REVIEW 4 major objections 5 minor 18 references
S-VOTE: Similarity-based Voting for Client Selection in Decentralized Federated Learning
T0 review · 4 major / 5 minor · reviewed 2026-08-09 · deepseek-v4-flash
Pith's one-line read S-VOTE selects aggregation partners by model-weight similarity, claiming up to 21% lower communication costs, 4-6% faster convergence, 9-17% higher accuracy in some configurations, and 14-24% energy savings in decentralized federated…
desk verdict A plausible DFL client-selection heuristic with a coherent idea, but an unreported core hyperparameter and selective baselines keep it from being reproducible as written. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The central mechanism is the combination of cosine similarity between model weight vectors (Eq. 3) with a statistical threshold selector (Eq. 4) that keeps only clients whose similarity to the local model exceeds mu + tau * sigma, where mu and sigma are the mean and standard deviation of similarities to all neighbors. A voting rule (Eq. 5) then gates local training: a client with at least Vmin = N/2 votes trains normally, while under-voted clients fall back to random training with a gradually increasing probability (Eq. 6). This machinery selects a per-client subset of aggregation partners at each round, reducing the number of model transmissions and the energy spent on communication and aggregation.
What would settle it
Measure the correlation between pairwise cosine similarity of model weight vectors and a direct measure of label-distribution similarity (e.g., Jensen-Shannon divergence of class proportions) across the Dirichlet(0.1) and Dirichlet(0.5) splits used in the paper; if the correlation is weak or negative in some regimes, S-VOTE would be selecting the wrong neighbors and its accuracy improvements would not replicate.
Extended reading notes
Core claim
S-VOTE establishes that a purely local, similarity-based voting rule can replace full-neighbor aggregation in DFL without sacrificing model quality, while substantially reducing resource usage. After an initial FedAvg phase and a short divergence phase, each client computes the cosine similarity between its own model weight vector and those of all neighbors, selects the neighbors whose similarity is at least the mean plus tau times the standard deviation of all pairwise similarities, and aggregates only the models of those selected clients. The voting step then decides which clients continue training: a client trains normally if it receives at least N/2 votes (or has at most two neighbors), otherwise it trains randomly with probability p that starts at 0.1 and increases by 0.1 each round it remains under-voted. The paper claims this mechanism enhances generalization to unseen data and stabilizes training in non-IID environments, as evidenced by experiments on MNIST, FashionMNIST, EMNIST, and CIFAR-10 with Dirichlet-split data.
Load-bearing premise
The whole mechanism rests on the assumption that the cosine similarity between two clients' model weight vectors reliably reflects how similar their local data distributions are, so that picking the neighbors above the mean-plus-tau threshold actually picks the most useful aggregation partners.
Editorial extensions
If this is right
- Deploying S-VOTE in a peer-to-peer federation reduces the number of model exchanges per round, directly lowering communication bandwidth and energy consumption without requiring a central coordinator.
- In highly non-IID settings (Dirichlet alpha = 0.1), S-VOTE improves F1 scores by up to 37 points over FedAvg in some configurations, suggesting that similarity-based selection mitigates the negative effects of data heterogeneity.
- The conditional training rule keeps underutilized clients from dropping out entirely, allowing rare data distributions to contribute occasionally and thereby improving generalization to unseen classes.
- The method's gains are consistent across fully connected and random topologies, indicating it scales to partial connectivity which is more realistic in decentralized deployments.
Reading between the lines
- Because the paper never reports the value of tau used to set the selection threshold, the exact operating point of the mechanism is underdetermined; a practitioner would need to tune tau per federation, and the claimed gains may be sensitive to this choice.
- The cosine-similarity assumption could be tested directly: if two clients have similar model weights but very different label distributions (e.g., disjoint classes that happen to produce similar weight directions), S-VOTE would select the wrong aggregation partners, so the mechanism's success hinges on the empirical correlation between weight similarity and data-distribution similarity.
- The energy savings reported likely depend on the communication cost model; in settings where local computation dominates over communication, the 14-24% reduction would shrink, while in bandwidth-constrained networks it might be larger.
- The voting and threshold ideas could be transplanted to centralized federated learning, where a server could apply the same similarity filter to select participating clients, eliminating the need for a local voting step while retaining the communication savings.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper presents S-VOTE, a decentralized federated learning client-selection mechanism in which each client selects aggregation partners via cosine similarity of model weight vectors (Eq. 3), applies a threshold μ + τ·σ (Eq. 4), and votes for selected peers; clients that receive fewer than Vmin votes perform probabilistic local training. The authors evaluate S-VOTE on MNIST, FashionMNIST, EMNIST, and CIFAR-10 under Dirichlet α∈{0.1,0.5}, comparing against FedAvg, SCAFFOLD, and FedProx. They report F1 scores, elapsed time, energy consumption, and byte-transmission costs, claiming up to 21% lower communication costs, 4-6% faster convergence, 9-17% higher accuracy, and 14-24% lower energy consumption.
Significance. The proposed mechanism is simple and plausible: using model-similarity to restrict aggregation is a reasonable heuristic for non-IID DFL, and the conditional-training rule is intuitive. The paper evaluates on four datasets, two topologies, and two heterogeneity levels, using real network communications and an energy-monitoring module, which is a strength. If the algorithm is fully specified and the gains survive a more systematic comparison, this would be a useful contribution to resource-efficient DFL. However, the current manuscript is not yet reproducible because the key threshold τ is unreported, and the empirical support is selective with several configurations where a baseline outperforms S-VOTE by a wide margin.
major comments (4)
- [III-D / Algorithm 1] Eq. (4) defines the selection rule with a threshold μ + τ·σ, and Algorithm 1 lists τ as an input, but Section IV-A never states the value of τ used in any of the experiments. Since τ controls how many neighbors are selected (τ=0 selects about half, large τ selects almost none), the reported F1 scores, convergence times, byte counts, and energy figures in Tables II-IV are functions of an unspecified hyperparameter. The paper must report τ for every configuration or provide a sensitivity analysis; without this, the central selection rule cannot be reproduced and the reported gains cannot be attributed to a well-defined algorithm.
- [IV-B.1 / Table II] The claim of 9-17% higher accuracy is not representative of Table II. On MNIST with α=0.1, FedProx achieves 0.97-0.98 while S-VOTE achieves 0.86-0.94 across the four topologies; on EMNIST and CIFAR-10, SCAFFOLD beats S-VOTE in most configurations, e.g., EMNIST α=0.5 fully-connected 20 clients: 0.84 vs 0.77, and CIFAR-10 α=0.5 fully-connected 20 clients: 0.70 vs 0.65. Also, the reported standard deviations (±0.2-0.3) are comparable to or larger than the differences in most configurations. The paper should report how often S-VOTE wins, loses, and ties per configuration, and quantify statistical significance, rather than highlighting only favorable cases.
- [IV-B.2 / Tables III and IV] Resource-efficiency comparisons are made only against FedAvg, while the abstract says 'compared to baseline methods'. SCAFFOLD and FedProx are excluded from the time/energy/communication tables because they 'introduce additional computational overhead' (Section IV-B.2), but this is not a sufficient justification when the abstract makes general claims. Furthermore, Table III reports elapsed time reduction (5-6%), not 'faster convergence' in terms of rounds, so the 4-6% faster convergence claim is not directly supported. The energy and communication gains also do not account for the extra cost of computing pairwise cosine similarities and voting messages.
- [Algorithm 1, Steps 3-6] Line 25 states that each client shares its updated model with all neighbors, yet Section IV-B.2 attributes the byte-count reductions to clients 'not waiting for updates from all their neighbors'. If models are always broadcast to all neighbors, the number of bytes sent per client should be similar to FedAvg; the reductions in Table IV require that non-selected or non-training clients either do not send or send less often. The paper should clarify the exact communication protocol and reconcile Algorithm 1 with Table IV.
minor comments (5)
- [Table I] The dataset name 'CIFACR10' is a typo and should be 'CIFAR10'.
- [Section III-C, Eq. (3)] The norms of the weight vectors are written as |wi| and |wj|; they should be double bars, e.g., ||wi|| and ||wj||.
- [Title and Algorithm 1] The title contains 'V oting' with an unusual space; the same spacing issue appears in the Algorithm 1 heading. The text should be 'Voting'.
- [Section IV-A] The statement that Vmin = N/2 'optimally stabilizes training' is not accompanied by a sensitivity analysis or comparison with other Vmin values, so this claim is not supported by the reported experiments.
- [References] Reference [7] contains 'F oundations' with an unusual space, and reference [4] is an arXiv preprint rather than a peer-reviewed publication; the authors should verify the formatting of all references.
Circularity Check
No circular derivation: S-VOTE's reported gains are measured against external benchmarks rather than derived from fitted inputs; the unreported τ is a reproducibility gap, not a circular step.
full rationale
S-VOTE is presented as an empirical heuristic, not as a derived bound or prediction. Equations (1)-(6) specify the aggregation, cosine-similarity selection, voting, and conditional local training; the reported improvements are direct comparisons against FedAvg, SCAFFOLD, and FedProx on MNIST, FashionMNIST, EMNIST, and CIFAR-10. No reported metric is defined as a function of a fitted parameter: τ is an input to Eq. (4), Vmin is set to N/2, and neither is calibrated against the reported accuracy, byte, or energy numbers. The main deficiency is that the value of τ used in Algorithm 1 is never stated, which makes the selection rule hard to reproduce exactly, but that is a specification and reproducibility concern rather than circularity. The self-citations in the paper ([8], [12], [17], [18]) appear in background and experimental-infrastructure roles; none is used as the justification for the claimed performance gains, and the core evaluations are external benchmarks. Therefore no circular step is found.
Assumptions & free parameters
free parameters (4)
- tau (similarity threshold multiplier) =
not reported
- Vmin (vote threshold) =
N/2
- p (initial random training probability) =
0.1, increasing by 0.1 per skipped round
- Tinit and n (initial and divergence round counts) =
not reported
assumptions (4)
- domain assumption Cosine similarity of model weight vectors between clients reflects similarity of their local data distributions.
- ad hoc to paper A global threshold mu + tau * sigma on cosine similarity selects an aggregation set that improves performance.
- ad hoc to paper Vote threshold Vmin = N/2 and probabilistic skip-training with p=0.1 increments preserve model quality while saving resources.
- domain assumption FedAvg is an upper bound on resource consumption, so comparisons against FedAvg alone establish resource efficiency.
Cite this review
Pith. "Pith review of S-VOTE: Similarity-based Voting for Client Selection in Decentralized Federated Learning." pith.science (2026). https://pith.science/paper/AWE7NNTK
@misc{pith2026250119279,
author = {Pith},
title = {Pith review of: S-VOTE: Similarity-based Voting for Client Selection in Decentralized Federated Learning},
year = {2026},
howpublished = {\url{https://pith.science/paper/AWE7NNTK}},
note = {Machine review of arXiv:2501.19279}
}
read the original abstract
Decentralized Federated Learning (DFL) enables collaborative, privacy-preserving model training without relying on a central server. This decentralized approach reduces bottlenecks and eliminates single points of failure, enhancing scalability and resilience. However, DFL also introduces challenges such as suboptimal models with non-IID data distributions, increased communication overhead, and resource usage. Thus, this work proposes S-VOTE, a voting-based client selection mechanism that optimizes resource usage and enhances model performance in federations with non-IID data conditions. S-VOTE considers an adaptive strategy for spontaneous local training that addresses participation imbalance, allowing underutilized clients to contribute without significantly increasing resource costs. Extensive experiments on benchmark datasets demonstrate the S-VOTE effectiveness. More in detail, it achieves lower communication costs by up to 21%, 4-6% faster convergence, and improves local performance by 9-17% compared to baseline methods in some configurations, all while achieving a 14-24% energy consumption reduction. These results highlight the potential of S-VOTE to address DFL challenges in heterogeneous environments.
Figures
Figures from the paper (1 more)
Reference graph
Works this paper leans on
-
[1]
Find Your Friends: Personalized Federated Learning with the Right Collaborators
Y . Sui, J. Wen, Y . Lau, B. L. Ross, and J. C. Cresswell, “Find your friends: Personalized federated learning with the right collaborators,” arXiv preprint arXiv:2210.06597, 2022
work page Pith review arXiv 2022
-
[2]
A. Sadiev et al. , “Decentralized personalized federated learning: Lower bounds and optimal algorithm for all per- sonalization modes,” EURO Journal on Computational Optimization, vol. 10, p. 100041, 2022
work page 2022
-
[3]
Decentral- ized personalized federated learning,
S. Kharrat, M. Canini, and S. Horvath, “Decentral- ized personalized federated learning,” arXiv preprint arXiv:2406.06520, 2024
arXiv 2024
-
[4]
Decentralized Personalized Federated Learning based on a Conditional Sparse-to-Sparser Scheme
Q. Long, Q. Wang, C. Anagnostopoulos, and D. Bi, “Decentralized personalized federated learning based on a conditional sparse-to-sparser scheme,” arXiv preprint arXiv:2404.15943, 2024
work page Pith review arXiv 2024
-
[5]
Smart Sampling: Helping from Friendly Neighbors for Decentralized Federated Learning
L. Wang, Y . Chen, Y . Guo, and X. Tang, “Smart sam- pling: Helping from friendly neighbors for decentralized federated learning,” arXiv preprint arXiv:2407.04460 , 2024
work page Pith review arXiv 2024
-
[6]
A review of ap- plications in federated learning,
L. Li, Y . Fan, M. Tse, and K.-Y . Lin, “A review of ap- plications in federated learning,” Computers & Industrial Engineering, vol. 149, p. 106854, 2020
2020
-
[7]
Advances and open problems in federated learning,
P. Kairouz et al. , “Advances and open problems in federated learning,” F oundations and trends® in machine learning, vol. 14, no. 1–2, pp. 1–210, 2021
work page 2021
-
[8]
E. T. M. Beltr ´an et al., “Decentralized federated learning: Fundamentals, state of the art, frameworks, trends, and challenges,” IEEE COMST , 2023
work page 2023
Show all 18 references
-
[9]
The non-iid data quagmire of decentralized ma- chine learning,
K. Hsieh, A. Phanishayee, O. Mutlu, and P. Gib- bons, “The non-iid data quagmire of decentralized ma- chine learning,” in International Conference on Machine Learning. PMLR, 2020, pp. 4387–4398
2020
-
[10]
Decentralized feder- ated learning: Balancing communication and computing costs,
W. Liu, L. Chen, and W. Zhang, “Decentralized feder- ated learning: Balancing communication and computing costs,” IEEE Transactions on Signal and Information Processing over Networks , vol. 8, pp. 131–143, 2022
2022
-
[11]
Dart: A solution for decentralized federated learning model robustness analysis,
C. Feng, A. H. Celdr ´an, J. V on der Assen, E. T. M. Beltr´an, G. Bovet, and B. Stiller, “Dart: A solution for decentralized federated learning model robustness analysis,” Array, p. 100360, 2024
2024
-
[12]
Profe: Communication-efficient decentralized federated learning via distillation and pro- totypes,
P. M. S ´anchez S ´anchez, E. Tom ´as Mart ´ınez Beltr ´an, M. Fern ´andez Llamas, G. Bovet, G. Mart ´ınez P ´erez, and A. Huertas Celdr´an, “Profe: Communication-efficient decentralized federated learning via distillation and pro- totypes,” arXiv preprint arXiv:2412.11207 , 2024
2024 arXiv
-
[13]
Client selection in federated learning: Principles, challenges, and opportunities,
L. Fu, H. Zhang, G. Gao, M. Zhang, and X. Liu, “Client selection in federated learning: Principles, challenges, and opportunities,” IEEE Internet of Things Journal , 2023
2023
-
[14]
Adaptive client clustering for efficient federated learning over non- iid and imbalanced data,
B. Gong, T. Xing, Z. Liu, W. Xi, and X. Chen, “Adaptive client clustering for efficient federated learning over non- iid and imbalanced data,” IEEE Transactions on Big Data, 2022
2022
-
[15]
A compre- hensive survey on client selections in federated learning,
A. Gouissem, Z. Chkirbene, and R. Hamila, “A compre- hensive survey on client selections in federated learning,” Innovation and Technological Advances for Sustainabil- ity, pp. 417–428, 2024
2024
-
[16]
A comprehensive survey on client selection strategies in federated learning,
J. Li, T. Chen, and S. Teng, “A comprehensive survey on client selection strategies in federated learning,” Com- puter Networks , p. 110663, 2024
2024
-
[17]
Fedstellar: A platform for decentralized federated learning,
E. T. M. Beltr ´an et al. , “Fedstellar: A platform for decentralized federated learning,” Expert Systems with Applications, vol. 242, p. 122861, 2024
2024
-
[18]
Design and implementation of environmental sustainability module for decentralized federated learning,
X. Cheng, “Design and implementation of environmental sustainability module for decentralized federated learning,” Master’s thesis, 2024. [Online]. Available: https://files.ifi.uzh.ch/CSG/staff/feng/external/ theses/MA XiaoChen.pdf
2024
Reviewed August 9, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.