The paper introduces Layer-wise Budgeted Adaptive Transmission (LBAT), a method that treats federated learning communication as a resource allocation problem under tight uplink budgets. LBAT estimates the transmission value of each layer using local training signals and then applies a byte‑level dynamic programming allocator to choose optimal rank and bit configurations for each layer. Experiments on heterogeneous federated tabular prediction and data generation tasks show that LBAT outperforms uniform compression baselines, achieving better communication‑utility tradeoffs while maintaining distributional fidelity.
By Pengfei Li, Mohammad Khalil
arXiv:2508. 06692v2 Announce Type: replace Abstract: Federated learning systems typically allocate gradient compression by link speed.
By Md. Akmol Masud, Md Abrar Jahin, Mahmud Hasan
arXiv:2610.07466v1 Announce Type: new
Abstract: Multimodal systems often encode every available input, even when a subset suffices for prediction. Adaptive acquisition can reduce this cost by using p...
By Payal Mohapatra, Haodong Yang, Yueyuan Sui, Stephen Xia, Benjamin Lundell, Qi Zhu
LightSplit introduces a lightweight orthogonal random projection at the split learning cut layer to reduce communication overhead and limit information exposure. By sending low‑dimensional projections instead of raw activations, the method preserves end‑to‑end differentiability while discarding irreversibly part of the representation, thereby mitigating reconstruction attacks. Experiments on seven benchmarks show up to 32× compression with minimal accuracy loss and strong resistance to state‑of‑the‑art inversion attacks.
By Mert Cihangiroglu, Alessandro Pegoraro, Phillip Rieger, Antonino Nocera, Ahmad-Reza Sadeghi
The paper introduces EMMI, a framework that enables communication‑efficient inference of multimodal large language models (MLLMs) on edge devices. EMMI encodes each sensor modality separately, fuses the representations, and compresses them into a compact latent vector that is transmitted to a server for high‑capacity reasoning. Experiments on a multimodal benchmark show that EMMI can cut the communication payload by 32× while keeping accuracy comparable, achieving up to a 3.4× reduction in end‑to‑end inference latency under bandwidth‑constrained conditions.
By Motahare Mounesan, Irfan Khan
arXiv:2607. 06651v1 Announce Type: new Abstract: Federated learning (FL) over mobile and edge devices increasingly involves multimodal models in which clients differ in both sensing capability and computational capacity.
By Quoc Bao Phan, Tuy Tan Nguyen
arXiv:2609.34653v2 Announce Type: replace
Abstract: On-device streaming omni-modal inference safeguards user privacy and eliminates prohibitive per-token API costs, but faces a critical bottleneck: t...
By Zongshang Shen, Wangsong Yin, Daliang Xu, Mengwei Xu, Xuanzhe Liu
arXiv:2608.23018v1 Announce Type: cross
Abstract: Federated fine-tuning of on-device large language models (LLMs) faces a significant computing burden. To overcome this limitation, split learning (SL...
By Tao Li, Yulin Tang, Qi Guo, Xianhao Chen
The paper introduces Importance-Aware Class-Balanced Sparsification (ICS) for wireless split learning, a lightweight method that ranks feature channels on the server using Grad‑CAM scores from the true‑class logit. By aggregating per‑class scores into a class‑balanced importance vector, clients can retain the top‑N feature channels without extra forward or backward passes, reducing communication overhead while mitigating head‑class bias under label skew. The authors provide a non‑asymptotic convergence bound, analyze overhead against baselines, and extend the approach to parallel split learning and transformer models, demonstrating consistent performance gains, especially under severe non‑i.i.d. data partitions.
By Bumjun Kim, Yoon Huh, Wan Choi
The paper introduces GaugeLasso, a method that applies symmetric group‑lasso penalties to transformer channels during training, enabling entire tensor slices to be zeroed out while maintaining dense tensors for GPU efficiency. By calibrating channel penalties based on inference utility per compute, the network self‑organizes into depth‑dependent structural profiles that can be dramatically smaller than the original architecture, achieving up to 255‑fold compression on a polynomial division task and outperforming hand‑designed baselines on language modeling and autoencoding benchmarks. The approach also accelerates training and reveals over‑provisioned axes that guide subsequent design iterations.
By Jed A. Duersch, Na\"im Es-Sebbani, Nathana\"el Haas, Zied Bouraoui
arXiv:2608. 09250v1 Announce Type: new Abstract: Federated learning (FL) must serve devices with varying computational capabilities.
By Bostan Khan, Masoud Daneshtalab
arXiv:2607. 18343v1 Announce Type: cross Abstract: Federated fine-tuning is bottlenecked by communication: FedAvg and pseudo-gradient schemes transmit a payload that scales with the model, and gradient compression shrinks it by only a constant factor.
By Radhakrishna Achanta, Will Reed