LATIDIA · Ciberseguridad
La geometría representativa como diagnóstico de la robustez adversaria
arXiv:2611.17684v1 Tipo de anuncio: cruzado Resumen: El entrenamiento adversarial es una técnica habitual para hacer que las redes neuronales sean más robustas frente a pequeñas perturbaciones de entrada creadas intencionadamente, y la robustez se suele medir
WhatsApp ↗Telegram ↗
La noticia
arXiv:2611.17684v1 Announce Type: cross Abstract: Adversarial training is a common technique for making neural networks more robust to small, intentionally crafted input perturbations, and robustness is commonly measured through accuracy under attack. By exploring inter-network representations across different adversarial training objectives, we can analyze whether multiple networks organize their latent representations similarly. We study Multilayer Perceptrons trained on MNIST under three adversarial training regimes and compare the resulting population training dynamics and geometry to those of a vanilla network. We use M-PHATE to embed data samples across training epochs and examine how the geometry of the data changes during training under adversarial objectives. On penultimate-layer representations we measure displacement between