← Back ◬ AI & Machine Learning —

Learnability and Privacy Vulnerability are Entangled in a Few Critical Weights

arXiv Security Archived Mar 16, 2026 ✓ Full text saved

arXiv:2603.13186v1 Announce Type: cross Abstract: Prior approaches for membership privacy preservation usually update or retrain all weights in neural networks, which is costly and can lead to unnecessary utility loss or even more serious misalignment in predictions between training data and non-training data. In this work, we observed three insights: i) privacy vulnerability exists in a very small fraction of weights; ii) however, most of those weights also critically impact utility performance

Full text archived locally

✦ AI Summary · Claude Sonnet

Computer Science > Machine Learning [Submitted on 13 Mar 2026] Learnability and Privacy Vulnerability are Entangled in a Few Critical Weights Xingli Fang, Jung-Eun Kim Prior approaches for membership privacy preservation usually update or retrain all weights in neural networks, which is costly and can lead to unnecessary utility loss or even more serious misalignment in predictions between training data and non-training data. In this work, we observed three insights: i) privacy vulnerability exists in a very small fraction of weights; ii) however, most of those weights also critically impact utility performance; iii) the importance of weights stems from their locations rather than their values. According to these insights, to preserve privacy, we score critical weights, and instead of discarding those neurons, we rewind only the weights for fine-tuning. We show that, through extensive experiments, this mechanism exhibits outperforming resilience in most cases against Membership Inference Attacks while maintaining utility. Comments: ICLR 2026 Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR) Cite as: arXiv:2603.13186 [cs.LG] (or arXiv:2603.13186v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2603.13186 Focus to learn more Submission history From: Jung-Eun Kim [view email] [v1] Fri, 13 Mar 2026 17:20:12 UTC (3,534 KB) Access Paper: HTML (experimental) view license Current browse context: cs.LG < prev | next > new | recent | 2026-03 Change to browse by: cs cs.AI cs.CR References & Citations NASA ADS Google Scholar Semantic Scholar Export BibTeX Citation Bookmark Bibliographic Tools Bibliographic and Citation Tools Bibliographic Explorer Toggle Bibliographic Explorer (What is the Explorer?) Connected Papers Toggle Connected Papers (What is Connected Papers?) Litmaps Toggle Litmaps (What is Litmaps?) scite.ai Toggle scite Smart Citations (What are Smart Citations?) Code, Data, Media Demos Related Papers About arXivLabs Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)

💬 Team Notes