Problem 1: The Wigner-entropy bound
Let be the set of states of a single-mode continuous-variable (CV) bosonic system with nonnegative Wigner function . For define the Wigner entropy
also known as the Wigner–Shannon entropy, and define
Call a prepare-and-measure fragment of quantum theory a CV quadrature fragment if its preparations are single-mode CV bosonic states and its measurements are quadrature measurements, including any classical post-processing of their outcomes. For a set of single-mode states, denote by the CV quadrature fragment whose preparations are all convex mixtures of states in and whose measurements are all quadrature measurements. We call the Wigner-positive subtheory.
The Wigner-entropy bound problem is the following two-part problem:
(a) Does adjoining any Wigner-negative state, together with all convex mixtures of the enlarged preparation set, necessarily make the resulting CV quadrature fragment contextual? In other words, for every single-mode state decide whether
(b) What is the value of ?
- Status
- open since 3 Oct 2026
- Updated
- 8 Nov 2026
- Posed by
- Rafael Wagner, Tobias Haas, Zacharie Van Herstraeten (?), Nicolas J. Cerf (?) and Ulysse Chabaud, 3 Oct 2026
Background
While generalized noncontextuality is well understood for discrete-variable (DV) systems, its extension to continuous variables (CV) is comparatively underdeveloped. For the Kochen-Specker (KS) formulation of noncontextuality [22] there are CV scenarios in which the notion can be made precise, quantified, and experimentally investigated [6]. Furthermore, in such scenarios KS contextuality can be shown to be in one-to-one correspondence with the negativity of the Wigner function [7]. This is the CV counterpart of a relation first established for DV systems, where KS contextuality (with respect to stabilizer measurements) is equivalent to the negativity of a suitable discrete Wigner function, established for rebits [23] and for qudits of odd dimension [24].
In order to formulate generalized contextuality in the CV case, more than one proposal has been put forward [8, 9], and the picture is more nuanced than in the DV case. These proposals are part of a broader literature on the connection between quasiprobability representations and classical explainability in the sense of generalized noncontextuality [20, 21]. This literature builds on Spekkens' observation that generalized noncontextuality and the existence of a nonnegative quasiprobability representation are equivalent notions of classicality [19]. References [20, 21] made this relation more precise, having classified in detail all assumptions that such quasiprobability representations needed to satisfy.
Constructing positive quasiprobabilistic representations is a clean way of constructing noncontextual ontological models for operational-probabilistic theories. An early result in this literature regarding CV systems is that Gaussian quantum mechanics is reproduced by Liouville mechanics with an epistemic restriction [11]. This implies the Gaussian subtheory, whose pure states are exactly the pure Wigner-positive ones, admits of a generalized-noncontextual ontological model. More recently, following a complementary path, Ref. [8] proposed a (set of) definition(s) of generalized contextuality for CV quantum theory and showed that it can be revealed with a single projective measurement, a result already known in the DV case [25]. As for the connection between generalized contextuality and Wigner negativity in quantum theory, a close relationship has been established for DV systems [10] and for some of the proposed CV definitions [9]. Notwithstanding these advances, an agreed understanding of contextuality in the CV setting, and of its full connection with the nonclassicality of quasiprobability distributions, is still largely work in progress.
For a bosonic mode with quadratures and , a state has a Wigner quasiprobability representation on phase space which, unlike a probability distribution, may take negative values. The states for which it does not, the Wigner-positive states , have a well-defined differential entropy , the Wigner entropy [1]. Proposed extensions of this functional to arbitrary states are not unique. One option is the complex-valued Wigner entropy of Ref. [26], the analytic continuation of the differential entropy, whose imaginary part is proportional to the negative volume of the Wigner function. Since is invariant under Gaussian transformations and the uncertainty principle forbids arbitrarily sharp joint localization in and , a state-independent lower bound is natural to expect.
Several works have articulated a conjectured lower bound on the Wigner entropy of Wigner-positive states, which became known as the Wigner-entropy conjecture [1, 17] and is sometimes also referred to as Anaelle's conjecture [27, 28, ]. The bound reads
with equality for Gaussian pure states, the only pure Wigner-positive states by Hudson's theorem. The WEC refines the Bialynicki-Birula-Mycielski entropic uncertainty relation [12]
for the marginal position and momentum distributions, since lower-bounds the sum of the marginal entropies whenever , by subadditivity.
Recently, however, three independent works (some of which through human-AI collaboration) provided counterexamples to the WEC [3, 4, 5], showing that the true infimum of the Wigner entropy lies strictly \emph{below} that of Gaussian pure states, without determining its exact value. It is known that [17, 3, 4, 5].
Although not valid in arbitrary cases, the conjecture had been established for certain sets of states. These include passive states [1], Wigner-positive mixtures of the first three Fock states [2], general mixed states on the Fock subspace [13], general mixtures of Fock states [14], beam-splitter states [15], and states with purities below [16]. Moreover, a R'enyi-entropy variant was proven for , with pure Gaussian states as unique minimizers [17].
Some remarks on the problem's formulation
Since a noncontextual model of a fragment restricts to any of its sub-fragments [32], part (a) or Problem 1 is equivalent to the following per-state statement: for every single-mode state , admits a generalized-noncontextual ontological model if and only if .
The claim in (a) is an equivalence between the fragment admitting a noncontextual (NC) ontological model and all of its states being Wigner-positive. The two directions are of very different difficulty.
(⇐) If , the fragment is the Wigner-positive subtheory itself. The Wigner representation of its states is nonnegative by definition, the Wigner representation of every quadrature effect, including post-processed ones, is nonnegative, and the representation is linear in both states and effects. These three properties are exactly what is required of a noncontextual ontological model, with phase space as the ontic state space. This direction is therefore settled.
(⇒) This is the substantive direction. One must show that whenever admits some NC ontological model, necessarily ; equivalently, that adding any Wigner-negative state to the Wigner-positive subtheory makes it contextual.
If (a) holds, then every state of every CV quadrature fragment admitting a generalized-noncontextual ontological model lies in and hence satisfies
so is the tightest entropic bound respected by noncontextual CV quadrature fragments. Determining its value is then exactly part (b).
The per-state statement would be false, for uninteresting reasons, if the fragment contained alone. A fragment with a single preparation has no preparation equivalences and a noncontextual model exists regardless of the sign of [31]. Including among the preparations supplies the operational equivalences that a noncontextual model must respect, and it is the natural choice since is exactly the set whose entropy is bounded in (b).
Relevance
The Wigner-entropy bound problem is motivated by three established connections in the literature: between KS noncontextuality and Wigner positivity in CV scenarios [7], between generalized noncontextuality and Wigner positivity in the setting of Ref.[9], and between generalized noncontextuality and the Gaussian subtheory of quantum theory [11]. Together with the recent refutation of the WEC, which leaves the true infimum undetermined, these connections make it timely to ask whether the correct entropic bound on Wigner-positive states is a noncontextuality inequality. Furthermore, a concrete inequality is also a way of putting the competing CV definitions of generalized contextuality [8, 9] under scrutiny.
The problem traces to Goal I, Reconceptualization of classicality, on two of the three fronts highlighted in our companion manuscript: \textit{subsumption} and \textit{stringency}. Since positive quasiprobability representations and the Gaussian subtheory are among the notions of classicality that generalized noncontextuality is meant to recover, understanding how positive Wigner representations fit into this picture falls squarely within the subsumption front of Goal I. Relatedly, since the problem bears on whether uncertainty relations, which alone do not imply contextuality, can nonetheless be recast as noncontextuality inequalities, it also speaks to the stringency front.
Significance
A resolution would advance Goal I in at least three foreseeable ways. First, and of independent interest to researchers working on Wigner representations, progress on the problem may lead to further advances in understanding the Wigner entropy itself, including the value of . Second, should the bound turn out to be a noncontextuality inequality, its connection to entropic uncertainty relations, together with the established link between contextuality and the nonclassicality captured by uncertainty relations [18], would significantly strengthen the case for a broader relationship between generalized noncontextuality and uncertainty relations. Since progress requires linking generalized noncontextuality, as a notion of classicality, to the positivity of quasiprobability representations, it would also extend the structure theorems of Refs. [19, 20, 21] to the CV setting. Third, such a link would hint at a connection between contextuality and the even more fundamental Wigner majorization conjecture in phase space [14, 2], which seeks to express the uncertainty principle in the most general terms. A negative answer is informative too: it would locate a natural CV entropic bound that noncontextuality alone does not explain.
Furthermore, the routes by which this problem might be solved could themselves open directions for future work. The idea of noncontextuality inequalities that are not linear functionals of the states, but entropic or otherwise nonlinear functionals, is still underdeveloped, and a connection with entropic functionals would be a natural way of fostering research in that direction. Relatedly, understanding how entropy behaves for signed, or even complex-valued, quasiprobability representations of quantum theory [26, 27] could have an impact beyond quantum foundations, consolidating connections that have been drawn between classicality and the negativity of quasiprobabilities in entropy production, as in Ref. [30].
Overall, investigating this problem could also motivate further work on determining the actual value of . In particular, a partial result would be to show that there exists some for which the inequality is a valid generalized noncontextuality inequality with respect to a well-posed notion of noncontextuality for CV systems, one that subsumes the results on the Gaussian formalism, the expectations carried over from the DV case, and the connection to Leibniz's principle of the identity of indiscernibles.
Permanent link:open-context.jan-bulling.com/contribution/1
References
- Z. Van Herstraeten and N. J. Cerf, Quantum Wigner entropy, Phys. Rev. A 104, 042211 (2021)
- Z. Van Herstraeten, M. G. Jabbour and N. J. Cerf, Continuous majorization in quantum phase space, Quantum 7, 1021 (2023)
- Z. He, Wigner entropy below vacuum: physical counterexamples and stability limits (2026) arXiv:2609.13312
- A.-X. Liu and C.-F. Qiao, Radial coarse graining restores vacuum majorization in Wigner phase space (2026) arXiv:2609.23277
- Z. Van Herstraeten, N. J. Cerf and U. Chabaud, Wigner-positive quantum states can have lower entropy than the vacuum (2026) arXiv:2609.24670
- R. S. Barbosa, T. Douce, P.-E. Emeriau, E. Kashefi and S. Mansfield, Continuous-variable nonlocality and contextuality, Commun. Math. Phys. 391, 1047 (2022)
- R. I. Booth, U. Chabaud and P.-E. Emeriau, Contextuality and Wigner negativity are equivalent for continuous-variable quantum measurements, Phys. Rev. Lett. 129, 230401 (2022)
- P. Jokinen, M. Weilenmann, M. Plávala, J.-P. Pellonpää, J. Kiukas and R. Uola, Generalised contextuality of continuous variable quantum theory can be revealed with a single projective measurement (2026) arXiv:2601.14067
- J. Haferkamp and J. Bermejo-Vega, Equivalence of contextuality and Wigner function negativity in continuous-variable quantum optics (2021) arXiv:2112.14788
- D. Schmid, H. Du, J. H. Selby and M. F. Pusey, Uniqueness of noncontextual models for stabilizer subtheories, Phys. Rev. Lett. 129, 120403 (2022)
Progress and history
- 8 Nov 2026EditAccepted a recent suggestion made by Yuije Zhang (moderator)
- 3 Oct 2026PosedEntry created from Problem I of the companion paper.
Contributions
No contributions recorded yet.
Contribute to this problem
You need to be signed in. Every contribution is reviewed by a moderator before it appears, and is credited to the people you name.
Discussion
Sign in to join the discussion.
Rafael Wagner · 18 Nov 2026
[Note] This question will be changed as both statements independent of one another. So we'll split it.