Metraipolis › archive › dc_aaff89ffc496
TANGO: Watermarking Masked Diffusion Language Models in Token Pairs — arXiv:2609.35224
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.35224
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-07T21:54:19.028Z
- SHA-256 of the archived text
15a5c180931eca1e661597776683285a02b2ece4cf9f78abfc19f31e91ef873c- Stable URL of this page
https://metraipolis.com/documento/dc_aaff89ffc496
How to cite
TANGO: Watermarking Masked Diffusion Language Models in Token Pairs — arXiv:2609.35224 — arXiv (en), https://arxiv.org/abs/2609.35224 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_aaff89ffc496 (Metraipolis archive dc_aaff89ffc496, sha256 15a5c180931eca1e661597776683285a02b2ece4cf9f78abfc19f31e91ef873c)
BibTeX
@misc{dcaaff89ffc496,
title = {TANGO: Watermarking Masked Diffusion Language Models in Token Pairs — arXiv:2609.35224},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.35224},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_aaff89ffc496; sha256 15a5c180931eca1e661597776683285a02b2ece4cf9f78abfc19f31e91ef873c},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.35224. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Masked-diffusion language models fill in masked positions in parallel and in no fixed order. Most practical text watermarks assume left-to-right generation. They key each token to the tokens before it, and in a diffusion model those tokens may still be masked. A fixed green list needs no such context, but it favors the same tokens at every position, so these tokens appear more often in watermarked text. An attacker who compares token frequencies in watermarked and unwatermarked text can recover the list and forge text that the provider's own detector accepts. We present TANGO, a watermark for masked-diffusion language models that keys each new token to a nearby token that is already unmasked. A secret key splits the vocabulary into color classes, and TANGO biases the new token toward a color determined by the key and the nearby token's color. The watermark is therefore embedded in pairs of tokens. Because the favored color changes from position to position, token frequencies stay much closer to those of unwatermarked text than under a fixed green list. Detection needs only the text and the key, and it does not assume any unmasking order. On two masked-diffusion models, TANGO detects nearly all unedited watermarked texts and most edited ones, and frequency attacks that forge the fixed green list fail against it. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-29 in cs.CL). Categorie arXiv: Machine Learning (cs.LG) ; Computation and Language (cs.CL); Cryptography and Security (cs.CR). Pagina del lavoro: https://arxiv.org/abs/2609.35224 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-07T21:27:08.841Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-09-29?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).