arXiv:2609.28367
The Minimal Dimension of Entangled-Noise Advantage
Xiao-Ke Wang, Zi-Yuan Liu, Ming-Yang Li, Shengjun Wu, Zeng-Bing Chen
Abstract
Can entangled noise erase entanglement more efficiently than separable noise? Standard robustness restricts the added noise to separable states; generalized robustness allows any state. We prove that the two costs coincide for every two-qubit state and give an explicit full-rank qubit--qutrit state with a strict gap. Positivity under partial transpose (PPT) characterizes separability in both dimensions, so the boundary is not caused by a failure of the PPT criterion. Instead, product-vector geometry permits a rank-one bridge between the two-qubit dual optimizations and supplies a two-dimensional completely entangled subspace for the qubit--qutrit separation. Local isometric embeddings complete the finite-dimensional bipartite classification, also for any fixed multipartite cut. Combined with known three-qubit separation, the result classifies universal equality relative to full separability in all finite multipartite systems.
AI Summary
The paper asks when entangled noise can erase a state’s entanglement at a lower admixture cost than separable noise. It compares standard robustness $R_{\mathrm s}$, which restricts the noise to separable states, with generalized robustness $R_{\mathrm g}$, which allows arbitrary states. It proves $R_{\mathrm s}(\rho)=R_{\mathrm g}(\rho)$ for every two-qubit state, including existence of a separable generalized-optimal noise. The proof reduces a standard-robustness dual optimizer to rank one by subtracting product-vector projectors, using the fact that every two-dimensional two-qubit subspace contains a product vector. The rank-one partial-transpose spectrum then makes this optimizer feasible for the generalized dual. An explicit full-rank qubit–qutrit state satisfies $R_{\mathrm g}\leq1/8<2/15\leq R_{\mathrm s}$, certifying a gap of at least $1/120$. These are bounds, not exact optimal values. Feasible entangled noise and a separable-state witness establish the separation analytically. PPT characterizes separability in both dimensions; the construction instead uses a two-dimensional completely entangled subspace. Local isometric embeddings show that, among nontrivial finite bipartite dimensions, universal equality holds only for $2\otimes2$. For full multipartite separability, universal equality holds only with at most one nontrivial party or exactly two qubits. Explicit GHZ certificates give $R_{\mathrm g}=1$ and $R_{\mathrm s}=2^{n-2}$, with formulas also supplied for white-noise mixtures. The results concern exact single-copy robustness, not asymptotic or smoothed tasks. Equality of values does not imply that every generalized-optimal noise is separable.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
- Comments
- 32 pages, 4 figures