Run inference on synthetic data and explore how adaptive thresholds improve anomaly detection
Lower = more sensitive (more detections), Higher = stricter (fewer false positives)
Accepts: timestamp, feature_1...feature_N columns. Download a sample CSV to see the format.