Provenance-Aware Transformers is proposed, a provenance-aware defense that makes application-supplied source labels actionable inside the model and shows that exposing provenance as a first-class architectural signal can shift LLM safety alignment from brittle pattern matching toward explicit trust separation.
Abstract
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.
Indirect prompt injection can persist through retrieval, generation, and revision after the visible payload disappears. Existing defenses primarily detect suspicious inputs or constrain immediate actions, but they provide less evidence about which source influenced which final-artifact span through a temporally valid r...
This work introduces PROSE (Provenance through Relational Organization of Semantic Expression), replacing fixed query sets with a target semantical domain and brittle response keys with semantic structures internalized as domain-conditioned response behavior.
Zhong-Rui Sun, Jia-Hao Chen, Ou-Bo Ma et al.· 0 citations
A recorded dependency graph cannot certify that it contains no omitted edge. For learned artifacts, graph-scoped invalidation therefore cannot by itself justify admission after authority changes, especially when output testing misses a provenance-stale derivation. We separate recorded lineage, which proposes impact sco...
Witness Overlap is proposed, a prompt-free, training-free white-box test for directional provenance, and on 176 LLM checkpoints from 16 families, the one-witness test orients 95.3\% of parent-child decisions using Frobenius cosine.
Si-Yuan Li, Hao-Xuan Zeng, Xin Luo et al.· 0 citations
LLM agents can produce harmful effects through sequences of ordinary operations. Judging such actions requires establishing both the authority that permits them and the origin of the data they carry. We present AGATE, an authorization and data-provenance gate at instrumented agent-harness boundaries. Operator declarati...
Xiao-Ru Zhang, Zhuo-Ran Cheng, Kai-Lin Liu et al.· arXiv.org· 5 citations
Tool-using large language model (LLM) agents turn generated text into real side effects, so poisoned tool metadata, retrieved pages, memory, and reusable skills can steer the next call. Vetting an artifact before admission does not settle this. A safe variant and a leaking variant can produce the same admission evidenc...
Feng-Peng Li, Qi-Zhou Wang, Yu-Ke Hu et al.· 1 citation
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.