Beyond Any-Shot Adaptation: Predicting Optimization Outcome for Robustness Gains without Extra Pay

Open in new window