LATIDIA · Ciberseguridad

Marca de agua de código guiada por gramática con temperatura verde

arXiv: 2610.05323v1Tipo de anuncio: nuevo Resumen: La marca de agua del modelo de lenguaje grande incrusta señales estadísticas detectables durante la decodificación, pero los cambios resultantes en las probabilidades del token pueden degradar la calidad de la generación.

La noticia

arXiv:2610.05323v1 Announce Type: new Abstract: Large language model watermarking embeds detectable statistical signals during decoding, but the resulting changes to token probabilities can degrade generation quality. This trade-off is particularly important for code, where small changes in token selection can break syntax or alter program behavior. Existing code watermarking methods mitigate this risk through entropy-based insertion or syntax-aware token selection, but they do not directly construct the watermark over the set of continuations admitted by the current grammar state. We propose Grammar-Guided Code Watermarking with Green Temperature (GTCW), which integrates grammar-constrained decoding with probability-aware watermarking. At each decoding step, GTCW restricts the candidate set to grammar-admissible tokens and partitions this