Logical Hardy-type Paradoxes in Quantum Contexts

• Logical Hardy-type Paradoxes are inequality-free proofs of quantum nonlocality and contextuality that derive contradictions solely from the pattern of possible and impossible measurement outcomes.
• They extend Hardy’s original bipartite paradox to multipartite and high-dimensional scenarios, achieving quantifiable success probabilities that serve as benchmarks in device-independent quantum tests.
• Their framework unifies logical, geometric, and graph-theoretic approaches to nonclassicality, enabling robust experimental self-testing and randomness amplification in quantum systems.

Logical Hardy-type Paradoxes are inequality-free proofs of quantum nonlocality and contextuality that derive contradictions between classical (local-realistic or noncontextual) theories and quantum predictions based solely on the existence/absence of certain measurement outcomes. They generalize Hardy’s original 1992 bipartite paradox to arbitrary multipartite, high-dimensional, and contextuality scenarios, providing a unified language for logical and possibilistic proofs of nonclassicality, and serve as pivotal tools in device-independent quantum information and foundational investigations.

1. Formal Structure of Logical Hardy-type Paradoxes

Hardy-type paradoxes are defined within the event-based framework as finite sets of measurement outcome events, with logical constraints dictating which combinations of outcomes are possible or impossible. Formally, let $(\mathcal A, p)$ be a system where $\mathcal A$ is a finite exclusive partial Boolean algebra of events and $p$ is a state (probability assignment). The paradox consists of:

• Logical contradiction in classical embedding: There exists a family $\{e_1,\dots,e_n\}\subseteq\mathcal A$ such that $e_1^c\land e_2^c\land\cdots\land e_n^c = \bot$ in the classical event algebra $\mathcal A^c$, meaning the joint realization of all $e_i$ is impossible in classical logic.
• Quantum witness: In the quantum model, $p(e_i)=1$ for $i\ne k$ and $p(e_k)>0$ for some $k$, i.e., all but one event certainly occur and the remaining one occurs with strictly positive probability.

The success probability (SP) of the paradox is $p(e_k)$. In the bipartite Bell $(2,2,2)$ scenario, Hardy’s original constraints are:

$$\begin{aligned} P(0,1|0,1) &= 0,\ P(1,0|1,0) &= 0,\ P(0,0|1,1) &= 0,\ P(0,0|0,0) &> 0. \end{aligned}$$

Classically, the first three conditions imply the fourth must be zero, but quantumly they can all be satisfied with $P(0,0|0,0)>0$ (Ramanathan et al., 2018, Liu et al., 4 Jan 2026).

2. Hardy-type Paradoxes, Contextuality, and Strong Contextuality

It is established that the existence of a logical Hardy-type paradox in any finite scenario (Bell or Kochen–Specker) is equivalent to logical contextuality: the impossibility for any classical model to match the zeros/nonzeros of the quantum support table (Liu et al., 4 Jan 2026). Strong contextuality occurs when no deterministic assignment can satisfy the support constraints, yielding SP=1; this is realized, for example, in the GHZ-Mermin scenario and generalized $n$-party "perfect" Hardy-type paradoxes (Jiang et al., 2017, Patra et al., 18 Dec 2025, Walleghem et al., 2024).

Hardy's original paradox and its ladder generalizations capture all forms of possibilistic nonlocality in $(2,2,d)$ and $(2,k,2)$ Bell scenarios, with the occurrence of any ladder paradox implying the presence of a Hardy pattern (Mansfield et al., 2011). This universality does not extend to arbitrary outcome numbers or measurement settings (e.g., $(2,3,3)$ scenario yields new patterns beyond Hardy) (Mansfield et al., 2011, Liu et al., 4 Jan 2026).

3. General Frameworks and Extensions

3.1. Bell Inequality Mapping

Any Bell inequality written in probability form

$\mathcal{I} = \sum_{j=1}^N f_j P_j \le L,$

can be mapped to a Hardy-type paradox by identifying Hardy constraints among $N-1$ probabilities and singling out one for the quantum success event (Yang et al., 2018). The construction provides a systematic translation between violation of Bell inequalities and all-versus-nothing (AVN) Hardy-type logical contradiction.

3.2. Multi-Setting, High-Dimensional, and Multipartite Paradoxes

The formalism extends to $(k,d)$ settings/outcome bipartite scenarios and multipartite scenarios, often yielding significantly enhanced SPs:

• For $(k,d)$, the ladder proof structure achieves $S_{k,d} \to 1/2$ for large $k$ in qubit systems and up to $S_{5,3}\approx0.40184$ in spin-1 systems (Meng et al., 2018).
• In multipartite settings, the maximal SP grows as $1/2^{n-1}$ for $n$ parties with generalized Hardy constraints, often outperforming standard constructions (Jiang et al., 2017). For $n=3$, extremal paradoxes with SP $=1/4$ are achievable.

General formulations cover both CLL-type and FTI-type paradoxes. The latter, associated with failure of transitivity of implications, yield higher degrees of success and enhanced robustness against experimental errors (Chen et al., 2023).

3.3. Contextuality and Kochen–Specker Connections

Hardy-type logical structures generalize to contextuality scenarios (KCBS, n-cycle), with exclusivity graph approaches and hypergraph "gadgets". The KCBS scenario admits Hardy-type paradoxes with maximal SP $\approx 10.56\%$; the method unifies all inequality-free nonlocality and contextuality proofs (Liu et al., 4 Jan 2026, Sohbi et al., 2019, Svozil, 2020).

Kochen–Specker proofs, especially via "01-gadgets" (small orthogonality graphs), provide systematic means to design Hardy-type paradoxes with arbitrary SP over $(0,1]$ by varying distinguished projectors and gadgets (Ramanathan et al., 2018).

4. Experimental Realizations and Self-Testing

Hardy-type paradoxes are experimentally realized in photonic systems and other platforms, achieving close agreement between ideal and measured SPs (Yang et al., 2018). The logical paradox can be used for device-independent randomness amplification: observing Hardy-type paradoxes certifies randomness under no-signalling assumptions, enabling amplification of arbitrarily weak sources of entropy via Raz’s two-source extractor, with the final min-entropy scaling linearly with the Hardy violation parameter (Ramanathan et al., 2018).

In multipartite scenarios, the quantum correlations saturating the Hardy-type paradox define exposed extremal points in the quantum set, serving as strong self-tests, e.g., for the tripartite GHZ state. Maximal Hardy-type paradoxes coincide with maximal violations of associated Bell inequalities (Mermin, CHSH), unifying logical and geometric approaches (Patra et al., 18 Dec 2025).

5. Logical Contextuality, Exclusivity Graphs, and Computational Aspects

Logical Hardy-type paradoxes interconnect with exclusivity graphs and hypergraphs representing event structure and exclusivity relations (Sohbi et al., 2019, Svozil, 2020). The logical contradiction is visualized as the impossibility of coloring or assigning binary values in such graphs subject to exclusivity and completeness constraints. Extensions to true-implies-false (TIFS), true-implies-true (TITS), and equality gadgets enforce more general relational properties among quantum observables.

In $(2,2,d)$ and $(2,k,2)$ scenarios, detection of Hardy-type nonlocality is algorithmically efficient (polynomial), while in higher-dimensional cases (e.g., $(2,3,3)$) complexity grows, and full classification may be computationally hard (NP-complete) (Mansfield et al., 2011).

6. Alternate Semantics and Resolution of Paradoxes

Analysis of Hardy-type paradoxes from non-classical semantic perspectives (partial/gappy, many-valued, weak-value semantics) demonstrates that classical contradictions evaporate if the underlying logic is relaxed. In partial or many-valued semantics, certain propositions acquire undefined or in-between truth values, blocking the logical inference underpinning the paradox (Bolotin, 2018). Weak-value semantics further decouple pre- and post-selected truth values, ensuring no contradiction arises.

7. Impact, Open Questions, and Significance

Logical Hardy-type paradoxes provide a unifying framework for all logical and possibilistic proofs of quantum nonlocality and contextuality. They underpin experimental certification of quantum phenomena, robust device-independent self-tests, and efficient randomness amplification, and reveal deep connections with Kochen-Specker theory, exclusivity principles, and graph-theoretic approaches.

Ongoing research addresses monotonicity of success probability with settings or outcome number (Chen et al., 2023), systematic construction of strong (SP=1) paradoxes, extension of self-testing methods to higher $N$-party scenarios (Patra et al., 18 Dec 2025), and computational protocols for detecting logical contextuality in general scenarios (Liu et al., 4 Jan 2026, Mansfield et al., 2011). The equivalence theorem between logical contextuality and Hardy-type paradoxes consolidates the logical foundation for all inequality-free contextuality and nonlocality proofs.