LATIDIA · Ciberseguridad

Sobre la relación entre la cuantificación del modelo y los ataques de inversión del modelo

arXiv: 2610.00382v1Announce Type: new Resumen: La cuantificación del modelo reduce la precisión numérica de los pesos y activaciones de la red neuronal para reducir los costos de almacenamiento y computación. Los ataques de inversión de modelo recuperan o re

La noticia

arXiv:2610.00382v1 Announce Type: new Abstract: Model quantization reduces the numerical precision of neural network weights and activations to lower storage and computational costs. Model inversion attacks recover or reconstruct sensitive training data or inference inputs from model outputs or intermediate features, so quantization may also alter their effectiveness. However, two questions remain unresolved: How does model quantization affect model inversion? How do data characteristics influence this relationship? To address the first, we bound quantization-induced changes in mutual information between inputs and a categorical variable defined by prediction probabilities, distinguishing informational effects from attack optimization obstacles. To address the second, we identify data-dependent changes in feature distributions and inversion outcomes, with