LATIDIA · Ciberseguridad

Lógicas de ataque, no resultados: Ataques basados ​​en conceptos para la eliminación de atajos en redes neuronales

arXiv:2610.34923v2 Anuncio Tipo: reemplazar Resumen: Las redes neuronales profundas (NN) para visión artificial son vulnerables a ataques adversarios, de modo que cambios maliciosos minúsculos en las entradas pueden inducir salidas poco intuitivas. En

La noticia

arXiv:2610.34923v2 Announce Type: replace Abstract: Deep neural networks (NNs) for computer vision are vulnerable to adversarial attacks, such that minuscule malicious changes to inputs may induce unintuitive outputs. One key approach to verify and mitigate such robustness issues is to falsify expected output behavior, which enables local robustness guarantees and adversarial training. Due to the black-box nature of NNs, current attacks only falsify a class of the final output, such as flipping from $\text{stop_sign}$ to $\neg \text{stop_sign}$. In this work we argue that robustness analysis should focus more generally on logical inconsistencies in concept-level reasoning rather than solely on output misclassification. We formalize this perspective through concept-based properties, i.e., constraints