A Unified Framework for LLM Watermarks
Thibaud Gloaguen, Robin Staab, Nikola Jovanović, Martin Vechev · 2026
Read A Unified Framework for LLM Watermarks by Thibaud Gloaguen, Robin Staab, Nikola Jovanović, Martin Vechev (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-06754.txt.
Description
Thibaud Gloaguen, Robin Staab, Nikola Jovanović, and Martin Vechev bring order to the crowded field of language model watermarking. A hidden signal is planted in generated text so its origin can later be traced. Many schemes exist, each designed in its own way, and no shared formulation has tied them together. The paper shows most of them fall out of one constrained optimization problem.
Writing the schemes in that common form exposes the hidden trade-off among text quality, diversity, and detection power that each one strikes. The framework also works in reverse: state a requirement, such as a quality proxy, and derive a scheme that is optimal for it. Experiments confirm that schemes derived from a given constraint do maximize detection power under that constraint.
Catalog record
| CREATOR | Thibaud Gloaguen, Robin Staab, Nikola Jovanović, Martin Vechev |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 84 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-06754