Metraipolis › archive › dc_632406c97537
Origin Is All You Need: Provenance-Aware Transformers for Structural Trust-Boundary Separation — arXiv:2609.21088
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.21088
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-10T01:43:42.698Z
- SHA-256 of the archived text
eafcefe6863cf8245fc0dba2e5909c746644139e8ae5bb8f81c95156c7ff4f19- Stable URL of this page
https://metraipolis.com/documento/dc_632406c97537
How to cite
Origin Is All You Need: Provenance-Aware Transformers for Structural Trust-Boundary Separation — arXiv:2609.21088 — arXiv (en), https://arxiv.org/abs/2609.21088 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_632406c97537 (Metraipolis archive dc_632406c97537, sha256 eafcefe6863cf8245fc0dba2e5909c746644139e8ae5bb8f81c95156c7ff4f19)
BibTeX
@misc{dc632406c97537,
title = {Origin Is All You Need: Provenance-Aware Transformers for Structural Trust-Boundary Separation — arXiv:2609.21088},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.21088},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_632406c97537; sha256 eafcefe6863cf8245fc0dba2e5909c746644139e8ae5bb8f81c95156c7ff4f19},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.21088. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Indirect prompt injection (IPI) remains a central safety and security challenge for large language model (LLM) systems because standard transformers lack architectural notion of source authority. Retrieved documents, user inputs, and system instructions are all processed through the same undifferentiated attention mechanism, forcing the model to infer from wording alone what should be obeyed and what should be treated as data. We propose Provenance-Aware Transformers, a provenance-aware defense that makes application-supplied source labels actionable inside the model. Each input token is assigned a ring ID encoding its origin, and the model is augmented with origin embeddings, a learnable origin attention bias, and a learnable origin scale that preserves provenance under normalization. The resulting architecture enforces a structural boundary between authoritative and non-authoritative sources during generation. To instantiate this architecture on released pretrained models, we propose a two-stage fine-tuning pipeline to teach the model origin semantics and task behavior under ring constraints. Evaluation shows that Provenance-Aware Transformers maintain robust resistance to IPI both in-distribution and out-of-distribution while preserving utility comparable to the base pretrained model. More broadly, our work shows that exposing provenance as a first-class architectural signal can shift LLM safety alignment from brittle pattern matching toward explicit trust separation. Tema degli esploratori: fiducia digitale e crittografia (retro-catalogo arXiv: annunci del 2026-09-21 in cs.CR). Categorie arXiv: Cryptography and Security (cs.CR). Pagina del lavoro: https://arxiv.org/abs/2609.21088 — 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 2026-10-10T01:13:42.007Z dalla pagina https://arxiv.org/catchup/cs.CR/2026-09-21?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).