WorksPhysics and Astronomy
Quantifying extractable imaginarity with qubit triangles
Abstract
Two qubit states with the same robustness of imaginarity can differ in the robustness attainable by real filtering on a successful outcome. We represent the input robustness and the largest conditional robustness by the base of a right triangle and the slope of its hypotenuse, respectively. Twice the area, , defines a faithful, convex measure that is monotone on average under real instruments. In arbitrary finite dimensions, maximizing the mean normalized area of the extracted qubits gives the pointwise smallest strongly monotone extension of the qubit measure. We derive computable bounds on this area by determining the maximum mean squared robustness of the output qubits. This maximum equals the Bures quantum divergence between the input and its real part. A projection onto orthogonal real subspaces attains the maximum and, when its outcome is retained, preserves the quantum Fisher information at the midpoint of a mixture of the input and its complex conjugate. We obtain exact areas for states with commuting real and imaginary parts and for pure states mixed with white noise. Two eight-dimensional states have identical , , and optimal mean squared robustness, yet their different areas rule out deterministic conversion by real operations in one direction.