Continuity of Regularized Channel Rényi Divergences
Recovering the regularized relative entropy as the threshold for quantum channel discrimination requires continuity of the regularized, stabilized sandwiched Rényi channel divergence at order one. Continuity for each fixed input and block length is not enough, because the divergence is optimized over entangled inputs and arbitrarily large blocks.
Left continuity follows from monotonicity in α and from interchanging suprema. Right continuity is shown by contradiction using the channel hockey-stick divergence: Gour's Stinespring approximation bound and a Schatten norm estimate turn an asymptotic bound below one into exponential decay at higher threshold rates. Theorem 1 is formalized in Lean, and Comparator checked the theorem statement and replayed the proofs in the Lean kernel.
For finite-dimensional channels, the limit of D_α^reg as α→1 equals D^reg. For channel pairs with finite max-relative entropy, this yields exponential strong converses for parallel and adaptive discrimination, a sharp zero–one testing law, and the subchannel asymptotic equipartition property.
