LATIDIA · Ciberseguridad

Los puntos ciegos de la detección: ataques de perturbación adversarios contra los sistemas de vigilancia en la inferencia LLM distribuida.

arXiv:2611.01097v1 Anuncio Tipo: nuevo Resumen: La descentralización de la inferencia de modelos de lenguaje grandes (LLM) a través de nodos web no confiables requiere una verificación robusta para garantizar la integridad computacional. Existente basada en la activación

La noticia

arXiv:2611.01097v1 Announce Type: new Abstract: Decentralizing large language model (LLM) inference across untrusted Web nodes requires robust verification to ensure computational integrity. Existing activation-based verification and detection mechanisms validate intermediate outputs against reference values, permitting bounded numerical deviations to accommodate benign cross-hardware discrepancies. However, a critical security question remains: what can a malicious worker achieve while operating within these tolerance bounds? In this work, we introduce Hush-Up, a stealthy availability attack against distributed LLM inference, where a single malicious worker halts generation while evading attribution. The attacker first profiles its local execution to model an empirical bound of benign hardware errors. Using this error bound as a feasibility constraint, the