AI RESEARCH
Why the Maximum Second Derivative of Activations Matters for Adversarial Robustness
arXiv CS.LG
•
ArXi:2603.23860v1 Announce Type: new This work investigates the critical role of activation function curvature -- quantified by the maximum second derivative $\max|\sigmain adversarial robustness. Using the Recursive Curvature-Tunable Activation Family (RCT-AF), which enables precise control over curvature through parameters $\alpha$ and $\beta$, we systematically analyze this relationship.