LATIDIA · Ciberseguridad

Marcas de agua con reconocimiento de bits mayoritario para modelos de lenguaje grandes

arXiv:2508.03829v3 Announce Type: replace-cross Resumen: El creciente despliegue de Modelos de Lenguaje Grande (LLM) ha generado preocupaciones sobre su uso indebido en la generación de contenido dañino o engañoso. Para abordar este problema, w

La noticia

arXiv:2508.03829v3 Announce Type: replace-cross Abstract: The growing deployment of Large Language Models (LLMs) has raised concerns about their misuse in generating harmful or deceptive content. To address this issue, watermarking methods have been proposed to embed identifiable multi-bit messages into generated text for misuse tracing. However, existing methods often suffer from a fundamental trade-off between text quality and decoding accuracy. In particular, they have to restrict the size of the preferred token set (i.e., green list) during encoding to maintain a detectable watermark signal for decoding, which inevitably degrades generation quality. To improve this trade-off, we propose a novel message encoding paradigm called \textit{majority bit-aware encoding}, which relaxes the watermark