Code of our paper - "WARDEN: Multi-Directional Backdoor Watermarks for Embedding-as-a-Service Copyright Protection" (Accepted to ACL2024 (Main Proceedings)).
Hi, Anudeex. A very nice work. I have a question:
(1) How to obtain the p-value in Table 3 from five results? For example, the p-values from five results are [0.1, 0.15, 0.12, 0.2, 0.08]. So, what is the final p-value? p>0.08(minimus value), p>0.15(maximus value) or p>0.13(mean value)?
In addition, I have two questions about your WET (WET: Overcoming Paraphrasing Vulnerabilities in Embeddings-as-a-Service with Linear Transformation Watermark) work.
(1) Is \delta_cos in Table 3 correct? They seem a bit large.
(2) Typo in Page 6: presentsthe
Hi, Anudeex. A very nice work. I have a question: (1) How to obtain the p-value in Table 3 from five results? For example, the p-values from five results are [0.1, 0.15, 0.12, 0.2, 0.08]. So, what is the final p-value? p>0.08(minimus value), p>0.15(maximus value) or p>0.13(mean value)? In addition, I have two questions about your WET (WET: Overcoming Paraphrasing Vulnerabilities in Embeddings-as-a-Service with Linear Transformation Watermark) work. (1) Is \delta_cos in Table 3 correct? They seem a bit large. (2) Typo in Page 6: presentsthe