Abstract
Scaling decentralized learning changes not only the number of clients $N$, but also the dynamics of information propagation and consensus. We argue that the effect of increasing $N$ cannot be understood in isolation, because data allocation, topology-dependent mixing, and communication capacity may change simultaneously. We study these coupled effects on CIFAR-10 with $N\in\{10,50,100,200\}$, comparing a degree-two Ring, a Static Random graph, and Local-First Heuristic Evolution (LFHE), a locally adaptive topology process based on friend-of-friend discovery. The Ring provides an analytically transparent failure mode: its Metropolis spectral gap decays as $Θ(N^{-2})$, implying progressively slower contraction of model disagreement as the population grows. Experiments show that holding the nominal local dataset size fixed substantially reduces the apparent population penalty observed when a fixed total dataset is divided among more clients. The remaining degradation depends strongly on communication structure: Ring enters a high-disagreement regime, whereas Static Random and LFHE remain close to consensus. Increasing LFHE's degree threshold further improves accuracy and consensus, but at a substantially higher model-transmission cost. These results show that decentralized scaling is governed by coupled learning and communication dynamics, rather than by the number of clients alone.
Keywords
Publication details
- Journal
- Not available
- Open access
- Green open access
Cite this article
APA 7
Liang, Y. K., Gao, Y., & Long, Y. (2026). Population Scaling or Data Dilution? Dynamics of Local Topology Evolution in Decentralized Learning. https://omanscience.com/en/articles/population-scaling-or-data-dilution-dynamics-of-local-topology-evolution-in-decentralized-learning
MLA 9
Liang, Yin-Kuan, et al. "Population Scaling or Data Dilution? Dynamics of Local Topology Evolution in Decentralized Learning." https://omanscience.com/en/articles/population-scaling-or-data-dilution-dynamics-of-local-topology-evolution-in-decentralized-learning.
Chicago (author–date)
Liang, Yin-Kuan, Yan Gao, and Yang Long. 2026. "Population Scaling or Data Dilution? Dynamics of Local Topology Evolution in Decentralized Learning." https://omanscience.com/en/articles/population-scaling-or-data-dilution-dynamics-of-local-topology-evolution-in-decentralized-learning.
Harvard
Liang, Y. K., Gao, Y. and Long, Y. (2026) 'Population Scaling or Data Dilution? Dynamics of Local Topology Evolution in Decentralized Learning', Available at: https://omanscience.com/en/articles/population-scaling-or-data-dilution-dynamics-of-local-topology-evolution-in-decentralized-learning.
Vancouver
Liang YK, Gao Y, Long Y. Population Scaling or Data Dilution? Dynamics of Local Topology Evolution in Decentralized Learning. https://omanscience.com/en/articles/population-scaling-or-data-dilution-dynamics-of-local-topology-evolution-in-decentralized-learning
IEEE
Y. K. Liang, Y. Gao, and Y. Long, "Population Scaling or Data Dilution? Dynamics of Local Topology Evolution in Decentralized Learning," https://omanscience.com/en/articles/population-scaling-or-data-dilution-dynamics-of-local-topology-evolution-in-decentralized-learning.