Top Slot Killing Tensor Fields

• Top slot Killing tensor fields are defined by the maximal-vanishing condition in the Taylor expansion of the polynomial first integral, ensuring the first nonzero term appears at the highest jet order.
• They convert complex Killing tensor PDEs into finite-dimensional, curvature-equivariant algebraic problems, offering a clear framework for classification on symmetric spaces and in Hamiltonian settings.
• In quadratic cases, top slot tensors yield a unique constant tensor that distinguishes irreducible from decomposable structures, with implications for models such as HP^m, OP^2, and the Calogero system.

Searching arXiv for recent and foundational papers on top slot Killing tensor fields and closely related formulations. Top slot Killing tensor fields are a distinguished class of Killing tensors singled out by a maximal-vanishing condition on the lower terms of their geodesic-normal-coordinate expansion at a point. On a Riemannian symmetric space, a rank-$d$ Killing tensor is called top slot at a point $y$ if, in the Taylor expansion of the corresponding polynomial first integral $\mathcal K(X,P)$ near $y$, no term of degree $<d$ in the normal coordinate $X$ appears; at the base point $o$, the associated space is denoted $\mathcal L^d$ (Matveev et al., 30 Apr 2026). In this sense, “top slot” refers not to tractor composition series, but to the highest possible initial position in the jet filtration compatible with the finite-type Killing equation (Matveev et al., 30 Apr 2026). Closely related but distinct usages occur elsewhere: in prolongation-based analyses, the “top slot” is the leading symmetric tensor component of a prolonged solution, while in Hamiltonian approaches it is the highest-rank coefficient of a polynomial first integral in momenta [(Eastwood, 2023); (Igata et al., 2010)]. Across these settings, the unifying theme is that the highest-order component is the decisive datum for existence, classification, or reducibility.

1. Definition and basic framework

A Killing tensor field of rank $d$ on a Riemannian manifold $(M,g)$ is a symmetric covariant tensor

$y$0

satisfying the Killing equation

$y$1

with respect to the Levi-Civita connection (Matveev et al., 30 Apr 2026). Equivalently, along every affinely parametrized geodesic $y$2, the quantity

$y$3

is constant (Matveev et al., 30 Apr 2026). In Hamiltonian form, a contravariant symmetric tensor $y$4 of rank $y$5 defines a homogeneous polynomial $y$6 of degree $y$7 on $y$8, and $y$9 is Killing iff

$\mathcal K(X,P)$0

(Matveev et al., 30 Apr 2026).

For symmetric spaces $\mathcal K(X,P)$1, the Lie algebra decomposes as

$\mathcal K(X,P)$2

and the curvature at the base point is encoded by

$\mathcal K(X,P)$3

(Matveev et al., 30 Apr 2026). This pointwise Lie-theoretic description is what permits the reduction of Killing-tensor PDEs to algebraic equations on $\mathcal K(X,P)$4 (Matveev et al., 30 Apr 2026).

The algebra of contravariant Killing tensors,

$\mathcal K(X,P)$5

is graded-commutative under symmetric product $\mathcal K(X,P)$6. Its subalgebra $\mathcal K(X,P)$7, generated by Killing vector fields, consists of the decomposable Killing tensors; indecomposable tensors are those not lying in this span (Matveev et al., 30 Apr 2026). This decomposable/indecomposable distinction is separate from top-slotness: a top slot tensor may be decomposable or indecomposable depending on the geometry (Matveev et al., 30 Apr 2026).

2. Normal-coordinate characterization on symmetric spaces

Fix $\mathcal K(X,P)$8 and identify a neighborhood via $\mathcal K(X,P)$9, writing points and cotangent vectors as $y$0. For a rank-$y$1 Killing tensor, the corresponding function $y$2 has a Taylor expansion in $y$3. Depending on parity under geodesic reflection at $y$4, it takes the form

$y$5

where $y$6 in the even case and $y$7 in the odd case (Matveev et al., 30 Apr 2026).

The defining condition for a top slot Killing tensor is that the expansion begin at degree exactly $y$8: $y$9 Thus a rank-$<d$0 tensor is top slot at $<d$1 when all lower jet components vanish (Matveev et al., 30 Apr 2026). Since a rank-$<d$2 Killing tensor is determined by its $<d$3-jet at one point, this is the latest possible starting order in normal coordinates, and therefore the “top” allowed slot in the jet filtration (Matveev et al., 30 Apr 2026).

A decisive simplification is Proposition $<d$4 of (Matveev et al., 30 Apr 2026): if $<d$5, then the whole Taylor expansion collapses to a single term,

$<d$6

where $<d$7 is a constant tensor on $<d$8, symmetric in the first $<d$9 and in the last $X$0 arguments (Matveev et al., 30 Apr 2026). This means that top slot Killing tensors are completely encoded by one constant $X$1-type tensor at a point. The same proposition gives necessary and sufficient algebraic conditions: $X$2 for all $X$3 and all $X$4 (Matveev et al., 30 Apr 2026). This is the paper’s explicit description of top slot Killing tensor fields.

This suggests a general principle: on symmetric spaces, top slot Killing tensors convert a finite-type overdetermined PDE into finite-dimensional curvature-equivariant linear algebra at a point (Matveev et al., 30 Apr 2026).

3. Quadratic theory and rank-one symmetric spaces

The rank-$X$5 case is especially explicit. If $X$6, then

$X$7

where $X$8 is a constant $X$9-tensor symmetric in the first pair and in the second pair (Matveev et al., 30 Apr 2026). Such a tensor defines a quadratic Killing tensor iff

$o$0

$o$1

$o$2

for all $o$3 (Matveev et al., 30 Apr 2026).

A further structural consequence is that $o$4 has the algebraic symmetries of a curvature tensor. In particular, the cyclic sum over any three arguments vanishes, and

$o$5

(Matveev et al., 30 Apr 2026). Thus the ambient module for quadratic top slot tensors is essentially the algebraic-curvature-tensor module (Matveev et al., 30 Apr 2026).

For compact irreducible symmetric spaces, the major theorem is that every quadratic Killing tensor is spanned by top-slot ones: $o$6 (Matveev et al., 30 Apr 2026). This is a vector-space spanning statement, not merely a statement about products (Matveev et al., 30 Apr 2026). It reduces classification of quadratic Killing tensors to classification of top-slot solutions at points.

On compact rank-one symmetric spaces of nonconstant curvature, the top-slot conditions simplify further. If

$o$7

then $o$8 is Killing iff

$o$9

or equivalently

$\mathcal L^d$0

(Matveev et al., 30 Apr 2026). Thus on rank-one spaces the top-slot problem is governed by two Poisson commutation relations.

This framework completes the quadratic classification on $\mathcal L^d$1 and $\mathcal L^d$2. On $\mathcal L^d$3, explicit top-slot tensors arise from the families

$\mathcal L^d$4

and

$\mathcal L^d$5

(Matveev et al., 30 Apr 2026). The first family yields decomposable tensors, while the second yields the indecomposable family from earlier work (Matveev et al., 30 Apr 2026). On $\mathcal L^d$6, a computer-assisted analysis shows that the space of top-slot quadratic tensors at a base point has dimension $\mathcal L^d$7, of which $\mathcal L^d$8 are decomposable and the remaining $\mathcal L^d$9 are precisely the indecomposable ones from earlier work (Matveev et al., 30 Apr 2026).

4. Prolongation, representation theory, and other meanings of “top slot”

A different but closely related use of the term occurs in prolongation theory. On $d$0 with the Fubini–Study metric, Killing tensors of arbitrary rank are classified using a Kählerian tractor connection (Eastwood, 2023). There, the prolonged parallel section has several components, and the first or highest symmetric component is exactly the original Killing tensor $d$1 (Eastwood, 2023). Theorem 2 of (Eastwood, 2023) identifies the solution space with tensors of Young type $d$2 that are $d$3-trace-free and annihilated by the derivation action of the complex structure: $d$4 (Eastwood, 2023). After complexification, this means the top slot has type $d$5 (Eastwood, 2023). In this model, the top slot is the leading symmetric covariant component of a prolonged solution, constrained representation-theoretically by $d$6 and $d$7-trace-freeness (Eastwood, 2023).

An even more explicit prolongation picture appears in the Young-symmetrizer treatment of the ordinary Killing equation. For a rank-$d$8 Killing tensor $d$9, the prolonged variables

$(M,g)$0

have Young symmetry $(M,g)$1, and the prolongation bundle is

$(M,g)$2

(Houri et al., 2017). In this representation-theoretic setting, the final prolongation variable $(M,g)$3, of shape $(M,g)$4, is the natural highest slot (Houri et al., 2017). For $(M,g)$5, all lower-slot integrability conditions vanish identically, and only the top-slot integrability condition survives, landing in Young type $(M,g)$6 (Houri et al., 2017). This identifies the highest irreducible derivative component as the decisive obstruction in the finite-type theory (Houri et al., 2017).

A different Hamiltonian usage appears in the theory of relativistic particles in external fields. If a conserved quantity is expanded as

$(M,g)$7

then the top slot is the highest-rank coefficient $(M,g)$8 (Igata et al., 2010). The generalized Killing hierarchy implies that this top coefficient satisfies a conformal Killing tensor equation,

$(M,g)$9

so the highest slot is conformal Killing in general and Killing only when $y$00 (Igata et al., 2010). This makes the top slot the leading symbol of a conserved quantity, but not by itself sufficient for the existence of the full invariant (Igata et al., 2010).

These usages are distinct. On symmetric spaces, top slot refers to maximal vanishing of lower normal-coordinate terms (Matveev et al., 30 Apr 2026). In prolongation theory, it denotes the highest irreducible component of a prolonged solution (Houri et al., 2017, Eastwood, 2023). In Hamiltonian hierarchies, it is the highest-rank momentum coefficient (Igata et al., 2010). The common structure is that one privileged highest-order piece controls the rest.

5. Reducibility, nonexistence, and rank-three phenomena

The top-slot viewpoint is especially effective in reducibility and nonexistence problems. In Weyl’s class of static axially symmetric vacuum spacetimes, every degree-$y$01 integral is reducible: $y$02 Equivalently, every valence-$y$03 Killing tensor can be written via symmetrized products of Killing vectors and quadratic Killing tensors (Vollmer, 2015). The key decomposition in the reduced Hamiltonian $y$04 is

$y$05

where $y$06 is the leading cubic part in the reduced momenta (Vollmer, 2015). In this setting, the “top slot” is precisely that leading cubic piece $y$07, which must be a metric integral on the reduced two-manifold: $y$08 (Vollmer, 2015). The analysis shows that in the generic nonflat case the parameter $y$09 controlling the would-be top component must vanish unless a strong compatibility condition holds, and when $y$10 one gets contradiction (Vollmer, 2015). Thus the leading cubic component is either absent or reducible (Vollmer, 2015).

For the Zipoy–Voorhees family, the statement is sharper: $y$11 Hence every valence-$y$12 Killing tensor is generated by the metric and Killing vectors alone (Vollmer, 2015). This is an explicit example where the top cubic slot carries no irreducible information.

Related nonexistence results were established computationally for stationary axisymmetric vacuum spacetimes. For selected Tomimatsu–Sato, C-metric, and Zipoy–Voorhees metrics, no additional independent Killing tensors exist up to valence $y$13, $y$14, and $y$15, respectively (Vollmer, 2016). Although (Vollmer, 2016) does not use the term “top slot,” its prolongation-and-parity decomposition

$y$16

isolates the highest reduced-momentum component in precisely the way a top-slot analysis would (Vollmer, 2016).

On two-dimensional tori, the closest analogue of a top slot is the highest trace-free harmonic of a symmetric tensor. A rank-$y$17 Killing tensor is determined by its higher harmonic uniquely up to lower-rank Killing data, and on the $y$18-torus this highest harmonic is explicitly controlled in isothermal coordinates (Sharafutdinov, 2014). For rank $y$19, existence reduces to the scalar condition

$y$20

(Sharafutdinov, 2020). This reveals the cubic “top component” as a holomorphic-harmonic datum subject to strong spectral and compatibility constraints (Sharafutdinov, 2020). A rigidity theorem shows that if both $y$21 and $y$22 satisfy the equation, then $y$23 must be one-dimensional, so the cubic integral is reducible through a Killing vector (Sharafutdinov, 2020).

These results correct a common misconception: higher-rank Killing tensors are not generically expected once lower-rank symmetries are present. In several important geometric classes, the highest-order slot is forced either to vanish or to factor through lower-order symmetries (Vollmer, 2015, Sharafutdinov, 2020, Vollmer, 2016).

6. Existence of genuinely irreducible highest-rank tensors and broader significance

The nonexistence and reducibility results are not universal. The Eisenhart lift of the rational Calogero model yields an $y$24-dimensional Lorentzian spacetime

$y$25

admitting irreducible Killing tensors of every rank up to $y$26 (Galajinsky, 2012). The rank assignment comes from the polynomial integrals

$y$27

of the Calogero system, each of degree $y$28 in the momenta (Galajinsky, 2012). In this family, the rank-$y$29 tensor associated with $y$30 is the natural highest-rank member and is explicitly stated to be irreducible (Galajinsky, 2012). This provides a counterpoint to reducibility theorems: highest-rank hidden symmetric tensors can arise as genuinely new objects.

Another useful comparison is $y$31, where every rank-$y$32 Killing tensor is generated by Killing fields (Eastwood, 2023). Eastwood’s surjectivity result

$y$33

shows that higher-rank Killing tensors on $y$34 introduce no new primitive generators beyond rank $y$35 (Eastwood, 2023). This is analogous to the Weyl-class reducibility picture, but established in a homogeneous Kähler setting and for all ranks (Eastwood, 2023).

The recent symmetric-space theory places these phenomena into a systematic framework. First, the study of Killing tensors on arbitrary symmetric spaces reduces to compact irreducible ones, via the product decomposition

$y$36

and duality of complexified Killing tensor algebras (Matveev et al., 30 Apr 2026). Second, top slot tensors provide a canonical finite-dimensional local model governed by curvature at one point (Matveev et al., 30 Apr 2026). Third, for quadratic tensors they span the entire solution space (Matveev et al., 30 Apr 2026). A plausible implication is that top-slot analysis isolates the genuinely new local degrees of freedom before decomposable and global symmetry-generated structures are reintroduced. The paper explicitly leaves open whether an analogous spanning result holds for all ranks $y$37 (Matveev et al., 30 Apr 2026).

In this sense, top slot Killing tensor fields serve as a local normal form for the hidden-symmetry problem on symmetric spaces (Matveev et al., 30 Apr 2026). They are not synonymous with indecomposable tensors, nor with tractor top components, nor with highest momentum coefficients, but they provide a common organizing principle across these contexts: identify the highest admissible or highest irreducible datum, impose the curvature or compatibility equations it must satisfy, and then determine whether the full tensor exists, is generated by lower-order symmetries, or yields genuinely new conserved quantities [(Matveev et al., 30 Apr 2026); (Houri et al., 2017); (Eastwood, 2023); (Igata et al., 2010)].