AI SOC Alert Fatigue: 2D Signal-Detection & ROC Explorer

This is a 2D-native counterpart to the 3D "AI SOC Alert Fatigue" simulator: instead of watching individual particles fly down a 3D corridor toward a detection plane, this build makes the underlying signal-detection-theory math directly visible. A live plot overlays the benign and intrusion Gaussian score distributions with the true/false-positive tail areas shaded at your chosen threshold — drag right on the plot itself to slide θ. A dwell-time board tracks every currently active intrusion as an independent filling bar, turning green when an analyst catches it or red if it survives long enough to breach. A third panel draws the full parametric ROC curve for the current attacker stealth level Δμ, with your live operating point and the analytically exact AUC. Tune detection threshold, attacker signal separation, traffic volume, and analyst capacity to see the exact same precision/recall and alert-fatigue trade-offs the 3D version models, now with the statistics laid bare instead of inferred from motion.