Authors: Amit Samanta and Mohammad Abrarul Hasanat (University of Utah), Jason Ludmir (Rice University), Ryan Stutsman (University of Utah), Tirthak Patel (Rice University), and Rohan Basu Roy (University of Utah)
Abstract: Variational quantum algorithms (VQAs) suffer from unstable optimization due to temporal noise drift and static qubit mappings that distort gradient signals across iterations. We present BAHAMAS, an online control framework that stabilizes noise exposure by adaptively selecting physical mappings via consensus-based fidelity estimation, without requiring simulators, offline training, or prior executions. Across real quantum devices, BAHAMAS improves optimization reliability and supports inference-time retargeting under drift through robust, per-iteration control.