Newest papers

Auditor-Assisted Summary-Channel Verification for Hosted LLM Identity Substitution

Ziyi Zhang, Ziyao Wang, Guoheng Sun, Ang Li, Jian Li · 2026 · arXiv (Cornell University)

DOI: 10.48550/arxiv.2610.25982 · cited by 0 works (OpenAlex) · open-access PDF

Spanish

Los servicios de LLM de sumo confiable están exponiendo cada vez más la identidad del modelo como una reclamación de producto, pero los usuarios externos y auditores a menudo no pueden inspeccionar qué configuración del modelo sirvió a una solicitud. Estudiamos un régimen asistido por auditor en el que un auditor de confianza registra las identidades reclamadas y prepara el material de verificación antes de la implementación, mientras que la verificación del tiempo de auditoría utiliza solo la identidad reclamada y la salida publicada, sin acceso en tiempo de ejecución a los internos del modelo, los registros de inferencia, las decisiones de enrutamiento o los metadatos del proveedor. Proponemos SUMMARK, un marco de marcado de agua de canal sumario que utiliza los resúmenes de razonamiento visibles al usuario como una superficie de auditoría clave. Durante la

Machine translation (Meta M2M100); the English original follows.

Abstract

Hosted LLM services increasingly expose model identity as a product claim, but external users and auditors often cannot inspect which model configuration served a request. We study an auditor-assisted regime in which a trusted auditor enrolls claimed identities and prepares verification material before deployment, while audit-time verification uses only the claimed identity and released output, without runtime access to model internals, inference logs, routing decisions, or provider metadata. We propose SUMMARK, a summary-channel watermarking framework that uses user-visible reasoning summaries as a keyed audit surface. During setup, SUMMARK constructs identity-specific vocabulary carriers, trains lightweight summary adapters, and calibrates per-identity thresholds. At deployment time, a keyed statistical test assesses whether the released summary is consistent with the claimed identity and, when needed, attributes the most likely enrolled identity. Across model families, scales, and task domains, SUMMARK reliably detects cross-family and cross-scale substitutions at low false-positive rates while preserving task-facing answer quality. We further study FP16->INT8/INT4 precision-consistency diagnostics and robustness under bounded post-processing, rewriting stress tests, and text-only spoofing, revealing both strong performance and clear failure boundaries. Our code is publicly available at https://github.com/trinitr0toluene/Summark.

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Topics: Computer science, Identity (music), Rewriting, Computer security, Robustness (evolution), Audit

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange