LATIDIA · Ciberseguridad

Repensar el razonamiento: una encuesta sobre puertas traseras basadas en el razonamiento en LLM

arXiv: 2510.07697v2Tipo de anuncio: reemplazar Resumen: Con el aumento de las capacidades de razonamiento avanzado, los modelos de lenguaje grandes (LLM) están recibiendo una atención cada vez mayor. Sin embargo, aunque el razonamiento mejora el rendimiento de los LLM

La noticia

arXiv:2510.07697v2 Announce Type: replace Abstract: With the rise of advanced reasoning capabilities, large language models (LLMs) are receiving increasing attention. However, although reasoning improves LLMs' performance on downstream tasks, it also introduces new security risks, as adversaries can exploit these capabilities to conduct backdoor attacks. Existing surveys on backdoor attacks and reasoning security offer comprehensive overviews but lack in-depth analysis of backdoor attacks and defenses targeting LLMs' reasoning abilities. In this paper, we take the first step toward providing a comprehensive review of reasoning-based backdoor attacks in LLMs by analyzing their underlying mechanisms, methodological frameworks, and unresolved challenges. Specifically, we introduce a new taxonomy that offers a unified perspective for