Nils Lukas
33 Arbeiten318 Zitationen
Relevante Arbeiten
Meistzitierte Publikationen im Bereich Gesundheit & MedTech
SoK: How Robust is Image Classification Deep Neural Network Watermarking?
2022 · 63 Zit. · 2022 IEEE Symposium on Security and Privacy (SP)
Analyzing Leakage of Personally Identifiable Information in Language Models
2023 · 13 Zit. · ArXiv.org
Pick your Poison: Undetectability versus Robustness in Data Poisoning Attacks
2023 · 0 Zit. · arXiv (Cornell University)
DP-Fusion: Token-Level Differentially Private Inference for Large Language Models
2025 · 0 Zit. · ArXiv.org