LATIDIA · Ciberseguridad

Análisis basado en grafos de la comparabilidad, la escala y la atribución en la inferencia de transformadores privados.

arXiv:2610.19607v1 Anuncio Tipo: nuevo Resumen: Las crecientes preocupaciones sobre la privacidad en torno a los grandes modelos de lenguaje han acelerado la investigación en inferencia privada de Transformer (PTI). Sintetizando las afirmaciones empíricas en esta línea de investigación

La noticia

arXiv:2610.19607v1 Announce Type: new Abstract: Rising privacy concerns around large language models have accelerated research in private Transformer inference (PTI). Synthesizing the empirical claims in this line of research, however, is hindered by multi-dimensional cryptographic progress and inconsistent evaluation contracts. In this work, we systematize the field by constructing an empirical graph of 62 PTI systems, mapping system attributes to node features and evaluation compatibility to edges. A three-tiered analysis of this graph illuminates the state of private inference. Node-level attribution shows that cryptographic progress drives performance gains across most systems, yet always co-occurs with other factors (model fidelity, hardware, cost accounting). Edge-level auditing shows that only 1 of 1891