Fubini–Study Tensor

• Fubini–Study tensor is the projectively induced curvature object that converts homogeneous coordinates and state vectors into intrinsic geometric structures.
• It underpins classical complex geometry, Bergman theory, and quantum state metrics, providing a unified framework for Kähler forms and Berry curvature.
• The tensor’s applications extend to current approximation in orbifolds and noncommutative settings, highlighting its role in diverse geometric and physical theories.

Searching arXiv for the supplied papers to ground the article in current literature. The Fubini–Study tensor is the canonical geometric structure attached, in different but closely related senses, to complex projective space, projective embeddings, and the projective Hilbert space of quantum states. In classical complex geometry it is the Kähler form or associated Kähler metric on $\mathbb{CP}^n$, induced by the hyperplane line bundle $\mathcal O_{\mathbb{CP}^n}(1)$; in geometric quantization and Bergman theory it appears through pullbacks, Fubini–Study metrics, and Fubini–Study currents associated with spaces of sections; and in quantum theory it is the quantum geometric tensor whose real part is the Riemannian metric measuring quantum distance and whose imaginary part is Berry curvature (Wolff, 2024, Cheng, 2010). Contemporary work treats these manifestations as instances of a common theme: projective geometry induces a distinguished tensor that converts linear data—homogeneous coordinates, holomorphic sections, or state vectors—into intrinsic metric, symplectic, or current-theoretic geometry.

1. Classical projective-geometric definition

On complex projective space $\mathbb{CP}^n$, the Fubini–Study tensor usually means the standard Kähler form $\omega_{FS}$, or equivalently the associated Kähler metric induced by $\mathcal O_{\mathbb{CP}^n}(1)$. In homogeneous coordinates $[z_0:\dots:z_n]$, a standard potential is

$\Phi(z)=\log\sum_{j=0}^n |z_j|^2,$

and the corresponding form is

$$\omega_{FS}=\frac{i}{2\pi}\,\partial\bar\partial \log \sum_{j=0}^n |z_j|^2.$$

Its cohomology class is $c_1(\mathcal O_{\mathbb{CP}^n}(1))$, and equivalently

$$c_1\bigl(\mathcal O_{\mathbb{CP}^n}(1),h_{FS}\bigr)=\omega_{FS}.$$

This is the model example in which the Fubini–Study tensor is the curvature form of the canonical Hermitian metric on $\mathcal O_{\mathbb{CP}^n}(1)$0 (Wolff, 2024).

On affine charts this becomes the familiar local Kähler potential. On $\mathcal O_{\mathbb{CP}^n}(1)$1, with $\mathcal O_{\mathbb{CP}^n}(1)$2, one has

$\mathcal O_{\mathbb{CP}^n}(1)$3

and the metric tensor is $\mathcal O_{\mathbb{CP}^n}(1)$4. More generally, if $\mathcal O_{\mathbb{CP}^n}(1)$5 is a positive definite Hermitian matrix, the paper on Chebyshev potentials considers the family

$\mathcal O_{\mathbb{CP}^n}(1)$6

with

$\mathcal O_{\mathbb{CP}^n}(1)$7

which are Fubini–Study metrics up to automorphisms and scaling (Jin et al., 2022).

The same tensor also appears in explicit affine form on $\mathcal O_{\mathbb{CP}^n}(1)$8. There,

$\mathcal O_{\mathbb{CP}^n}(1)$9

and can be written as

$\mathbb{CP}^n$0

This formula exhibits the Fubini–Study form as the flat Kähler form plus a rank-one correction term involving the tensor

$\mathbb{CP}^n$1

(Nakahara, 2013).

A common misconception is that “Fubini–Study tensor” names a single object with identical meaning across all areas. The classical literature already uses it ambiguously for the Kähler form, the Hermitian metric tensor, or the curvature of $\mathbb{CP}^n$2. The underlying datum is the same projectively induced geometry, but the tensorial object being emphasized varies with context.

2. Pullbacks, Kodaira maps, and Fubini–Study currents

For a very ample line bundle $\mathbb{CP}^n$3 on a smooth projective complex variety, choosing sections $\mathbb{CP}^n$4 defines an embedding

$\mathbb{CP}^n$5

Pulling back the hyperplane bundle identifies $\mathbb{CP}^n$6, and the induced Fubini–Study metric on $\mathbb{CP}^n$7 is given locally by

$\mathbb{CP}^n$8

with associated Kähler form

$\mathbb{CP}^n$9

Thus projective embeddings convert finite-dimensional linear data into intrinsic Kähler geometry (Fang, 2022).

In Bergman-theoretic settings, especially for general line bundles and singular metrics, the same mechanism is encoded by a Fubini–Study current. If $\omega_{FS}$0 is a holomorphic line bundle with Hermitian metric $\omega_{FS}$1, and $\omega_{FS}$2 is an orthonormal basis of the Bergman space $\omega_{FS}$3, then locally, writing $\omega_{FS}$4, the Fubini–Study current is

$\omega_{FS}$5

It is the pullback of the Fubini–Study form by the Kodaira map

$\omega_{FS}$6

so that $\omega_{FS}$7. The key identity is

$\omega_{FS}$8

where $\omega_{FS}$9 is the Bergman kernel function (Wolff, 2024).

This current-theoretic formulation extends to orbifolds. For an orbifold line bundle $\mathcal O_{\mathbb{CP}^n}(1)$0 with singular Hermitian metric $\mathcal O_{\mathbb{CP}^n}(1)$1, the orbifold Fubini–Study current $\mathcal O_{\mathbb{CP}^n}(1)$2 associated with $\mathcal O_{\mathbb{CP}^n}(1)$3 satisfies

$\mathcal O_{\mathbb{CP}^n}(1)$4

and, under semipositivity and positivity hypotheses, one has weak convergence

$\mathcal O_{\mathbb{CP}^n}(1)$5

on the positivity locus. In this sense, Fubini–Study tensors or currents are projectively induced approximants to curvature currents (Coman et al., 2012).

This suggests a useful unifying description: the Fubini–Study tensor is often a “projectively induced curvature object,” realized either as a smooth Kähler form or as a positive closed current depending on regularity.

3. Quantum-state geometry and the quantum geometric tensor

In quantum theory, the Fubini–Study tensor is the quantum geometric tensor on the projective Hilbert space of pure states. For a smooth family of normalized states $\mathcal O_{\mathbb{CP}^n}(1)$6, it is

$\mathcal O_{\mathbb{CP}^n}(1)$7

with decomposition

$\mathcal O_{\mathbb{CP}^n}(1)$8

The real part $\mathcal O_{\mathbb{CP}^n}(1)$9 is the gauge-invariant Riemannian metric measuring quantum distance, while the imaginary part is the Berry curvature up to a conventional factor (Cheng, 2010).

Equivalently, for nearby states,

$[z_0:\dots:z_n]$0

and the overlap expansion yields

$[z_0:\dots:z_n]$1

In adiabatic problems with $[z_0:\dots:z_n]$2, the tensor may be written as

$[z_0:\dots:z_n]$3

which makes its gauge invariance manifest and exhibits its singular behavior near degeneracies (Cheng, 2010).

For parameterized quantum circuits preparing pure states

$[z_0:\dots:z_n]$4

the Fubini–Study metric tensor is written componentwise as

$[z_0:\dots:z_n]$5

It defines the local quadratic form

$[z_0:\dots:z_n]$6

and its pseudoinverse $[z_0:\dots:z_n]$7 enters the quantum natural gradient update

$[z_0:\dots:z_n]$8

The paper “L2O-$[z_0:\dots:z_n]$9” embeds this geometry into a learned optimizer via the effective preconditioner

$\Phi(z)=\log\sum_{j=0}^n |z_j|^2,$0

interpolating between natural-gradient and Euclidean directions (Huang et al., 2024).

A second source of ambiguity therefore arises in quantum information: some authors use “Fubini–Study metric” for the real part alone, while others use “Fubini–Study tensor” or “quantum geometric tensor” for the full complex Hermitian tensor. The data support both usages, but the distinction between metric and full tensor is technically important (Cheng, 2010).

4. Canonical geometry on projective and homogeneous spaces

On $\Phi(z)=\log\sum_{j=0}^n |z_j|^2,$1, the Fubini–Study metric is the canonical homogeneous Kähler metric. In the h-projective rigidity literature, it is characterized by the fact that complex geodesics are projective lines and by the existence of a large solution space to the h-projective PDE

$\Phi(z)=\log\sum_{j=0}^n |z_j|^2,$2

A principal result states that if a closed connected Kähler manifold has degree of mobility $\Phi(z)=\log\sum_{j=0}^n |z_j|^2,$3, then either all h-projectively equivalent metrics are affinely equivalent, or the metric is, up to scaling, the Fubini–Study metric on $\Phi(z)=\log\sum_{j=0}^n |z_j|^2,$4 (Fedorova et al., 2010).

From the viewpoint of curvature, the Fubini–Study metric on $\Phi(z)=\log\sum_{j=0}^n |z_j|^2,$5 is normalized in several papers by the real-index formula

$\Phi(z)=\log\sum_{j=0}^n |z_j|^2,$6

expressing constant holomorphic sectional curvature in terms of the metric tensor and Kähler form. This formula underlies analyses of Killing tensors, h-projective equivalence, and X-ray transforms on complex projective space (Eastwood, 2023, Eastwood et al., 2011).

The same nomenclature extends beyond projective space itself. On complex Grassmannians $\Phi(z)=\log\sum_{j=0}^n |z_j|^2,$7, the Fubini–Study metric is the canonical homogeneous Kähler–Einstein metric induced from the standard Hermitian form on $\Phi(z)=\log\sum_{j=0}^n |z_j|^2,$8. In local coordinates $\Phi(z)=\log\sum_{j=0}^n |z_j|^2,$9, with

$$\omega_{FS}=\frac{i}{2\pi}\,\partial\bar\partial \log \sum_{j=0}^n |z_j|^2.$$0

its components are

$$\omega_{FS}=\frac{i}{2\pi}\,\partial\bar\partial \log \sum_{j=0}^n |z_j|^2.$$1

and it satisfies

$$\omega_{FS}=\frac{i}{2\pi}\,\partial\bar\partial \log \sum_{j=0}^n |z_j|^2.$$2

Recent work shows that when $$\omega_{FS}=\frac{i}{2\pi}\,\partial\bar\partial \log \sum_{j=0}^n |z_j|^2.$$3 is odd, the Fubini–Study metric on $$\omega_{FS}=\frac{i}{2\pi}\,\partial\bar\partial \log \sum_{j=0}^n |z_j|^2.$$4 is rigid despite the existence of infinitesimal Einstein deformations (Hall, 2024).

These results show that the Fubini–Study tensor is not merely a local formula: it is a canonical invariant tensor on compact Hermitian symmetric spaces, often singled out by strong rigidity properties.

5. Asymptotics, approximation, and singular settings

A classical theme is that Fubini–Study tensors obtained from finite-dimensional spaces of sections approximate continuous geometric data. For sequences of line bundles $$\omega_{FS}=\frac{i}{2\pi}\,\partial\bar\partial \log \sum_{j=0}^n |z_j|^2.$$5 on a compact Kähler manifold, the Fubini–Study currents $$\omega_{FS}=\frac{i}{2\pi}\,\partial\bar\partial \log \sum_{j=0}^n |z_j|^2.$$6 satisfy, under appropriate hypotheses,

$$\omega_{FS}=\frac{i}{2\pi}\,\partial\bar\partial \log \sum_{j=0}^n |z_j|^2.$$7

where $$\omega_{FS}=\frac{i}{2\pi}\,\partial\bar\partial \log \sum_{j=0}^n |z_j|^2.$$8 is the equilibrium metric and $$\omega_{FS}=\frac{i}{2\pi}\,\partial\bar\partial \log \sum_{j=0}^n |z_j|^2.$$9 is the scaling factor determined by auxiliary positivity data. Under additional cohomological control,

$c_1(\mathcal O_{\mathbb{CP}^n}(1))$0

This places Fubini–Study currents at the interface of Bergman kernel asymptotics, pluripotential theory, and equilibrium metrics (Wolff, 2024).

On orbifolds, analogous convergence holds for singular Hermitian line bundles: if $c_1(\mathcal O_{\mathbb{CP}^n}(1))$1 and is strictly positive on a set $c_1(\mathcal O_{\mathbb{CP}^n}(1))$2, then the Fubini–Study currents $c_1(\mathcal O_{\mathbb{CP}^n}(1))$3 associated with $c_1(\mathcal O_{\mathbb{CP}^n}(1))$4 satisfy

$c_1(\mathcal O_{\mathbb{CP}^n}(1))$5

on $c_1(\mathcal O_{\mathbb{CP}^n}(1))$6, while

$c_1(\mathcal O_{\mathbb{CP}^n}(1))$7

This extends Tian-type approximation to orbifold and singular settings and connects Fubini–Study currents with equidistribution of zeros of random sections (Coman et al., 2012).

A more singular regime is that of punctured Riemann surfaces with Poincaré-type metrics. If $c_1(\mathcal O_{\mathbb{CP}^n}(1))$8 is a punctured Riemann surface, $c_1(\mathcal O_{\mathbb{CP}^n}(1))$9 is a holomorphic line bundle, and near each puncture one has the model

$$c_1\bigl(\mathcal O_{\mathbb{CP}^n}(1),h_{FS}\bigr)=\omega_{FS}.$$0

then the pullback Fubini–Study forms by the Kodaira maps of $$c_1\bigl(\mathcal O_{\mathbb{CP}^n}(1),h_{FS}\bigr)=\omega_{FS}.$$1 satisfy

$$c_1\bigl(\mathcal O_{\mathbb{CP}^n}(1),h_{FS}\bigr)=\omega_{FS}.$$2

Thus near cusp singularities the ratio of the induced Fubini–Study form to the Poincaré form grows at most polynomially in the tensor power (Apredoaei et al., 6 Jun 2025).

These asymptotic regimes clarify a frequent misconception: Fubini–Study tensors do not always converge smoothly to the target curvature object. In singular, orbifold, or noncompact settings, weak convergence, current-theoretic convergence, or polynomial control near singularities is the relevant notion.

6. Generalizations: mixed states, quantum projective spaces, and non-Archimedean geometry

Beyond pure states, a purification-based mixed-state generalization defines a local $$c_1\bigl(\mathcal O_{\mathbb{CP}^n}(1),h_{FS}\bigr)=\omega_{FS}.$$3-gauge invariant Fubini–Study metric for density operators $$c_1\bigl(\mathcal O_{\mathbb{CP}^n}(1),h_{FS}\bigr)=\omega_{FS}.$$4. With $$c_1\bigl(\mathcal O_{\mathbb{CP}^n}(1),h_{FS}\bigr)=\omega_{FS}.$$5, the tensor is

$$c_1\bigl(\mathcal O_{\mathbb{CP}^n}(1),h_{FS}\bigr)=\omega_{FS}.$$6

Its real part defines the line element, it satisfies a quantum Cramér–Rao bound, and under a monotonicity condition it reduces to the square-root derivative quantum Fisher information. The same framework yields $$c_1\bigl(\mathcal O_{\mathbb{CP}^n}(1),h_{FS}\bigr)=\omega_{FS}.$$7-metrics as a family of generalized Fubini–Study metrics for mixed states (Mondal, 2015).

In noncommutative geometry, analogues of the Fubini–Study metric on quantum projective spaces are constructed as genuine two-tensors. For the Heckenberger–Kolb calculus $$c_1\bigl(\mathcal O_{\mathbb{CP}^n}(1),h_{FS}\bigr)=\omega_{FS}.$$8, the quantum Fubini–Study metric is

$$c_1\bigl(\mathcal O_{\mathbb{CP}^n}(1),h_{FS}\bigr)=\omega_{FS}.$$9

with

$\mathcal O_{\mathbb{CP}^n}(1)$00

It is symmetric in the sense $\mathcal O_{\mathbb{CP}^n}(1)$01, real in the sense $\mathcal O_{\mathbb{CP}^n}(1)$02, admits an inverse bimodule pairing, and possesses a Levi–Civita connection that is torsion free, cotorsion free, and satisfies $\mathcal O_{\mathbb{CP}^n}(1)$03 (Matassa, 2020).

In non-Archimedean geometry, a strict Cartesian ultrametric norm on a vector space $\mathcal O_{\mathbb{CP}^n}(1)$04 induces a Fubini–Study metric on $\mathcal O_{\mathbb{CP}^n}(1)$05. If $\mathcal O_{\mathbb{CP}^n}(1)$06 is a projective subvariety, the associated Fubini–Study metric on $\mathcal O_{\mathbb{CP}^n}(1)$07 determines a non-Archimedean Monge–Ampère measure and a Monge–Ampère polytope

$\mathcal O_{\mathbb{CP}^n}(1)$08

The metric is critical precisely when

$\mathcal O_{\mathbb{CP}^n}(1)$09

which is equivalent to residual semistability of the Chow form under a non-Archimedean Kempf–Ness criterion (Fang, 2022).

Taken together, these generalizations suggest that the Fubini–Study tensor is best understood not as a single formula but as a functorial construction: from linear or projective data one obtains a canonical tensorial geometry, and the ambient category—complex analytic, quantum, or non-Archimedean—determines what “tensor,” “metric,” or “curvature” means.