Particle-Preserving Commutant

• Particle-Preserving Commutant is a higher-order invariant algebra defined by the particle-number-conserving subgroup of fermionic Gaussian unitaries on replicated Fock space.
• It is generated by generalized copy-hopping operators that realize a representation of 𝔲(t) through Howe duality, enabling a clear block decomposition into U(t)-irreps.
• The algebra’s structure is explicitly quantified via dimension formulas and a Gelfand–Tsetlin basis construction, providing critical tools for analyzing fermionic correlations and resource quantification.

Searching arXiv for the cited paper and related commutant context. arXiv search query: (Braccia et al., 19 Mar 2026) In fermionic representation theory, the particle-preserving commutant is the $t$-th order commutant associated with the particle-number-conserving subgroup of fermionic Gaussian unitaries on $n$ fermionic modes. For the particle-preserving (PP) subgroup, it is the algebra

$$C^{\rm PP}_{t,n} = \left\{ M\in \operatorname{End}(H^{\otimes t}) \;:\; [M,R(U)^{\otimes t}]=0\ \forall\,U\in U(n) \right\},$$

equivalently the algebra of operators on $H^{\otimes t}$ invariant under the diagonal $U(n)$ action. This object governs Haar averages over the PP Gaussian group and therefore enters fermionic randomized protocols, invariant theory, and resource quantification. Its defining structural feature is that it is a Howe-dual commutant for the pair $(U(n),U(t))$, generated by generalized copy-hopping operators. It is distinct from the commutant of the full fermionic Gaussian group, which is generated by generalized quadratic Majorana bilinears together with parity (Braccia et al., 19 Mar 2026).

1. Algebraic setting

The PP commutant is formulated on the $n$-mode fermionic Fock space

$$F_n \cong \Lambda(\mathbb C^n)=\bigoplus_{r=0}^n \Lambda^r(\mathbb C^n),$$

which decomposes into fixed-particle-number sectors

$H_r \cong \Lambda^r(\mathbb C^n).$

The PP subgroup is the subgroup of fermionic Gaussian unitaries that preserve particle number, namely the image of $U(n)$ under second quantization. Its action on creation and annihilation operators is

$n$0

This formulation makes the PP commutant a replicated invariant algebra: it consists of all operators on $n$1 tensor copies that are invisible to the diagonal second-quantized $n$2 action. In this sense, it is the natural invariant algebra for number-conserving fermionic Gaussian dynamics. The construction also fixes the precise meaning of “particle-preserving”: the symmetry group is not the full fermionic Gaussian group, but its particle-number-conserving subgroup.

2. Copy-hopping generators

The main structural theorem identifies the PP commutant with an algebra generated by generalized copy-hopping operators. These are built from dressed fermionic operators

$n$3

$n$4

where

$n$5

is the parity operator on the $n$6-th copy. The dressing ensures canonical anticommutation across different copies: $n$7

The generalized copy-hopping operators are

$n$8

They commute with the diagonal $n$9 action and satisfy

$$C^{\rm PP}_{t,n} = \left\{ M\in \operatorname{End}(H^{\otimes t}) \;:\; [M,R(U)^{\otimes t}]=0\ \forall\,U\in U(n) \right\},$$0

so they furnish a representation of $$C^{\rm PP}_{t,n} = \left\{ M\in \operatorname{End}(H^{\otimes t}) \;:\; [M,R(U)^{\otimes t}]=0\ \forall\,U\in U(n) \right\},$$1.

The PP commutant is generated by nearest-neighbor copy-hopping operators together with one diagonal generator: $$C^{\rm PP}_{t,n} = \left\{ M\in \operatorname{End}(H^{\otimes t}) \;:\; [M,R(U)^{\otimes t}]=0\ \forall\,U\in U(n) \right\},$$2 This gives the PP case an especially explicit generator-level description. Rather than being presented abstractly as an invariant algebra, it is realized concretely through operators that move particles between tensor copies while preserving total particle number.

3. Howe duality and block decomposition

The generator theorem is equivalent to a skew Howe duality statement. On replicated Fock space one has

$$C^{\rm PP}_{t,n} = \left\{ M\in \operatorname{End}(H^{\otimes t}) \;:\; [M,R(U)^{\otimes t}]=0\ \forall\,U\in U(n) \right\},$$3

and the commuting pair $$C^{\rm PP}_{t,n} = \left\{ M\in \operatorname{End}(H^{\otimes t}) \;:\; [M,R(U)^{\otimes t}]=0\ \forall\,U\in U(n) \right\},$$4 acts multiplicity-freely as

$$C^{\rm PP}_{t,n} = \left\{ M\in \operatorname{End}(H^{\otimes t}) \;:\; [M,R(U)^{\otimes t}]=0\ \forall\,U\in U(n) \right\},$$5

Consequently,

$$C^{\rm PP}_{t,n} = \left\{ M\in \operatorname{End}(H^{\otimes t}) \;:\; [M,R(U)^{\otimes t}]=0\ \forall\,U\in U(n) \right\},$$6

This is the central algebraic description of the particle-preserving commutant. Each partition $$C^{\rm PP}_{t,n} = \left\{ M\in \operatorname{End}(H^{\otimes t}) \;:\; [M,R(U)^{\otimes t}]=0\ \forall\,U\in U(n) \right\},$$7 labels a block, and each block is a full matrix algebra on the corresponding $$C^{\rm PP}_{t,n} = \left\{ M\in \operatorname{End}(H^{\otimes t}) \;:\; [M,R(U)^{\otimes t}]=0\ \forall\,U\in U(n) \right\},$$8-irrep. The multiplicity-free nature of the decomposition is what makes the structure clean: there is no further multiplicity algebra to resolve beyond the $$C^{\rm PP}_{t,n} = \left\{ M\in \operatorname{End}(H^{\otimes t}) \;:\; [M,R(U)^{\otimes t}]=0\ \forall\,U\in U(n) \right\},$$9 representation spaces themselves.

A useful consequence is that the PP commutant is not merely a collection of conserved quantities attached to low-order observables. It is the full higher-order invariant algebra selected by the diagonal $H^{\otimes t}$0 action on $H^{\otimes t}$1 replicas. This provides a precise algebraic counterpart to replicated fermionic states and higher-order fermionic invariants.

4. Dimension formulas and growth

The dimension of the PP commutant admits a closed product formula: $H^{\otimes t}$2

The low-order checks are

$H^{\otimes t}$3

The asymptotic regimes given in the same work are: $H^{\otimes t}$4 and

$H^{\otimes t}$5

These formulas quantify the size of the invariant algebra available to particle-preserving Gaussian dynamics. In particular, the fixed-$H^{\otimes t}$6 scaling shows polynomial growth of degree $H^{\otimes t}$7 in the number of modes, while the fixed-$H^{\otimes t}$8 scaling shows the rapid proliferation of higher-replica invariants as $H^{\otimes t}$9 increases.

5. Constructive Gelfand–Tsetlin bases

Beyond the abstract decomposition, the PP commutant admits an explicit basis construction through a Gelfand–Tsetlin (GT) procedure along the chain

$U(n)$0

A GT pattern for a partition $U(n)$1 is a chain

$U(n)$2

subject to the interlacing inequalities

$U(n)$3

Each GT pattern $U(n)$4 labels a basis vector

$U(n)$5

obtained by applying an ordered product $U(n)$6 of lowering operators to a highest-weight vector. The matrix units spanning the commutant are

$U(n)$7

This construction is important because it converts the representation-theoretic description into an explicit orthonormal operator basis. The result is not only a classification of invariant operators, but also a concrete method for writing them down block by block.

6. Low-order structure

For $U(n)$8, the PP commutant reduces to the algebra of functions of the total number operator

$U(n)$9

The spectral projectors are

$(U(n),U(t))$0

An equivalent basis is furnished by symmetric $(U(n),U(t))$1-string operators

$(U(n),U(t))$2

which are related to $(U(n),U(t))$3 through Krawtchouk polynomials: $(U(n),U(t))$4

For $(U(n),U(t))$5, the algebra is essentially $(U(n),U(t))$6, with generators

$(U(n),U(t))$7

It is convenient to introduce

$(U(n),U(t))$8

The $(U(n),U(t))$9 irreps are labeled by $n$0, with spin

$n$1

In this case the GT labels admit the standard spin-multiplet interpretation. The basis elements are built from highest-weight “copy-singlet” and “copy-polarized” states, and the GT basis vectors are realized by moving particles between copies with $n$2. The paper explicitly constructs the $n$3-, $n$4-, and $n$5-particle blocks. These examples show that the simplification to number-operator functions is specific to $n$6; already at $n$7, the commutant is organized by nontrivial copy-space representation theory.

7. Invariants, correlations, and broader significance

Because $n$8-th order commutants are exactly the objects that appear in Haar averages over the group, the PP commutant provides an exact algebraic handle on replicated fermionic observables, randomized protocols, and invariant polynomials (Braccia et al., 19 Mar 2026). In the PP setting, the copy-hopping generators $n$9 define a hierarchy of invariants refining standard diagnostics such as one-body purities and generalized Plücker constraints. The same framework also clarifies the structure of replicated fermionic states and connects naturally to measures of fermionic correlations, generalized Plücker-type constraints, and the stabilizer entropy of fermionic Gaussian states.

At $$F_n \cong \Lambda(\mathbb C^n)=\bigoplus_{r=0}^n \Lambda^r(\mathbb C^n),$$0, the paper relates overlaps with commutant elements to purities of reduced density matrices through

$$F_n \cong \Lambda(\mathbb C^n)=\bigoplus_{r=0}^n \Lambda^r(\mathbb C^n),$$1

This makes the commutant a concrete tool for quantifying how non-free a fermionic state is under PP Gaussian dynamics. A plausible implication is that the invariant algebra is not merely classificatory; it provides operational observables for resource quantification.

The terminology also sits within a wider commutant-algebra program. In related analyses of number-conserving fermionic Hamiltonian families, commutants are used to preserve fixed-$$F_n \cong \Lambda(\mathbb C^n)=\bigoplus_{r=0}^n \Lambda^r(\mathbb C^n),$$2 sectors and more refined spin or pseudospin decompositions (Moudgalya et al., 2022). In the PP fermionic Gaussian setting, however, the notion is more specific: it is the higher-order diagonal $$F_n \cong \Lambda(\mathbb C^n)=\bigoplus_{r=0}^n \Lambda^r(\mathbb C^n),$$3-invariant algebra on replicated Fock space, with generators, dimensions, and bases determined explicitly by skew Howe duality.