Metraipolis › archive › dc_6504f61cd3bc

Towards Computational Provenance: Carrying Causal-State Evidence in Generated Text — arXiv:2608.16868

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.16868
License (as archived)
CC0-1.0
Archived
2026-10-18T18:07:41.606Z
SHA-256 of the archived text
e9ac2db0d86b7ec74ed62ba23f36d7497863d10637e7e92dd7df484cfba0c746
Stable URL of this page
https://metraipolis.com/documento/dc_6504f61cd3bc

How to cite

Towards Computational Provenance: Carrying Causal-State Evidence in Generated Text — arXiv:2608.16868 — arXiv (en), https://arxiv.org/abs/2608.16868 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_6504f61cd3bc (Metraipolis archive dc_6504f61cd3bc, sha256 e9ac2db0d86b7ec74ed62ba23f36d7497863d10637e7e92dd7df484cfba0c746)
BibTeX
@misc{dc6504f61cd3bc,
  title = {Towards Computational Provenance: Carrying Causal-State Evidence in Generated Text — arXiv:2608.16868},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.16868},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_6504f61cd3bc; sha256 e9ac2db0d86b7ec74ed62ba23f36d7497863d10637e7e92dd7df484cfba0c746},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.16868. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

A language model's output does not by itself provide verifiable evidence about the internal computation that produced it. We study computational provenance: whether generated text can carry detectable evidence of which causally relevant internal state occurred. We test a bounded form of this idea in two controlled architectures: a modular feed-forward neural network and a transformer-based model. Both architectures are trained on the same arithmetic task with a mandatory pathway through two discrete intermediate states, allowing different internal paths to produce the same answer. We deliberately switch between these paths, authenticate the state actually used, and let that verified state determine a subtle statistical pattern in the generated text that can later be detected. The feed-forward and transformer systems each passed all 128 matched pairs in both their public and separately sealed protected end-to-end evaluations, with the detector recovering the signal associated with the authenticated internal state. The required causal computation also reproduced across five independently trained feed-forward models and three independently trained transformers. In a separate answer-only transformer experiment, our linear probes did not recover a naturally learned intermediate state. These results provide a controlled proof of concept that information about a verified, causally relevant internal state can be preserved in generated text even when the answer is unchanged.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-18 in cs.AI).
Categorie arXiv: Computation and Language (cs.CL) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2608.16868 — 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 2026-10-18T16:29:14.165Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-08-18?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).