Differentially Private SGD Without Clipping Bias: An Error-Feedback
Approach

v1v2 (latest)

Differentially Private SGD Without Clipping Bias: An Error-Feedback Approach

24 November 2023

Zhiwei Steven Wu

ArXiv (abs)PDF HTML

Papers citing "Differentially Private SGD Without Clipping Bias: An Error-Feedback Approach"

4 / 4 papers shown

Title
DiSK: Differentially Private Optimizer with Simplified Kalman Filter for Noise Reduction Xinwei Zhang Zhiqi Bu Borja Balle Mingyi Hong Meisam Razaviyayn Vahab Mirrokni 141 2 0 04 Oct 2024
Clip Body and Tail Separately: High Probability Guarantees for DPSGD with Heavy Tails Haichao Sha Yang Cao Yong Liu Yuncheng Wu Ruixuan Liu Hong Chen 83 2 0 27 May 2024
Privacy-Preserving Instructions for Aligning Large Language Models Da Yu Peter Kairouz Sewoong Oh Zheng Xu 115 25 0 21 Feb 2024
DPZero: Private Fine-Tuning of Language Models without Backpropagation Liang Zhang Bingcong Li K. K. Thekumparampil Sewoong Oh Niao He 92 15 0 14 Oct 2023