What is hard red-list language model watermarking?
Definition
Hard red-list language model watermarking deterministically partitions the vocabulary for each generation step using a secret key and contextual seed. The decoder excludes red-listed tokens and samples from the green list, creating a strong pattern that a detector with the key can reproduce.
The restriction improves detectability but can reduce quality. When the best continuation is red-listed, the model must choose a less suitable token, and the effect is especially noticeable when only a few continuations are semantically or factually correct.
Acronyms and aliases
hard red-list watermarking varianthard token-list watermark variant
General terms
Frequently asked questions
Why is hard red-list watermarking easy to detect?
The generator strongly favors the keyed green list, so a sufficiently long passage contains far more allowed tokens than chance predicts.
Why can hard red-list language model watermarking reduce quality?
It can ban the most appropriate next token, forcing the model to choose a weaker continuation solely to preserve the watermark.
Videos explaining hard red-list language model watermarking
