Do you have an open-source repository where one can test/experiment with watermarked/fingerprinted models?
I have a general understanding of what was described but can't distinguish if this is a "perfect solution", such as asymmetric cryptography, or an additional layer such as those added to circumvent prompt injection.
Do you have an open-source repository where one can test/experiment with watermarked/fingerprinted models?
I have a general understanding of what was described but can't distinguish if this is a "perfect solution", such as asymmetric cryptography, or an additional layer such as those added to circumvent prompt injection.
Perfect solutions are always theoretical, real world solutions have trade offs, so does the solution sketched above :)
We will be open sourcing our solutions very soon!