A.1 Proof Theorem(Algorithmic Fixt2Pt. inequalityhA(, S) learned USuniformly A tolossL: E