Learning-Based Shrinking Disturbance-Invariant Tubes for State- and Input-Dependent Uncertainty Paper • 2601.11426 • Published Jan 16 • 1