arXiv:2609.23946
A constructive violation of additivity of minimum output von Neumann entropy
Laura Shou, Alexey V. Gorshkov
Abstract
We give an explicit non-random example of nonadditivity of minimum output von Neumann entropy. The proof uses a finite-dimensional construction which imitates free Haar unitary behavior. The same channel gives a violation of additivity for the minimum output Rényi-$p$ entropy for any $p\in[1,\infty]$. Additionally, given any $p_0>0$ and $\mathfrak g>0$, we extend the construction to obtain an explicit non-random channel which has entropy gap $\ge\mathfrak g$ uniformly over $p\in[p_0,\infty]$.
AI Summary
The paper addresses the construction of a specific closed-form, non-random quantum channel violating additivity of minimum output von Neumann entropy. Random counterexamples and deterministic polynomial-time construction algorithms were already known; the remaining target here is an explicitly specified example. It constructs a real channel $\Phi:M_d(\mathbb C)\to M_k(\mathbb C)$ with $k=2^{32}$ and $d=3\cdot2^{5268390106224805281791998}$, proving $2H_{\min}(\Phi)-H_{\min}(\Phi\otimes\Phi)>3/17179869184$. The same channel violates minimum output Rényi-$p$ entropy additivity uniformly for every $p\in[1,\infty]$, with the same lower bound. An extended construction gives, for any $p_0>0$ and prescribed gap, an explicit channel achieving that gap uniformly over $p\ge p_0$. Its quantitative bound is $(4\ell/5-6)\log2$, with output dimension $2^{17\ell}$. The construction uses permutation unitaries from finite quotients of a free group, obtained through explicit $\mathrm{SL}_2(\mathbb Z)$ generators. Their normalized trace moments match free Haar-unitary moments to a sufficiently high finite order. A damping operator, defined using high powers summed over an explicit net, suppresses exceptional subspaces while remaining close to the identity in normalized trace. Haagerup-type inequalities then lower-bound single-use output entropy; a maximally entangled input upper-bounds two-use entropy. Product-group variants yield the larger gaps. The stated dimensions require roughly five septillion input qubits. Dimension and gap optimization are not attempted, and the construction does not provide deterministic strong convergence to free Haar unitaries.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
- Comments
- 22 pages; v2 added Theorem 1.3/Appendix B on extension to large entropy gap and p\ge p_0>0