ELF: Federated Langevin Algorithms with Primal, Dual and Bidirectional Compression

Abstract

Federated sampling algorithms have recently gained great popularity in the community of machine learning and statistics. This paper proposes a new federated sampling algorithm called Error Feedback Langevin algorithms (ELF). In particular, we analyze the combinations of EF21 and EF21-P with the federated Langevin Monte-Carlo. We propose three algorithms, P-ELF, D-ELF, and B-ELF, that use primal, dual, and bidirectional compressors. We analyze the proposed methods under Log-Sobolev inequality and provide non-asymptotic convergence guarantees. Simple experimental results support our theoretical findings.

Cite

Text

Karagulyan and Richtárik. "ELF: Federated Langevin Algorithms with Primal, Dual and Bidirectional Compression." Proceedings of the Forty-first Conference on Uncertainty in Artificial Intelligence, 2025.

Markdown

[Karagulyan and Richtárik. "ELF: Federated Langevin Algorithms with Primal, Dual and Bidirectional Compression." Proceedings of the Forty-first Conference on Uncertainty in Artificial Intelligence, 2025.](https://mlanthology.org/uai/2025/karagulyan2025uai-elf/)

BibTeX

@inproceedings{karagulyan2025uai-elf,
  title     = {{ELF: Federated Langevin Algorithms with Primal, Dual and Bidirectional Compression}},
  author    = {Karagulyan, Avetik and Richtárik, Peter},
  booktitle = {Proceedings of the Forty-first Conference on Uncertainty in Artificial Intelligence},
  year      = {2025},
  pages     = {1965-1989},
  volume    = {286},
  url       = {https://mlanthology.org/uai/2025/karagulyan2025uai-elf/}
}