Absolutely Wigner-Positive States

• Absolutely Wigner-positive states are quantum states defined by universally nonnegative Wigner functions in a chosen phase-space representation.
• They connect representation-dependent phenomena—such as Hudson's theorem for continuous variables and discrete qudit criteria—to efficient classical simulation and magic resource theories.
• Their study leverages convex geometry, spectral analysis, and simulation frameworks, with ongoing challenges in mixed state characterization and relativistic extensions.

Absolutely Wigner-positive states are quantum states whose Wigner function is non-negative at every phase-space point in the representation being used. In continuous-variable and related settings, this commonly means pointwise positivity of the standard scalar Wigner function. In odd-prime-dimensional qudit theory, the term is used more stringently for states whose discrete Wigner function remains nonnegative after conjugation by every unitary, so that the entire unitary orbit stays inside the Wigner-positive region. The concept is therefore inseparable from the chosen phase-space formalism—continuous, discrete, spin, or relativistic—and its significance ranges from Hudson-type structure theorems to resource theories of magic and classical simulability (Mari et al., 2012, Zurel et al., 25 Feb 2026, Campos et al., 2014).

1. Definition and representational dependence

For a phase-space representation $W_\rho$, Wigner positivity is the pointwise condition

$W_\rho(r)\ge 0$

for all phase-space points $r$. In the Mari–Eisert framework, this defines the positive-Wigner sector in a unified formalism covering continuous variables and odd-prime discrete systems; a positive Wigner function then behaves as a genuine probability distribution, and positive-Wigner channels act through stochastic kernels on phase space (Mari et al., 2012).

The adjective “absolute” is used in more than one way. In one usage, it refers simply to everywhere nonnegative Wigner functions. In the odd-prime qudit setting, however, the definition is explicitly unitary-orbit based: $$W_{U\rho U^\dagger}(u)\ge 0\quad \forall u,\ \forall U\in U(d^n),$$ equivalently

$$\mathrm{AWP}:=\bigcap_{U\in U(d^n)} U\,WP\,U^\dagger,$$

where $WP$ is the Wigner polytope of Wigner-positive states (Zurel et al., 25 Feb 2026). In spin systems the analogous generalization is “absolutely Wigner-bounded,” where the full unitary orbit is required to satisfy a lower bound $W(\Omega)\ge W_{\min}$, with absolute Wigner positivity recovered as the special case $W_{\min}=0$ (Denis et al., 2023).

Relativistic quantum mechanics introduces a further representational shift. For Dirac spinors, the relevant object is a matrix-valued Wigner function $W(t,x,p)$, and positivity is typically discussed for the scalar component

$$W^0(t,x,p)=\frac14 \operatorname{Tr}[W(t,x,p)\gamma^0].$$

The same phrase “Wigner-positive” therefore refers there to positivity of $W_\rho(r)\ge 0$0, not of the full Wigner matrix (Campos et al., 2014).

2. Continuous-variable theory

In the standard continuous-variable setting, the structural benchmark is Hudson’s theorem: for pure states on $W_\rho(r)\ge 0$1, the Wigner function is everywhere nonnegative if and only if the state is Gaussian. Hence the only pure continuous-variable Wigner-positive states are Gaussian pure states, up to the usual phase-space transformations, while any non-Gaussian pure state necessarily has Wigner negativity (Corney et al., 2014).

Mixed states are much subtler. The set of Wigner-positive mixed states strictly exceeds the convex hull of Gaussian pure states, and its full characterization remains open. A functional-analytic refinement is provided by Feichtinger states, namely mixed states whose components lie in modulation spaces $W_\rho(r)\ge 0$2; for these states the Wigner distribution is absolutely integrable, has the correct marginals, and for $W_\rho(r)\ge 0$3 has well-defined covariance matrix. This identifies a large class in which positivity is mathematically well behaved, but it does not imply positivity by itself (Gosson et al., 2021).

The distinction between exact and approximate positivity is important. For the single-mode anharmonic oscillator, the truncated Wigner method propagates an everywhere positive phase-space density and can reproduce non-Gaussian cumulants while preserving the phase-space purity condition $W_\rho(r)\ge 0$4, but this positive distribution is only an approximation to the exact Wigner function. In the exact theory, once a pure state becomes non-Gaussian, Hudson’s theorem forces Wigner negativity (Corney et al., 2014).

Wigner positivity also supports an entropy theory. For Wigner-positive states, the Wigner entropy is defined as the Shannon differential entropy of the Wigner function,

$W_\rho(r)\ge 0$5

and is invariant under symplectic transformations. It is conjectured to satisfy

$W_\rho(r)\ge 0$6

for single-mode Wigner-positive states in the convention $W_\rho(r)\ge 0$7, with equality at Gaussian pure states; this bound is proved on the subset of passive states (Herstraeten et al., 2021). A more general lower bound valid for positive Wigner functions in arbitrary dimension is

$W_\rho(r)\ge 0$8

while for single-mode Wigner-non-negative states under the minimal constraints $W_\rho(r)\ge 0$9, $r$0, and $r$1, the purity conditions

$r$2

are sufficient to guarantee the conjectured lower bound $r$3 in the convention used there (Dias et al., 2023, Qian et al., 23 Jan 2026).

3. Relativistic Dirac states

The relativistic case sharply departs from the nonrelativistic Hudson picture. For Dirac spinors, the scalar relativistic Wigner function $r$4 has the correct position and momentum marginals and reproduces expectation values of phase-space observables, but mathematically it behaves like the Wigner function of a mixed nonrelativistic state obtained by summing over spinor components. This weakens the purity restriction that underlies Hudson’s theorem (Campos et al., 2014).

The central result is that there exist pure, non-Gaussian Dirac spinors with everywhere nonnegative $r$5. One explicit example is the non-Gaussian spinor

$r$6

whose Wigner function is

$r$7

Other explicit positive examples include multi-peaked superpositions of separated wavepackets whose scalar Wigner functions remain strictly positive because the interference terms vanish in the $r$8-weighted spinor trace (Campos et al., 2014).

These states are not classical in any straightforward sense. The analytic examples are coherent superpositions of positive- and negative-energy components, and their dynamics displays zitterbewegung. Free Dirac evolution can generate negative regions of $r$9 even from positive initial data, and superpositions of separated packets can remain Wigner-positive without the interference fringes that are unavoidable in the nonrelativistic scalar case (Campos et al., 2014).

The status of particle-only states remains open. The authors conjecture that pure positive-energy Dirac spinors with strictly positive scalar Wigner function exist, and they give numerical evidence by projecting out antiparticles directly in phase space; for one parameter choice, the filtered Wigner function is numerically nonnegative up to round-off at the level of $$W_{U\rho U^\dagger}(u)\ge 0\quad \forall u,\ \forall U\in U(d^n),$$0. No closed-form classification analogous to Hudson’s theorem is known in the Dirac setting (Campos et al., 2014).

4. Odd-prime qudits and the formal absolute notion

For $$W_{U\rho U^\dagger}(u)\ge 0\quad \forall u,\ \forall U\in U(d^n),$$1 qudits of odd prime local dimension $$W_{U\rho U^\dagger}(u)\ge 0\quad \forall u,\ \forall U\in U(d^n),$$2, the discrete Wigner function is defined with respect to the Gross/Wootters phase-space point operators $$W_{U\rho U^\dagger}(u)\ge 0\quad \forall u,\ \forall U\in U(d^n),$$3, $$W_{U\rho U^\dagger}(u)\ge 0\quad \forall u,\ \forall U\in U(d^n),$$4, via

$$W_{U\rho U^\dagger}(u)\ge 0\quad \forall u,\ \forall U\in U(d^n),$$5

A state is Wigner-positive when $$W_{U\rho U^\dagger}(u)\ge 0\quad \forall u,\ \forall U\in U(d^n),$$6 for all $$W_{U\rho U^\dagger}(u)\ge 0\quad \forall u,\ \forall U\in U(d^n),$$7, equivalently when it lies in the Wigner polytope $$W_{U\rho U^\dagger}(u)\ge 0\quad \forall u,\ \forall U\in U(d^n),$$8. Absolute Wigner positivity is then the unitary-invariant condition

$$W_{U\rho U^\dagger}(u)\ge 0\quad \forall u,\ \forall U\in U(d^n),$$9

so membership depends only on the spectrum of $$\mathrm{AWP}:=\bigcap_{U\in U(d^n)} U\,WP\,U^\dagger,$$0 (Zurel et al., 25 Feb 2026).

The decisive simplification is spectral. For odd prime $$\mathrm{AWP}:=\bigcap_{U\in U(d^n)} U\,WP\,U^\dagger,$$1 and any $$\mathrm{AWP}:=\bigcap_{U\in U(d^n)} U\,WP\,U^\dagger,$$2,

$$\mathrm{AWP}:=\bigcap_{U\in U(d^n)} U\,WP\,U^\dagger,$$3

where $$\mathrm{AWP}:=\bigcap_{U\in U(d^n)} U\,WP\,U^\dagger,$$4 are the eigenvalues in nonincreasing order. Equivalently, AWP spectra form a polytope in the simplex whose vertices are the permutations of

$$\mathrm{AWP}:=\bigcap_{U\in U(d^n)} U\,WP\,U^\dagger,$$5

and

$$\mathrm{AWP}:=\bigcap_{U\in U(d^n)} U\,WP\,U^\dagger,$$6

This gives a complete unitary-orbit characterization of absolute Wigner positivity in odd-prime dimensions (Zurel et al., 25 Feb 2026).

The AWP set is strictly larger than the absolutely stabilizer set. In particular, for odd-prime systems there exist absolutely Wigner-positive states that are not absolutely stabilizer, giving a unitarily invariant version of bound magic: some states never develop Wigner negativity under any unitary, yet some unitary can still move them outside the stabilizer polytope (Zurel et al., 25 Feb 2026).

These spectral results dovetail with the simulation framework of positive Wigner circuits. In the Mari–Eisert formalism, if the initial states, channels, and local POVM elements all have positive Wigner functions, then the circuit can be sampled in time $$\mathrm{AWP}:=\bigcap_{U\in U(d^n)} U\,WP\,U^\dagger,$$7, and Wigner negativity is thereby necessary for quantum computational advantage within that model. In the same framework, pure Wigner-positive states are exactly Gaussian in continuous variables and stabilizer states in odd-prime discrete systems (Mari et al., 2012).

The odd-prime AWP polytope also admits sharp Hilbert–Schmidt radius statements. Its inradius is

$$\mathrm{AWP}:=\bigcap_{U\in U(d^n)} U\,WP\,U^\dagger,$$8

which coincides with the inradius of the absolutely stabilizer set. Consequently,

$$\mathrm{AWP}:=\bigcap_{U\in U(d^n)} U\,WP\,U^\dagger,$$9

is a sufficient condition for both absolute stabilizerhood and absolute Wigner positivity. The circumradius is

$WP$0

so

$WP$1

is a necessary condition for not being AWP (Zurel et al., 25 Feb 2026).

5. Convex geometry and finite-dimensional variants

Recent continuous-variable work approaches mixed Wigner-positive states through convex geometry. The set of Wigner-positive states is treated via its extreme points, and a constructive procedure generates large classes of extreme Wigner-positive states by first characterizing extreme phase-invariant Wigner-positive quasi-states, then applying the Vertigo map, and finally using extremality-preserving maps to relax phase invariance. This construction generates all extreme Wigner-positive states of low dimension and highlights beam-splitter states $WP$2 as a fundamental class of extreme Wigner-positive states (Herstraeten et al., 16 Dec 2025).

For qubits, the usual odd-prime discrete formalism does not directly apply, and a distinct Wigner–Weyl–Moyal construction based on three Grassmann generators is used instead. In that framework, qubit stabilizer states have non-negative Wigner functions, Clifford gates act as state-independent permutations of phase-space points, and stabilizer states are exactly the absolutely Wigner-positive states under Clifford dynamics. The same formalism makes clear why the $WP$3-gate lies outside the positive-Wigner, noncontextual sector (Kocia et al., 2017).

Spin systems furnish a further absolute notion. For spin-$WP$4, absolute Wigner boundedness with lower cutoff $WP$5 is characterized by the eigenvalue inequality

$WP$6

where $WP$7 are the ordered eigenvalues of the Stratonovich–Weyl kernel. The corresponding polytope reduces to absolute Wigner positivity at $WP$8. The largest Hilbert–Schmidt ball centered at the maximally mixed state and contained in the spin AWP polytope has radius

$WP$9

In the infinite-spin contraction toward the bosonic limit, the outer AWP radius collapses to zero, which strongly supports the conclusion that absolutely Wigner-positive bosonic states do not exist in the analogous unitary-orbit sense (Denis et al., 2023).

6. Operational meaning, nonclassicality, and open problems

Positive Wigner functions are strongly linked to efficient classical simulation, but they do not coincide with classicality in any universal operational sense. In the Mari–Eisert setting, positivity of all state, gate, and measurement Wigner representations yields an efficient stochastic simulation and generalizes Gottesman–Knill. This establishes Wigner negativity as a necessary resource for quantum speedup in that framework, but only relative to the adopted representation and gate set (Mari et al., 2012).

Several results show that Wigner positivity does not preclude stronger nonclassical effects. Weakly squeezed Gaussian states with everywhere positive Wigner functions can violate CHSH and Clauser–Horne inequalities in stochastic simulations based on positive Wigner sampling, because the trajectory-level variables corresponding to normally ordered observables acquire the $W(\Omega)\ge W_{\min}$0 corrections associated with symmetric ordering and can become negative. In the low-gain limit, the simulations recover the Tsirelson value $W(\Omega)\ge W_{\min}$1 (Lantz et al., 2020). Likewise, single-mode squeezed vacuum and displaced squeezed states, despite their positive Wigner functions, can violate non-contextuality inequalities by a factor $W(\Omega)\ge W_{\min}$2 for odd $W(\Omega)\ge W_{\min}$3 when suitably embedded pseudo-spin observables are used (Roy, 2018).

A recent operational criterion links Wigner negativity to coherent-state superpositions. Within a QND-based direct-measurement framework, the absence of coherent superpositions in the coherent-state basis is a sufficient condition for Wigner positivity. For Schrödinger-cat states this becomes necessary and sufficient: if the residual coherence parameter is $W(\Omega)\ge W_{\min}$4, positivity holds exactly when

$W(\Omega)\ge W_{\min}$5

For higher-order cat states on a circle, an analogous criterion is derived in the limit of a large number of densely packed coherent states (Solinas et al., 22 Apr 2026).

The main open problems remain structural. A complete characterization of mixed Wigner-positive states is still unavailable even in the single-mode continuous-variable case, despite recent progress through extreme-point methods (Herstraeten et al., 16 Dec 2025). The sharp Shannon lower bound $W(\Omega)\ge W_{\min}$6 for positive Wigner functions is still conjectural in full generality, although nontrivial universal lower bounds and purity-based sufficient conditions are known (Dias et al., 2023, Qian et al., 23 Jan 2026). In the relativistic Dirac setting, the existence of strictly positive scalar Wigner functions for pure positive-energy states is conjectured but unproved, and no relativistic Hudson-type classification is known (Campos et al., 2014). These unresolved points underscore the central lesson of the subject: absolute Wigner positivity is not a single theorem but a family of representation-dependent convex and spectral constraints whose physical meaning depends on dynamics, observables, and the chosen quasiprobability framework.