Introduction
In quantum mechanics, the Wigner function provides a powerful way to represent states as distributions over phase space. While these functions often take negative values—a signature of non-classicality—they offer an intuitive framework for analyzing symmetries and measurement statistics. However, traditional Wigner functions are deeply rooted in the mathematical structure of Hilbert spaces. This research extends the concept to General Probabilistic Theories (GPTs), a broad class of operational theories that includes classical physics, quantum mechanics, and post-quantum theories like "Boxworld."
The GPT Framework and Wigner Representations
The authors utilize convex geometry to describe physical systems. In this framework, states are represented as points in a compact convex set, and measurements are affine functions on that set. A Wigner representation is defined as a map from this state space to a space of "signed probability measures"—distributions that sum to one but allow for negative values.
A central requirement for these representations is that their marginal distributions must match the statistics of a set of chosen physical observables (e.g., position and momentum). This allows researchers to describe the theory in terms of fixed, operationally meaningful observables rather than abstract state vectors.
Symmetries: Lifted vs. Transported
A key contribution of this work is the formalization of symmetries within Wigner representations. The authors distinguish between two types:
- Lifted Symmetries: Transformations acting on the measurement outcomes or the "phase space" of the representation.
- Transported Symmetries: Physical transformations that act directly on the underlying state space.
The paper explores the "symmetry gap"—instances where a transformation in the phase space does not correspond to a physical transformation of the state. This is analogous to thermodynamics, where macroscopic changes (like pressure) may not map uniquely to specific microscopic molecular configurations.
Key Theoretical Insights
Negativity and Incompatibility
The study proves that a Wigner representation can remain strictly positive (resembling a standard probability distribution) if and only if the observables used to construct it are compatible, or jointly measurable. If the observables are incompatible—such as position and momentum in quantum mechanics—negativity is a mathematical necessity. This reinforces the idea that negativity in phase space is a fundamental marker of non-classicality.
Uniqueness and Complementarity
One of the paper's most significant results (Theorem 4) addresses the uniqueness of these representations. The authors demonstrate that if the chosen observables are complementary (where knowledge of one implies total uncertainty of the other), surjective, and informationally complete, there is at most one "covariant" Wigner representation. This explains why the standard quantum Wigner function is often viewed as the "natural" choice: it is essentially the unique representation that respects the symmetries of the measurement outcomes.
Applications and Future Directions
The framework successfully recovers standard Wigner representations for qubits and qudits and extends to Mutually Unbiased Bases (MUBs). It also provides insights into "Boxworld," a non-quantum theory with a square state space, demonstrating the framework's versatility across different physical axioms.
While currently limited to finite-dimensional systems, the authors anticipate that these results will generalize to infinite-dimensional cases. Future research will focus on incorporating dynamics to determine if physical laws require time-evolution to be a transported symmetry and exploring connections to complex-valued distributions like the Kirkwood-Dirac distribution.
Summary updated