QML-PipeGuard: Drift-Aware Behavioral Fingerprinting for Quantum Machine Learning Pipeline Integrity

QML-PipeGuard detects both hardware drift and adversarial channel substitution in quantum ML pipelines by fingerprinting circuits via expectation value vectors under structured measurement families, distinguishing benign recalibration noise from mathematically distinct channel swaps.