News & Announcements
- (2023-11) — We have a new paper on training ML models with differential privacy using data augmentations such as mixup. This work will appear at NeurIPS 2023!
- (2023-10) — We wrote an SoK on Memorization in Large Language Models (LLMs)!
Selected (Recent) Publications
- DP-Mix: Mixup-based Data Augmentation for Differentially Private Learning [PDF]In Neural Information Processing Systems (2023)
- SoK: Memorization in General-Purpose Large Language Models [PDF]In arXiv preprint (2023)
- EMI-LiDAR: Uncovering Vulnerabilities of LiDAR Sensors in Autonomous Driving Setting using Electromagnetic Interference [PDF (External Link)]In ACM Conference on Security and Privacy in Wireless and Mobile Networks (2023)
- Enhanced Membership Inference Attacks against Machine Learning Models [PDF]In ACM SIGSAC Conference on Computer and Communications Security (2022)
- Analyzing the Monetization Ecosystem of Stalkerware [PDF]In Privacy Enhancing Technologies Symposium (2022)
- Privacy accounting εconomics: Improving differential privacy composition via a posteriori bounds [PDF]In Privacy Enhancing Technologies Symposium (2022)
- Demystifying Limited Adversarial Transferability in Automatic Speech Recognition Systems [PDF]In International Conference on Learning Representations (2022)