LATIDIA · Ciberseguridad

Calibración consciente de la estructura de pruebas de estrés de redes neuronales de gráficos de malware bajo cambio de tipo

arXiv: 2609.28517v2Tipo de anuncio: reemplazar Resumen: Los calibradores de malware post-hoc pueden condicionar la confianza en la estructura del gráfico, pero sus entradas estructurales pueden dejar el soporte representado por los datos de validación bajo malware

La noticia

arXiv:2609.28517v2 Announce Type: replace Abstract: Post-hoc malware calibrators can condition confidence on graph structure, but their structural inputs may leave the support represented by validation data under malware-type shift. We study this risk on MalNet-Tiny by holding out each of four malware types across three seeds, freezing a graph isomorphism network, and fitting post-hoc mappings only on known-type data. On known-type tests, generic and combined structural calibrators have negative log likelihood (NLL) of 0.4341 and 0.4628. Their mean held-out-type gap widens to 0.1369, with a four-type 95% interval of [-0.0125, 0.2864] and improvement in one of four types. Eight label-independent nuisance variables produce a smaller mean NLL increase of 0.0415,