LATIDIA · Ciberseguridad

RoCoMark: Marca de agua de código robusta contra ataques de reescritura de preservación de funcionalidad basados en LLM

arXiv: 2611.11399v1Tipo de anuncio: nuevo Resumen: los modelos de lenguaje grandes (LLM) presentan desafíos significativos para el rastreo de fuentes del código generado, ya que la reescritura de LLM puede alterar sustancialmente la cara sintáctica y estructural

La noticia

arXiv:2611.11399v1 Announce Type: new Abstract: Large Language Models (LLMs) present significant challenges to source tracing of generated code, as LLM rewriting can substantially alter the syntactic and structural characteristics of watermarked code. However, existing watermarking methods often suffer severe performance degradation under LLM-based functionality-preserving rewriting attacks. We attribute this vulnerability not merely to the intensity of rewriting, but to a fundamental distribution mismatch between the perturbations encountered during watermark training and the global, non-deterministic rewriting behaviors induced by LLMs. In order to deal with this challenge, we propose RoCoMark, a Robust Code waterMarking framework that explicitly learns and adapts to LLM rewriting behaviors. RoCoMark is a multi-round minimax game between