Research map / Federated learning and privacy / Personalized and heterogeneous FL
Data-Free Client Contribution Estimation via Logit Maximization for Federated Learning
Asim Ukaye, Nurbek Tastan, Mubarak Abdu-Aguye, Karthik Nandakumar
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 CELM class-wise contribution weights instead of uniform FedAvg | FedAvg · CIFAR-10 PLS | accuracy: 56.82 -> 59.11 | Improved |
| use CELM weights instead of FedAvg | FedAvg · FashionMNIST PLS | accuracy: 80.62 -> 83.70 | Improved |
| use CELM weights instead of FedAvg | FedAvg · CIFAR-10 Dirichlet 0.10 | accuracy: 68.98 -> 69.21 | No clear effect |
| use CELM instead of CFFL on step label skew | CFFL · FashionMNIST SLS | accuracy: 88.13 vs 87.00 | Hurt |
| use CELM on natural clinical partition | FedAvg · FedISIC | balanced accuracy: 61.25 -> 70.18 | Improved |
| use class-wise normalized evidence instead of similarity-to-average (CGSV) | CGSV · Maverick split | rare-class accuracy: 0.00 -> 61.21 | Improved |
| use CELM under Maverick split | FedAvg · CIFAR-10 Maverick | rare-class accuracy: 39.95 -> 61.21 | Improved |
| combine CELM server aggregation with FedRS restricted softmax | FedRS · CIFAR-10 PLS | accuracy: 62.90 -> 64.05 | Improved |
| combine CELM aggregation with FedRS on Maverick split | CELM · CIFAR-10 Maverick | accuracy: 68.60 -> 63.74 | Hurt |
| use longer warm-up horizon (5% -> 30%) (ablation) | CELM · FashionMNIST Dirichlet 0.05 | accuracy: 83.64 -> 82.70 | Hurt |
| use longer warm-up horizon (5% -> 30%) (ablation) | CELM · FashionMNIST PLS | accuracy: 83.70 -> 84.14 | Improved |
| use CELM for free-rider plus Maverick detection | CFFL/CGSV · FashionMNIST FRM | AUROC: 0.79 -> 1.00 | Improved |
Closest papers by meaning
- Data-Free Black-Box Federated Learning via Zeroth-Order Gradient Estimation (AAAI 2025 · similarity 0.88)
- FLUX: Efficient Descriptor-Driven Clustered Federated Learning under Arbitrary Distribution Shifts (NeurIPS 2025 · similarity 0.88)
- FedQS: Optimizing Gradient and Model Aggregation for Semi-Asynchronous Federated Learning (NeurIPS 2025 · similarity 0.87)
- FedLMT: Tackling System Heterogeneity of Federated Learning via Low-Rank Model Training with Theoretical Guarantees (ICML 2024 · similarity 0.87)
- FedAWA: Adaptive Optimization of Aggregation Weights in Federated Learning Using Client Vectors (CVPR 2025 · similarity 0.87)
- A Bayesian Approach for Personalized Federated Learning in Heterogeneous Settings (NeurIPS 2024 · similarity 0.87)
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.
