A General Class of Functionals for Certifying Quantum Incompatibility
Kuan-Yi Lee, Jhen-Dong Lin, Adam Miranowicz, Yueh-Nan Chen · 2026
Read A General Class of Functionals for Certifying Quantum Incompatibility by Kuan-Yi Lee, Jhen-Dong Lin, Adam Miranowicz, Yueh-Nan Chen (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2601-02239.txt.
Introduction
Quantum steering, measurement mismatch, and instrument mismatch look like separate effects. They are now understood as faces of one resource: quantum incompatibility. Kuan-Yi Lee and three co-authors build a general way to certify that resource. Their nonlinear witnesses need no tuning step.
The witnesses come from convex functions. They are nontrivial exactly when the function is non-affine on extremal points such as pure states. For pure bipartite states, they lower-bound entanglement measures. They beat most linear steering bounds in that regime. The same construction certifies measurement and instrument mismatch too. There, the witnesses behave as genuine monotones of the resource. Two examples show the framework at work: a skew-information function and a coherence function.
Catalog record
| CREATOR | Kuan-Yi Lee, Jhen-Dong Lin, Adam Miranowicz, Yueh-Nan Chen |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 57 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2601-02239
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
A General Class of Functionals for Certifying Quantum Incompatibility Kuan-Yi Lee, Jhen-Dong Lin, Adam Miranowicz, Yueh-Nan Chen Abstract Quantum steering, measurement incompatibility, and instrument incompatibility have recently been recognized as unified manifestations of quantum incompatibility. Building on this perspective, we develop a general framework for constructing optimization-free, nonlinear incompatibility witnesses based on convex functionals, valid in arbitrary dimensions. We prove that these witnesses are nontrivial precisely when the underlying functional is non-affine on extremal points (e.g., pure states for ensembles). For pure bipartite states, the witnesses yield lower bounds on entanglement measures, thereby outperforming most linear steering inequalities in the pure-state regime. Moreover, the construction extends in full generality to certify measurement and instrument incompatibility, where the witnesses act as genuine incompatibility monotones. We demonstrate the versatility of our approach with two operationally relevant functionals: the Wigner-Yanase skew information and an ℓ 2 \ell_{2} -type coherence functional. Introduction. —Incompatibility is a foundational quantum resource. In classical physics, any prescribed family of operations can be reproduced by a single device together with classical post-processing Heinosaari et al. 2016 ; Gühne et al. 2023a . By contrast, many quantum families admit no such joint simulation, i.e., compatibility fails. This structural constraint, already exemplified by the position–momentum trade-off, powers steering Gallego and Aolita 2015 ; Xiang et al. 2022 ; Gühne et al. 2023b , Bell nonlocality Buscemi 2012 ; Brunner et al. 2014 , cryptography Gisin et al. 2002 ; Branciard et al. 2012 , and precision advantages in estimation and discrimination Piani and Watrous 2015 ; Takagi et al. 2019 ; Skrzypczyk and Linden 2019 ; Skrzypczyk et al. 2019 ; Tan et al. 2021 ; Flatt et al. 2022 ; Lee et al. 2023 . An increasing number of studies emphasize a unified compatibility viewpoint for states, measurements, and instruments Uola et al. 2018 ; Gühne et al. 2023a . In this formulation, quantum steering is the failure of a local-hidden-state model Wiseman et al. 2007 (the compatibility of state assemblage), its link to measurement incompatibility ties (un)steerability to (joint) measurability Quintino et al. 2014 ; Uola et al. 2014 ; Uola et al. 2015 ; Ku et al. 2022a . Related developments extend the same framework to temporal and channel scenarios Chen et al. 2014 ; Bartkiewicz et al. 2016 ; Chen et al. 2016a ; Buscemi et al. 2020 ; Leppäjärvi and Sedlák 2024 ; Ji and Chitambar 2024 , allowing assemblages across settings to be analyzed on comparable footing. For certification, two approaches dominate: first, optimization-based measures (e.g., robustness and compatibility via semi-definite programs) are informative but computationally demanding Cavalcanti and Skrzypczyk 2016 ; Designolle et al. 2019 ; and, second, linear or non-linear witnesses (e.g., moment-matrix criteria Kogias et al. 2015 ; Chen et al. 2016b , uncertainty-based inequalities Pramanik et al. 2014 ; Kriváchy et al. 2018 ; Costa et al. 2018 , and tests based on the Clauser-Horne-Shimony-Holt inequality Wolf et al. 2009 ; Cavalcanti et al. 2015 ) are easy to implement yet sometimes weak or lacking a clear task-related interpretation. This gap motivates simple, model-independent tests that avoid optimization while retaining operational meaning. In this work, we develop a convex-analytic framework to assemblages and show that any convex functional yields a class of optimization-free, non-linear, and dimension-independent incompatibility witnesses. We prove a necessary and sufficient criterion: a witness of quantum incompatibility is nontrivial if and only if the functional is non-affine on extremal points. The framework delivers two concrete, operationally meaningful witnesses–instantiated with the Wigner–Yanase skew information (WYSI) Wigner and Yanase 1963 ; Luo 2003 and an \ell_{2} -type coherence functional Baumgratz et al. 2014 –and encompasses earlier metrology- Yadin et al. 2021 and coherence-based Lee et al. 2025 tasks within a single framework. For pure states, the witness lower-bounds an entanglement measure and is saturated only by maximally entangled states. Via the steering–equivalent-observable map and minimal Stinespring dilation Uola et al. 2018 , the construction applies uniformly to measurements and instruments, recovering the known joint-measurability thresholds Heinosaari et al. 2008 and illustrating distinctive channel behavior. A unified framework of quantum incompatibility test. —It is known that quantum steering, measurement incompatibility, and instrument incompatibility can be investigated within a single unified framework in semi-device-independent settings Uola et al. 2018 ; Gühne et al. 2023b ; Gühne et al. 2023a . We refer to this framework as quantum incompatibility. Specifically, as depicted in Fig. 1 (a) , we consider a quantum device with multiple classical inputs labeled by x , each corresponding to a button. When a particular input is chosen, the device stochastically produces a classical output a with probability p(a|x) , along with a corresponding quantum object O_{a|x} . We note that this quantum object can generally represent a quantum state, a quantum measurement, or a quantum instrument with corresponding notations in Tab. 1 . We temporally leave it unspecified to maintain the generality. Spaces Assemblages Witnesses States \bm{\sigma} \mathcal{S}_{g}(\bm{\sigma}) Measurements \bm{M} \mathcal{M}_{g}(\bm{M}) Instruments \bm{\Lambda} \mathcal{I}_{g}(\bm{\Lambda}) Table 1: Notations of assemblages and witnesses. This device can now be fully characterized by an assemblage \bm{O}=\{p(a|x)O_{a|x}\}_{a,x} , which allows us to introduce the notion of quantum incompatibility. We regroup the assemblage according to each input x , defining \bm{O}_{x}=\{p(a|x)O_{a|x}\}_{a} , and ask whether \bm{O}_{x} are mutually compatible. If this is the case, the assemblage can be classically simulated by a hidden-object (HO) model as depicted in Fig. 1 (b) . This model consists of predetermined hidden parent objects \{O_{\lambda}\}_{\lambda} drawn from a distribution p_{\lambda} and a classical stochastic map p(a|x,\lambda) such that \displaystyle p(a|x)O_{a|x}=\sum_{\lambda}p(\lambda)p(a|x,\lambda)O_{\lambda}. (1) An assemblage is quantum incompatible, if it cannot be simulated by the HO model. Figure 1: Schematic illustration of compatibility tasks. We are now in position to present our main result, in which we propose a framework comprising a general class of information tasks to certify quantum incompatibility. The central ingredient of this approach relies on the concept of real-valued convex functionals of quantum objects. Specifically, let \mathcal{O} be the space of all quantum objects. A convex functional g:\mathcal{O}\to\mathbb{R}^{+} satisfies \sum_{i}p_{i}g(O_{i})\geq g(\sum_{i}p_{i}O_{i}) with p_{i}\geq 0 and \sum_{i}p_{i}=1 . Operationally, most resource quantifiers satisfy convexity, which implies that classical randomness and mixture cannot create non-classical resources. For every setting x , we define a functional of \bm{O}_{x} as g^{\rm as}(\bm{O}_{x}):=\sum_{a}p(a|x)g(O_{a|x}), (2) where superscript “as” refers to a functional of assemblage. In addition, we define the roof extension of g as F_{g}(O):=\sup_{\{p_{i},\psi_{i}\}_{i}}\sum_{i}p_{i}g(\psi_{i}) that satisfies \sum_{i}p_{i}\psi_{i}=O and \{\psi_{i}\}_{i}\subset\text{ext}(\mathcal{O}) , where \rm ext(\mathcal{O}) denotes the set of extreme points of \mathcal{O} , e.g., pure states for a state space. Note that the roof extension is achievable provided the supremum is attained by a physically realizable decomposition. Due to the convexity of g in \mathcal{O} , the roof extension serves as an upper bound of g . Similarly, we define F^{\rm as}_{g}(\bm{O}_{x}):=\sum_{a}p(a|x)F_{g}(O_{a|x}) . Following the standard convex-theoretic analysis, we obtain the following general criterion for certifying quantum incompatibility. Result 1 . If an assemblage \bm{O} admits the HO model, then g^{\rm as}(\bm{O}_{x})\leq F_{g}^{\rm as}(\bm{O}_{x^{\prime}})~~\forall x,x^{\prime}. (3) The details of our proof can be found in the Appendix A , where the central idea relies on the convexity (concavity) of g ( F_{g} ) by showing that g^{\rm as}(\bm{O}_{x})\leq\sum_{\lambda}p(\lambda)g(O_{\lambda}) [F^{\rm as}_{g}(\bm{O}_{x})\geq\sum_{\lambda}p(\lambda)F_{g}(O_{\lambda})] , respectively. If