LATIDIA · Ciberseguridad

El atacante de LLM: un marco unificado de equipo rojo para la evaluación de la privacidad de los sistemas LLM divididos y distribuidos

arXiv:2610.22722v1 Announce Type: new Resumen: Los modelos de lenguaje dividido mantienen el texto de entrada en un cliente pero envían representaciones intermedias a un servidor. Los ataques pueden reconstruir el texto de estos mensajes. Comparación de la eva de privacidad

La noticia

arXiv:2610.22722v1 Announce Type: new Abstract: Split language models keep input text on a client but send intermediate representations to a server. Attacks can reconstruct text from these messages. Comparing privacy evaluations is difficult when attackers see different messages, know different model weights or use different datasets and recovery measures. We present LLM-Attacker, a framework that records these choices and lets researchers check, compare and extend an evaluation. Compatible attacks reuse recorded messages, and their predictions are saved for independent scoring. The scoring rule requires a linked undefended reference before a defence can pass. Reports also identify attacks that remain untested. In a controlled GPT-2 experiment, adding gradients returned during training