LATIDIA · Ciberseguridad

Modelo de Lenguaje Grande Federado Bizantino-Resiliente a través del Paso de Mensajes Hipergráficos

arXiv:2610.12711v1 Anuncio Tipo: cruzado Resumen: Los modelos de lenguaje grandes federados (FedLLM) prometen ofrecer inteligencia colaborativa que preserve la privacidad. El proceso de entrenamiento descentralizado es vulnerable a Byz

La noticia

arXiv:2610.12711v1 Announce Type: cross Abstract: Federated Large Language Models (FedLLMs) are promising to deliver privacy-preserving, collaborative intelligence. The decentralized training process is vulnerable to Byzantine attacks that corrupt semantic consistency and factual resilience through malicious model updates. This paper presents a new resilience strategy that prevents semantically corrupted knowledge from propagating during federated aggregation. A defense problem is formulated to jointly maximize global accuracy and suppress manipulated model updates, enabling trustworthy knowledge aggregation without accessing private local data. To strengthen the resilience of FedLLMs, we develop a novel Hypergraph Message Passing Graph Autoencoder (HMP-GAE) framework, where low-rank model updates are represented as nodes in a hypergraph and high-order semantic relationships