Data as of Jul 25, 2026 · Based on 2,718,867 AI responses across 9,511 prompts · See how Parse measures this
XAttnMarkliuyixin-louis.github.io · Low evidenceXAttnMark is a robust audio watermarking system that uses cross attention and temporal conditioning to embed imperceptible, traceable marks into audio. It achieves state-of-the-art detection and attribution accuracy, even against generative editing and standard audio transformations.
Parse Score