Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

arXiv:2411.12208 · Physical Review A

Extremal maximal entanglement

Wanchen Zhang, Yu Ning, Fei Shi, Xiande Zhang

Published May 7, 2025• 5 citations
arXivHTMLPDF

Abstract

A pure multipartite quantum state is called absolutely maximally entangled if all reductions of no more than half of the parties are maximally mixed. However, an $n$-qubit absolutely maximally entangled state exists only when $n$ equals 2, 3, 5, and 6. A natural question arises when it does not exist: Which $n$-qubit pure state has the largest number of maximally mixed $\ensuremath{\lfloor}n/2\ensuremath{\rfloor}$-party reductions? We denote this number by $\mathrm{Qex}(n)$. It was shown that $\mathrm{Qex}(4)=4$ by Higuchi and Sudbery [Phys. Lett. A 273, 213 (2000)] and $\mathrm{Qex}(7)=32$ by Huber et al. [Phys. Rev. Lett. 118, 200502 (2017)]. In this paper, we give a general upper bound of $\mathrm{Qex}(n)$ by linking the well-known Tur\'an problem in graph theory and provide lower bounds using constructive and probabilistic methods. In particular, we show that $\mathrm{Qex}(8)=56$, which is the third known value for this problem.

DOI
10.1103/physreva.111.052410

Related papers

Robust Quantum Extremal Numbers

Wanchen Zhang, Zicheng Han, Xiande Zhang

Absolutely maximally entangled states require every reduction of at most half of the parties to be maximally mixed, a condition that is both rigid and often impossible for qubit systems. Previous work introduced the quantum extremal number, which maximizes the number of exactly maximally mixed half-body marginals, and determined the exact value Qex(8,4)=56. The present work develops a robust extension of this extremal problem. For a subsystem $A$, the marginal maximal-mixing defect is defined by \[ D_A=2^{|A|}\operatorname{Tr}(ρ_A^2)-1 =2^{|A|}\left\|ρ_A-\frac{I_A}{2^{|A|}}\right\|_2^2, \] and $Q_{\mathrm{ex},\varepsilon}^{D}(n,k)$ is defined as the maximum number of $k$-body marginals satisfying $D_A\leq\varepsilon$ in an $n$-qubit pure state. This counting problem differs from approximate $k$-uniformity, which requires all $k$-body marginals to obey a common error bound. For pure states on $4m$ qubits, the following local stability inequality is established: \[ \sum_{i\in T}D_{T\setminus\{i\}}\geq1 \qquad (|T|=2m+1). \] It follows that, whenever $\varepsilon<1/(2m+1)$, the hypergraph of $\varepsilon$-good $2m$-subsets is $K_{2m+1}^{(2m)}$-free. Combined with the known exact eight-qubit construction, this yields the stability plateau \[ Q_{\mathrm{ex},\varepsilon}^{D}(8,4)=56, \qquad 0\leq\varepsilon<\frac15. \] For odd systems of $2k+1$ qubits, the exact forbidden hypergraph $H_k$ is used to derive explicit finite-error stability radii. In particular, $Q_{\mathrm{ex},\varepsilon}^{D}(9,4)\leq120$ for $0\leq\varepsilon<1/17$. These results turn exact quantum Turán obstructions into quantitative robustness statements and identify intervals on which quantum extremal numbers are stable under imperfect marginal mixedness.

quant-pharXiv:2608.13907•Aug 17, 2026•0 citations

Explicit Absolutely Maximally Entangled states in dimensions twice a prime power

Palash Pandya

Absolutely maximally entangled (AME) states are pure multipartite states with maximally mixed reductions on every set of at most half the parties. We construct explicit families in local dimension $2q$, twice an odd prime power $q$, using an established coupling of qubits and $q$-level systems. The $q$-level states remove interference between branches, so the qubit states need maximally mixed reductions only on average. This gives AME$(9,2q)$ and AME$(11,2q)$ for every odd prime power $q\geq5$, AME$(13,2q)$ for $q\geq17$, AME$(15,2q)$ for $q\geq9$, and AME$(17,2q)$ for $q\geq67$. For every odd $N\geq3$, the same coupling gives AME$(N,2q)$ under the sufficient condition $q\geq\max\{N,2^{N-2}\}$ on the odd prime power $q$. Within this coupling, the smallest such $q$ must grow exponentially with $N$. Combined with known inputs and products, these families complete the nine-party classification: for integer $d\geq2$, an AME$(9,d)$ state exists if and only if $d\geq3$. Choosing the branch pairing also gives AME$(15,2p)$ states with a cyclic shift of the parties compensated by local unitaries, for every prime $p\equiv1\pmod{15}$.

arXiv:2611.17216•Nov 12, 2026•0 citations

Exploring pure quantum states with maximally mixed reductions

Ludovic Arnaud, Nicolas J. Cerf

We investigate multipartite entanglement for composite quantum systems in a pure state. Using the generalized Bloch representation for $n$-qubit states, we express the condition that all $k$-qubit reductions of the whole system are maximally mixed, reflecting maximum bipartite entanglement across all $k$ vs $n\ensuremath{-}k$ bipartitions. As a special case, we examine the class of balanced pure states, which are constructed from a subset of the Pauli group ${\mathcal{P}}_{n}$ that is isomorphic to ${\mathbb{Z}}_{2}^{n}$. This makes a connection with the theory of quantum error-correcting codes and provides bounds on the largest allowed $k$ for fixed $n$. In particular, the ratio $k/n$ can be lower and upper bounded in the asymptotic regime, implying that there must exist multipartite entangled states with at least $k=\ensuremath{\lfloor}0.189\phantom{\rule{0.16em}{0ex}}n\ensuremath{\rfloor}$ when $n\ensuremath{\rightarrow}\ensuremath{\infty}$. We also analyze symmetric states as another natural class of states with high multipartite entanglement and prove that, surprisingly, they cannot have all maximally mixed $k$-qubit reductions with $k>1$. Thus, measured through bipartite entanglement across all bipartitions, symmetric states cannot exhibit large entanglement. However, we show that the permutation symmetry only constrains some components of the generalized Bloch vector, so that very specific patterns in this vector may be allowed even though $k>1$ is forbidden. This is illustrated numerically for a few symmetric states that maximize geometric entanglement, revealing some interesting structures.

arXiv:1211.4118•Physical Review A•Jan 18, 2013•90 citations

Absolutely Maximally Entangled States of $2q$ Parties in Every Odd Prime-Power Dimension $q$

Mykhailo Hontarenko, Karol Życzkowski

Absolutely maximally entangled (AME) states represent an extreme form of multipartite entanglement: every reduced system containing at most half of the parties is maximally mixed. These states provide perfect tensors and optimal quantum error-correcting codes, yet their existence is known only in restricted parameter regimes. For every odd prime power $q=p^e\ge3$, we construct a stabilizer $\mathrm{AME}(2q,q)$ state whose normalized one-party projection yields a stabilizer $\mathrm{AME}(2q-1,q)$ state. A closed-form $q\times q$ bordered-circulant matrix $A_q$ over $\mathbb{F}_{q^2}$ generates a Hermitian self-dual maximum distance separable (MDS) code $[2q,q,q+1]_{q^2}$, which lies outside the extended Reed-Solomon classes. In suitable bases, the amplitude tensors define normalized $q$-unitary complex Hadamard matrices of order $q^q$ with $p$th-root phases. Additional constructions yield $\mathrm{AME}(q+3,q)$ states and families at intermediate particle numbers through explicit rescalings of selected submatrices. We also provide nine explicit parent matrices and the corresponding one-party projections.

quant-pharXiv:2609.11796•Sep 11, 2026•1 citation

Bounds on absolutely maximally entangled states from shadow inequalities, and the quantum MacWilliams identity

Felix Huber, Christopher Eltschka, Jens Siewert +1 more

Abstract A pure multipartite quantum state is called absolutely maximally entangled (AME), if all reductions obtained by tracing out at least half of its parties are maximally mixed. Maximal entanglement is then present across every bipartition. The existence of such states is in many cases unclear. With the help of the weight enumerator machinery known from quantum error correction and the shadow inequalities, we obtain new bounds on the existence of AME states in dimensions larger than two. To complete the treatment on the weight enumerator machinery, the quantum MacWilliams identity is derived in the Bloch representation. Finally, we consider AME states whose subsystems have different local dimensions, and present an example for a 2 × 3 × 3 × 3 system that shows maximal entanglement across every bipartition.

arXiv:1708.06298•Journal of Physics A Mathematical and Theoretical•Feb 8, 2018•95 citations

Absolutely Maximally Entangled States of Seven Qubits Do Not Exist

Felix Huber, Otfried Gühne, Jens Siewert

Pure multiparticle quantum states are called absolutely maximally entangled if all reduced states obtained by tracing out at least half of the particles are maximally mixed. We provide a method to characterize these states for a general multiparticle system. With that, we prove that a seven-qubit state whose three-body marginals are all maximally mixed, or equivalently, a pure ((7,1,4))_{2} quantum error correcting code, does not exist. Furthermore, we obtain an upper limit on the possible number of maximally mixed three-body marginals and identify the state saturating the bound. This solves the seven-particle problem as the last open case concerning maximally entangled states of qubits.

arXiv:1608.06228•Physical Review Letters•May 17, 2017•165 citations
A pure multipartite quantum state is called absolutely maximally entangled if all reductions of no more than half of the parties are maximally mixed. However, an nnn-qubit absolutely maximally entangled state exists only when nnn equals 2, 3, 5, and 6. A natural question arises when it does not exist: Which nnn-qubit pure state has the largest number of maximally mixed \ensuremath⌊n/2\ensuremath⌋\ensuremath{\lfloor}n/2\ensuremath{\rfloor}\ensuremath⌊n/2\ensuremath⌋-party reductions? We denote this number by Qex(n)\mathrm{Qex}(n)Qex(n). It was shown that Qex(4)=4\mathrm{Qex}(4)=4Qex(4)=4 by Higuchi and Sudbery [Phys. Lett. A 273, 213 (2000)] and Qex(7)=32\mathrm{Qex}(7)=32Qex(7)=32 by Huber et al. [Phys. Rev. Lett. 118, 200502 (2017)]. In this paper, we give a general upper bound of Qex(n)\mathrm{Qex}(n)Qex(n) by linking the well-known Tur'an problem in graph theory and provide lower bounds using constructive and probabilistic methods. In particular, we show that Qex(8)=56\mathrm{Qex}(8)=56Qex(8)=56, which is the third known value for this problem.
Absolutely maximally entangled states require every reduction of at most half of the parties to be maximally mixed, a condition that is both rigid and often impossible for qubit systems. Previous work introduced the quantum extremal number, which maximizes the number of exactly maximally mixed half-body marginals, and determined the exact value Qex(8,4)=56. The present work develops a robust extension of this extremal problem. For a subsystem AAA, the marginal maximal-mixing defect is defined by DA=2∣A∣Tr⁡(ρA2)−1=2∣A∣∥ρA−IA2∣A∣∥22, D_A=2^{|A|}\operatorname{Tr}(ρ_A^2)-1 =2^{|A|}\left\|ρ_A-\frac{I_A}{2^{|A|}}\right\|_2^2, DA​=2∣A∣Tr(ρA2​)−1=2∣A∣​ρA​−2∣A∣IA​​​22​, and Qex,εD(n,k)Q_{\mathrm{ex},\varepsilon}^{D}(n,k)Qex,εD​(n,k) is defined as the maximum number of kkk-body marginals satisfying DA≤εD_A\leq\varepsilonDA​≤ε in an nnn-qubit pure state. This counting problem differs from approximate kkk-uniformity, which requires all kkk-body marginals to obey a common error bound. For pure states on 4m4m4m qubits, the following local stability inequality is established: ∑i∈TDT∖{i}≥1(∣T∣=2m+1). \sum_{i\in T}D_{T\setminus\{i\}}\geq1 \qquad (|T|=2m+1). ∑i∈T​DT∖{i}​≥1(∣T∣=2m+1). It follows that, whenever ε<1/(2m+1)\varepsilon<1/(2m+1)ε<1/(2m+1), the hypergraph of ε\varepsilonε-good 2m2m2m-subsets is K2m+1(2m)K_{2m+1}^{(2m)}K2m+1(2m)​-free. Combined with the known exact eight-qubit construction, this yields the stability plateau Qex,εD(8,4)=56,0≤ε<15. Q_{\mathrm{ex},\varepsilon}^{D}(8,4)=56, \qquad 0\leq\varepsilon<\frac15. Qex,εD​(8,4)=56,0≤ε<51​. For odd systems of 2k+12k+12k+1 qubits, the exact forbidden hypergraph HkH_kHk​ is used to derive explicit finite-error stability radii. In particular, Qex,εD(9,4)≤120Q_{\mathrm{ex},\varepsilon}^{D}(9,4)\leq120Qex,εD​(9,4)≤120 for 0≤ε<1/170\leq\varepsilon<1/170≤ε<1/17. These results turn exact quantum Turán obstructions into quantitative robustness statements and identify intervals on which quantum extremal numbers are stable under imperfect marginal mixedness.
Absolutely maximally entangled (AME) states are pure multipartite states with maximally mixed reductions on every set of at most half the parties. We construct explicit families in local dimension 2q2q2q, twice an odd prime power qqq, using an established coupling of qubits and qqq-level systems. The qqq-level states remove interference between branches, so the qubit states need maximally mixed reductions only on average. This gives AME(9,2q)(9,2q)(9,2q) and AME(11,2q)(11,2q)(11,2q) for every odd prime power q≥5q\geq5q≥5, AME(13,2q)(13,2q)(13,2q) for q≥17q\geq17q≥17, AME(15,2q)(15,2q)(15,2q) for q≥9q\geq9q≥9, and AME(17,2q)(17,2q)(17,2q) for q≥67q\geq67q≥67. For every odd N≥3N\geq3N≥3, the same coupling gives AME(N,2q)(N,2q)(N,2q) under the sufficient condition q≥max⁡{N,2N−2}q\geq\max\{N,2^{N-2}\}q≥max{N,2N−2} on the odd prime power qqq. Within this coupling, the smallest such qqq must grow exponentially with NNN. Combined with known inputs and products, these families complete the nine-party classification: for integer d≥2d\geq2d≥2, an AME(9,d)(9,d)(9,d) state exists if and only if d≥3d\geq3d≥3. Choosing the branch pairing also gives AME(15,2p)(15,2p)(15,2p) states with a cyclic shift of the parties compensated by local unitaries, for every prime p≡1(mod15)p\equiv1\pmod{15}p≡1(mod15).
We investigate multipartite entanglement for composite quantum systems in a pure state. Using the generalized Bloch representation for nnn-qubit states, we express the condition that all kkk-qubit reductions of the whole system are maximally mixed, reflecting maximum bipartite entanglement across all kkk vs $n\ensuremath{-}k$ bipartitions. As a special case, we examine the class of balanced pure states, which are constructed from a subset of the Pauli group Pn{\mathcal{P}}_{n}Pn​ that is isomorphic to Z2n{\mathbb{Z}}_{2}^{n}Z2n​. This makes a connection with the theory of quantum error-correcting codes and provides bounds on the largest allowed kkk for fixed nnn. In particular, the ratio k/nk/nk/n can be lower and upper bounded in the asymptotic regime, implying that there must exist multipartite entangled states with at least $k=\ensuremath{\lfloor}0.189\phantom{\rule{0.16em}{0ex}}n\ensuremath{\rfloor}$ when $n\ensuremath{\rightarrow}\ensuremath{\infty}$. We also analyze symmetric states as another natural class of states with high multipartite entanglement and prove that, surprisingly, they cannot have all maximally mixed kkk-qubit reductions with k>1k>1k>1. Thus, measured through bipartite entanglement across all bipartitions, symmetric states cannot exhibit large entanglement. However, we show that the permutation symmetry only constrains some components of the generalized Bloch vector, so that very specific patterns in this vector may be allowed even though k>1k>1k>1 is forbidden. This is illustrated numerically for a few symmetric states that maximize geometric entanglement, revealing some interesting structures.
Absolutely Maximally Entangled States of 2q2q2q Parties in Every Odd Prime-Power Dimension qqq
Absolutely maximally entangled (AME) states represent an extreme form of multipartite entanglement: every reduced system containing at most half of the parties is maximally mixed. These states provide perfect tensors and optimal quantum error-correcting codes, yet their existence is known only in restricted parameter regimes. For every odd prime power q=pe≥3q=p^e\ge3q=pe≥3, we construct a stabilizer AME(2q,q)\mathrm{AME}(2q,q)AME(2q,q) state whose normalized one-party projection yields a stabilizer AME(2q−1,q)\mathrm{AME}(2q-1,q)AME(2q−1,q) state. A closed-form q×qq\times qq×q bordered-circulant matrix AqA_qAq​ over Fq2\mathbb{F}_{q^2}Fq2​ generates a Hermitian self-dual maximum distance separable (MDS) code [2q,q,q+1]q2[2q,q,q+1]_{q^2}[2q,q,q+1]q2​, which lies outside the extended Reed-Solomon classes. In suitable bases, the amplitude tensors define normalized qqq-unitary complex Hadamard matrices of order qqq^qqq with pppth-root phases. Additional constructions yield AME(q+3,q)\mathrm{AME}(q+3,q)AME(q+3,q) states and families at intermediate particle numbers through explicit rescalings of selected submatrices. We also provide nine explicit parent matrices and the corresponding one-party projections.
Abstract A pure multipartite quantum state is called absolutely maximally entangled (AME), if all reductions obtained by tracing out at least half of its parties are maximally mixed. Maximal entanglement is then present across every bipartition. The existence of such states is in many cases unclear. With the help of the weight enumerator machinery known from quantum error correction and the shadow inequalities, we obtain new bounds on the existence of AME states in dimensions larger than two. To complete the treatment on the weight enumerator machinery, the quantum MacWilliams identity is derived in the Bloch representation. Finally, we consider AME states whose subsystems have different local dimensions, and present an example for a 2 × 3 × 3 × 3 system that shows maximal entanglement across every bipartition.
Pure multiparticle quantum states are called absolutely maximally entangled if all reduced states obtained by tracing out at least half of the particles are maximally mixed. We provide a method to characterize these states for a general multiparticle system. With that, we prove that a seven-qubit state whose three-body marginals are all maximally mixed, or equivalently, a pure ((7,1,4))_{2} quantum error correcting code, does not exist. Furthermore, we obtain an upper limit on the possible number of maximally mixed three-body marginals and identify the state saturating the bound. This solves the seven-particle problem as the last open case concerning maximally entangled states of qubits.