Research map / Federated learning and privacy / Differential privacy
Differentially Private Learning with Adaptive Clipping
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| adaptive clipping to the median instead of best fixed clip chosen in hindsight | DP-FedAvg · CIFAR-100, EMNIST-AE, Shakespeare | test performance | Helped |
| adaptive clipping to the median instead of best fixed clip | DP-FedAvg · EMNIST-CR, SO-NWP | test performance | No clear effect |
| adaptive clipping to the median instead of best fixed clip | DP-FedAvg · SO-LR | recall@5 | Hurt |
| adaptive clipping (no noise) instead of no clipping | FedAvg · CIFAR-100, EMNIST-AE, SO-NWP | performance | Helped |
| adaptive clipping (no noise) instead of no clipping | FedAvg · Shakespeare, SO-LR | performance | Hurt |
| use next larger fixed clip 27.3 instead of 16.0 (ablation) | DP-FedAvg fixed clip · SO-LR dev set | recall@5: 55.1 -> 51.8 | Hurt |
Closest papers by meaning
- Online Sensitivity Optimization in Differentially Private Learning (AAAI 2024 · similarity 0.88)
- Mitigating the Privacy–Utility Trade-off in Decentralized Federated Learning via f-Differential Privacy (NeurIPS 2025 · similarity 0.87)
- Noise-Aware Algorithm for Heterogeneous Differentially Private Federated Learning (ICML 2024 · similarity 0.87)
- Differentially Private Relational Learning with Entity-level Privacy Guarantees (NeurIPS 2025 · similarity 0.86)
- Enabling Differentially Private Federated Learning for Speech Recognition: Benchmarks, Adaptive Optimizers, and Gradient Clipping (NeurIPS 2025 · similarity 0.86)
- One-shot Empirical Privacy Estimation for Federated Learning (ICLR 2024 · similarity 0.86)
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.