Mixed-Dimensional Shadow Identity

• Mixed-Dimensional Shadow Identity is a framework demonstrating precise equivalences between formulations defined in different dimensions, as seen in 4-manifold state sums, quantum state geometry, and celestial OPE data.
• It connects the simplicial Crane–Yetter model to Turaev’s shadow model by matching tensor network contractions and normalization factors, yielding identical state-sum invariants for closed 4-manifolds.
• In quantum and coding theory contexts, the identity equates mixed-state numerical shadows with enlarged operator shadows and extends to heterogeneous Hilbert spaces, enforcing positivity constraints vital for error correction and AME state existence.

Searching arXiv for the cited papers and closely related work on mixed-dimensional shadow identities. “Mixed-dimensional shadow identity” is not a single uniformly standardized term across the literature. In its most explicit use, it denotes the exact equality between two apparently different semisimple $4$-manifold state-sum theories, the simplicial Crane–Yetter model and Turaev’s shadow model, for any closed oriented PL $4$-manifold $X$ and coordinated premodular category $C$: $\int_X^{CY} C=\int_X^{sh} C$ (Guu, 2022). Closely related usages appear elsewhere as structural identities rather than identically named theorems: the mixed-state numerical shadow identity $P^{\mu_K}_{A}(z)=P_{A\otimes {\mathbbm 1}_K}(z)$ in quantum-state geometry (Dunkl et al., 2011), the shadow-enumerator formalism constraining absolutely maximally entangled states and quantum codes (Huber et al., 2017), the explicitly named mixed-dimensional shadow identity for heterogeneous Hilbert spaces (González-Lociga et al., 28 Apr 2026), and universal shadow-transformation rules for celestial operator product expansion coefficients (Himwich et al., 14 May 2025). The shared motif is an exact relation between data carried by a “shadow” description and data defined in a space of different apparent dimension, support profile, or locality structure.

1. Terminological scope and recurrent pattern

The literature uses the expression non-uniformly. In some works it is a theorem title or an identified central formula; in others it is a natural descriptive label for the main structural equivalence. This suggests that the term is best read as denoting a family of shadow correspondences rather than a single formalism.

Domain	Identity or correspondence	Status in source
$4$-manifold state sums	$\int_X^{CY} C=\int_X^{sh} C$	Explicit central theorem (Guu, 2022)
Numerical shadow theory	$P^{\mu_K}_{A}(z)=P_{A\otimes {\mathbbm 1}_K}(z)$	Main mixed-state structural identity (Dunkl et al., 2011)
AME / QECC, uniform local dimension	Shadow inequalities and shadow transform of enumerators	Core machinery, not a separately new “mixed-dimensional” theorem (Huber et al., 2017)
AME / QECC, heterogeneous local dimension	$S_{MN}(\vec x,\vec y)=A_{MN}(\cdots)$	Explicit theorem in mixed dimensions (González-Lociga et al., 28 Apr 2026)
Celestial CFT	Shadow-transformed OPE coefficients via OPE blocks	Structural identity for celestial and shadow primaries (Himwich et al., 14 May 2025)

Across these settings, “shadow” does not have a single meaning. It may refer to a shadowed polyhedron with gleams, a probability distribution on a numerical range, a shadow enumerator, or a two-dimensional conformal shadow transform. Likewise, “mixed-dimensional” may refer to a $4$0-dimensional theory encoded by $4$1-dimensional data, to mixed-state versus pure-state enlargements, or to heterogeneous local dimensions such as qubits combined with qutrits and higher-dimensional qudits.

2. The $4$2-manifold state-sum identity

The most direct use of the term occurs in the equivalence between the Crane–Yetter and shadow models of a closed oriented PL $4$3-manifold (Guu, 2022). The theorem states:

$4$4

Here the Crane–Yetter model is fully simplicial. One begins with a triangulation $4$5 of $4$6, a coloring $4$7 of oriented $4$8-simplices by simple objects of $4$9 satisfying $X$0, and a total ordering of vertices. Its state sum is

$X$1

with $X$2, $X$3 the numbers of vertices and edges, the product over $X$4-faces $X$5, and $X$6 the large contraction of the tensor network. The local data are $X$7-symbols assigned to oriented $X$8-simplices.

The shadow model is lower-dimensional in appearance. It starts from a shadowed simple $X$9-polyhedron $C$0 with regions decorated by gleams, a coloring $C$1 of oriented regions by simple objects of $C$2, and local tensor data attached to singular sets. For closed $C$3, the state sum becomes a number:

$C$4

The normalization factor is

$C$5

In this expression, $C$6 is the second Betti number of the shadow, $C$7 is the nullity of the gleam quadratic form, $C$8 is the Euler characteristic of a region $C$9, and $\int_X^{CY} C=\int_X^{sh} C$0 is its gleam. Gleams encode $\int_X^{CY} C=\int_X^{sh} C$1-dimensional framing or self-intersection data, so the theory is not merely $\int_X^{CY} C=\int_X^{sh} C$2-dimensional despite its polyhedral input.

The proof proceeds by converting a triangulation into a canonical shadow derived from the dual handle decomposition, then comparing the two local tensor networks. The paper gives a dictionary between labels on triangulated $\int_X^{CY} C=\int_X^{sh} C$3-faces and labels on shadow regions, with the same multiplicity spaces

$\int_X^{CY} C=\int_X^{sh} C$4

and the same duality convention under orientation reversal:

$\int_X^{CY} C=\int_X^{sh} C$5

In the shadow picture, contraction maps $\int_X^{CY} C=\int_X^{sh} C$6, canonical elements $\int_X^{CY} C=\int_X^{sh} C$7, and local $\int_X^{CY} C=\int_X^{sh} C$8-symbol data at tetrahedral points replace the more compact $\int_X^{CY} C=\int_X^{sh} C$9-symbol notation of the simplicial theory. The central local statement is that, after passing to the canonical shadow, the shadow contributions combine into the same $P^{\mu_K}_{A}(z)=P_{A\otimes {\mathbbm 1}_K}(z)$0-symbol expression, up to a normalization factor that matches globally. A particularly important special case is the degenerate $P^{\mu_K}_{A}(z)=P_{A\otimes {\mathbbm 1}_K}(z)$1-symbol identity

$P^{\mu_K}_{A}(z)=P_{A\otimes {\mathbbm 1}_K}(z)$2

which collapses parts of the network when a trivial color appears.

The proof also uses the sphere-shadow normalization

$P^{\mu_K}_{A}(z)=P_{A\otimes {\mathbbm 1}_K}(z)$3

together with the Biedenharn–Elliott identity, the orthonormality relation, and the Racah identity. These are the shadow-side analogues of Pachner-move identities in the simplicial setting.

The same work recalls that both theories degenerate to the $P^{\mu_K}_{A}(z)=P_{A\otimes {\mathbbm 1}_K}(z)$4D Witten–Reshetikhin–Turaev theory in the relevant special case. The equivalence therefore supplies another proof that WRT is the boundary theory of Crane–Yetter. It also surveys Turaev’s shadow construction, including stable shadows and the statement that every closed $P^{\mu_K}_{A}(z)=P_{A\otimes {\mathbbm 1}_K}(z)$5-manifold has a stable shadow. The author further suggests once again that the semisimple models have reached their limits.

3. Numerical shadows, projections, and the mixed-state identity

In quantum-state geometry, the relevant shadow identity is not usually phrased as “mixed-dimensional shadow identity,” but the central structural relation has that role (Dunkl et al., 2011). For a matrix $P^{\mu_K}_{A}(z)=P_{A\otimes {\mathbbm 1}_K}(z)$6 of order $P^{\mu_K}_{A}(z)=P_{A\otimes {\mathbbm 1}_K}(z)$7, the numerical range is

$P^{\mu_K}_{A}(z)=P_{A\otimes {\mathbbm 1}_K}(z)$8

and equivalently

$P^{\mu_K}_{A}(z)=P_{A\otimes {\mathbbm 1}_K}(z)$9

where

$4$0

The standard numerical shadow is the probability distribution

$4$1

supported on $4$2, where $4$3 and $4$4 is the unitarily invariant Fubini–Study measure. The paper emphasizes the geometric identification that the images of orthogonal projections of $4$5 onto a two-plane are similar to the numerical ranges of matrices of order $4$6.

This is formulated in the Hilbert–Schmidt geometry

$4$7

with real inner product

$4$8

For normal matrices, numerical ranges correspond to planar images of the classical simplex $4$9; for arbitrary matrices, they correspond to planar projections of the full quantum-state body $\int_X^{CY} C=\int_X^{sh} C$0.

The mixed-state extension is

$\int_X^{CY} C=\int_X^{sh} C$1

with $\int_X^{CY} C=\int_X^{sh} C$2 a unitarily invariant measure on $\int_X^{CY} C=\int_X^{sh} C$3. The paper concentrates on induced measures $\int_X^{CY} C=\int_X^{sh} C$4, obtained by taking a random pure state on $\int_X^{CY} C=\int_X^{sh} C$5 and tracing out the $\int_X^{CY} C=\int_X^{sh} C$6-dimensional subsystem. Their eigenvalue density is

$\int_X^{CY} C=\int_X^{sh} C$7

The central mixed-state shadow identity is

$\int_X^{CY} C=\int_X^{sh} C$8

It states that the mixed-state shadow of $\int_X^{CY} C=\int_X^{sh} C$9 with respect to $P^{\mu_K}_{A}(z)=P_{A\otimes {\mathbbm 1}_K}(z)$0 is exactly the ordinary pure-state numerical shadow of the enlarged operator $P^{\mu_K}_{A}(z)=P_{A\otimes {\mathbbm 1}_K}(z)$1. For $P^{\mu_K}_{A}(z)=P_{A\otimes {\mathbbm 1}_K}(z)$2, $P^{\mu_K}_{A}(z)=P_{A\otimes {\mathbbm 1}_K}(z)$3 coincides with the flat Hilbert–Schmidt measure on $P^{\mu_K}_{A}(z)=P_{A\otimes {\mathbbm 1}_K}(z)$4; for $P^{\mu_K}_{A}(z)=P_{A\otimes {\mathbbm 1}_K}(z)$5, the induced measure reduces to the Fubini–Study measure and the mixed-state shadow reduces to the usual pure-state shadow. The paper also uses the same projection framework to study unitary dynamics through trajectories of the form $P^{\mu_K}_{A}(z)=P_{A\otimes {\mathbbm 1}_K}(z)$6 and the mixed-state analogue $P^{\mu_K}_{A}(z)=P_{A\otimes {\mathbbm 1}_K}(z)$7.

4. Shadow inequalities, AME states, and the quantum MacWilliams identity

In the theory of quantum error correction and multipartite entanglement, “shadow” refers to Rains’s generalized shadow inequalities and the associated shadow enumerator (Huber et al., 2017). The setting is an $P^{\mu_K}_{A}(z)=P_{A\otimes {\mathbbm 1}_K}(z)$8-partite system of equal local dimension $P^{\mu_K}_{A}(z)=P_{A\otimes {\mathbbm 1}_K}(z)$9. A pure state is absolutely maximally entangled if every reduction onto at most half the parties is maximally mixed, equivalently

$S_{MN}(\vec x,\vec y)=A_{MN}(\cdots)$0

For equal local dimensions,

$S_{MN}(\vec x,\vec y)=A_{MN}(\cdots)$1

The shadow-inequality machinery packages subsystem purities into linear constraints. For positive semidefinite Hermitian operators $S_{MN}(\vec x,\vec y)=A_{MN}(\cdots)$2 and any fixed $S_{MN}(\vec x,\vec y)=A_{MN}(\cdots)$3,

$S_{MN}(\vec x,\vec y)=A_{MN}(\cdots)$4

Defining

$S_{MN}(\vec x,\vec y)=A_{MN}(\cdots)$5

the shadow coefficients are

$S_{MN}(\vec x,\vec y)=A_{MN}(\cdots)$6

and the shadow enumerator polynomial is

$S_{MN}(\vec x,\vec y)=A_{MN}(\cdots)$7

The generalized shadow inequalities are precisely the positivity conditions $S_{MN}(\vec x,\vec y)=A_{MN}(\cdots)$8 for all $S_{MN}(\vec x,\vec y)=A_{MN}(\cdots)$9.

A central technical result of the same paper is the quantum MacWilliams identity in the Bloch representation. Using a local orthonormal operator basis $4$00 satisfying $4$01, and the Shor–Laflamme enumerators

$4$02

one obtains

$4$03

The shadow transform of enumerators is

$4$04

For AME states $4$05, the unitary coefficients simplify to

$4$06

and hence

$4$07

with $4$08 the Krawtchouk polynomial. Negative shadow coefficients rule out candidate AME states. The paper recovers known bounds like Scott’s bound and strengthens it in many higher-dimensional cases, including nonexistence results for $4$09, $4$10, and $4$11 listed explicitly in the source. A concrete example is the four-qubit obstruction

$4$12

which rules out a four-qubit AME state.

The same work also considers mixed-dimensional systems operationally. It adopts the generalized notion that a pure multipartite state is maximally entangled across every bipartition if, for each bipartition, the smaller side is maximally mixed. Within that perspective it analyzes four-party qubit–qutrit configurations, excludes $4$13 and $4$14 by shadow inequalities, and gives an explicit state on $4$15 that is maximally entangled across every bipartition.

5. Heterogeneous Hilbert spaces and the explicit mixed-dimensional shadow identity

A fully explicit mixed-dimensional shadow identity appears in the heterogeneous-system extension of the quantum MacWilliams formalism (González-Lociga et al., 28 Apr 2026). The basic space is

$4$16

with subsystem dimensions $4$17 not necessarily equal. For $4$18,

$4$19

and the paper replaces scalar support weight by the dimension multiset

$4$20

If $4$21, then

$4$22

where $4$23 is the multiplicity of dimension $4$24, $4$25, and $4$26 is the set of distinct local dimensions.

The multiset Shor–Laflamme coefficients are

$4$27

$4$28

while the unitary multiset coefficients are

$4$29

$4$30

The associated multivariate polynomials use variables $4$31 indexed by local dimension $4$32. The mixed-dimensional quantum MacWilliams identity is

$4$33

From this the paper derives the mixed-dimensional shadow identity. First,

$4$34

and then, substituting the relation between $4$35 and $4$36,

$4$37

This is the theorem explicitly named the mixed-dimensional shadow identity. The shadow coefficients themselves are

$4$38

and generalized shadow inequalities imply their nonnegativity for positive semidefinite Hermitian $4$39.

The paper emphasizes what is new in mixed dimensions: composition-sensitive counting, componentwise dimension-dependent transforms, and generalized combinatorics through factors such as

$4$40

These identities support a linear program for code feasibility with constraints including

$4$41

along with the dimensional minimum-distance conditions stated in the source. They also yield generalized bounds. The mixed-dimensional quantum Hamming bound is

$4$42

and the mixed-dimensional Singleton bound states that if $4$43 with

$4$44

then necessarily

$4$45

For pure mixed-dimensional codes the paper gives a tighter bound with no homogeneous analogue in the same form:

$4$46

The same machinery is applied to heterogeneous AME states, defined by the condition that every subsystem $4$47 with $4$48 is maximally mixed. The paper derives closed forms for AME coefficients, proves that $4$49 for $4$50, computes shadow coefficients to exclude concrete heterogeneous systems, and introduces a combinatorial grid method for explicit tripartite constructions such as $4$51 and $4$52.

6. Celestial shadow transformations and OPE data

In celestial conformal field theory, the relevant shadow identity concerns the shadow transformation of operator product expansion data rather than state-sum or enumerator equivalence (Himwich et al., 14 May 2025). The setting begins with two families of highest-weight solutions to the $4$53D massless wave equation, related by a $4$54D shadow transform. The celestial primaries are Mellin-transformed scattering states, while the shadow primaries are given by

$4$55

The paper emphasizes that shadow operators are nonlocal on the celestial sphere, so a naive shadowing of only the leading OPE term is insufficient. The required object is the OPE block, which packages a primary together with its global descendants:

$4$56

Shadow-transforming the entire conformal-family contribution yields universal formulas for OPE coefficients with shadowed external or exchanged operators. For example, with the normalization choice

$4$57

the coefficient after shadowing operator $4$58 is

$4$59

shadowing operator $4$60 gives

$4$61

and shadowing the exchanged primary yields

$4$62

With the same normalization one has the involution property

$4$63

These formulas determine three-point coefficients involving any combination of celestial and shadow primaries. The paper applies them to $4$64 currents, stress tensors, and a chiral current algebra of soft gluons. In each case, a naive interpretation of the shadowed OPE coefficient as a fully local OPE can be misleading. For example, the transformed current coefficient implies $4$65, and the transformed stress-tensor coefficient implies $4$66. The paper treats these signs as evidence that shadowed operators have additional nonlocal singularity structure beyond the shadowed OPE coefficient alone.

Taken together, these literatures show that “mixed-dimensional shadow identity” names a recurring algebraic phenomenon rather than a unique theorem schema. In $4$67-manifold topology it is an exact equality between a $4$68-simplex $4$69-symbol state sum and a $4$70-dimensional shadow state sum with gleams. In numerical shadow theory it identifies mixed-state shadows with ordinary shadows of enlarged operators. In coding theory it becomes a shadow-enumerator transform whose positivity constrains codes and AME states, and in heterogeneous systems this takes the explicit multivariate form indexed by dimension multisets. In celestial CFT it becomes a controlled transformation law for OPE and three-point data, valid only after whole conformal families are organized into OPE blocks. The common lesson is that shadow data, though apparently lower-dimensional, nonlocal, or compositionally reindexed, can encode the same invariant information as a more direct formulation.