Clifford Entropy in Quantum Resource Theory

• Clifford entropy is a measure that defines the non-Cliffordness of a unitary by quantifying deviations from ideal Clifford operations, with applications in quantum error correction and classical simulation.
• It establishes a direct mapping to stabilizer entropy for Choi states, thereby quantifying the cost of magic and guiding resource assessments in quantum circuit design.
• Analytical bounds, subadditivity, and concentration effects of Clifford entropy offer key insights into circuit complexity and efficient simulation strategies in quantum architectures.

Clifford entropy is a resource-theoretic measure introduced to quantify how far an arbitrary unitary operator is from being a Clifford unitary, generalizing the stabilizer entropy for quantum states. The Clifford group plays a foundational role in quantum information as the normalizer of the Weyl–Heisenberg (Pauli) group, with Clifford operations and states integral to classical simulation, quantum error correction, and benchmarking quantum resources beyond stabilizer circuits. Clifford entropy provides a systematic framework for comparing gate sets, quantifying the resource cost of nonstabilizerness (commonly called “magic”), and bounding circuit complexity in architectures augmented with non-Clifford elements (Cuffaro et al., 28 Dec 2025).

1. Formal Definition and Basic Properties

The α-Clifford entropy $S_C^{(\alpha)}(U)$ is defined for any unitary %%%%1%%%%, with $d=2^n$ for $n$ qubits. Let $\{D_a\}_{a\in \mathbb Z_d^2}$ be a fixed Weyl–Heisenberg frame on the Hilbert space $\mathcal H_d$. The characteristic matrix,

$$C_{ab}(U) = \frac{1}{d} \mathrm{Tr}[D_a^\dagger U D_b U^\dagger],$$

is used to construct the bistochastic matrix $\mathfrak D_{ab}(U) = |C_{ab}(U)|^2$. Any order-$\alpha$ Tsallis entropy can then be evaluated on rows and columns of $\mathfrak D(U)$. The symmetric choice,

$$S_C^{(\alpha)}(U) = H_\alpha(U) = \frac{1}{\alpha-1} \left( 1 - \frac{1}{d^2} \sum_{a,b} [\mathfrak D_{ab}(U)]^\alpha \right),$$

specializes for $\alpha=2$ to the computationally useful quadratic form,

$$H_2(U) = 1 - \frac{1}{d^6} \sum_{a,b} |\mathrm{Tr}[D_a^\dagger U D_b U^\dagger]|^4.$$

The Clifford entropy satisfies several fundamental properties:

• Faithfulness: $H_\alpha(U) = 0$ if and only if $U$ is Clifford. For Clifford unitaries, $C_{ab}(U)$ is monomial (up to phase) and thus $\mathfrak D_{ab}(U)$ is a permutation matrix.
• Clifford-invariance: $H_\alpha(C_1 U C_2) = H_\alpha(U)$ for any Clifford $C_1, C_2$.
• Subadditivity: $$H_\alpha(U\otimes V) \leq H_\alpha(U) + H_\alpha(V)$$, with precise correction terms expressed for general $\alpha$ (Cuffaro et al., 28 Dec 2025).

2. Relation to Stabilizer Entropy and Choi States

Clifford entropy for a unitary $U$ admits an exact correspondence with the order-$\alpha$ stabilizer entropy $M_\alpha$ of its channel Choi state $\rho_\Phi$ associated to conjugation $X\mapsto U X U^\dagger$: $$\chi_{ab}(\rho_\Phi) = \frac{1}{d^2} |\mathrm{Tr}[D_a^\dagger\,U D_b U^\dagger]|^2 = \frac{1}{d^2} \mathfrak D_{ab}(U).$$ The stabilizer entropy is

$$M_\alpha(\rho_\Phi) = \frac{1}{1-\alpha} \ln\left( \sum_{a,b} \chi_{ab}(\rho_\Phi)^\alpha \right) - 2\ln d,$$

and the relationship

$$H_\alpha(U) = \frac{1}{\alpha-1}\left[ 1 - \exp\left( -(\alpha-1) M_\alpha(\rho_\Phi) \right) \right ]$$

demonstrates the direct mapping to the state-based resource theory of magic. For quantum channels and CPTP maps, extending this correspondence constitutes an active research direction (Cuffaro et al., 28 Dec 2025).

3. Analytical Bounds, Statistical Behavior, and Non-tightness

Upper bounds on Clifford entropy are inherited from known extremal bounds on stabilizer entropy for pure states. For $\alpha=2$,

$H_2(U) \leq 1 - \frac{2}{d^2 + 1}.$

However, this bound is not tight: achieving it would require a Choi fiducial that is both a symmetric informationally complete (SIC) state and maximally entangled, which is forbidden by the structure of bipartite Weyl–Heisenberg-covariant SICs (Cuffaro et al., 28 Dec 2025).

For small dimensions ($d=2,\ldots,9$), direct numerical optimization yields: $$\max_U H_2(U) \approx \{0.50,\,0.79,\,0.89,\,0.93,\,0.95,\,0.96,\,0.97,\,0.98\}.$$ The Haar average over unitaries, obtained via Weingarten calculus, approaches unity as $d\to\infty$: $$\mathbb E_{U\sim\mathrm{Haar}}[\,H_2(U)\,] = 1 - O(d^{-2}),$$ with exact formulas given for general $d$ (Cuffaro et al., 28 Dec 2025). Empirically, this average rapidly tracks the maximum as dimension increases, indicating the concentration of Clifford entropy for typical (Haar-random) unitaries near the universal bound.

4. Subadditivity, Concentration, and Circuit Complexity Bounds

Subadditivity of Clifford entropy under composition,

$G(U,V) = H_2(UV) - H_2(U) - H_2(V),$

is a high-probability event for Haar-random unitaries: $\Pr_{U,V}[G(U,V)\ge 0]$ decays exponentially in $d^2$. The function $G$ is Lipschitz with constant $O(d^{-1/2})$.

Operationally, for circuits generated by Clifford and a fixed magic gate $T$, the minimal $T$-count required to realize $U$ is bounded by

$t(U) \geq \frac{H_2(U)}{H_2(T)}$

except with exponentially small failure probability. For Haar-random $U$, $H_2(U)\approx1$ yields $t(U)=\Omega(1/H_2(T))$, quantifying the circuit depth cost in terms of Clifford entropy. Numerical experiments confirm vanishing rate ($<10^{-6}$) of subadditivity violation even at modest $d$ (Cuffaro et al., 28 Dec 2025).

5. Clifford Entropy Polytopes, Entanglement Quantization, and Cones

In the context of state-based entanglement, the stabilizer (Clifford) entropy cone $c^S_n$ captures the set of entropy vectors $\vec S$ realizable by stabilizer (Clifford) states. For $n$-qubit states, all membership constraints in $c^S_n$ include subadditivity, strong subadditivity, and the Ingleton/related Kinser inequalities, but not e.g., monogamy of mutual information (MMI), which can be violated by Clifford states (Schnitzer, 2020).

In SU(N)$_1$ Chern–Simons theory with $N$ odd prime, it is established that $c^t_n(\mathrm{SU}(N)_1) = c^S_n$; the set of topologically constructible entropies coincides precisely with the Clifford entropy cone. For SU(N)$_K$ with $K\ge2$, the topological cone is strictly contained in $c^S_n$, as not all generalized Pauli (and thus Clifford) operations are synthesizable (Schnitzer, 2020).

For concrete qubit counts, the set of entanglement values is highly discrete within the Clifford group. For four-qubit Clifford orbits, bipartite von Neumann entropies take only the values $\{0,\,2/3,\,1,\,4/3,\,5/3\}$, reflecting the coarse-grained complexity and efficient simulatability of Clifford circuits (Latour et al., 2020).

6. Graph-Theoretic Structure and Entropic Diversity in Clifford Circuits

Contracted graphs formalize the action of Clifford circuits on entropy vectors. Each vertex corresponds to an equivalence class of quantum states sharing the same entropy pattern, defined by double cosets of the Clifford group modulo the stabilizer subgroup and the entropy-preserving subgroup of local Cliffords. The number of vertices in the contracted graph provides an upper bound on the entropic diversity accessible from a given initial state using Clifford circuits.

Explicitly, for $n$-qubit Clifford circuits, the number of reachable entropy vectors cannot exceed

$$\frac{|\mathcal C_n|}{24^n} = 2^{n^2 - n} \prod_{j=1}^n (4^j - 1),$$

where $|\mathcal C_n|$ is the order of the Clifford group (Keeler et al., 2023). For small $n$, these counts match explicit contraction enumerations. The diameter of the contracted graph gives a measure of circuit depth needed to traverse entropic patterns, with implications for quantum error correction, state synthesis, and connections to holographic duality constraints (Keeler et al., 2023).

7. Spreading and Transport of Magic and Clifford Entropy

Stabilizer Rényi entropy (SRE), effectively the Clifford entropy for states, quantifies the degree of nonstabilizerness ("magic"). In brickwork random Clifford circuits, SRE for single-qubit reduced density matrices, after initializing with local magic, exhibits emergent transport phenomena: profiles broaden diffusively within a strictly causal (ballistic) light cone. When restricting to certain Clifford-only architectures (e.g., CNOT plus local H, S), superdiffusive broadening is observed. Global SRE is preserved under Clifford evolution, but its local redistribution is governed by hydrodynamic-like equations without actual conservation laws (Maity et al., 11 Nov 2025).

These results establish that Clifford entropy and its associated measures are essential in capturing the operational magic/complexity in Clifford+T circuits, circuit synthesis, entanglement dynamics, and topological quantum computing. The resource-theoretic and geometric perspectives provided by Clifford entropy unify several contemporary threads in quantum information science, from classical simulability to bounds on magic state injection and connections to topological field theory.

Key research directions include generalizations to arbitrary CPTP channels, monotonicity properties under stabilizer superchannels, statistical properties in generic circuits, and deeper connections to phase-space scrambling and out-of-time-order correlator (OTOC) diagnostics (Cuffaro et al., 28 Dec 2025).