Metraipolis › archive › dc_7cea18b7233f

FissionGuard: Testing Local Watermark Proportions with FDR Control — arXiv:2610.11625

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.11625
License (as archived)
CC0-1.0
Archived
2026-10-09T06:27:44.158Z
SHA-256 of the archived text
7b8fbb4ca5fec6c636ea222596a44c41f41859261fc80061554ed7ce11153477
Stable URL of this page
https://metraipolis.com/documento/dc_7cea18b7233f

How to cite

FissionGuard: Testing Local Watermark Proportions with FDR Control — arXiv:2610.11625 — arXiv (en), https://arxiv.org/abs/2610.11625 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_7cea18b7233f (Metraipolis archive dc_7cea18b7233f, sha256 7b8fbb4ca5fec6c636ea222596a44c41f41859261fc80061554ed7ce11153477)
BibTeX
@misc{dc7cea18b7233f,
  title = {FissionGuard: Testing Local Watermark Proportions with FDR Control — arXiv:2610.11625},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.11625},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_7cea18b7233f; sha256 7b8fbb4ca5fec6c636ea222596a44c41f41859261fc80061554ed7ce11153477},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.11625. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

A watermark detector can flag a text even when only a small fraction of its tokens are watermarked. We study sequential auditing to identify local passages whose watermark proportions exceed a prespecified threshold. Each alarm fixes a candidate passage's endpoint, while its start is chosen from earlier tokens. Reusing the same evidence to locate and test a passage can invalidate inference. We propose FissionGuard, which uses data fission to construct an alarm value and a validation value from each token-level $p$-value. The alarm values guide alarms and localization. The validation values yield passage-level $p$-values through partial conjunction, excluding as many of the strongest signals as the proportion null allows. These $p$-values are used to test the corresponding hypotheses at levels assigned by an online multiple-testing procedure. Under the stated watermark generation model, allowing dependence among token-level $p$-values at watermarked positions, we prove finite-sample false discovery rate (FDR) control among reported passages at every fixed finite horizon. We also derive a finite-sample lower bound on power for testing nonnull candidates. Experiments on synthetic streams and Mistral and Qwen generations across four marking layouts show gains in watermarked-token coverage over sample splitting, with empirical FDR below the target.

Tema degli esploratori: statistica e scienza dei dati (elenco arXiv stat.ME, nuovi lavori del giorno).
Categorie arXiv: Methodology (stat.ME).
Pagina del lavoro: https://arxiv.org/abs/2610.11625 — 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 2026-10-09T05:51:46.708Z dalla pagina https://arxiv.org/list/stat.ME/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).