constant across the rows of Table I.

Circularity Check

1 steps flagged · score 6.0 of 10

The headline 3000 TPS result is arithmetic by construction: Sec. VI-D-2 defines L2 throughput as batch size times L1 throughput, so the claimed 'achieved throughput' reduces to chosen input parameters rather than a measured system capability.

  1. self definitional [Section VI-D-2 (L2 vs L1 Performance), paragraph after Table I]
    "To calculate L2 throughput, the batch size is multiplied by the L1 throughput. For example, with a batch size of 20 transactions and a L1 throughput of 150 (TPS), AutoDFL can achieve a throughput of 20 ×150 = 3000 (TPS)."

    The paper's central quantitative claim of 'over 3000 TPS' is not a measured result of the AutoDFL proof of concept; it is the output of the paper's own defining formula: L2 throughput = batch size × L1 throughput. The batch size (20) and L1 throughput (150 TPS) are chosen inputs, so the claim 'AutoDFL can achieve 3000 TPS' is exactly the product of those inputs by definition. No empirical L2 throughput measurement, sequencer capacity test, or proof-generation bottleneck analysis is provided. The paper's own latency table (Table II) shows end-to-end L2 times of about 2.3–3.8 s for 20 calls, which implies far lower throughput, further indicating that the 3000 TPS figure is a definitional calculation rather than an observed result. Thus the headline prediction reduces to its own construction.

full rationale

The only step that meets the circularity standard is the throughput claim. Section VI-D-2 explicitly defines the L2 throughput calculation as multiplying batch size by L1 throughput, and then presents the resulting product as the throughput AutoDFL 'can achieve.' This is a self-definitional reduction: the claimed number is exactly the arithmetic product of two parameters the paper selects, so the 'prediction' has no independent empirical content. The paper reports real gas measurements on an L1/L2 setup (Table I) and real latency figures (Table II), but those measured values are not used to derive the 3000 TPS headline; the headline comes from the multiplier formula. This is not a case of self-citation being load-bearing: the references to the authors' prior RollupTheCrowd work [11] are contextual and do not justify the throughput claim. The reputation simulation in Fig. 3 confirms the behavior of the proposed reputation equations rather than predicting external outcomes, but that is a self-contained demonstration of the model's own dynamics, not a circular derivation of a separate result. The security claim of maintaining L1 security is a standard zk-rollup property asserted rather than derived from experiments, but it is not circular in the sense of reducing to the paper's inputs. Overall, one central advertised result is forced by its own definition, while most of the system evaluation remains independent; a score of 6 reflects this partial circularity.

Assumptions & free parameters 8 free parameters · 5 assumptions · 0 invented entities

The central scalability claim rests on the measured L1 throughput and the assumed linear scaling to L2. The reputation model rests on multiple configurable parameters whose values are not disclosed. No new physical or system entities are introduced beyond standard roles. Domain assumptions about honest node thresholds, differential privacy, FedAvg convergence, subjective logic, and L2 security are asserted without proof.

free parameters (8)
  • tau (distance penalty threshold) = not specified
    Eq. 2; adjustable, suggested as average of distances among trainers; simulation value not reported.
  • sigma (uncertainty weight) = not specified
    Eq. 7; configurable parameter controlling subjective reputation; value not reported.
  • gamma (objective vs subjective weight) = not specified
    Eq. 8; configurable; value not reported.
  • lambda (reputation update rate) = not specified
    Eq. 10; set by consortium operating system; value not reported.
  • theta (good/poor behavior weight) = not specified
    Eq. 6; impact factor for subjective logic; value not reported.
  • Rmin (trust threshold) = not specified
    Eq. 9; threshold below which a trainer is untrusted; value not reported.
  • Rinit (initial reputation) = not specified
    Section IV-D; initial reputation for new participants; value not reported.
  • zk-rollup batch size = 20
    Section VI-D-2; used to multiply L1 throughput into the 3000 TPS claim; not empirically varied.
assumptions (5)
  • domain assumption At least 2/3 of blockchain and DON nodes behave correctly
    Threat model (Section III-B) assumes this for security; Quorum BFT and DON integrity rely on it.
  • domain assumption Differential privacy noise (w' = w + n) preserves model accuracy largely intact while preventing inference
    Section III-D step 3 states this without quantifying privacy budget or accuracy loss.
  • domain assumption Score-weighted FedAvg aggregation (Eq. 1) converges similarly to standard FedAvg
    Section III-D step 5 adapts FedAvg but provides no convergence or robustness proof.
  • domain assumption Subjective logic opinions (b, d, u) model trust in this setting
    Section IV-B adapts Josang's subjective logic; the representation is assumed valid for the FL reputation scenario.
  • domain assumption zk-rollups provide the same security as the underlying L1
    Section III-C-3 asserts this without a security analysis; the L1 is a permissioned Quorum chain.

how reviews work

0 comments
Cite this review

Pith. "Pith review of AutoDFL: A Scalable and Automated Reputation-Aware Decentralized Federated Learning." pith.science (2026). https://pith.science/paper/AKY6ZSVV

@misc{pith2026250104331,
  author       = {Pith},
  title        = {Pith review of: AutoDFL: A Scalable and Automated Reputation-Aware Decentralized Federated Learning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/AKY6ZSVV}},
  note         = {Machine review of arXiv:2501.04331}
}
read the original abstract

Blockchained federated learning (BFL) combines the concepts of federated learning and blockchain technology to enhance privacy, security, and transparency in collaborative machine learning models. However, implementing BFL frameworks poses challenges in terms of scalability and cost-effectiveness. Reputation-aware BFL poses even more challenges, as blockchain validators are tasked with processing federated learning transactions along with the transactions that evaluate FL tasks and aggregate reputations. This leads to faster blockchain congestion and performance degradation. To improve BFL efficiency while increasing scalability and reducing on-chain reputation management costs, this paper proposes AutoDFL, a scalable and automated reputation-aware decentralized federated learning framework. AutoDFL leverages zk-Rollups as a Layer-2 scaling solution to boost the performance while maintaining the same level of security as the underlying Layer-1 blockchain. Moreover, AutoDFL introduces an automated and fair reputation model designed to incentivize federated learning actors. We develop a proof of concept for our framework for an accurate evaluation. Tested with various custom workloads, AutoDFL reaches an average throughput of over 3000 TPS with a gas reduction of up to 20X.

Figures

Figures reproduced from arXiv: 2501.04331 by the authors.

Figure 1
Figure 1. System Architecture. ASC refers to the Access Smart Contract, [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. The base layer operates as a permissioned blockchain [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 2
Figure 2. Dual-Layered Blockchain Design: L1 is an Ethereum Virtual [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: L1 Throughput and latency comparison under different [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Average Throughput Comparison: Single-Layered BFL vs [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

33 extracted references · 24 canonical work pages

  1. [1]

    Federated learning for internet of things: A comprehensive survey,

    D. C. Nguyen, M. Ding, P. N. Pathirana, A. Seneviratne, J. Li, and H. V . Poor, “Federated learning for internet of things: A comprehensive survey,” IEEE Communications Surveys & Tutorials , vol. 23, no. 3, pp. 1622–1658, 2021

    2021

  2. [2]

    Data Distribution- Aware Model Aggregation for non-IID Data in a Federated Learning Framework,

    D. Kushwaha, A. Mehrotra, and R. M. Hegde, “Data Distribution- Aware Model Aggregation for non-IID Data in a Federated Learning Framework,” in NOMS IEEE Network Operations and Management Symposium. IEEE, 2024, pp. 1–7

    work page 2024

  3. [3]

    Poisoning attacks against federated learning in load forecasting of smart energy,

    N. B. S. Qureshi, D.-H. Kim, J. Lee, and E.-K. Lee, “Poisoning attacks against federated learning in load forecasting of smart energy,” in NOMS IEEE/IFIP Network Operations and Management Symposium . IEEE, 2022, pp. 1–7

    2022

  4. [4]

    Source inference attacks: Beyond membership inference attacks in federated learning,

    H. Hu, X. Zhang, Z. Salcic, L. Sun, K.-K. R. Choo, and G. Dobbie, “Source inference attacks: Beyond membership inference attacks in federated learning,” IEEE Transactions on Dependable and Secure Computing, 2023

    work page 2023

  5. [5]

    Federated Learning for Generalization, Robustness, Fairness: A Survey and Benchmark,

    W. Huang, M. Ye, Z. Shi, G. Wan, H. Li, B. Du, and Q. Yang, “Federated Learning for Generalization, Robustness, Fairness: A Survey and Benchmark,” IEEE Transactions on Pattern Analysis and Machine Intelligence, pp. 1–20, 2024

    work page 2024

  6. [6]

    Decentralized federated learning: A survey and perspective,

    L. Yuan, Z. Wang, L. Sun, S. Y . Philip, and C. G. Brinton, “Decentralized federated learning: A survey and perspective,” IEEE Internet of Things Journal, 2024

    work page 2024

  7. [7]

    Blockchain- empowered federated learning: Challenges, solutions, and future direc- tions,

    J. Zhu, J. Cao, D. Saxena, S. Jiang, and H. Ferradi, “Blockchain- empowered federated learning: Challenges, solutions, and future direc- tions,” ACM Computing Surveys , vol. 55, no. 11, pp. 1–31, 2023

    work page 2023

  8. [8]

    Decentralized federated learning under free-riders: Credibility analysis,

    L. Zhang, S. Qin, G. Feng, and Y . Peng, “Decentralized federated learning under free-riders: Credibility analysis,” in IEEE INFOCOM - IEEE Conference on Computer Communications Workshops (INFOCOM WKSHPS), 2024, pp. 01–06

    work page 2024

Show all 33 references
  1. [9]

    CDFed: Contribution-based Dynamic Federated Learning for Managing System and Statistical Heterogeneity,

    Y . Qiao, M. S. Munir, A. Adhikary, A. D. Raha, and C. S. Hong, “CDFed: Contribution-based Dynamic Federated Learning for Managing System and Statistical Heterogeneity,” in NOMS IEEE/IFIP Network Operations and Management Symposium , 2023, pp. 1–5

    2023

  2. [10]

    Federated learning framework to decentralize mobility forecasting in smart cities,

    R. Valente, C. Senna, P. Rito, and S. Sargento, “Federated learning framework to decentralize mobility forecasting in smart cities,” inNOMS IEEE/IFIP Network Operations and Management Symposium . IEEE, 2023, pp. 1–5

    2023

  3. [11]

    RollupTheCrowd: Leveraging ZkRollups for a Scalable and Privacy-Preserving Reputation-Based Crowdsourcing Platform,

    A. M. Rafik Bendada, M. A. Bouchiha, M. Rabah, and Y . Ghamri- Doudane, “RollupTheCrowd: Leveraging ZkRollups for a Scalable and Privacy-Preserving Reputation-Based Crowdsourcing Platform,” in IEEE 48th Annual Computers, Software, and Applications Conference (COMPSAC), 2024, p...

    2024

  4. [12]

    A credible and fair federated learning framework based on blockchain,

    L. Chen, D. Zhao, L. Tao, K. Wang, S. Qiao, X. Zeng, and C. W. Tan, “A credible and fair federated learning framework based on blockchain,” IEEE Transactions on Artificial Intelligence , pp. 1–15, 2024

    2024

  5. [13]

    Trustworthy federated learning against malicious attacks in web 3.0,

    Z. Yuan, Y . Tian, Z. Zhou, T. Li, S. Wang, and J. Xiong, “Trustworthy federated learning against malicious attacks in web 3.0,” IEEE Trans- actions on Network Science and Engineering , vol. 11, no. 5, pp. 3969– 3982, 2024

    2024

  6. [14]

    BE- SIFL: Blockchain-empowered secure and incentive federated learning paradigm in IoT,

    Y . Xu, Z. Lu, K. Gai, Q. Duan, J. Lin, J. Wu, and K.-K. R. Choo, “BE- SIFL: Blockchain-empowered secure and incentive federated learning paradigm in IoT,” IEEE Internet of Things Journal , vol. 10, no. 8, pp. 6561–6573, 2021

    2021

  7. [15]

    FGFL: A blockchain-based fair incentive governor for Federated Learning,

    L. Gao, L. Li, Y . Chen, C. Xu, and M. Xu, “FGFL: A blockchain-based fair incentive governor for Federated Learning,” Journal of Parallel and Distributed Computing, vol. 163, pp. 283–299, 2022

    2022

  8. [16]

    Blockchain empowered reli- able federated learning by worker selection: A trustworthy reputation evaluation method,

    Q. Zhang, Q. Ding, J. Zhu, and D. Li, “Blockchain empowered reli- able federated learning by worker selection: A trustworthy reputation evaluation method,” in IEEE wireless communications and networking conference workshops (WCNCW) . IEEE, 2021, pp. 1–6

    2021

  9. [17]

    An Efficient Blockchain As- sisted Reputation Aware Decentralized Federated Learning Framework,

    H. Kasyap, A. Manna, and S. Tripathy, “An Efficient Blockchain As- sisted Reputation Aware Decentralized Federated Learning Framework,” IEEE Transactions on Network and Service Management, vol. 20, no. 3, pp. 2771–2782, 2023

    2023

  10. [18]

    High-quality model aggregation for blockchain-based federated learning via reputation- motivated task participation,

    J. Qi, F. Lin, Z. Chen, C. Tang, R. Jia, and M. Li, “High-quality model aggregation for blockchain-based federated learning via reputation- motivated task participation,” IEEE Internet of Things Journal , vol. 9, no. 19, pp. 18 378–18 391, 2022

    2022

  11. [19]

    BlockFLow: Decentralized, privacy-preserving, and accountable federated machine learning,

    V . Mugunthan, R. Rahman, and L. Kagal, “BlockFLow: Decentralized, privacy-preserving, and accountable federated machine learning,” in Blockchain and Applications: 3rd International Congress . Springer, 2022, pp. 233–242

    2022

  12. [20]

    Fairness, integrity, and privacy in a scalable blockchain-based federated learning system,

    T. R ¨uckel, J. Sedlmeir, and P. Hofmann, “Fairness, integrity, and privacy in a scalable blockchain-based federated learning system,” Computer Networks, vol. 202, p. 108621, 2022

    2022

  13. [21]

    IPFS-content addressed, versioned, P2P file system,

    J. Benet, “IPFS-content addressed, versioned, P2P file system,” arXiv preprint arXiv:1407.3561, 2014

    2014 arXiv

  14. [22]

    A survey on blockchain sharding,

    X. Liu, H. Xie, Z. Yan, and X. Liang, “A survey on blockchain sharding,” ISA transactions, vol. 141, pp. 30–43, 2023

    2023

  15. [23]

    Multi-tasking federated learning meets blockchain to foster trust and security in the Metaverse,

    H. Moudoud and S. Cherkaoui, “Multi-tasking federated learning meets blockchain to foster trust and security in the Metaverse,” Ad Hoc Networks, vol. 150, p. 103264, 2023

    2023

  16. [24]

    ScaleSFL: a shard- ing solution for blockchain-based federated learning,

    E. Madill, B. Nguyen, C. K. Leung, and S. Rouhani, “ScaleSFL: a shard- ing solution for blockchain-based federated learning,” in Proceedings of the Fourth ACM International Symposium on Blockchain and Secure Critical Infrastructure, 2022, pp. 95–106

    2022

  17. [25]

    Analysing and improving shard allocation protocols for sharded blockchains,

    R. Han, J. Yu, and R. Zhang, “Analysing and improving shard allocation protocols for sharded blockchains,” in Proceedings of the 4th ACM Conference on Advances in Financial Technologies, 2022, pp. 198–216

    2022

  18. [26]

    HotStuff: BFT consensus in the lens of blockchain,

    M. Yin, D. Malkhi, M. K. Reiter, G. G. Gueta, and I. Abraham, “HotStuff: BFT consensus in the lens of blockchain,” arXiv preprint arXiv:1803.05069, 2018

    2018 arXiv

  19. [27]

    Blockchain scaling using rollups: A comprehensive survey,

    L. T. Thibault, T. Sarry, and A. S. Hafid, “Blockchain scaling using rollups: A comprehensive survey,” IEEE Access , vol. 10, pp. 93 039– 93 054, 2022

    2022

  20. [28]

    A comprehensive survey on local differential privacy,

    X. Xiong, S. Liu, D. Li, Z. Cai, and X. Niu, “A comprehensive survey on local differential privacy,” Security and Communication Networks , vol. 2020, no. 1, p. 8829523, 2020

    2020

  21. [29]

    On the convergence of FedAvg on non-IID data,

    X. Li, K. Huang, W. Yang, S. Wang, and Z. Zhang, “On the convergence of FedAvg on non-IID data,” arXiv preprint arXiv:1907.02189 , 2019

    1907 arXiv

  22. [30]

    Trust network analysis with subjective logic,

    A. Josang, R. Hayward, and S. Pope, “Trust network analysis with subjective logic,” in Conference Proceedings of the 29th Australasian Computer Science Conference, ACSW, Tasmania, Australia. Australian Computer Society, 2006, pp. 85–94

    2006

  23. [31]

    The MNIST database of handwritten digits,

    Y . LeCun and C. Cortes, “The MNIST database of handwritten digits,”

  24. [32]

    Improved LeNet-5 model based on handwritten numeral recognition,

    S. Tan and Z. Tan, “Improved LeNet-5 model based on handwritten numeral recognition,” in Chinese Control And Decision Conference (CCDC). IEEE, 2019, pp. 6396–6399

    2019

  25. [2005]

    Available: https://api.semanticscholar.org/CorpusID: 60282629

    [Online]. Available: https://api.semanticscholar.org/CorpusID: 60282629

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.