Research map / Federated learning and privacy / Communication-efficient FL
FedLTN: Federated Learning for Sparse and Personalized Lottery Ticket Networks
Vaikkunth Mugunthan, Eric Lin, Vignesh Gokul, Christian Lau, Lalana Kagal, Steve Pieper
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 |
|---|---|---|---|
| postprune without rewinding to round 0 (ablation) | FedLTN(0.9) · CIFAR-10 | average client test accuracy (%): 71.5 -> 82.6 | Improved |
| postprune without rewinding to round 0 (ablation) | FedLTN(0.9) · Tiny ImageNet | average client test accuracy (%): 71.8 -> 74.5 | Improved |
| postprune without rewinding to round 0 (ablation) | FedLTN(0.9) · CIFAR-10 | minimum client test accuracy (%): 53.0 -> 63.0 | Improved |
| add Jump-Start local pretraining (ablation) | FedLTN(0.9) · CIFAR-10 | communication cost (MB): 3,753.9 -> 1,846.0 | Improved |
| add Jump-Start local pretraining (ablation) | FedLTN(0.9) · Tiny ImageNet | average client test accuracy (%): 74.5 -> 83.1 | Improved |
| add Jump-Start local pretraining (ablation) | FedLTN(0.9) · CIFAR-10 | average client test accuracy (%): 82.6 -> 82.2 | No clear effect |
| raise target pruning rate from 0.1 to 0.9 (ablation) | FedLTN · CIFAR-10 | average client test accuracy (%): 74.5 -> 82.6 | Improved |
| raise target pruning rate from 0.1 to 0.9 (ablation) | FedLTN · Tiny ImageNet | average client test accuracy (%): 68.4 -> 74.5 | Improved |
| use FedLTN(0.5) instead of LotteryFL(0.5) with Dirichlet skew | FedLTN · CIFAR-10 Dirichlet alpha=0.5 | average client test accuracy (%): 61.19 -> 67.20 | Improved |
| use FedLTN(0.9) instead of LotteryFL(0.5) with Dirichlet skew | FedLTN · CIFAR-10 Dirichlet alpha=0.7 | average client test accuracy (%): 54.00 -> 55.60 | Mixed |
| use BN-preserved aggregation (ResNet18 vs custom CNN without BN) | FedLTN · CIFAR-10 | average client test accuracy (%): 82.6 -> 75.4 | Improved |
| use FedLTN(0.9) instead of LotteryFL(0.5) in 100-client setting | FedLTN · CIFAR-10 (100 clients, custom CNN) | average client test accuracy (%): 74.9 -> 75.4 | Improved |
Closest papers by meaning
- Deconstructing Lottery Tickets: Zeros, Signs, and the Supermask (NeurIPS 2019 · similarity 0.87)
- Efficient Lottery Ticket Finding: Less Data is More (ICML 2021 · similarity 0.87)
- Lottery Pools: Winning More by Interpolating Tickets without Increasing Training or Inference Cost (AAAI 2023 · similarity 0.87)
- FedSelect: Personalized Federated Learning with Customized Selection of Parameters for Fine-Tuning (CVPR 2024 · similarity 0.86)
- Weisfeiler and Leman Go Gambling: Why Expressive Lottery Tickets Win (ICML 2025 · similarity 0.85)
- SparsyFed: Sparse Adaptive Federated Learning (ICLR 2025 · similarity 0.85)
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.
