CoeF-SFL: Preserving Collaborative Server-Client Learning with Enhanced Communication Efficiency
Abstract
Split Federated Learning (SFL) enables resource-constrained clients to participate in collaborative training, but vanilla SFL exchanges smashed data and gradients at every batch, which incurs significant communication overhead. Recent methods reduce this overhead with an auxiliary network at the client-side cut layer. However, we identify that this approach makes the client optimize a local objective that differs from the end-to-end objective, which fundamentally limits the collaborative training between the client and the server. We propose Compensated Feedback based SFL (CoeF-SFL), a communication-efficient framework that retains the end-to-end objective without any auxiliary network. In CoeF-SFL, the client and the server exchange the smashed data and the gradients once per round and reuse them during local training. Since this reuse makes the gradients stale on the client side, we compensate them with a curvature-based correction in the activation space and develop two variants. CoeF-D approximates the Hessian with a diagonal gradient outer product, while CoeF-J exploits the tractable Jacobian-based Hessian of a surrogate loss that upper-bounds the true loss. We provide the theoretical background of each method, characterizing its compensation. Across vision and language tasks, model capacities, cut layers, and data distributions, CoeF-SFL significantly outperforms auxiliary-network-based methods under the same communication frequency, and the improvement is most substantial on vision tasks. Code is available at https://anonymous.4open.science/r/CoeF-SFL-2686/README.md
Related
- Embedding Information in Policy Randomness: Distribution-Preserving Implicit Communication in MDPs
- Linear maps preserving the matrix discriminant. An enhanced approach
- Ask to Learn: Query-Based Collaborative Reinforcement Learning
- Causality-enhanced and symmetry-preserving PINNs for damped elastodynamic PDEs
- Data-Driven Free-Energy Learning from Trajectories via Structure-Preserving Exponential Time Differencing
BibTeX
@misc{bae2026coefsfl,
title = {{CoeF-SFL: Preserving Collaborative Server-Client Learning with Enhanced Communication Efficiency}},
author = {Junwoo Bae and Jin-Hyun Ahn},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.34360},
eprint = {2609.34360},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.34360},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - CoeF-SFL: Preserving Collaborative Server-Client Learning with Enhanced Communication Efficiency
AU - Junwoo Bae
AU - Jin-Hyun Ahn
PY - 2026
DA - 2026/09/28
PB - arXiv
DO - 10.48550/arxiv.2609.34360
UR - https://arxiv.org/abs/2609.34360
AB - Split Federated Learning (SFL) enables resource-constrained clients to participate in collaborative training, but vanilla SFL exchanges smashed data and gradients at every batch, which incurs significant communication overhead. Recent methods reduce this overhead with an auxiliary network at the client-side cut layer. However, we identify that this approach makes the client optimize a local objective that differs from the end-to-end objective, which fundamentally limits the collaborative training between the client and the server. We propose Compensated Feedback based SFL (CoeF-SFL), a communication-efficient framework that retains the end-to-end objective without any auxiliary network. In CoeF-SFL, the client and the server exchange the smashed data and the gradients once per round and reuse them during local training. Since this reuse makes the gradients stale on the client side, we compensate them with a curvature-based correction in the activation space and develop two variants. CoeF-D approximates the Hessian with a diagonal gradient outer product, while CoeF-J exploits the tractable Jacobian-based Hessian of a surrogate loss that upper-bounds the true loss. We provide the theoretical background of each method, characterizing its compensation. Across vision and language tasks, model capacities, cut layers, and data distributions, CoeF-SFL significantly outperforms auxiliary-network-based methods under the same communication frequency, and the improvement is most substantial on vision tasks. Code is available at https://anonymous.4open.science/r/CoeF-SFL-2686/README.md
ER -