Picture for Michael Kearns

Michael Kearns

Reconstruction Attacks on Machine Unlearning: Simple Models are Vulnerable

Add code
May 30, 2024
Viaarxiv icon

Model Ensembling for Constrained Optimization

Add code
May 27, 2024
Viaarxiv icon

Oracle-Efficient Reinforcement Learning for Max Value Ensembles

Add code
May 27, 2024
Viaarxiv icon

Diversified Ensembling: An Experiment in Crowdsourced Machine Learning

Add code
Feb 16, 2024
Viaarxiv icon

Membership Inference Attacks on Diffusion Models via Quantile Regression

Add code
Dec 08, 2023
Figure 1 for Membership Inference Attacks on Diffusion Models via Quantile Regression
Figure 2 for Membership Inference Attacks on Diffusion Models via Quantile Regression
Figure 3 for Membership Inference Attacks on Diffusion Models via Quantile Regression
Figure 4 for Membership Inference Attacks on Diffusion Models via Quantile Regression
Viaarxiv icon

Scalable Membership Inference Attacks via Quantile Regression

Add code
Jul 07, 2023
Figure 1 for Scalable Membership Inference Attacks via Quantile Regression
Figure 2 for Scalable Membership Inference Attacks via Quantile Regression
Figure 3 for Scalable Membership Inference Attacks via Quantile Regression
Figure 4 for Scalable Membership Inference Attacks via Quantile Regression
Viaarxiv icon

Balanced Filtering via Non-Disclosive Proxies

Add code
Jul 05, 2023
Figure 1 for Balanced Filtering via Non-Disclosive Proxies
Figure 2 for Balanced Filtering via Non-Disclosive Proxies
Figure 3 for Balanced Filtering via Non-Disclosive Proxies
Figure 4 for Balanced Filtering via Non-Disclosive Proxies
Viaarxiv icon

Replicable Reinforcement Learning

Add code
May 24, 2023
Figure 1 for Replicable Reinforcement Learning
Figure 2 for Replicable Reinforcement Learning
Viaarxiv icon

AI Model Disgorgement: Methods and Choices

Add code
Apr 07, 2023
Figure 1 for AI Model Disgorgement: Methods and Choices
Viaarxiv icon

Improved Differentially Private Regression via Gradient Boosting

Add code
Mar 06, 2023
Figure 1 for Improved Differentially Private Regression via Gradient Boosting
Figure 2 for Improved Differentially Private Regression via Gradient Boosting
Figure 3 for Improved Differentially Private Regression via Gradient Boosting
Figure 4 for Improved Differentially Private Regression via Gradient Boosting
Viaarxiv icon