LATIDIA · Ciberseguridad

Miscalibración de seguridad controlada por recuperabilidad en modelos de lenguaje grandes

arXiv: 2610.22425v1Tipo de anuncio: nuevo Resumen: se espera que los modelos de lenguaje alineados con la seguridad rechacen las solicitudes dañinas y sigan siendo útiles para las benignas. Sin embargo, debido a que muchas evaluaciones califican solo la respuesta final

La noticia

arXiv:2610.22425v1 Announce Type: new Abstract: Safety-aligned language models are expected to refuse harmful requests while remaining useful for benign ones. However, because many evaluations score only the final response, they often conflate intent understanding with response-level safety behavior. To address this, we introduce RiskCueBench, a recoverability-controlled benchmark that separates these two stages by first measuring whether a model can successfully recover a high-level intent and then evaluating its safety behavior on the same prompt. RiskCueBench contains 160 latent intents and 640 abstract non-operational surface prompts across eight risk categories. The benchmark covers 25 completed model groups and 32,000 successful model outputs with final rubric-adjudicated labels. When conditioning on pairs where