Metraipolis › archive › dc_871ebdc99825

Constructing Budgeted KGW Evasion through Token-Event Evidence Replacement — arXiv:2611.00726

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.00726
License (as archived)
CC0-1.0
Archived
2026-11-03T18:22:39.442Z
SHA-256 of the archived text
47f3c9d2a80befa318064f22e26c0ee7f80c1c79bb6c7938c44f117eddbdbd9c
Stable URL of this page
https://metraipolis.com/documento/dc_871ebdc99825

How to cite

Constructing Budgeted KGW Evasion through Token-Event Evidence Replacement — arXiv:2611.00726 — arXiv (en), https://arxiv.org/abs/2611.00726 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_871ebdc99825 (Metraipolis archive dc_871ebdc99825, sha256 47f3c9d2a80befa318064f22e26c0ee7f80c1c79bb6c7938c44f117eddbdbd9c)
BibTeX
@misc{dc871ebdc99825,
  title = {Constructing Budgeted KGW Evasion through Token-Event Evidence Replacement — arXiv:2611.00726},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.00726},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_871ebdc99825; sha256 47f3c9d2a80befa318064f22e26c0ee7f80c1c79bb6c7938c44f117eddbdbd9c},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.00726. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

KGW watermarks bias sampling to accumulate keyed detection evidence. We construct budgeted evasion through token-event evidence replacement using ordinary phrase rewrites. Our construction jointly selects compositions of LLM-proposed alternatives from an independent model and measures global token-edit distance on complete outputs. At the largest attainable cost within the budget, exact coverage search maximizes removal of original scored token pairs. It uses a known tokenizer without the key, detector feedback or queries to the watermarked model. Exact accounting separates source evidence removed, evidence created and length effects, with initial strength setting the required reduction. Separate signed-event analysis found that mean suppression arose mainly from removing watermark-enriched source events. In controlled, occurrence-counted, one-token-context KGW tests on short passages, 39 of 82 qualified, initially detected sources yielded rewrites with terminal detector score Z < 4 that passed blinded source-relative model review, at a mean actual token-edit cost of 16.0% under a 20% allowance. To examine whether this initial requirement can be inferred, a separate historical 95-passage study estimated initial detector score Z0 from generation traces with correlation 0.793 and an in-sample association beyond length. This aggregate information did not establish improved allocation, and the separate no-trace stopping test did not reliably exploit cheaper reviewed crossings in the candidate bank. Comparisons did not establish an advantage over simpler selection and ordinary constrained rewriting. The construction makes budgeted evidence replacement executable while exposing limits in signed reduction and recognition of sufficient editing.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.CL, nuovi lavori del giorno).
Categorie arXiv: Computation and Language (cs.CL) ; Cryptography and Security (cs.CR).
Pagina del lavoro: https://arxiv.org/abs/2611.00726 — 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 2026-11-03T18:10:07.898Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).