Jing Shang (Beijing Jiaotong University), Jian Wang (Beijing Jiaotong University), Kailun Wang (Beijing Jiaotong University), Jiqiang Liu (Beijing Jiaotong University), Nan Jiang (Beijing University of Technology), Md Armanuzzaman (Northeastern University), Ziming Zhao (Northeastern University)

Model pruning is a technique for compressing deep learning models, and using an iterative way to prune the model can achieve better compression effects with lower utility loss. However, our analysis reveals that iterative pruning significantly increases model memorization, making the pruned models more vulnerable to membership inference attacks (MIAs). Unfortunately, the vast majority of existing defenses against MIAs are designed for original and unpruned models. In this paper, we propose a new framework WeMem to weaken memorization in the iterative pruning process. Specifically, our analysis identifies two important factors that increase memorization in iterative pruning, namely data reuse and inherent memorability. We consider the individual and combined impacts of both factors, forming three scenarios that lead to increased memorization in iteratively pruned models. We design three defense primitives based on these factors' characteristics. By combining these primitives, we propose methods tailored to each scenario to weaken memorization effectively. Comprehensive experiments under ten adaptive MIAs demonstrate the effectiveness of the proposed defenses. Moreover, our defenses outperform five existing defenses in terms of privacy-utility tradeoff and efficiency. Additionally, we enhance the proposed defenses to automatically adjust settings for optimal defense, improving their practicability.

View More Papers

Retrofitting XoM for Stripped Binaries without Embedded Data Relocation

Chenke Luo (Wuhan University), Jiang Ming (Tulane University), Mengfei Xie (Wuhan University), Guojun Peng (Wuhan University), Jianming Fu (Wuhan University)

Read More

The Kids Are All Right: Investigating the Susceptibility of...

Elijah Bouma-Sims (Carnegie Mellon University), Lily Klucinec (Carnegie Mellon University), Mandy Lanyon (Carnegie Mellon University), Julie Downs (Carnegie Mellon University), Lorrie Faith Cranor (Carnegie Mellon University)

Read More

Throwaway Accounts and Moderation on Reddit

Cheng Guo (Clemson University), Kelly Caine (Clemson University)

Read More

Revealing the Black Box of Device Search Engine: Scanning...

Mengying Wu (Fudan University), Geng Hong (Fudan University), Jinsong Chen (Fudan University), Qi Liu (Fudan University), Shujun Tang (QI-ANXIN Technology Research Institute; Tsinghua University), Youhao Li (QI-ANXIN Technology Research Institute), Baojun Liu (Tsinghua University), Haixin Duan (Tsinghua University; Quancheng Laboratory), Min Yang (Fudan University)

Read More