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:2610.20988

Classification of quantum states through conserved operators

Federico Petrovich, R. Rossignoli, M. Cerezo

quant-phSubmitted Oct 14, 2026• Announced Oct 16, 2026• 1 citation
arXivHTMLPDF

Abstract

Determining whether quantum states can be interconverted by a prescribed set of transformations is a fundamental problem in quantum information. Here we develop a general classification framework based on conserved operators (COs), i.e., not necessarily hermitian matrices having the state of interest as an eigenstate. We first show that COs belonging to the algebra generating the allowed transformations can be obtained from the kernel of a quantum covariance matrix. Their number determines the dimension of the accessible orbit, while their algebraic structure and eigenvalues provide further constraints on state convertibility. We establish general bounds on the number of COs and prove that highest-weight states, like separable states in the case of $n$ qudits, attain the maximum. We also show how these operators generate parent Hamiltonians having the characterized state as an exact ground state. We then also consider positive zero-eigenvalue COs, which encode forbidden directions of reduced density matrices and incorporate rank-based constraints into the same framework. Applying this approach to multiqubit SLOCC, we obtain complementary classifications from algebraic and rank COs. While for $n\leq 3$ qubits both lead to the usual classes, for $n=4$ qubits they distinguish 17 and 20 types separately, and 29 when combined. The combined classification yields 287 fixed types and $4$ continuous families for $n=5$ qubits, and 4926 fixed types and 147 continuous families for $n=6$ qubits, providing, to our knowledge, the most refined exhaustive classification of SLOCC-convertibility classes based on such data.

AI Summary

The paper studies necessary conditions for pure-state interconvertibility under invertible transformations generated by a complex reductive Lie algebra. It defines conserved operators (COs) by $Q|\psi\rangle=\lambda|\psi\rangle$, allowing non-Hermitian $Q$. Within the transformation algebra, these are projective-stabilizer generators, obtained as the kernel of an unsymmetrized quantum covariance matrix. Their dimension fixes the orbit dimension: $\dim G\cdot[\psi]=\dim\mathfrak g-\dim\mathfrak p_\psi$. Algebra conjugacy and eigenvalue characters provide further constraints. The general bound is $\max\{0,\dim\mathfrak g+1-\dim\mathcal H\}\leq\dim\mathfrak p_\psi\leq\dim\mathfrak p_{\rm hw}$, with the upper bound attained precisely on the highest-weight orbit—product states for local qudit transformations. The proof uses rank–nullity and the Borel fixed-point theorem. Positive zero-eigenvalue COs additionally encode reduced-density-matrix kernels and transform by congruence. COs also yield parent Hamiltonians through positive quadratic combinations. For qubit SLOCC, combining algebraic COs with two-qubit reduced-state kernels gives 29 data types at four qubits, versus 17 algebraic-only and 20 kernel-only types. The combined classification gives 287 fixed types and four continuous families at five qubits, and 4,926 fixed types and 147 continuous families at six. Enumeration uses adapted Lie-algebra forms, simultaneous local equivalence, and polynomial rank conditions to test exact realizability; four-qubit results are cross-checked against an existing stabilizer classification. All classification inputs require only one- and two-body correlations. The classification is explicitly coarser than SLOCC equivalence: distinct orbits can share these data, including trivial algebraic stabilizers and full-rank pair reductions. Higher-body and multicopy COs are proposed as refinements.

Little sign of AI-generated textAI use disclosedUnclear noveltyPossible error
The authors acknowledge the use of ChatGPT Astra for the deterministic code that performs the full classification.

Error found: The local-qudit lower bound in Eq. (quditbound) is missing −n, producing contradictions with the paper’s own one-qudit and three-qubit examples. The claim that the lower bound is nonpositive for every four-party system with local dimensions at least two is false: dimensions (2,2,2,10) require at least 29 algebraic conserved operators even under the corrected bound.

Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.

Comments
14+12 pages, 2, figures, 2+1 tables, one supplemental data file with the six qubit classification

Related papers

Method for classifying multiqubit states via the rank of the coefficient matrix and its application to four-qubit states

Xiangrong Li, Dafa Li

We construct coefficient matrices of size ${2}^{\ensuremath{\ell}}$ by ${2}^{n\ensuremath{-}\ensuremath{\ell}}$ associated with pure $n$-qubit states and prove the invariance of the ranks of the coefficient matrices under stochastic local operations and classical communication (SLOCC). The ranks give rise to a simple way of partitioning pure $n$-qubit states into inequivalent families and distinguishing degenerate families from one another under SLOCC. Moreover, the classification scheme via the ranks of coefficient matrices can be combined with other schemes to build a more refined classification scheme. To exemplify we classify the nine families of four qubits introduced by Verstraete et al. [Phys. Rev. A 65, 052112 (2002)] further into inequivalent subfamilies via the ranks of coefficient matrices, and as a result, we find 28 genuinely entangled families and all the degenerate classes can be distinguished up to permutations of the four qubits. We also discuss the completeness of the classification of four qubits into nine families.

arXiv:1201.2229•Physical Review A•Oct 23, 2012•25 citations

Classification of four qubit states and their stabilisers under SLOCC operations

Heiko Dietrich, Willem Adriaan de Graaf, Alessio Marrani C +1 more

Abstract We classify four qubit states under SLOCC operations, that is, we classify the orbits of the group S L ( 2 , C ) 4 on the Hilbert space H 4 = ( C 2 ) ⊗ 4 . We approach the classification by realising this representation as a symmetric space of maximal rank. We first describe general methods for classifying the orbits of such a space. We then apply these methods to obtain the orbits in our special case, resulting in a complete and irredundant classification of S L ( 2 , C ) 4 -orbits on H 4 . It follows that an element of ( C 2 ) ⊗ 4 is conjugate to an element of precisely 87 classes of elements. Each of these classes either consists of one element or of a parameterised family of elements, and the elements in the same class all have equal stabiliser in S L ( 2 , C ) 4 . We also present a complete and irredundant classification of elements and stabilisers up to the action of S y m 4 ⋉ S L ( 2 , C ) 4

arXiv:2111.05488•Journal of Physics A Mathematical and Theoretical•Jan 13, 2022•10 citations

Classifying entanglement by algebraic geometry

Masoud Gharahi

Quantum Entanglement is one of the key manifestations of quantum mechanics that separate the quantum realm from the classical one. Characterization of entanglement as a physical resource for quantum technology became of uppermost importance. While the entanglement of bipartite systems is already well understood, the ultimate goal to cope with the properties of entanglement of multipartite systems is still far from being realized. This paper covers characterization of multipartite entanglement using algebraic-geometric tools. First, we establish an algorithm to classify multipartite entanglement by [Formula: see text]-secant varieties of the Segre variety and [Formula: see text]-multilinear ranks that are invariant under Stochastic Local Operations with Classical Communication (SLOCC). We present a fine-structure classification of multiqubit and tripartite entanglement based on this algorithm. Another fundamental problem in quantum information theory is entanglement transformation that is quite challenging regarding to multipartite systems. It is captivating that the proposed entanglement classification by algebraic geometry can be considered as a reference to study SLOCC and asymptotic SLOCC interconversions among different resources based on tensor rank and border rank, respectively. In this regard, we also introduce a new class of tensors that we call persistent tensors and construct a lower bound for their tensor rank. We further cover SLOCC convertibility of multipartite systems considering several families of persistent tensors.

arXiv:2408.12265•International Journal of Quantum Information•Dec 3, 2023•6 citations

Classification of General n -Qubit States under Stochastic Local Operations and Classical Communication in Terms of the Rank of Coefficient Matrix

Xiangrong Li, Dafa Li

We solve the entanglement classification under stochastic local operations and classical communication (SLOCC) for general n-qubit states. For two arbitrary pure n-qubit states connected via local operations, we establish an equation between the two coefficient matrices associated with the states. The rank of the coefficient matrix is preserved under SLOCC and gives rise to a simple way of partitioning all the pure states of n qubits into different families of entanglement classes, as exemplified here. When applied to the symmetric states, this approach reveals that all the Dicke states |ℓ,n> with ℓ=1,…,[n/2] are inequivalent under SLOCC.

arXiv:1106.6105•Physical Review Letters•May 3, 2012•53 citations

Unitary States and Gates of Three-Qubit Quantum Systems

Mars B. Gabbassov, Tolybay Z. Kuanov, Berik I. Tuleuov +1 more

Pure states of three-qubit quantum systems are represented by eight-dimensional complex vectors, whereas their unitary transformations are described by $8 \times 8$ complex matrices. Motivated by this correspondence, we extend our algebraic model for two-qubit systems to eight-dimensional spaces endowed with commutative and associative multiplication. To each eight-dimensional number we assign a matrix and prove that the resulting correspondence is a homomorphic bijection. We derive conditions under which the associated matrix is unitary and hence represents a three-qubit quantum gate. This leads to a specific algebraic notion of a unitary state: a normalized state whose multiplication matrix is unitary. We construct 16,384 basis spaces; within each fixed space, multiplication generates a commuting family of quantum gates. For any two unitary states in the class considered, we give an explicit transition-gate construction within one of these algebraic families. For real components, the unitary states in a representative basis space form a union of thirty-two four-dimensional manifolds in \(\mathbb{R}^{8}\). The construction therefore yields new structured classes of three-qubit unitary operators and invariant state sets, while its extension to larger multiqubit systems remains open.

•Oct 20, 2026•0 citations

Probability-Preserving Reparametrizations of Quantum States and Effects

Kaibin Liu

The empirical content of quantum theory is entirely encoded in the Born probability pairing between states and effects. In finite dimension, we study the following structural problem: which reparametrizations of states and effects preserve all Born probabilities? We consider bijective affine maps on the state space and effect space that leave the Born pairing invariant for all state–effect pairs, and provide a complete classification of all such transformations. We prove that every probability-preserving reparametrization is necessarily induced by a Jordan *-automorphism of the matrix algebra B(H). In finite dimension, this restricts the transformation to the form A→UAU† or A→UA⊤U†, for some unitary operator U. In particular, equivalence classes of states under probability-preserving reparametrizations coincide with unitary orbits and are completely characterized by the spectrum of the density operator. The result shows that unitary and antiunitary conjugations exhaust all probability-preserving affine reparametrizations of the state–effect framework.

•Mar 22, 2026•1 citation
Determining whether quantum states can be interconverted by a prescribed set of transformations is a fundamental problem in quantum information. Here we develop a general classification framework based on conserved operators (COs), i.e., not necessarily hermitian matrices having the state of interest as an eigenstate. We first show that COs belonging to the algebra generating the allowed transformations can be obtained from the kernel of a quantum covariance matrix. Their number determines the dimension of the accessible orbit, while their algebraic structure and eigenvalues provide further constraints on state convertibility. We establish general bounds on the number of COs and prove that highest-weight states, like separable states in the case of nnn qudits, attain the maximum. We also show how these operators generate parent Hamiltonians having the characterized state as an exact ground state. We then also consider positive zero-eigenvalue COs, which encode forbidden directions of reduced density matrices and incorporate rank-based constraints into the same framework. Applying this approach to multiqubit SLOCC, we obtain complementary classifications from algebraic and rank COs. While for n≤3n\leq 3n≤3 qubits both lead to the usual classes, for n=4n=4n=4 qubits they distinguish 17 and 20 types separately, and 29 when combined. The combined classification yields 287 fixed types and 444 continuous families for n=5n=5n=5 qubits, and 4926 fixed types and 147 continuous families for n=6n=6n=6 qubits, providing, to our knowledge, the most refined exhaustive classification of SLOCC-convertibility classes based on such data.
The paper studies necessary conditions for pure-state interconvertibility under invertible transformations generated by a complex reductive Lie algebra. It defines conserved operators (COs) by Q∣ψ⟩=λ∣ψ⟩Q|\psi\rangle=\lambda|\psi\rangleQ∣ψ⟩=λ∣ψ⟩, allowing non-Hermitian QQQ. Within the transformation algebra, these are projective-stabilizer generators, obtained as the kernel of an unsymmetrized quantum covariance matrix. Their dimension fixes the orbit dimension: dim⁡G⋅[ψ]=dim⁡g−dim⁡pψ\dim G\cdot[\psi]=\dim\mathfrak g-\dim\mathfrak p_\psidimG⋅[ψ]=dimg−dimpψ​. Algebra conjugacy and eigenvalue characters provide further constraints.
The general bound is max⁡{0,dim⁡g+1−dim⁡H}≤dim⁡pψ≤dim⁡phw\max\{0,\dim\mathfrak g+1-\dim\mathcal H\}\leq\dim\mathfrak p_\psi\leq\dim\mathfrak p_{\rm hw}max{0,dimg+1−dimH}≤dimpψ​≤dimphw​, with the upper bound attained precisely on the highest-weight orbit—product states for local qudit transformations. The proof uses rank–nullity and the Borel fixed-point theorem. Positive zero-eigenvalue COs additionally encode reduced-density-matrix kernels and transform by congruence. COs also yield parent Hamiltonians through positive quadratic combinations.
For qubit SLOCC, combining algebraic COs with two-qubit reduced-state kernels gives 29 data types at four qubits, versus 17 algebraic-only and 20 kernel-only types. The combined classification gives 287 fixed types and four continuous families at five qubits, and 4,926 fixed types and 147 continuous families at six. Enumeration uses adapted Lie-algebra forms, simultaneous local equivalence, and polynomial rank conditions to test exact realizability; four-qubit results are cross-checked against an existing stabilizer classification.
All classification inputs require only one- and two-body correlations. The classification is explicitly coarser than SLOCC equivalence: distinct orbits can share these data, including trivial algebraic stabilizers and full-rank pair reductions. Higher-body and multicopy COs are proposed as refinements.
We construct coefficient matrices of size ${2}^{\ensuremath{\ell}}$ by ${2}^{n\ensuremath{-}\ensuremath{\ell}}$ associated with pure nnn-qubit states and prove the invariance of the ranks of the coefficient matrices under stochastic local operations and classical communication (SLOCC). The ranks give rise to a simple way of partitioning pure nnn-qubit states into inequivalent families and distinguishing degenerate families from one another under SLOCC. Moreover, the classification scheme via the ranks of coefficient matrices can be combined with other schemes to build a more refined classification scheme. To exemplify we classify the nine families of four qubits introduced by Verstraete et al. [Phys. Rev. A 65, 052112 (2002)] further into inequivalent subfamilies via the ranks of coefficient matrices, and as a result, we find 28 genuinely entangled families and all the degenerate classes can be distinguished up to permutations of the four qubits. We also discuss the completeness of the classification of four qubits into nine families.
Abstract We classify four qubit states under SLOCC operations, that is, we classify the orbits of the group S L ( 2 , C ) 4 on the Hilbert space H 4 = ( C 2 ) ⊗ 4 . We approach the classification by realising this representation as a symmetric space of maximal rank. We first describe general methods for classifying the orbits of such a space. We then apply these methods to obtain the orbits in our special case, resulting in a complete and irredundant classification of S L ( 2 , C ) 4 -orbits on H 4 . It follows that an element of ( C 2 ) ⊗ 4 is conjugate to an element of precisely 87 classes of elements. Each of these classes either consists of one element or of a parameterised family of elements, and the elements in the same class all have equal stabiliser in S L ( 2 , C ) 4 . We also present a complete and irredundant classification of elements and stabilisers up to the action of S y m 4 ⋉ S L ( 2 , C ) 4
Quantum Entanglement is one of the key manifestations of quantum mechanics that separate the quantum realm from the classical one. Characterization of entanglement as a physical resource for quantum technology became of uppermost importance. While the entanglement of bipartite systems is already well understood, the ultimate goal to cope with the properties of entanglement of multipartite systems is still far from being realized. This paper covers characterization of multipartite entanglement using algebraic-geometric tools. First, we establish an algorithm to classify multipartite entanglement by [Formula: see text]-secant varieties of the Segre variety and [Formula: see text]-multilinear ranks that are invariant under Stochastic Local Operations with Classical Communication (SLOCC). We present a fine-structure classification of multiqubit and tripartite entanglement based on this algorithm. Another fundamental problem in quantum information theory is entanglement transformation that is quite challenging regarding to multipartite systems. It is captivating that the proposed entanglement classification by algebraic geometry can be considered as a reference to study SLOCC and asymptotic SLOCC interconversions among different resources based on tensor rank and border rank, respectively. In this regard, we also introduce a new class of tensors that we call persistent tensors and construct a lower bound for their tensor rank. We further cover SLOCC convertibility of multipartite systems considering several families of persistent tensors.
We solve the entanglement classification under stochastic local operations and classical communication (SLOCC) for general n-qubit states. For two arbitrary pure n-qubit states connected via local operations, we establish an equation between the two coefficient matrices associated with the states. The rank of the coefficient matrix is preserved under SLOCC and gives rise to a simple way of partitioning all the pure states of n qubits into different families of entanglement classes, as exemplified here. When applied to the symmetric states, this approach reveals that all the Dicke states |ℓ,n> with ℓ=1,…,[n/2] are inequivalent under SLOCC.
Pure states of three-qubit quantum systems are represented by eight-dimensional complex vectors, whereas their unitary transformations are described by 8×88 \times 88×8 complex matrices. Motivated by this correspondence, we extend our algebraic model for two-qubit systems to eight-dimensional spaces endowed with commutative and associative multiplication. To each eight-dimensional number we assign a matrix and prove that the resulting correspondence is a homomorphic bijection. We derive conditions under which the associated matrix is unitary and hence represents a three-qubit quantum gate. This leads to a specific algebraic notion of a unitary state: a normalized state whose multiplication matrix is unitary. We construct 16,384 basis spaces; within each fixed space, multiplication generates a commuting family of quantum gates. For any two unitary states in the class considered, we give an explicit transition-gate construction within one of these algebraic families. For real components, the unitary states in a representative basis space form a union of thirty-two four-dimensional manifolds in R8\mathbb{R}^{8}R8. The construction therefore yields new structured classes of three-qubit unitary operators and invariant state sets, while its extension to larger multiqubit systems remains open.
The empirical content of quantum theory is entirely encoded in the Born probability pairing between states and effects. In finite dimension, we study the following structural problem: which reparametrizations of states and effects preserve all Born probabilities? We consider bijective affine maps on the state space and effect space that leave the Born pairing invariant for all state–effect pairs, and provide a complete classification of all such transformations. We prove that every probability-preserving reparametrization is necessarily induced by a Jordan *-automorphism of the matrix algebra B(H). In finite dimension, this restricts the transformation to the form A→UAU† or A→UA⊤U†, for some unitary operator U. In particular, equivalence classes of states under probability-preserving reparametrizations coincide with unitary orbits and are completely characterized by the spectrum of the density operator. The result shows that unitary and antiunitary conjugations exhaust all probability-preserving affine reparametrizations of the state–effect framework.