Boundary Geometry Turns Entanglement into Steering
Yu-Xuan Zhang, Jing-Ling Chen · 2026
Read Boundary Geometry Turns Entanglement into Steering by Yu-Xuan Zhang, Jing-Ling 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-2605-21245.txt.
Introduction
Yu-Xuan Zhang and Jing-Ling Chen identify a boundary mechanism that turns quantum entanglement into Einstein-Podolsky-Rosen steering. Entanglement alone does not imply steering, but a state meeting the boundary of the trusted state space in a nondegenerate way can close the gap. A projective assemblage with first-order tangential coherence but only second-order inward defect cannot be reproduced by any finite-measure local-hidden-state model. In finite dimensions this yields a support-kernel criterion, with the two-qubit product-null sector as the minimal example.
Catalog record
| CREATOR | Yu-Xuan Zhang, Jing-Ling Chen |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 26311 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.21245
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Boundary Geometry Turns Entanglement into Steering Yu-Xuan Zhang, Jing-Ling Chen Abstract Quantum entanglement does not necessarily imply Einstein-Podolsky-Rosen steering. We identify a boundary mechanism that closes this gap when an entangled state meets the boundary of the trusted state space in a nondegenerate way. The mechanism is local: a projective assemblage that approaches a boundary contact with first-order tangential coherence but only second-order inward defect cannot be reproduced by any finite-measure local-hidden-state model. In finite dimensions this yields a support-kernel criterion for an arbitrary fixed steering cut: if a trusted conditional state is rank-deficient and its first-order block couples its support to its kernel, then the state is both NPT (negative partial transpose) and projectively steerable. The two-qubit product-null sector provides the minimal transparent realization. A product vector in the kernel gives the pure boundary contact, and a single coherence entry simultaneously serves as the tangential displacement, the NPT minor, and the steering obstruction. Hence, on this boundary, entanglement collapses to two-way projective steering — including the genuinely mixed rank-three branch and the rank-two case as a low-dimensional instance. We also provide a filtered standard form, an explicit rank-three Cholesky parametrization, and a compact boundary certificate. Introduction. — Einstein-Podolsky-Rosen (EPR) steering is a form of quantum nonlocality in which measurements by an untrusted party remotely prepare conditional states for a trusted party in a way that cannot be explained by a local-hidden-state (LHS) model EPR1935 ; Schrodinger1935 ; Wiseman2007 ; Jones2007 ; Uola2020 ; Cavalcanti2016 . It occupies an intermediate position in the hierarchy {\rm Bell\ nonlocality}\Rightarrow{\rm steering}\Rightarrow{\rm entanglement} , while the reverse implications fail in general Bell1964 ; Werner1989 ; Horodecki2009 ; Wiseman2007 ; Jones2007 ; Uola2020 . In particular, steering is directional, and entanglement alone is not usually a steering certificate Werner1989 ; Bowles2014 . The question addressed here is when the boundary geometry of the state itself leaves no room for an LHS simulation. This viewpoint complements steering inequalities, semidefinite-programming methods, measurement incompatibility, and ellipsoid techniques Pusey2013 ; Skrzypczyk2014 ; Nguyen2019 ; Quintino2014 ; Uola2014 ; Uola2015 . The mechanism is easiest to visualize in its minimal two-qubit realization. A product vector in the kernel is not merely a rank defect: it forces a trusted conditional state to acquire a zero direction. For a trusted qubit this is a pure boundary contact with the Bloch sphere, and in the steering-ellipsoid picture it is a tangency between the steering ellipsoid and the Bloch sphere Jevtic2014 ; Jevtic2015 ; NguyenVu2016 ; SongBakerWiseman2023 . The contact becomes nondegenerate when nearby conditional states leave the boundary with first-order tangential coherence while positivity fills the orthogonal direction only to second order. The resulting boundary scaling is incompatible with any fixed finite hidden-state measure. A hidden atom at the contact point is harmless, since it has no tangential component; after the contact face is removed, the punctured neighborhoods that could supply the tangential motion shrink to the empty set for a finite measure. This local geometry is illustrated in Fig. 1 . The same mechanism has a finite-dimensional support-kernel form. For an arbitrary fixed cut X|Y , where Y may be high-dimensional or multipartite, a rank-deficient trusted conditional state defines a boundary face of the positive cone. If the first-order block of the state couples the support of this boundary state to its kernel, then the same two-dimensional compression gives the boundary scaling obstruction and, at the same time, a negative partial-transpose minor. Thus the two-qubit product-null sector should be read as the minimal explicit realization of a support-kernel boundary mechanism, not as the source of the mechanism itself. On the two-qubit product-null boundary this general idea becomes especially sharp. After local unitaries the product-null vector can be written as |01\rangle , and a single matrix entry h_{02}=\langle 00|\rho|11\rangle controls both entanglement and steering. It is the leading tangential coherence at the boundary contact and also the off-diagonal entry of the NPT minor. Hence the usual hierarchy collapses on this boundary stratum: entanglement is equivalent to two-way projective steering. We further identify a steering-safe filtered standard form, give an explicit Cholesky parametrization of the genuinely mixed rank-three branch, and record the high-dimensional support-kernel criterion as the general form of the same boundary principle. Figure 1: Boundary contact and quadratic inward defect. (a) Product-null boundary contact at N , with shrinking caps \Gamma_{t} . (b) The normalized inward defect \delta(u) is bounded by a quadratic curve Cu^{2} for a representative product-null family. For two-qubit product-null states, the general support-kernel obstruction becomes exact on the corresponding boundary stratum. By local unitaries one may place the product-null vector at |01\rangle . The tangency is then controlled by a single tangential coherence, h_{02}=\langle 00|\rho|11\rangle. This same matrix element has two roles. First, it is the first-order coefficient that makes Bob’s conditional states leave the contact point tangentially and hence triggers the boundary-contact scaling obstruction. Second, it gives a negative principal minor of the partial transpose. By the Peres-Horodecki criterion, PPT is equivalent to separability for two qubits Peres1996 ; Horodecki1996 . Therefore, on the product-null boundary, the same tangential coherence detects entanglement and certifies projective steering. This yields a sharp collapse of the usual hierarchy on the product-null boundary: {\rm entanglement}\Longleftrightarrow{\rm two\mbox{-}way\ projective\ steering}. As a consequence, every entangled two-qubit rank-two state is two-way projectively steerable, because the two-dimensional kernel of a rank-two two-qubit state always contains a product vector. This rank-two consequence recovers the previously established rank-two steering theorem from a different viewpoint ZhangChen2025Fundamental . The present proof is different: it identifies the product-null tangency as the minimal realization of the local boundary-contact scaling obstruction. The same mechanism also covers the rank-three branch whose null space is spanned by a product vector. This boundary result has a simple operational reading: once a candidate product-null direction is fixed, a boundary population test checks the contact, and the single tangential coherence directly determines the NPT minor. Thus the mechanism turns a boundary NPT witness into a projective-steering certificate without requiring an independent steering inequality. We also provide a steering-safe filtered standard form, an explicit Cholesky parametrization of the rank-three branch, and a compact boundary certificate. Finally, the support-kernel mechanism extends to arbitrary finite dimensions and multipartite trusted systems, where it remains a sufficient condition for NPT entanglement and projective steering across a fixed cut. The entanglement-steering gap therefore depends not only on the amount of entanglement but crucially on how the state meets the boundary of the trusted state space. Boundary-contact scaling obstruction. — We first isolate the local obstruction behind the proof. At this stage no special form of the bipartite state is used; the relevant object is only a two-dimensional trusted slice near a boundary contact. Let \sigma_{t} , 0<t<t_{0} , be unnormalized qubit conditional states. Choose the trusted basis so that the normalized states approach the pure boundary point |0\rangle\langle 0| as t\to 0 , and write \sigma_{t}=\left(\begin{smallmatrix}a_{t}&b_{t}\\ \bar{b}_{t}&d_{t}\end{smallmatrix}\right) . Assume that, for all sufficiently small t>0 , \displaystyle|b_{t}|\geq lt,\qquad d_{t}\leq ct^{2}, (1) with constants l>0 and c<\infty . The entry b_{t} is the support-kernel coherence in this two-dimensional slice, while d_{t} is the weight in the orthogonal direction. Thus Eq. ( 1 ) describes first-order tangential motion along the boundary and only second-order inward