Information Leakage of Sentence Embeddings via Generative Embedding Inversion Attacks
Published in Proceedings of the 48th International ACM SIGIR Conference on Research and Development in Information Retrieval, 2025
This paper explores security vulnerabilities in sentence embeddings through generative embedding inversion attacks, demonstrating how sensitive information can be leaked from embeddings.
Recommended citation: Tragoudaras, Antonios and Aslanidis, Theofanis and Lionis, Emmanouil Georgios and Orozco Gonzalez, Marina and Eustratiadis, Panagiotis (2025). "Information Leakage of Sentence Embeddings via Generative Embedding Inversion Attacks" Proceedings of the 48th International ACM SIGIR Conference on Research and Development in Information Retrieval. 3234-3243.
Download Paper