“Zero-Knowledge Watermarking

2026

Open publication workspace · Sign in to read the full PDF.

AI-generated summary

This paper introduces zkWM, a novel framework for watermarking Large Language Models that ensures both public verifiability and privacy.

* Presents a zero-knowledge watermarking framework (zkWM) for LLMs.
* Enables privacy-preserving and publicly verifiable watermark detection using SNARKs.
* Details the prover-verifier setup and the cryptographic proofs involved.

Tags: Zero-Knowledge, Watermarking, LLMs, Cryptography, Privacy

Check the original publication for accuracy and context.