arXiv:2609.31350
Sharp Data-Processing Region for Quantum Conditional Rényi Entropies
Milad M. Goodarzi, Yucheng Bao, Marco Tomamichel
Abstract
We study the three-parameter family of quantum conditional Rényi entropies $H_{α,z}^λ$ introduced in \textit{Commun. Math. Phys.} 407, 180 (2026). We prove that the previously established data-processing region is sharp. We then characterize the club-sandwiched family through a classical coarse-graining principle: within the full data-processing region, universal monotonicity under deterministic coarse-graining of the classical conditioned register holds if and only if $z = α$. Finally, we establish rigidity of entropic duality by showing that any universal duality relation for finite-dimensional tripartite pure states must satisfy the known parameter relations. Thus, the data-processing region, the club-sandwiched subfamily, and the duality transformation are each uniquely characterized by their respective universal properties.
AI Summary
The paper determines which parameter choices make the three-parameter quantum conditional Rényi entropies $H_{\alpha,z}^{\lambda}$ compatible with data processing, classical coarse-graining, and pure-state duality. It considers finite-dimensional systems with $\alpha>0$, $\alpha\ne1$, and $z>0$. Its first result proves necessity of the previously sufficient data-processing region. For $0<\alpha<1$, the exact conditions are $\max\{\alpha,1-\alpha\}\le z$ and $1-z/(1-\alpha)\le\lambda\le1$. For $\alpha>1$, they are $\max\{\alpha-1,\alpha/2\}\le z\le\alpha$ and $1-z/(\alpha-1)\le\lambda\le1$. Data processing here allows instruments on the conditioning system followed by outcome-dependent sub-unital channels on the conditioned system. The proof transfers conditional-entropy data processing to data processing of the corresponding $\alpha$-$z$ Rényi divergence, using perturbative classical–quantum states with fixed marginal. Known divergence bounds, classical restrictions, and pure-state transformations then supply the parameter constraints. Within this region, deterministic coarse-graining of the classical conditioned register satisfies $H(X|B)\ge H(f(X)|B)$ universally if and only if $z=\alpha$. Necessity follows from a two-dimensional matrix perturbation. Requiring coarse-graining for both an entropy and its dual further selects $\lambda=1$, with dual orders satisfying $1/\alpha+1/\hat\alpha=2$. Finally, explicit pure-state test families show that universal duality between parameter triples in the data-processing region forces exactly the previously known parameter relations. The coarse-graining and duality characterizations are restricted to that region; sufficiency relies on earlier results. No infinite-dimensional or $\alpha=1$ characterization is provided.
Error found: The abstract states unrestricted necessity of the duality parameter relations, but the paper proves necessity only when both parameter triples belong to the data-processing region D. The duality proof uses a pure-state formula requiring r < 1 without first excluding the r = 1 boundary admitted by its theorem, although the separate boundary formula supplies a short exclusion. The lower-λ-bound proof displays an incorrect Schatten norm index and exponent; the correct intermediate expression has norm index 2t and exponent 2z.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.