Research map / Federated learning and privacy / Personalized and heterogeneous FL
ScaleFL: Resource-Adaptive Federated Learning With Heterogeneous Clients
Fatih Ilhan, Gong Su, Ling Liu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use ScaleFL's resource-adaptive width-and-depth splitting with early exits and self-distillation instead of FedDF | ResNet110 federated training · CIFAR-10 (α=100) | global accuracy: 84.44 -> 85.53 | Improved |
| Use ScaleFL instead of HeteroFL | ResNet110 federated training · CIFAR-10 (α=1) | global accuracy: 79.91 -> 80.83 | Improved |
| Use ScaleFL instead of HeteroFL | ResNet110 federated training · CIFAR-100 (α=100) | global accuracy: 47.12 -> 49.94 | Improved |
| Use ScaleFL instead of FedDF | ResNet110 federated training · CIFAR-100 (α=1) | global accuracy: 44.50 -> 44.95 | Improved |
| Use ScaleFL instead of HeteroFL | MSDNet24 federated training · CIFAR-10 (α=100) | global accuracy: 83.02 -> 84.77 | Improved |
| Use ScaleFL instead of HeteroFL | MSDNet24 federated training · CIFAR-100 (α=1) | global accuracy: 43.00 -> 46.12 | Improved |
| Use ScaleFL instead of HeteroFL | EfficientNet-B4 federated training · ImageNet (α=100) | global accuracy: 46.61 -> 48.95 | Improved |
| Use ScaleFL instead of HeteroFL | EfficientNet-B4 federated training · ImageNet (α=1) | global accuracy: 43.59 -> 46.78 | Improved |
| Use ScaleFL instead of the strongest listed baseline | BERT federated training · SST-2 (α=100) | global accuracy: 88.95 -> 88.58 | Hurt |
| Use ScaleFL instead of FedDF | BERT federated training · SST-2 (α=1) | global accuracy: 82.95 -> 83.79 | Improved |
| Use ScaleFL instead of HeteroFL | BERT federated training · AgNews (α=100) | global accuracy: 91.51 -> 92.13 | Improved |
| Use ScaleFL instead of HeteroFL | BERT federated training · AgNews (α=1) | global accuracy: 90.85 -> 91.20 | Improved |
Later papers that compared with this method
2 later accepted papers report ScaleFL as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.
- FedLAGC: Towards High Performance System-Heterogeneous Federated Learning via Layer-Adaptive Submodel Extraction and Gradient Correction (AAAI 2026 · better on 7 of 7 shared results)
- FIARSE: Model-Heterogeneous Federated Learning via Importance-Aware Submodel Extraction (NeurIPS 2024 · better on 6 of 6 shared results)
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Low Precision Local Training is Enough for Federated Learning (NeurIPS 2024 · similarity 0.89)
- Federated Learning with Heterogeneous LLMs: Integrating Small Student Client Models with a Large Hungry Model (AAAI 2025 · similarity 0.89)
- Adaptive Latent-Space Constraints in Personalized Federated Learning (NeurIPS 2025 · similarity 0.89)
- FedQS: Optimizing Gradient and Model Aggregation for Semi-Asynchronous Federated Learning (NeurIPS 2025 · similarity 0.88)
- NTK-DFL: Enhancing Decentralized Federated Learning in Heterogeneous Settings via Neural Tangent Kernel (ICML 2025 · similarity 0.88)
- One-Shot Heterogeneous Federated Learning with Local Model-Guided Diffusion Models (ICML 2025 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
