LATIDIA · Ciberseguridad

Profile-DP: Privacidad diferencial para la publicación de perfiles controlados por el usuario en inferencia LLM personalizada

arXiv: 2610.27213v1Tipo de anuncio: nuevo Resumen: Los modelos personalizados de lenguaje grande (LLM) se basan cada vez más en la inferencia condicionada por el perfil, emparejando las solicitudes públicas con datos enriquecidos y específicos del usuario. Este paradigma expone

La noticia

arXiv:2610.27213v1 Announce Type: new Abstract: Personalized large language models (LLMs) increasingly rely on profile-conditioned inference, pairing public requests with rich, user-specific facts. This paradigm exposes the user profile itself, together with the user's explicit sharing preferences. We introduce Profile-DP, a policy-aware local differential privacy framework designed specifically for user-profile release in personalized LLM inference. Profile-DP provably bounds the release of user-withheld facts while treating public instructions, profile schemas, and authorized facts as public side information. By generating a privatized profile that composes with the public instruction prior to answer generation, our framework decouples formal privacy guarantees from downstream model behavior. We develop two finite-domain mechanisms satisfying Profile-DP: DP-SLOT independently perturbs