English
Related papers

Related papers: Private Fine-tuning of Large Language Models with …

200 papers

Recent works have shown that selecting an optimal model architecture suited to the differential privacy setting is necessary to achieve the best possible utility for a given privacy budget using differentially private stochastic gradient…

Machine Learning · Computer Science 2024-11-01 Anderson Santana de Oliveira , Caelin Kaplan , Khawla Mallat , Tanmay Chakraborty

Differentially private (stochastic) gradient descent is the workhorse of DP private machine learning in both the convex and non-convex settings. Without privacy constraints, second-order methods, like Newton's method, converge faster than…

Machine Learning · Computer Science 2023-05-23 Arun Ganesh , Mahdi Haghifam , Thomas Steinke , Abhradeep Thakurta

Traditional Differentially Private Stochastic Gradient Descent (DP-SGD) introduces statistical noise on top of gradients drawn from a Gaussian distribution to ensure privacy. This paper introduces the novel Differentially Private Block-wise…

Machine Learning · Computer Science 2025-01-22 David Zagardo

Per-example gradient clipping is a key algorithmic step that enables practical differential private (DP) training for deep learning models. The choice of clipping threshold R, however, is vital for achieving high accuracy under DP. We…

Machine Learning · Computer Science 2023-10-05 Zhiqi Bu , Yu-Xiang Wang , Sheng Zha , George Karypis

Ensuring the privacy of Large Language Models (LLMs) is becoming increasingly important. The most widely adopted technique to accomplish this is DP-SGD, which trains a model to guarantee Differential Privacy (DP). However, DP-SGD…

Cryptography and Security · Computer Science 2024-04-30 James Flemings , Meisam Razaviyayn , Murali Annavaram

Differential privacy (DP) is an essential technique for privacy-preserving. It was found that a large model trained for privacy preserving performs worse than a smaller model (e.g. ResNet50 performs worse than ResNet18). To better…

Machine Learning · Computer Science 2021-11-30 Yinchen Shen , Zhiguo Wang , Ruoyu Sun , Xiaojing Shen

Large language models (LLMs) frequently memorize sensitive or personal information, raising significant privacy concerns. Existing variants of differential privacy stochastic gradient descent (DPSGD) inject uniform noise into every gradient…

Machine Learning · Computer Science 2025-09-30 Manjiang Yu , Priyanka Singh , Xue Li , Yang Cao

Differentially Private Stochastic Gradient Descent (DP-SGD) is widely used to protect training data in machine learning. Its privacy guarantee is commonly analyzed through a security game in which an adversary infers whether a target record…

Cryptography and Security · Computer Science 2026-05-18 Wenhao Wang , Shujie Cui , Hui Cui , Xingliang Yuan

Differentially Private Stochastic Gradient Descent (DPSGD) is widely used to protect sensitive data during the training of machine learning models, but its privacy guarantee often comes at a large cost of model performance due to the lack…

Machine Learning · Computer Science 2026-01-16 Hao Liang , Wanrong Zhang , Xinlei He , Kaishun Wu , Hong Xing

Adaptive optimizers are the de facto standard in non-private training as they often enable faster convergence and improved performance. In contrast, differentially private (DP) training is still predominantly performed with DP-SGD,…

Machine Learning · Computer Science 2025-12-01 Mihaela Hudişteanu , Nikita P. Kalinin , Edwige Cyffers

The Noisy-SGD algorithm is widely used for privately training machine learning models. Traditional privacy analyses of this algorithm assume that the internal state is publicly revealed, resulting in privacy loss bounds that increase…

Machine Learning · Computer Science 2023-05-18 Shahab Asoodeh , Mario Diaz

Fine-tuning large pretrained language models (LLMs) is a cornerstone of modern NLP, yet its growing memory demands (driven by backpropagation and large optimizer States) limit deployment in resource-constrained settings. Zero-order (ZO)…

Machine Learning · Computer Science 2026-02-17 Valery Parfenov , Grigoriy Evseev , Andrey Veprikov , Nikolay Bushkov , Stanislav Moiseev , Aleksandr Beznosikov

We propose SPOF (Stabilization of Perturbed Loss Function), a differentially private training mechanism intended for multi-user local differential privacy (LDP). SPOF perturbs a stabilized Taylor expanded polynomial approximation of a…

Machine Learning · Computer Science 2025-08-22 Salman Habib , Remi Chou , Taejoon Kim

Protecting large language models from privacy leakage is becoming increasingly crucial with their wide adoption in real-world products. Yet applying differential privacy (DP), a canonical notion with provable privacy guarantees for machine…

Computation and Language · Computer Science 2022-10-28 Weiyan Shi , Ryan Shea , Si Chen , Chiyuan Zhang , Ruoxi Jia , Zhou Yu

Balancing convergence speed, generalization capability, and computational efficiency remains a core challenge in deep learning optimization. First-order gradient descent methods, epitomized by stochastic gradient descent (SGD) and Adam,…

When we enforce differential privacy in machine learning, the utility-privacy trade-off is different w.r.t. each group. Gradient clipping and random noise addition disproportionately affect underrepresented and complex classes and…

Machine Learning · Computer Science 2020-09-29 Depeng Xu , Wei Du , Xintao Wu

Large language models (LLMs) excel across various tasks, but standard first-order (FO) fine-tuning demands considerable memory, significantly limiting real-world deployment. Recently, zeroth-order (ZO) optimization stood out as a promising…

Machine Learning · Computer Science 2025-11-04 Qitao Tan , Jun Liu , Zheng Zhan , Caiwei Ding , Yanzhi Wang , Xiaolong Ma , Jaewoo Lee , Jin Lu , Geng Yuan

Many forms of sensitive data, such as web traffic, mobility data, or hospital occupancy, are inherently sequential. The standard method for training machine learning models while ensuring privacy for units of sensitive information, such as…

We consider stochastic convex optimization for heavy-tailed data with the guarantee of being differentially private (DP). Most prior works on differentially private stochastic convex optimization for heavy-tailed data are either restricted…

Machine Learning · Computer Science 2024-09-11 Chenhan Jin , Kaiwen Zhou , Bo Han , James Cheng , Tieyong Zeng

Fine-tuning large language models (LLMs) using standard first-order (FO) optimization often drives training toward sharp, poorly generalizing minima. Conversely, zeroth-order (ZO) methods offer stronger exploratory behavior without relying…

Machine Learning · Computer Science 2026-01-12 Feihu Jin , Ying Tan