JAX-Privacy: A library for differentially private machine learning
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866910027853332480 |
|---|---|
| author | McKenna, Ryan Andrew, Galen Balle, Borja Doroshenko, Vadym Ganesh, Arun Kong, Weiwei Kurakin, Alex McMahan, Brendan Pravilov, Mikhail |
| author_facet | McKenna, Ryan Andrew, Galen Balle, Borja Doroshenko, Vadym Ganesh, Arun Kong, Weiwei Kurakin, Alex McMahan, Brendan Pravilov, Mikhail |
| contents | JAX-Privacy is a library designed to simplify the deployment of robust and performant mechanisms for differentially private machine learning. Guided by design principles of usability, flexibility, and efficiency, JAX-Privacy serves both researchers requiring deep customization and practitioners who want a more out-of-the-box experience. The library provides verified, modular primitives for critical components for all aspects of the mechanism design including batch selection, gradient clipping, noise addition, accounting, and auditing, and brings together a large body of recent research on differentially private ML. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_17861 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | JAX-Privacy: A library for differentially private machine learning McKenna, Ryan Andrew, Galen Balle, Borja Doroshenko, Vadym Ganesh, Arun Kong, Weiwei Kurakin, Alex McMahan, Brendan Pravilov, Mikhail Machine Learning JAX-Privacy is a library designed to simplify the deployment of robust and performant mechanisms for differentially private machine learning. Guided by design principles of usability, flexibility, and efficiency, JAX-Privacy serves both researchers requiring deep customization and practitioners who want a more out-of-the-box experience. The library provides verified, modular primitives for critical components for all aspects of the mechanism design including batch selection, gradient clipping, noise addition, accounting, and auditing, and brings together a large body of recent research on differentially private ML. |
| title | JAX-Privacy: A library for differentially private machine learning |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2602.17861 |