Research map / Federated learning and privacy / Personalized and heterogeneous FL
Client2Vec: Improving Federated Learning by Distribution Shifts Aware Client Indexing
Yongxin Guo, Lin Wang, Xiaoying Tang, Tao Lin
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 |
|---|---|---|---|
| add Client2Vec (sampling+aggregation+local training, Federated index) | FedAvg · CIFAR10 | accuracy: 42.24 -> 59.29 | Improved |
| add Client2Vec (all three case studies, Global index) | FedDyn · CIFAR10 | accuracy: 37.22 -> 70.59 | Improved |
| add Client2Vec (all three case studies, Global index) | FedAvg · DomainNet | accuracy: 46.31 -> 57.43 | Improved |
| add orthogonal loss to index generation (ablation) | Client2Vec · CIFAR10 | accuracy: 15.32 -> 58.28 | Improved |
| add text align loss to index generation (ablation) | Client2Vec · CIFAR10 | accuracy: 54.35 -> 58.28 | Improved |
| add reconstruction loss to index generation (ablation) | Client2Vec · CIFAR10 | accuracy: 31.65 -> 58.28 | Improved |
| use DSA-IGN index instead of Feature Average index (ablation) | Client2Vec · CIFAR10 | accuracy: 11.40 -> 58.28 | Improved |
| use DSA-IGN index instead of Class Prototypes index (ablation) | Client2Vec · CIFAR10 | accuracy: 15.41 -> 58.28 | Improved |
| add label index to feature-only index (ablation) | Client2Vec · CIFAR10 | accuracy: 44.99 -> 58.28 | Improved |
| add orthogonal regularization to local training (ablation) | Client2Vec · CIFAR10 | accuracy: 43.71 -> 58.28 | Improved |
| add client-index-aware sampling only (Global, tau 1.0) (ablation) | FedAvg · CIFAR10 | accuracy: 42.24 -> 45.56 | Improved |
| vary sampling temperature tau from 1.0 to 0.1 (Global index) (ablation) | FedAvg · CIFAR10 | accuracy: 45.56 -> 41.28 | Hurt |
Closest papers by meaning
- FedImpro: Measuring and Improving Client Update in Federated Learning (ICLR 2024 · similarity 0.90)
- Enhancing Clustered Federated Learning: Integration of Strategies and Improved Methodologies (ICLR 2025 · similarity 0.89)
- FLUX: Efficient Descriptor-Driven Clustered Federated Learning under Arbitrary Distribution Shifts (NeurIPS 2025 · similarity 0.89)
- Heterogeneity-Guided Client Sampling: Towards Fast and Efficient Non-IID Federated Learning (NeurIPS 2024 · similarity 0.88)
- Federated Learning with Sample-level Client Drift Mitigation (AAAI 2025 · similarity 0.88)
- Low Precision Local Training is Enough for Federated Learning (NeurIPS 2024 · 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.
