Private Model

Private model research focuses on developing machine learning models that protect sensitive training data while maintaining high accuracy. Current efforts concentrate on techniques like differential privacy (DP), applied through algorithms such as DP-SGD and model-specific adaptations, and on leveraging public data or model ensembles to improve the privacy-utility trade-off. This field is crucial for responsible AI development, enabling the use of sensitive data in various applications while mitigating privacy risks and addressing fairness concerns arising from privacy-preserving methods.

Papers

June 7, 2024

Concept Drift Detection using Ensemble of Integrally Private Models
Ayush K. Varshney, Vicenc Torra
Neural Network Diverse Ensemble Concept Drift Drift Detection Private Model Concept Drift Detection

May 25, 2024

ModelLock: Locking Your Model With a Spell
Yifeng Gao, Yuhua Sun, Xingjun Ma, Zuxuan Wu, Yu-Gang Jiang
Full Model Private Model Diffusion Based Framework Model Protection Spell Variation

April 14, 2024

MAP: Model Aggregation and Personalization in Federated Learning with Incomplete Classes
Xin-Chun Li, Shaoming Song, Yinchuan Li, Bingshuai Li, Yunfeng Shao, Yang Yang, De-Chuan Zhan
Data Heterogeneity Model Aggregation Private Model Model Personalization mAP Decentralized Framework

February 28, 2024

Pre-training Differentially Private Models with Limited Public Data
Zhiqi Bu, Xinwei Zhang, Mingyi Hong, Sheng Zha, George Karypis
Differential Privacy Loss Function Limited Data Private Stochastic Private Model

February 8, 2024

Differentially Private Deep Model-Based Reinforcement Learning
Alexandre Rio, Merwan Barlier, Igor Colin, Albert Thomas
Offline Reinforcement Learning Differential Privacy Guarantee Private Model RL Agent Model Based Policy Private Reinforcement Learning

January 26, 2024

Training Differentially Private Ad Prediction Models with Semi-Sensitive Features
Lynn Chua, Qiliang Cui, Badih Ghazi, Charlie Harrison, Pritish Kamath, Walid Krichene, Ravi Kumar, Pasin Manurangsi, Krishna Giri Narra, Amer Sinha, Avinash Varadarajan, Chiyuan Zhang
Training Data Sensitive Attribute Differential Privacy SGD Private Model

October 2, 2023

LoFT: Local Proxy Fine-tuning For Improving Transferability Of Adversarial Attacks Against Large Language Model
Muhammad Ahmed Shah, Roshan Sharma, Hira Dhamyal, Raphael Olivier, Ankit Shah, Joseph Konan, Dareen Alharthi, Hazim T Bukhari, Massa Baali, Soham Deshmukh, Michael Kuhlmann, Bhiksha Raj, Rita Singh
Large Language Model Adversarial Attack Fine Tuning Task Transferability Transferable Attack Private Model Adversarial Suffix Proxy Model Proxy Tuning

September 28, 2023

Recent Advances of Differential Privacy in Centralized Deep Learning: A Systematic Survey
Lea Demelius, Roman Kern, Andreas Trügler
Differential Privacy Comprehensive Survey Recent Advance Privacy Guarantee Privacy Utility Private Model Private Generative Model

August 8, 2023

Accurate, Explainable, and Private Models: Providing Recourse While Minimizing Training Data Leakage
Catherine Huang, Chelse Swoopes, Christina Xiao, Jiaqi Ma, Himabindu Lakkaraju
PRIvacy Leakage Algorithmic Recourse Robust Recourse Private Model Training Data Leakage

July 19, 2023

The importance of feature preprocessing for differentially private linear optimization
Ziteng Sun, Ananda Theertha Suresh, Aditya Krishna Menon
Differential Privacy Importance Aware Content Based Feature Private Stochastic Private Model Private Optimization

May 20, 2023

Can Public Large Language Models Help Private Cross-device Federated Learning?
Boxin Wang, Yibo Jacky Zhang, Yuan Cao, Bo Li, H. Brendan McMahan, Sewoong Oh, Zheng Xu, Manzil Zaheer
Large Scale Language Model Data Distribution Privacy Utility Private Model Cross Device Federated Learning User Level Differential Privacy

May 19, 2023

On the Fairness Impacts of Private Ensembles Models
Cuong Tran, Ferdinando Fioretto
Differential Privacy Semi Supervised Private Model Private Aggregation Fairness Implication Ensemble Teacher Private Ensemble

April 22, 2023

Differentially Private Synthetic Data Generation via Lipschitz-Regularised Variational Autoencoders
Benedikt Groß, Gerhard Wunder
Generative Model Synthetic Data Differential Privacy Private Model Private Synthetic Data Generation

March 7, 2023

Bootstrap The Original Latent: Learning a Private Model from a Black-box Model
Shuai Wang, Daoan Zhang, Jianguo Zhang, Weiwei Zhang, Rui Li
Black Box Model Intermediate Latent Private Model Model Privacy

March 3, 2023

Exploring Machine Learning Privacy/Utility trade-off from a hyperparameters Lens
Ayoub Arous, Amira Guesmi, Muhammad Abdullah Hanif, Ihsen Alouani, Muhammad Shafique
Privacy Preserving Privacy Guarantee Hyper Parameter Private Stochastic Gradient Descent Private Model Learning Privacy Utility Tradeoff

March 2, 2023

Choosing Public Datasets for Private Machine Learning via Gradient Subspace Distance
Xin Gu, Gautam Kamath, Zhiwei Steven Wu
Gradient Descent Available Datasets Private Stochastic Private Machine Learning Private Model Gradient Subspace

February 28, 2023

Arbitrary Decisions are a Hidden Cost of Differentially Private Training
Bogdan Kulynych, Hsiang Hsu, Carmela Troncoso, Flavio P. Calmon
Differential Privacy Hidden CoST Privacy Preserving Machine Learning Private Model Private Model Training Predictive Multiplicity

February 14, 2023

Bounding Training Data Reconstruction in DP-SGD
Jamie Hayes, Saeed Mahloujifar, Borja Balle
Training Data Membership Inference Attack Private Model Dp SGD

January 28, 2023

Context-Aware Differential Privacy for Language Modeling
My H. Dinh, Ferdinando Fioretto
Language Model Differential Privacy Privacy Preserving Machine Learning Private Model

November 21, 2022

Private Ad Modeling with DP-SGD
Carson Denison, Badih Ghazi, Pritish Kamath, Ravi Kumar, Pasin Manurangsi, Krishna Giri Narra, Amer Sinha, Avinash V Varadarajan, Chiyuan Zhang
Private Stochastic Private Model Privacy Utility Trade