Sharp quasi-reverse Minkowski constant for exponents below two

Determine the sharp constant \(\widehat C_p\) for every exponent \(1<p<2\) such that, for all operators \(A,B\) on a finite-dimensional Hilbert space, \(\|A+B\|_p\leq\widehat C_p\bigl\||A|+|B|\bigr\|_p\).

Background

The paper proves that the proposed constant CpC_p, defined by xpp=2xp+1x_p^p=2x_p+1 and Cp=xp(xp+1)/(xpp+1)1/pC_p=\sqrt{x_p(x_p+1)}/(x_p^p+1)^{1/p}, is optimal for the two-summand Schatten-norm inequality when p2p\geq2. For $1

The authors first analyze a family of rank-one examples whose quotient is represented by Fp(c,d)F_p(c,d), and show that its maximum exceeds the proposed value CpC_p. They then give a higher-dimensional numerical example, specifically for p=6/5p=6/5, whose quotient exceeds the maximum of Fp(c,d)F_p(c,d). Consequently, identifying the universal optimal constant C^p\widehat C_p for each $1

References

For every 1<p<2 and A,B\in \mathcal{B(H)}. What is the sharp constant \widehat C_p that makes

|A+B|_p \leq\widehat C_p\bigl||A|+|B|\bigr|_p?

Sharp Quasi-Reverse Minkowski Inequality for Schatten Norms  (2608.17565 - Qiu, 18 Aug 2026) in Question labeled \ref{open}, Section 5, “Counterexamples for \(1<p<2\)”