Metraipolis › archive › dc_c7734aeb412d

SemTrace: Source-Grounded Semantic Signatures for Tracing LLM Exposure to Protected Documents — arXiv:2608.29575

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.29575
License (as archived)
CC0-1.0
Archived
2026-10-17T08:07:01.892Z
SHA-256 of the archived text
84a4b436ec7dc1fce4049356a312651c7a902a9a3e08d54d8b32f4ec5a269d9b
Stable URL of this page
https://metraipolis.com/documento/dc_c7734aeb412d

How to cite

SemTrace: Source-Grounded Semantic Signatures for Tracing LLM Exposure to Protected Documents — arXiv:2608.29575 — arXiv (en), https://arxiv.org/abs/2608.29575 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_c7734aeb412d (Metraipolis archive dc_c7734aeb412d, sha256 84a4b436ec7dc1fce4049356a312651c7a902a9a3e08d54d8b32f4ec5a269d9b)
BibTeX
@misc{dcc7734aeb412d,
  title = {SemTrace: Source-Grounded Semantic Signatures for Tracing LLM Exposure to Protected Documents — arXiv:2608.29575},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.29575},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_c7734aeb412d; sha256 84a4b436ec7dc1fce4049356a312651c7a902a9a3e08d54d8b32f4ec5a269d9b},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.29575. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Large language models are increasingly used to read documents and produce downstream text, creating a provenance problem when the document owner cannot control or inspect the model that performs the generation. We introduce SemTrace, a source-grounded semantic watermark for detecting whether a generated review was influenced by a known protected manuscript copy. Rather than biasing token probabilities or imposing surface-form patterns, SemTrace constructs a document-specific binary signature from factual propositions that are directly supported by the manuscript itself. A protected PDF invisibly carries a content contract that selects one fact from each binary pair and asks an instruction-following reviewer to express those facts in fixed review slots without changing its independent evaluation. A frozen natural language inference model then decodes the resulting semantic evidence with explicit erasures and scores the recovered bits against the codeword assigned to that copy. This design targets model-agnostic, assigned-copy exposure detection while keeping the watermark semantically tied to the source document.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-01 in cs.CL).
Categorie arXiv: Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2608.29575 — 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 2026-10-17T07:47:52.712Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-09-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).