 Language Models as Zero-shot Lossless Gradient Compressors: Towards General Neural Parameter Prior ModelsNeural Information Processing Systems (NeurIPS), 2024 |
 FedLAP-DP: Federated Learning by Sharing Differentially Private Loss
ApproximationsProceedings on Privacy Enhancing Technologies (PoPETs), 2023 |
 ProgFed: Effective, Communication, and Computation Efficient Federated
Learning by Progressive TrainingInternational Conference on Machine Learning (ICML), 2021 |
