Asynchrony
Cross-source consensus on Asynchrony from 1 sources and 4 claims.
1 sources · 4 claims
How it works
Benefits
Risks & contraindications
Comparisons
Highlighted claims
- In AFU-IC, retained clients continue training while the target client performs local unlearning. — Asynchronous Federated Unlearning with Invariance Calibration for Medical Imaging
- Most prior unlearning protocols are synchronous and require the server to wait while the target client erases its contribution. — Asynchronous Federated Unlearning with Invariance Calibration for Medical Imaging
- Synchronous unlearning can create downtime in heterogeneous cross-silo medical settings when a slow target client blocks retained clients. — Asynchronous Federated Unlearning with Invariance Calibration for Medical Imaging
- The asynchronous variant was much faster than the synchronous variant while maintaining nearly unchanged backdoor and clean accuracy. — Asynchronous Federated Unlearning with Invariance Calibration for Medical Imaging