# Importance-Aware Feature Sparsification for Wireless Split Learning

arXiv: https://arxiv.org/abs/2609.39194 · HTML: https://mlsanity.com/paper/2609.39194/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 28 of its 38 references that are in the map.
- First posted on arXiv 30 Sep 2026.

## Builds on (most cited first, 28 of 28)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localization](https://mlsanity.com/paper/1610.02391/) (2016)
- [Split learning for health: Distributed deep learning without sharing raw patient data](https://mlsanity.com/paper/1812.00564/) (2018)
- [Sparsified SGD with Memory](https://mlsanity.com/paper/1809.07599/) (2018)
- [SplitFed: When Federated Learning Meets Split Learning](https://mlsanity.com/paper/2004.12088/) (2020)
- [Distributed Learning in Wireless Networks: Recent Progress and Future Challenges](https://mlsanity.com/paper/2104.02151/) (2021)
- [Efficient Parallel Split Learning over Resource-constrained Wireless Edge Networks](https://mlsanity.com/paper/2303.15991/) (2023)
- [Split Learning over Wireless Networks: Parallel Design and Resource Management](https://mlsanity.com/paper/2204.08119/) (2022)
- [Split Learning in 6G Edge Networks](https://mlsanity.com/paper/2306.12194/) (2023)
- [AdaptSFL: Adaptive Split Federated Learning in Resource-constrained Edge Networks](https://mlsanity.com/paper/2403.13101/) (2024)
- [Hierarchical Split Federated Learning: Convergence Analysis and System Optimization](https://mlsanity.com/paper/2412.07197/) (2024)
- [Reducing Communication for Split Learning by Randomized Top-k Sparsification](https://mlsanity.com/paper/2305.18469/) (2023)
- [Communication-Efficient Split Learning via Adaptive Feature-Wise Compression](https://mlsanity.com/paper/2307.10805/) (2023)
- [ESFL: Efficient Split Federated Learning over Resource-Constrained Heterogeneous Wireless Devices](https://mlsanity.com/paper/2402.15903/) (2024)
- [FedLite: A Scalable Approach for Federated Learning on Resource-constrained Clients](https://mlsanity.com/paper/2201.11865/) (2022)
- [Federated Split Learning with Model Pruning and Gradient Quantization in Wireless Networks](https://mlsanity.com/paper/2412.06414/) (2024)
- [Federated Learning with Communication Delay in Edge Networks](https://mlsanity.com/paper/2008.09323/) (2020)
- [Tackling the Class Imbalance Problem of Deep Learning Based Head and Neck Organ Segmentation](https://mlsanity.com/paper/2201.01636/) (2022)
- [Feature Allocation for Semantic Communication with Space-Time Importance Awareness](https://mlsanity.com/paper/2401.14614/) (2024)
- [Universal Joint Source-Channel Coding for Modulation-Agnostic Semantic Communication](https://mlsanity.com/paper/2405.10749/) (2024)
- [Biased Federated Learning under Wireless Heterogeneity](https://mlsanity.com/paper/2503.06078/) (2025)
- [Convergence Analysis of Sequential Split Learning on Heterogeneous Data](https://mlsanity.com/paper/2302.01633/) (2023)
- [Privacy-Enhanced Over-the-Air Federated Learning via Client-Driven Power Balancing](https://mlsanity.com/paper/2410.05907/) (2024)
- [Feature Reconstruction Aided Federated Learning for Image Semantic Communication](https://mlsanity.com/paper/2508.02048/) (2025)
- [Mask-Encoded Sparsification: Mitigating Biased Gradients in Communication-Efficient Split Learning](https://mlsanity.com/paper/2408.13787/) (2024)
- [Two Stage Wireless Federated LoRA Fine-Tuning with Sparsified Orthogonal Updates](https://mlsanity.com/paper/2505.00333/) (2025)
- [Federated Learning Enhanced by Feature Reconstruction for Semantic Communication Module Updates of Agents](https://mlsanity.com/paper/2508.03248/) (2025)
- [Markov-Enforced Discrete Diffusion Model for Digital Semantic Symbol Error Correction](https://mlsanity.com/paper/2603.22983/) (2026)

## Abstract (arXiv)

Wireless split learning (SL) reduces on-device computation by offloading upper layers to a server, yet transmitting high-dimensional intermediate features at each iteration remains a major communication bottleneck. Existing methods select features at the client side using task-agnostic criteria such as magnitude, statistics, or clustering, which increases client-side processing and often degrades accuracy under non-independent and identically distributed (non-i.i.d.) client data. We propose importance-aware class-balanced sparsification (ICS), a lightweight approach in which the server ranks feature channels using Grad-CAM-based scores obtained from the true-class logit during backpropagation. The per-class scores are aggregated into a class-balanced, label-agnostic importance vector that mitigates head-class bias under label skew, and each client reuses this vector in the next round to retain the top-N feature channels, incurring no additional client-side forward or backward passes. We further derive a non-asymptotic convergence bound that isolates the sparsification-induced error and characterizes how the sparsification ratio and mini-batch size jointly affect convergence under a fixed communication budget, and we analyze the communication and computational overhead of ICS against representative baselines. Beyond sequential CNN-based SL, we extend ICS to parallel split learning and to transformer-based models. Experiments show that ICS consistently outperforms the baselines, with larger gains under severe non-i.i.d. partitions.