arXiv:2610.06787
Computational entanglement distillation beyond pure states: efficient protocols and fundamental limitations
Takeru Utsumi, Jacopo Rizzo, Ryuji Takagi, Lorenzo Leone
Abstract
The information-theoretic approach to entanglement typically relies on arbitrarily many input copies and unboundedly complex operations; however, the available resources are constrained in reality. This motivates the study of entanglement theory with limited resources. One of its central questions is which quantities characterize entanglement manipulations, and so far, only the idealized case of pure states has been well explored. In this work, we address entanglement distillation from mixed states and identify the Rényi coherent information as the quantity determining the optimal rate in the worst case under two distinct resource constraints. We first characterize the sample-efficient regime, in which unbounded computational power is allowed. Our universal protocol achieves this rate without knowledge of the input state under certain polynomial rank constraints. This rate is optimal among universal protocols: we construct states whose distillable entanglement is nearly extensive, yet from which no universal sample-efficient protocol attains a rate above their Rényi coherent information. We then push this characterization to the computationally efficient frontier: the same bound holds for any efficient protocol even when the input state is known, and another universal protocol achieves this rate efficiently under additional polynomial constraints on the marginal ranks. Overall, our results give the Rényi coherent information a novel operational meaning in two computational settings: universal distillation with limited copies and efficient distillation even with knowledge of the input state.
AI Summary
The paper asks which entropic quantities govern mixed-state entanglement distillation when the number of copies and, separately, computational resources are polynomial in the qubit count $n$. It establishes achievability guarantees and matching worst-case constructions involving sandwiched Rényi coherent information $\widetilde I_\beta(A\rangle B)$, for fixed $\beta\in[1/2,1)$. For inputs within trace distance $1/\log^{1+c}n$ of a polynomial-rank state $\hat\rho$, a universal one-way LOCC protocol uses polynomially many copies and achieves $\min\{\widetilde I_\beta(\hat\rho),C\log n\}-o(1)$ for any constant $C>0$, without a circuit-size bound. If $\hat\rho$ also has polynomial marginal ranks, another universal protocol achieves $\widetilde I_\beta(\hat\rho)-o(1)$ with polynomial-size circuits. The first extends Schur coding through enlarged invariant supports; the second combines collective local compression, tomography, decoupling, and an efficiently implemented Uhlmann decoder. Concatenated stabilizer codes reduce the output error. Converse constructions have information-theoretic distillable entanglement $\widetilde\Omega(n)$ but resource-limited rates capped, at suitable Rényi orders, by $\min\{\widetilde I_\beta,\log k\}+o(1)$. Statistical indistinguishability gives bounds for universal sample-efficient and state-dependent efficient protocols. Public-key pseudoentanglement yields analogous bounds for efficiently preparable states and description-aware compilers under superpolynomial LWE hardness. Constructed superpolynomial-rank families permit no efficient distillation. Other constructions show positive attainable rates despite negative conditional min-entropy, and exclude any single fixed Rényi order as a general characterization. Efficient achievability without the marginal-rank condition remains open. Optimality concerns constructed worst-case families, not arbitrary mixed states; rank alone does not imply inaccessibility, and the description-aware bounds require cryptographic assumptions.
Central claim may not hold: The abstract's sample-efficient rate characterization omits the logarithmic cap: under polynomial global rank, the body guarantees min{Ĩβ, C log n} − o(1), and for its high-Ĩβ hard families establishes only constant-factor optimality at a rate of order log k.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
- Comments
- 87 pages, 2 figures