AI RESEARCH

Geometry-Aware Localized Watermarking for Copyright Protection in Embedding-as-a-Service

arXiv CS.CL

ArXi:2604.11344v1 Announce Type: cross Embedding-as-a-Service (EaaS) has become an important semantic infrastructure for natural language and multimedia applications, but it is highly vulnerable to model stealing and To address this problem, we propose GeoMark, a geometry-aware localized watermarking framework for EaaS Experiments on four benchmark datasets show that GeoMark preserves downstream utility and geometric fidelity while maintaining robust